Corporate Turnaround And Corporate Governance - An Empirical Investigation Of The Role Of

Description
Corporate Turnaround And Corporate Governance - An Empirical Investigation Of The Role Of Ownership Structure In Corporate Turnarounds In Western European Firms

Master’s Thesis, 30 ECTS

M.Sc. in Economics and Business Administration,
Applied Economics and Finance (AEF)

Copenhagen Business School
November 12
th
, 2012





Corporate Turnaround
and Corporate Governance
-
An Empirical Investigation of the Role of
Ownership Structure in Corporate Turnarounds
in Western European Firms





Author: Anders Vest Hansen
CPR.no.: XXXXXX-XXXX

Academic Supervisor: Associate Professor, Bersant Hobdari
Institute: Department of International Economics and Management – The Center for
Corporate Governance (CCG)


Number of pages: 78
Number of characters incl. spaces: 180.259
Number of standard pages: 79.2
EXECUTIVE SUMMARY
In this thesis, I argue that the research of corporate turnarounds needs to look beyond the
classical elements in turnaround situations and instead look at governance functions in a firm to
fully understand the phenomenon and understand why some firms continue to underperform.
Therefore, I will empirically examine the relationship between ownership structure and
corporate turnaround performance and outcome.
In order to examine the suggested relationship, I employ a comprehensive sample procedure
to construct a heterogeneous panel dataset consisting of Western European firms experiencing a
genuine 6-year turnaround situation within the period from 1995 to 2010. I use fixed effect and
dynamic fixed effect panel models to test the relation between ownership structure and
turnaround performance, while I use logistic fixed effect model to test whether turnaround and
non-turnaround firms have significantly different ownership structures. In the empirical testing,
I also test the effect of both cost and asset retrenchment.
My results indicate that ownership concentration and turnaround performance and outcome
are not significantly related. I find dominant blockholdings to be weakly significant and
negatively related to turnarounds, while the entry of new blockholders and large shifts in the
ownership structure exert an insignificant effect on turnaround performance and outcome.
Turnaround performance is not affected by firm size, while turnaround firms are significantly
larger compared to non-turnaround firms. Although cost and asset retrenchment has a
pronounced position in the turnaround literature as the essential turnaround measure, I find cost
retrenchment to exert no influence on turnarounds. I find a negative relation between asset
retrenchment and turnaround performance and outcome, which is contrary to the advocated
effect. Generally, the findings are plagued by potential econometric issues, in which case the
findings are only suggestive. The results indicate the presence of endogeneity issues between
turnaround performance and the examined factors. In particular, the results indicate potential
endogeneity of ownership concentration.
In general, my thesis attempts to advance the understanding of governance arrangements in
the turnaround process by suggesting that unbalanced governance functions may be a
constraining factor in the turnaround effort. The results indicate that ownership structure is not
an effective governance mechanism in turnaround situations, thereby suggesting that governance
mechanisms have varying importance depending on the stage of a firm’s life-cycle.

Page 1 of 111
Table of Contents
1. INTRODUCTION .................................................................................................................... 4
1.1. Aim and relevance of this thesis ......................................................................................... 5
1.2. Problem statement ............................................................................................................... 6
1.3. Delimitations ....................................................................................................................... 7
1.4. Empirical approach ............................................................................................................. 8
1.5. Applicability ........................................................................................................................ 8
1.6. Thesis outline ...................................................................................................................... 9
2. THEORETICAL FRAMEWORK ....................................................................................... 10
2.1. Corporate Turnarounds ..................................................................................................... 10
2.1.1. Turnaround process .................................................................................................... 11
2.1.2. Two-stage turnaround model ...................................................................................... 13
2.1.3. The role of Turnaround models and strategies .......................................................... 14
2.2. Corporate Governance....................................................................................................... 16
2.2.1. Governance Life-cycle ................................................................................................ 17
3. BRIDGING THEORY AND EMPIRICAL FINDINGS IN THE HYPOTHESIS
DEVELOPMENT ...................................................................................................................... 20
3.1. Ownership structure .......................................................................................................... 20
3.1.1. Ownership concentration ........................................................................................... 21
3.1.2. Blockholder dominance .............................................................................................. 23
3.1.3. Change in ownership structure: Takeovers and block investments ........................... 26
3.1.4 Other aspects of ownership structure .......................................................................... 27
3.2. Structural differences in governance across countries ...................................................... 27
3.3. Summary of hypotheses .................................................................................................... 28
4. DATA AND METHODOLOGY ........................................................................................... 29
4.1. Turnaround cycle and measures ........................................................................................ 29
4.1.1. Turnaround cycle ........................................................................................................ 29
4.1.2. Turnaround measures ................................................................................................. 31
4.2. Sampling procedure........................................................................................................... 38
4.2.1. Sampling criteria ........................................................................................................ 38
4.2.2. Final Sample ............................................................................................................... 41
4.3. Data Sources and Sample Characteristics ......................................................................... 41
4.3.1. Validity and reliability of data .................................................................................... 43
4.4. Variables and measure definitions .................................................................................... 44
4.4.1. Performance measures: The dependent variables ...................................................... 45
4.4.2. Independent variables ................................................................................................. 45
4.4.3. Control variables ........................................................................................................ 47
4.4.4. Summary of variable definitions and data sources .................................................... 49
4.5. Descriptive statistics .......................................................................................................... 49
4.6. Empirical methodology and Econometric model specification ........................................ 53
4.6.1. Standard panel models ............................................................................................... 53
4.6.2. Dynamic models ......................................................................................................... 55
4.6.3. Logit models ............................................................................................................... 57
5. EMPIRICAL RESULTS AND ANALYSES ....................................................................... 60
5.1. Evidence from panel regressions: Estimation results ........................................................ 60
5.1.1. Fixed effect estimation results .................................................................................... 60
5.1.2. Dynamic panel estimation results ............................................................................... 63
Page 2 of 111
5.2. Robustness tests................................................................................................................. 65
5.2.1. Logistic estimation results .......................................................................................... 65
6. DISCUSSION ......................................................................................................................... 68
6.1. Firm turnaround performance and ownership structure – A question of endogeneity? .... 69
6.2. Corporate turnarounds: Too complex a phenomenon? ..................................................... 71
6.2.1. Retrenchment .............................................................................................................. 73
6.2.2. Firm size – Is size of importance? .............................................................................. 75
6.3. Econometric erosions, limitations and considerations ...................................................... 75
7. CONCLUSION ....................................................................................................................... 77
7.1. Future research .................................................................................................................. 78
LIST OF REFERENCES .......................................................................................................... 79
APPENDIXES ............................................................................................................................ 83

List of tables
Table 1: Summary of the hypothesized relation between ownership structure and turnarounds . 28
Table 2: Summary of the number of companies in the analysis .................................................. 41
Table 3: Sample description of industry group representation .................................................... 42
Table 4: Distribution of firms by country .................................................................................... 42
Table 5: Summary of explanatory and control variables ............................................................. 49
Table 6: Sample descriptive statistics .......................................................................................... 50
Table 7: Sample descriptive data represented for each year in the turnaround cycle period ....... 51
Table 8: Correlations between variables considered in this thesis ............................................... 52
Table 9: Fixed effect estimation results ....................................................................................... 61
Table 10: Results of dynamic panel regression with GMM and FE estimation .......................... 63
Table 11: Estimation results from logit fixed effects models of turnaround outcome ................. 66
Table 12: Summary of estimation results ..................................................................................... 68
Table 13: Illustration of the threshold levels for the Z-score models .......................................... 87
Table 14: Annual risk-free rates for each country........................................................................ 88
Table 15: Sample description of industry group representation in definition 2a ......................... 90
Table 16: Sample description of industry group representation in definition 2b ......................... 90
Table 17: Definition 2a: Descriptive statistics presented per year ............................................... 94
Table 18: Definition 2b: Descriptive statistics presented per year............................................... 95
Table 19: Variance inflated factors (VIF) tests ............................................................................ 96
Table 20: Means, standard deviations, and correlations for variables used in definition 2a ....... 96
Table 21: Means, standard deviations and correlation for variables used in definition 2b .......... 97
Table 22: Fixed effects panel estimation results without time dummies ..................................... 97
Table 23: Pooled regression estimation results ............................................................................ 98
Table 24: Random effects estimation results ............................................................................... 99
Table 25: Random effect estimation results with ROIC as dependent variable......................... 100
Table 26: Results of dynamic panel regression with GMM and FE estimation without time
dummies ..................................................................................................................................... 101
Table 27: Results of dynamic OLS regression ........................................................................... 102
Table 28: Dynamic panel GMM regression with ownership held exogenous ........................... 102
Table 29: Two-stage least squares estimation results of dynamic models ................................. 103
Table 30: Odds ratios from logit analysis of turnaround outcome ............................................. 104
Table 31: Estimation results from logit models of turnaround outcome with dummies ............ 104
Table 32: Odds ratios from logit analysis of turnaround outcome with dummies ..................... 105
Page 3 of 111
Table 33: Definition 2a: Pooled logistic regression ................................................................... 105
Table 34: Definition 2b: Pooled logistic regression ................................................................... 106
Table 35: Definition 2a: Logit estimation results presented yearly ........................................... 106
Table 36: Definition 2a: Yearly marginal effects ...................................................................... 107
Table 37: Definition 2b: Logit estimation results presented yearly ........................................... 107
Table 38: Definition 2b: Yearly marginal effects. ..................................................................... 108
Table 39: Table of the average top blockholders and ownership concentration ratio by country
.................................................................................................................................................... 108
Table 40: Comparing average retrenchment across industries .................................................. 109

List of figures
Figure 1: Turnaround model with consideration to the role of governance factors ..................... 19
Figure 2: Time structure of the panel data ................................................................................... 31
Figure 3: Illustration of the turnaround process including sampling criteria ............................... 41
Figure 4: Performance of sample firms during the turnaround cycle .......................................... 43
Figure 5: ROA of firms for definition 2a ..................................................................................... 91
Figure 6: ROA of firms for definition 2b ..................................................................................... 91
Figure 7: Z-score manufacturing firms, def. 2a ........................................................................... 92
Figure 8: Z-score non-manufacturing firms, def. 2a .................................................................... 92
Figure 9: Z-score manufacturing firms, def. 2b ........................................................................... 92
Figure 10: Z-score non-manufacturing firms, def. 2b .................................................................. 92
Figure 11: ROA plotted against ownership concentration ratio ................................................. 109
Figure 12: Distribution of ownership concentration in the sampling period ............................. 109
Figure 13: Level of asset and cost retrenchment during the turnaround period ......................... 110


Page 4 of 111
1. INTRODUCTION
Nearly every firm experience a stage in their life-cycle with declining performance threatening
firm survival. While some firms continue to decline and eventually fail others undergo
successful turnarounds and return to prosperity. From my perspective, what makes this aspect
interesting is the wide variation in responses to performance declines and in turnaround
outcomes across firms. The early turnaround literature (e.g. Hofer, 1980) state performance
decline as a strategic problem, which should be solved by management directing all resources
towards undertaking a strategic reorientation until the firm recover. However, following the
early ideas of Bibeault (1999), turnaround was argued to be much more than a strategic change
and was viewed as a process consisting of two phases; decline and recovery phase.
Following the classical study by Robbins and Pearce (1992), cost and asset retrenchment
was perceived to be the central key strategy in order to mitigate decline and ensure performance
recovery, and they argued this to be more effective than management’s selection of an
appropriate turnaround strategy. However, Pearce and Robbins (2008) later stressed
retrenchment as a component of turnaround strategy, where both strategic (e.g. repositioning in
the market, asset redeployment) and operational elements (e.g. cost and asset retrenchment)
could be combined in forming the overall strategy in the turnaround process.
Given the above consensus among the early turnaround researchers that these are important
elements when attempting to achieve turnaround, the question is why some firms continue to
underperform and never turn around. Suggested by the life-cycle theory, performance declines
can be viewed as an inevitable consequence of insufficient management, and thus insufficient
governance, over time resulting in misaligned strategy, structure, purpose, and financial
decisions that are increasingly at odds with the reality in the environment (Barker & Duhaime,
1997). Filatotchev and Toms (2006) questions the emphasis on retrenchment and state that
realignment of governance arrangements is a necessary condition for firms in a turnaround
situation, because misalignment may lead to managerial expropriation, performance
deterioration and declining shareholder value. That is that certain governance mechanisms may
suppress the management in taking the necessary actions in the turnaround process (Filatotchev
& Toms, 2006).
Large shareholders possess high influence on management decisions, which may impact
firm attributes such as performance, operational decisions and strategy (Fich & Slezak, 2008),
Page 5 of 111
suggesting that the right combination of corporate governance functions may help to mitigate
strategic and operational thresholds, and ensure the management undertake the needed
corrective measures in declining situations. As observed by Fich and Slezak (2008), the
governance structure of a firm is not uniformly effective in all situations, suggesting that
specific types of governance mechanisms may be more effectively than others on the turnaround
outcome.
Most studies investigate the role of governance on firm value, firm performance or
bankruptcy, where findings may not be applicable to turnaround situations. Few have directly
investigated the role of governance elements in turnaround situations (e.g. Filatotchev & Toms,
2006; Mueller & Barker, 1997). Most recent turnaround studies have differentiated the focus on
examining the role of top management (e.g. Abebe et al., 2011; Abebe, 2010), but very few have
made an direct effort to investigate the effect of governance aspects, and those who do include
governance factors have yet not reached a general consensus about the efficiency of governance
structure, why answers hereto remain ambiguous. Arogyaswamy et al. (1995) argue that
successful turnaround attempts must manage decline and recovery by changing firm strategy,
internal processes, and deal with causes of decline, whereas Filatotchev et al. (2006) advocate
governance arrangements must be aligned to ensure above actions and changes are
implemented. This perspective raises the question whether governance arrangements have an
impact on turnarounds?

1.1. Aim and relevance of this thesis
As a consequence of the above questions, this thesis seeks to identify the aspects of a firm’s
governance structure that affect the possibility of turnaround once the firm has been subject to a
severe and life-threatening performance decline. More specifically, I examine the influence of
ownership structure on turnaround performance and the probability of turnaround. That is to
what extent specific ownership characteristics set turnaround firms apart from firms which
continue to decline and/or eventually fail.
In investigating the above-mentioned governance aspect, I will draw on the existing
turnaround and governance literature. Most turnaround research has primarily been undertaken
in U.S. and U.K, thus holding a U.S.- and U.K.-perspective, leading to a limited knowledge
about turnarounds outside these contexts. In fact, very little turnaround research has been
undertaken in a Western European context, which motivates me to examine turnarounds in this
Page 6 of 111
setting. Most turnaround research has focused at retrenchment as a key element in the
turnaround strategy. Together with the fact that Filatotchev and Toms (2006) suggest that
governance arrangement may be a serious constraining factor in the turnaround effort, I assign
my attention to ownership structure and examine its role in the turnaround process and
relationship to successful turnarounds. Hence, understanding the role of governance aspects in
corporate turnarounds are important since misaligned governance functions may results in
continued poor performance or eventually end the existence of the firm.
Pandit (2000) notes that significant parts of the turnaround literature lack theoretical basis
or have been directly undertaken without connecting to theory afterwards. Consequently, the
research area corporate turnarounds do not entail a complete unifying theory (Pearce & Robbins,
1993). This raises the question whether research should (at all) be guided by theory. Fuglsang et
al. (2007) suggest that applied empirical research must be grounded in and build on extant
theory. Therefore, I will attempt to find a theoretical standpoint by drawing on existing agency
theory, life-cycle theory, resource-dependence theory, general governance perspectives, and
models within the turnaround literature.
The main motivation is that by addressing the role of ownership structure in turnarounds,
new aspects for advancing the understanding of the phenomenon of corporate turnarounds may
be provided.

1.2. Problem statement
Having introduced the considerations that make up my area of interest, the observations and
interest can be brought together in expressing a generic problem statement, which will serve as a
guiding tool for the theoretical direction, hypotheses development, methodology, sample
collection and analysis. As a result of the above attempt at problematising current scarce
empirical focus at how specific governance aspects play a role in corporate turnarounds in
Western Europe, the following is an appropriate overall foundation for an analysis of the topic:

“How does ownership structure influence corporate turnarounds in Western European
firms?”



Page 7 of 111
In order to address and answer the problem statement, I will address the following sub-question:
I. If ownership and corporate governance in general play a role in corporate turnarounds,
how is this expressed in the turnaround process?
II. How does ownership concentration influence turnaround firm performance and outcome?
III. How do large shareholders in the ownership structure exercise impact on the probability
of recovering from severe performance decline and successfully complete a turnaround?
Do dominant blockholders induce positive influence on turnaround performance?
IV. How do changes in the governance aspect influence the turnaround performance and
outcome of firms having experienced severe performance decline? For example, how does
the entry (or exit) of a dominant blockholder influence turnaround performance and
outcome. How does block investments by smaller blockholders impact turnarounds?
V. Is retrenchment a fundamental turnaround strategy among Western European firms?

1.3. Delimitations
There is possibly an infinite amount of factors other than the role of ownership structure that
may influence corporate turnarounds. Many different perspectives could be examined.
Corporate governance and corporate turnarounds are both two broad areas of literature, which
cannot be examined in one thesis. Hence, I focus at specific ownership characteristics and leave
other, yet interesting, characteristics for future studies.
To examine the relationship between ownership structure and corporate turnarounds, the
geographical scope of this thesis is restricted to companies across 15 countries in Western
Europe and the sample is gathered within 1995 to 2010. The countries considered as Western
European in this thesis are Denmark, Sweden, Norway, Germany, United Kingdom, Austria,
Belgium, Switzerland, Spain, Finland, France, Ireland, Italy, Netherlands, and Portugal.
Disclosure of ownership concentration became mandatory in late 1990’s and early 2000’ in
western European countries, whereas the information in many eastern European countries first
became available later. Hence, both countries and time period have been selected with
considerations to data availability.
It is not the scope of this thesis to examine the reasons of decline or the actual turnaround
measures taken during the turnaround process, but rather to examine the relation between
ownership aspects and corporate turnarounds by employing basic econometrics to the dataset. It
is not my attention to explain the underlying mathematics methods.
Page 8 of 111
Beside above delimitations that do not attend to raise any further questions, the remaining
delimitations will be presented and taken when necessary in order to allow for a more natural
reasoning and questioning of the reader, and thus not answer any questions that have not yet
been raised.

1.4. Empirical approach
Guided by the problem statement, this thesis draws on theoretical frameworks and prior
empirical findings to establish the foundation of the empirical analysis. I employ basic
econometric methodology to a panel dataset of financial and ownership information. I will
further assess the data by descriptive statistics and set up model specifications in order to
describe the relation between corporate turnarounds and ownership structure. The panel dataset
are better for identifying and measuring effects of the variables of interest that otherwise would
not be possible with cross-section and time series datasets.
I use three kinds of econometric methods; standard panel models to consider unobserved
firm heterogeneity, dynamic models to consider persistency in turnaround performance and
discrete choice (logit) models to estimate turnaround outcome. Compared to case studies, using
econometrics allows me to make inference concerning the relationship between turnaround and
ownership structure.

1.5. Applicability
In this thesis, I take a research philosophy that is linked to the positivistic approach. I will
attempt to create generalizable and objective findings, which are applicable to firms in a
turnaround situation. Further, this thesis takes a deductive approach by using theory as a starting
point and together with previous empirical findings, it moves towards giving answers to the
relation between ownership structure and turnarounds.
According to Fuglsang et al. (2007), the term validation should be connected to the research
to raise the question whether the research presents argumentations and answers to the actual
stipulated truth. That is if the research provides answers to what it intended. I argue that I by
enlightening the role of ownership structure in corporate turnarounds in Western European firms
can bring insight to the governance and turnaround literature, while contributing to broadening
the research area of turnarounds to include several countries and test theories in untested setting.
Page 9 of 111
As described by Fuglsang et al. (2007), the research results should not be affected by the
methodical approach or incidental factors. Fuglsang et al. (2007) define this as reliability,
meaning to which degree an identical study design investigating the same problem would find
corresponding results. A method is more reliable when it provides consistent results in different
environments. It is arguable whether this study would find similar results in a different setting.
However, if an equivalent study was conducted in the same context, it would presumably derive
the same findings and conclusions.
In terms of generalizability and robustness, the thesis will be not directly comparable to
previous turnaround research due to the uncommon/new geographical and industrial settings.
However, turnarounds do have relevance in a Western European setting, which hopefully will
lead other researchers to attempt the same approach and thereby enhancing the generalizability
of the findings. The thesis has its relevance by addressing unexplored questions and settings in
the turnaround literature, which hopefully add to advance the understanding of ownership
structure in turnaround situation and the role of ownership as a governance mechanism in the
decline life-cycle stage of Western European firms.

1.6. Thesis outline
The remainder of this thesis is organized by starting with section 2 that reviews the fundamental
theory of corporate governance and corporate turnaround. Section 3 looks at the empirical
literature while building the research hypotheses. Chapter 4 describes the considerations about
the sample procedure, methodology and sample, while reflecting on the data quality. Section 5
presents results from the empirical analysis, while section 6 discusses the results. Finally,
Section 7 presents a conclusion of the empirical findings to answer the raised problem
statement, while ending the section with future potential perspectives.


Page 10 of 111
2. THEORETICAL FRAMEWORK
I will in this theoretical section attempt to make a connection between the two theoretical
domains; corporate turnarounds and corporate governance. In doing so, I discuss the early
theory building of corporate turnaround and the latest attempts to conceptualize corporate
turnaround into a framework model. After that, I take the agency perspective of corporate
governance and discuss the relevant governance mechanisms for this study. In discussing the
theoretical domains, I draw upon theoretical ideas from life-cycle theory, which is built on the
resource-dependency and agency theory. I use life-cycle theory to connect the two theoretical
domains and get to a common level of theoretical understanding.

2.1. Corporate Turnarounds
A considerable amount of research studies the phenomenon corporate turnarounds with the
objective to distinguish firms that overcome severe performance decline and return to prosperity
from those who eventually fail to recover. An innumerable amount of factors has been suggested
as important influences on successful corporate turnarounds, e.g. management change,
retrenchment, strategy change, etc. Nevertheless, a definition of the phenomenon is appropriate
to start at a common level of understanding. Successful corporate turnaround can be defined as
when a firm undergoes an existence-threatening performance decline over a period of years but
are able to reverse the decline and adjust, end threat to survival, stabilise and make a substantial
and sustained positive change in performance to a more strong and thriving situation (e.g.
Barker & Duhaime, 1997; Bibeault, 1999; Robbins & Pearce, 1992; Bruton et al., 2003; Pandit,
2000). Importantly, firm survival would be doubtful without performance improvements (Hofer,
1980).
Despite researchers have had a common understanding of the term turnaround, turnaround
research have long been criticised for not being grounded in and built on existing theory (Pandit,
2000). Pandit (2000) argues that the consequence of the fact that early turnaround research has
been greatly based on case observations and have had a narrow focus on specific turnaround
aspect with no theoretical connections are that the potential opportunity of contributing to theory
building of the phenomenon turnaround has been missed. Thus, there has been no unifying
theory or single theoretic framework model to guide and/or generate questions to be asked and
answered. However, Bibeault (1999) and Pearce and Robbins (1993) were some of the first
Page 11 of 111
attempting to conceptualize the phenomenon by building a theoretical framework as a staged
model, helping to advance the understanding of turnarounds, while the more recently two-stage
theory was refined by Arogyaswamy et al. (1995) and later adjusted by Smith and Graves
(2005).
The turnaround literature can be divided into three associated areas: 1) cause and severity of
the turnaround situation, 2) successful turnaround response, and 3) the turnaround process
(Pearce & Robbins, 1993; Loui & Smith, 2006; Kazozcu, 2011). These three areas of literature
have provided the conceptual building blocks in forming the unified theoretical framework
models. The early models were more attended as an emergency and stabilisation attempt to stop
a firm’s financial crisis and improve strategy (Kazozcu, 2011), while later models additionally
emphasized the radical assessment of the cause of decline, severity of financial situation and
strategy, and organisational structure (Loui & Smith, 2006; Filatotchev & Toms, 2006). As the
models get more recent, they tend to incorporate additional aspects to the model that they
consider critical to each stage, which can be an indication of that prior models failed to capture
the complexity of the turnaround process (Arogyaswamy, 1995) or that the turnaround process
simply is unique to the individual firm but with a number of common turnaround measures.

2.1.1. Turnaround process
Hofer (1980) was among the first to divide recovery strategies in the turnaround process into
two different groups; operational (efficiency created through cost and asset retrenchment,
integration of production facilities) and entrepreneurial (innovation, asset reconfiguration,
repositioning in market). He argued that the type of strategy should be linked to the cause of
decline. If insufficient operations were the main reason of decline, then the company should
initiate efficiency-oriented recovery strategies. If the strategy was no longer appropriate, then
the company should remake the strategy to reflect the changes in the environment, i.e. an
entrepreneurial-oriented strategy (Graves & Smith, 2005). However, the later models viewed the
turnaround process as consisting of stages with both decline-stemming and recovery turnaround
actions (Bibeault, 1999; Pearce & Robbins, 1993; Arogyaswamy et al., 1995).
Bibeault (1999) is considered to be the first to approach turnaround as sequential processes
consisting of different phases. Based on his observations, Bibeault (1999) theorized four key
elements to make the turnaround effort work, which led to the introduction of a two-staged
model of turnaround. First, Bibeault (1999) stressed the importance of improvement of
Page 12 of 111
management processes, e.g. by a new and fully-supported management, and secondly, the firm
had to shrink back to profitable segments and its core competencies to ensure a financially
sound and competitively viable core business. Thirdly, bridge capital through debt
reconfiguration or negotiation was considered essential in order to provide sufficient financial
resources during the turnaround situation. Lastly, improving corporate culture and employee
motivation was deemed necessary to cultivate organisational momentum during the adversity.
These interdependent elements acted as the springboard for the two-staged approach.
Bibeault (1999) viewed the turnaround process as involving five interdependent phases,
which can be expressed by a two stage process. The first stage was directed towards the primary
objectives of addressing the survival-threatening performance decline by taking emergency
actions to end the bleeding, i.e. ensure a positive cash-flow. Concurrently, the firm should
initiate a stabilisation plan to create a viable core business, i.e. shrink back to those market
segments where the business can compete effectively and profitably. The means to achieve these
objectives include retrenchment, asset redeployment, working capital improvement, financial
restructuring, organisational restructuring, divestment, etc. (Bibeault, 1999). Further, the initial
stage also include the evaluation of current management and corporate board, as their
ineffectiveness and failure to recognize factors of decline often were suggested as being the
reason of failure to achieve successful turnaround (Bibeault, 1999). In accordance with Hofer
(1980), Bibeault stress the importance of turnaround actions being determined as a function of
the severity (liquidity crisis, negative cash-flow, potential bankruptcy, etc.) to ensure firm
survival. He also emphasize that turnaround actions should be balanced and adequate to the
situation, since an unbalanced combination of turnaround measures potentially could leave the
firm without the right critical resources to create a sound platform for recovery.
The subsequent stage was theorised by Bibeault (1999) to consist of a situation where the
firm had to decide whether to continue with its old strategy in a reduced and refined form, or
whether it should pursue a new strategic direction. Much in accordance with Hofer (1980),
Bibeault suggest strategic transformation as an essential alternative due to the fact that the
competitive landscape may have been permanently altered. Further, he argues a new strategy
could be necessary to align efforts towards the same objectives. Bibeault (1999) suggested
possible return-to-normal-growth/recovery strategies to encompass acquisitions, new
market/products, and focus on market share.
Page 13 of 111
Building on the concept composed by Bibeault (1999), Robbins and Pearce (1992) provided
evidence by investigating the textile manufacturing industry that retrenchment was a critical
element in the strategies undertaken by firms that achieved successful turnarounds. Further, they
found that the type and extent of retrenchment depended on the severity (measured by the
financial soundness of the firm) of the decline. They suggested that the retrenchment strategy
should progress from cost retrenchment to asset retrenchment as the severity of the turnaround
situation increases (Robbins & Pearce, 1992). They presented a multi-staged model consisting
of a retrenchment stage
1
and a recovery stage, and thus attempted to produce the theoretical
conclusion that retrenchment served as an essential tool both to stabilize the situation and to re-
ensure the viability of the firm. Retrenchment was regarded as an essential step in the initial
phase of the turnaround process, while strategic transformation (e.g. repositioning in the market)
was argued as the absolute last step that should ideally be initiated in the recovery phase (Pearce
& Robbins, 2008; Bruton et al., 2003).

2.1.2. Two-stage turnaround model
Consistent with Bibeault (1999), Pearce and Robbins (1993), Arogyaswamy et al. (1995), and
Graves and Smith (2005) viewed the turnaround process as consisting of a decline and recovery
stage and all proposed models of the turnaround process. They all had in common that they
viewed the crucial objective of the decline period as to stabilise the firm’s financial condition
and address cause of decline. They suggested that the firm should undertake decline stemming
strategies including turnaround actions such as improving efficiency by initiating cost
retrenchment, renewing the firm’s stakeholder support, supporting organisational motivation,
and stabilising internal environment (decision-making processes, responsibilities, and climate)
in order to achieve stabilisation, which is necessary for continuing with recovery strategies. In
their models, they stress that the decline-stemming strategies should be applied with
considerations to the severity of decline, size of the firm, and the level of available resources.
When stabilised, the firm should consider the causes of decline and competitive situation in
forming the recovery strategy. Before undertaking the recovery strategy, the firm should choose
whether to continue to pursue its current strategy in a reduced form or implement a more
growth-oriented (also mentioned as entrepreneurial-oriented) strategy.

1
The first initial stage in the turnaround process is named both the decline and retrenchment stage in the turnaround
literature. The two names describe the same stage.
Page 14 of 111
Previous model stress that decline stemming and recovery strategies should be executed
sequentially, although they accept that turnaround actions may be overlapping (e.g. Pearce &
Robbins, 1992). However, the major contribution of Smith and Graves’ (2005) model is that it
accepts that the two phases may be executed simultaneously due to firm-specific circumstances.
This was theorized based on case observations, where turnaround actions were observed to be
executed simultaneously in practice.

2.1.3. The role of Turnaround models and strategies
Along with the extensive portion of theoretical development in the turnaround literature, there
has also been a significant academic research emphasis on examining the turnaround process
and strategies in the decline and recovery stage. These strategies are generally considered as
operational (cost and asset retrenchment), strategic (repositioning), or entrepreneurial
(innovation) (Smith & Graves, 2005). These turnaround measures have been theorized to be key
factors in leading to successful turnaround outcome and thus play a significant role in the
theoretical framework models. Despite there is a consensus about the elements of the turnaround
process, there has been a significant debate about the importance and effect of the individual
turnaround strategies. Several studies examine the effect of retrenchment but find ambiguous
results (Loui & Smith, 2006). Barker and Mone (1994) ignited the debate about retrenchment by
arguing that retrenchment is not an essential element of any turnaround strategy. Rather, they
argued that retrenchment was a consequence of severe and rapid performance decline.
Supported by Barker and Duhaime (1997) and Arogyaswamy et al. (1995), they question the
value of retrenchment and argue that sole focus on retrenchment initiatives may obscure,
exacerbate and even reduce the chance of recovering successfully by reducing morale and
available resources. Especially Arogyaswamy et al. (1995) view retrenchment as to drastic and
detriment compared to other turnaround actions. Morrow et al. (2004) suggest that the effect of
retrenchment on turnaround performance is contingent on industry dynamics. This implies that
retrenchment initiatives even may be unnecessary and even counter-productive in some
situations (Filatotchev & Toms, 2006). In this connection, later theoretical building suggests that
strategic reorientation and actions may be sufficient to ensure successful turnaround.
The discussion about the significance of turnaround strategies and the continuous
incorporation of additional factors in the theoretical framework models may be a result of that
the turnaround process is more complex than yet modelled. Additionally, the turnaround
Page 15 of 111
outcome may be contingent upon an additional number of factors individual to each firm. This
suggest that there should be brought more focus to address the individual details of the many
factors that lead to successful turnaround outcome. In response, many researchers (e.g. Barker &
Mone, 1994; Bibeault, 1999; Abebe et al., 2011) suggest that managerial response or the actual
failure to respond to performance decline may exercise a significant role in the turnaround
process, indicating that management possess great influence on ability to successfully complete
a turnaround. Therefore, it is widely suggested that change in top management is an important
precondition of successful turnarounds (e.g. Bibeault, 1999). This perspective assumes that
incumbent top management during performance decline is not always able or willing to
undertaking necessary drastic changes and execute turnaround strategies. Managerial inaction
may be due to denial of the situation, lack of competencies, self-interests, or an attempt to retain
reputation. Similar, the replacement of managers is unlikely to happen in situations with weak
governance structure and where management is entrenched (Lai & Sudarsanam, 1997). As
observed, management change is often required to regain and reshape credibility, while ensuring
initiation of turnaround actions (Mueller & Barker, 1997; Abebe et al., 2012). However,
Filatotchev and Toms (2006) suggest that managerial inaction may not be due to insufficient and
poor management. They argue that governance functions may have a constraining effect on the
top management in strategic decisions and turnaround initiatives during the turnaround process.
Filatotchev and Toms (2006) extends the two-stage turnaround model presented by Robbins
and Pearce (1992) and suggest a governance-based model that includes governance factors.
Their model suggests that management may be heavily prevented from undertaking turnaround
actions by governance constraints, which may be a result of diverging and misaligned interests
of governance groups and arrangements (Filatotchev & Toms, 2006). They criticize prior
models for assuming that the firm without difficulty can stabilise decline and enter into the stage
of recovery. Building on observations by Bibeault (1999), Filatotchev and Toms (2006) argue
that necessary governance preconditions must be present before initiating turnaround actions.
Specifically, they suggest that there must be a consensus and alignment of objectives between
principal- and agent groups because if misaligned it may create significant constraints on the
management in the turnaround process, which may form significant insurmountable thresholds
to the firm’s turnaround. Thus, they argue that turnaround cannot be sensibly examined without
consideration to governance arrangements and interests of governance groups in turnaround
situations (Filatotchev & Toms, 2006).
Page 16 of 111
This perspective suggests discussing corporate turnarounds from a governance perspective.
The governance perspective can be guided by the governance life-cycle theory that suggests that
optimal governance arrangements differ during the stages of the firm life-cycle as a result of
changes in the firm’s strategic dynamics, e.g. competitive challenges, financial soundness,
performance, etc. (Filatotchev et al., 2006), which may help understanding the role of
governance aspects in turnarounds.

2.2. Corporate Governance
The agency theory originates from the early ideas developed by Berle and Means (Denis &
McConnell, 2003), which was later formalised by Jensen and Meckling (1976), describing the
potential benefits and conflicts arising from the separation between ownership and control.
According to Jensen and Meckling (1976), the firm is represented by a nexus of contracts
between agents (management) and principals (owners), where the agent is hired to execute
activities, i.e. control and decision-making of the firm, on the behalf of the owners. The central
premise of the agency theory is that this situation may give rise to potential interest and
preference conflicts between the agent and principal. Without the appropriate incentives,
managers may engage in self-interested and non-optimal behaviour that may be inconsistent
with the value-maximizing interests of the owners (Jensen & Meckling, 1976), and since
complete and costless monitoring and controlling is difficult, this can potentially create agency
problems such as insufficient efforts and managerial opportunism, e.g. empire-building,
expropriation, extensive self-dealing (Becht et al., 2003). To mitigate the agency problems, and
thus reduce the derived agency costs, shareholders may use a wide range of governance
mechanisms to induce self-interested managers to take actions that are more in their interests
(Denis & McConnell, 2003). From a theoretical perspective, these ideas have had by far the
most profound impact on the development of governance theory (Filatotchev et al., 2006).
Governance mechanisms hold the objective of minimizing the misalignment of interest
between management and shareholders, and constrain managerial opportunism. As summarized
by Denis and McConnell (2003), the mechanisms can be broadly categorized as either internal
or external to the firm. Internal mechanisms are primarily large blockholders, board of directors,
managerial remuneration and incentives contracts. The primary external mechanisms are the
external market for corporate control (takeovers), the managerial labour market and the legal
protection system. As suggested by Filatotchev et al. (2006), the right combinations of the
Page 17 of 111
firm’s governance arrangement (mechanisms) may reduce agency costs and align the interests of
agent and principal groups, helping the firm to overcome its strategic thresholds during its life-
cycle. Based on this point of view, different stages during the life-cycle may demand different
balances of governance arrangements.

2.2.1. Governance Life-cycle
Filatotchev et al. (2006) criticize that most governance studies has concentrated on the largest
mature firms and mainly has been guided by agency theory, which consequently has resulted in
a narrow theoretical perspective on governance. They argue that the neglected investigation has
lead to a limited understanding of the variation in agent-principal relationships and governance
arrangements throughout the entire life-cycle of firms. By building on the life-cycle theory that
has the central premise that the firm’s strategic dynamics vary across the different stages of the
life-cycle, they present a fundamental framework that integrates these strategic dynamics with
the changes in governance arrangements. The model extend the perspective of governance by
going beyond agency theory in suggesting that governance roles such as resource and strategy
functions may have equal important roles alongside monitoring and control functions in the
decision-making processes (Filatotchev et al., 2006)
2
.
The framework combines a resource-dependency
3
and agency perspective in describing
how balancing governance arrangements during life-cycle may help the firm in overcoming its
strategic thresholds that may exists as a result of changing firm dynamics, and that failure to
adapt its governance arrangements may create significant barriers to overcome adversity and
transit from on life-cycle stage to another (Filatotchev et al., 2006).
Filatotchev et al. (2006) illustrate that a firm will likely experience a dramatic shift in the
life-cycle stages from having a significant accumulated resource-base and several seize-able
business opportunities as it matures to be in a stage of decline, having exhausted its business
opportunities, possibly over-diversified into unrelated and non-core activities, over-expanded,

2
As stated by Filatotchev et al. (2006), governance is originally viewed as ensuring accountability and
responsibility of management and to minimize the downside shareholder risk. However, it is also about enabling
top management to seize positive business opportunities, allowing shareholders to also benefit from upside business
potential. They refer to this conception as wealth-creating (resource and strategy function) and wealth-protecting
(monitoring and control) aspects of governance.
3
In general and very basic, resource-dependence theory is concerned with the access to external resources and how
these affect the possibilities of the firm. Filatotchev et al. (2006) illustrate that governance structure and
arrangements may from a resource-based perspective affect the creation of a unique resource base that may give the
firm a competitive advantage.
Page 18 of 111
and with increased managerial rent-seeking opportunities. This will most likely lead to
performance deterioration. They stress that the value-protecting (monitoring and control) aspect
of governance becomes increasingly important for declining firms because firms at this life-
cycle stage may develop less effective and unbalanced governance functions, which potentially
generates severe agency problems, thus increasing the role of governance. This view is
consistent with the turnaround literature (e.g. Mueller & Barker, 1997; Bibeault, 1999).
Filatotchev et al. (2006) argue that the unbalanced governance functions in the decline life-
cycle stage are insufficient in preventing opportunistic behaviour by management or ensure
performance recovery. In the perspective of turnaround, this indicate that realignment and
rebalancing of governance arrangements and incentives may be a necessary condition in
ensuring that a firm obtain balanced and well-established governance structure efficient for its
given stage. Furthermore, this will help the firm to mitigate potential problems and overcome
the turnaround situation, thereby transit through its strategic and operational thresholds.
However, governance factors may also be a constraining factor, creating significant barriers to
the transition from one threshold to another, which may explain why some firms fail to
successfully complete a turnaround and continue to underperform.
According to Filatotchev et al. (2006), (outside) blockholders should enhance their role as
firms approach decline in performance since effectively monitored firms are more likely to
engage in refocusing and downsizing. However, the turnaround situation may be associated with
another threshold which affects the balance between the governance functions. In order to
reverse decline, the value-protecting (monitor and control) role may be less significant. Instead,
lower monitoring may be necessary to increase the flexibility of the firm in the turnaround
situation, while the value-creating role may increase in importance by providing the
management with resources (e.g. bridge capital, knowledge, and skills) and strategic advice in
the decision-making process (Filatotchev et al., 2006; Mueller & Barker, 1997). Thus, the
theoretical model establishes the foundation that corporate governance has significant different
roles during the stages of a firms life-cycle, and that blockholders hold an important role in
ensuring that declining firms overcome their operational and strategic barriers (Filatotchev et al.,
2006). Together with the statement from Filatotchev and Toms (2006), turnaround outcomes are
cannot be examined without consideration to governance arrangements and especially the
function of owners.
Page 19 of 111
Based on the discussed frameworks, my theoretical approach and the connection between
turnaround and governance can be illustrated in a simple framework model as below in Figure 1.
Figure 1: Turnaround model with consideration to the role of governance factors
TURNAROUND PROCESS
Turnaround situation Decline stage Recovery stage Outcome





The model is based on the ideas and theoretical two-stage framework models presented by Bibeault (1999), Robbins and Pearce (1992), Pearce and
Robbins (1993, Figure 1), Arogyaswamy et al. (1995, Figure 1), Graves and Smith (2005, Figure 1), Filatotchev and Toms (2006), and Filatotchev et al.
(2006, Figure 1 and Figure 2).

The simple turnaround-governance model illustrates continuously balanced governance
arrangements and functions as a prerequisite in the turnaround process, which otherwise may act
as a constraining factor.

Turnaround situation:
- Cause of decline
- Firm size.
- Severity (Z-score).
Decline-stemming
strategies:
- Stakeholder support.
- Efficiency through cost
and asset retrenchment.
- Stabilize internal climate
and decision-making.
Stability
for
recovery
strategies
,
Realignment and balance of governance arrangements and functions as a necessary factor
Recovery
strategies:
- Growth-oriented
- Entrepreneurial /
innovation.
- Efficiency-
oriented.

Extent of
recovery
/ Turn-
around
Page 20 of 111
3. BRIDGING THEORY AND EMPIRICAL FINDINGS IN THE
HYPOTHESIS DEVELOPMENT
The theoretical foundation above depicts how the extent of turnaround actions, turnaround
strategies and turnaround outcome are contingent upon an undetermined number of factors. The
governance life-cycle theory and the governance turnaround model suggest that (internal and
external) mechanisms of corporate governance exercise an important effect in the firm’s ability
to achieve a turnaround and must be aligned and rebalanced during the turnaround process.
I restrict my attention to one potentially important corporate governance mechanism:
Ownership. A theoretical understanding of the mechanism is necessary to understand its
potential constraining effects and influence on turnaround outcome. I now review empirical
literature on governance and turnarounds, and bring the theoretical and empirical implications
together in building the hypotheses necessary to answer my research questions.

3.1. Ownership structure
The importance of ownership and its complication for firm performance has been widely
examined, which especially are a consequence of the agency theory discussing the separation of
ownership and corporate control. Concentrated ownership among few shareholders – also
characterised as blockholders or large shareholders – often constitute a significant part of the
ownership structure. A blockholder can be an individual, organisation or entity, which as a
consequence of their large ownership position may possess the ability to influence decisions in
the firm and to provide effective oversight (Lai & Sudarsanam, 1997; Denis & McConnell,
2003). A shareholder holding 5 pct. or more of a firm’s equity is characterized as being a
blockholder (Holderness, 2003)
4
. In the following discussions, I use the term blockholders to
refer to large outside stockholder. Outside shareholders are distinct from inside shareholders,
e.g. managers. The two groups of blockholders are likely to have conflicting interests, and
outside blockholders are more likely to monitor the firm (Denis & McConnell, 2003).
Taking the perspective of the agency theory, Bethel and Liebeskind (1993) argue that a
manager’s wealth increases more through diversification than through maximisation of
shareholder value. Unless the management are compelled to take turnaround actions by

4
In discussing ownership, I use the term ownership to indicate shareholders cash-flow rights, while the actual
control depends on the voting rights held by the shareholder. When using agency theory in the discussions,
blockholders are assumed to hold equal ownership and control rights.
Page 21 of 111
blockholders, the willingness of the management to shrink back to the core business, restructure,
and undertake retrenchment activities may not be in their interest (Bethel & Liebeskind, 1993).
Hence, blockholders is deemed as an efficient governance mechanism for solving the agency
problems by ensuring initiation of turnaround strategies.
5
I focus at three elements of the
shareholder structure; concentrated ownership among a number of blockholders, dominant
ownership by a single blockholder, and changes in ownership structure.

3.1.1. Ownership concentration
As described by Holderness (2003), two aspects motivate blockholding by shareholders: 1)
shared benefits of control and 2) private benefits of control. Shared benefits of control arise
from the superior monitoring that blockholders can perform as a result of their concentrated
ownership and its accompanied rights (Holderness, 2003). Private benefits of control may arise
from blockholder holding the incentive to use their power and influence on management to
expropriate firm resources and enjoy private benefits at the expense of minority shareholders
(Holderness, 2003).
As a consequence of blockholders large interest in a firm’s equity, the agency theory
suggests that blockholders have the incentive and power to actively monitor management and
operations more effectively than minority shareholders. Even though this makes smaller
shareholders able to free-ride on the monitoring effort carried out by blockholders, blockholders
reduce the free-riding problem related to disperse ownership, which benefit all shareholders
(Holderness, 2003). Holderness (2003) argues that as the ownership position of blockholders
increases, the blockholder has an increasing incentive to ensure effectiveness of value-creating
and -enhancing activities, which is assumed to positively affect firm value. Larger ownership
concentration among several blockholders will reduce monitoring costs related with dispersed
ownership, increasing the effectiveness of ownership monitoring and controlling compared to
low ownership concentration (Shleifer & Vishny, 1997).
Blockholders has the ability to channel or pressure their opinions through the board by
appointing directors or by incorporate managers in the firm to represent their interests, allowing
blockholders to ultimately affect management decisions and firm activities for the better (Becht

5
In practice, many blockholders (e.g. institutions) may despite having a large ownership position in a firm actually
seek to diversify their investments and risk rather than taking an active monitoring and controlling role. For
example, many large European institutions are criticized to be passive in spite of their significant blockholdings and
not exercising their ownership rights (Nielsen, 2012).
Page 22 of 111
et al., 2003). Consequently, blockholders are likely to seek influence in turnaround situations,
where performance and firm value are declining and in jeopardy, to ensure the management
reverse decline and pursue viable turnaround strategies.
6
However, blockholders with no
connection to the board or management will have limited or no impact on the turnaround
process (Holderness, 2003). In context of the resource-dependence theory, blockholders may
also serve as a potentially valuable link to external and superior resources in the external
environment (e.g. to suppliers, customer groups, industry knowledge, information reducing
uncertainty, operational and strategic alliances, etc.). In return the firm may provide the
blockholder with influence in the decision process (Becht et al., 2003). The theoretical
perspectives suggest that blockholders are able to impact the turnaround process and have a
positive effect on turnarounds.
The empirical evidence on the topic of ownership concentration is extensively summarized
by Denis and McConnell (2003), who state that the empirical findings generally show a positive
relationship between performance (both in terms of market value and profitability) and
concentrated ownership in the hands of blockholders non-U.S. studies. For example, they
summarize a number of European studies that find a positive impact of ownership concentration
on performance. More importantly, their review reveals that firms with blockholders are more
prone and faster to restructure in response to performance decline, suggesting a positive
relationship between concentrated ownership and turnarounds. In addition, findings show that
blockholders are more likely to appoint independent directors.
7
Further, management turnover
are found to be higher in firms with more concentrated ownership structures, suggesting
blockholders are more likely to discipline poor top management (Denis & McConnell, 2003).
However, Denis & McConnell (2003) and Becht et al. (2003) summarize empirical findings
that also report conflicting evidence, which overall suggest no significant relationship between

6
Blockholders may seek influence in a turnaround situation by 1) seeking to change the ownership structure, e.g. by
increasing their holding, or to change firm activities (e.g. by altering board structure to avoid board members
become too entrenched and thereby undermine their role of monitoring, elect more independent directors, etc.), or
2) form collisions with other blockholders to enhance their collective influence. Two other options are, although
these do not result in influence, is to either 1) exit the ownership position or 2) accept the situation and take no
action.
7
Both in the resource-dependence theory and agency theory, board structure and composition affect firms’ ability
to recover from decline. Specifically, the two theoretical perspectives suggest that independent (outside) board
members bring value to the governance mechanism by limiting the managerial self-serving behaviour and provide
the firm with valuable key resources otherwise unavailable or limited to the firm. Thus, both theories propose that
the proportion of independent directors positively affect a firm’s ability to successfully complete a turnaround,
which suggest that blockholders may be positively associated with turnaround performance and outcome.
Page 23 of 111
ownership and firm performance measured by different performance measures. Denis and
McConnell (2003) summarize that the relationship between firm performance and the type of
blockholders (e.g. companies, institutions, families, private, etc.) largely depend on who the
largest blockholders are and the context of the study. Depending on the type of blockholder and
context, ownership concentration is showed to affect firm performance both adversely and
positively, while others find no relation at all. Based on their literature survey, Denis &
McConnell (2003) conclude that most non-U.S research find a positive relationship between
ownership concentration and firm performance, when the relationship is significant. Hence,
consensus is yet to be reached within the literature.
The part of the literature reporting insignificant results between ownership concentration
and firm performance may be a result of endogeneity. Initially hypothesised by Demsetz and
Lehn (1985) and shown in later studies (Denis & McConnell, 2003), ownership is endogenous.
Thus, a firm’s ownership structure will be firm-specific and adjusted to the most appropriate
state given the firm characteristics and situation. Thereby, it is difficult to observe any
significant relationship between ownership and performance.
Bringing the theoretical and empirical observations together, I embrace the theoretical
perspective and the part of the literature suggesting that concentrated ownership among
blockholders is to be considered as an effective governance mechanism positively related to
performance. The first hypothesis is formulated as follows:

H1: Turnaround firms will have a greater ownership concentration situated in the hands of
blockholders than non-turnaround firms, meaning that ownership concentration is positively
associated with turnaround performance and the extent of turnaround outcome.

3.1.2. Blockholder dominance
The selected turnaround strategy is likely to be determined based on the power and dominance
of the blockholder(s) present in the firm. The size of a blockholders ownership position is a
good indicator of the blockholders power and ability to exert influence or even dominate
decision-making (Lai & Sudarsanam, 1997; Jostarndt & Sautner, 2008). The ownership and
control structure may differ significantly as some blockholder may hold large voting rights but
relatively small cash-flow rights, giving them different incentives than blockholders with large
cash-flow rights. In this thesis, I only gather data on cash-flow rights. Therefore, I take the
Page 24 of 111
perspective of blockholders with incentives stemming from cash-flow rights rather than voting
rights.
In their review of ownership research, Denis and McConnell (2003) find a presence of high
ownership concentration in western European firms. The fact that multiple blockholder are
common is supported by Laeven and Levine (2008). Similar, they state that majority ownership
(determined by cash-flow rights) by a single blockholder is not uncommon. Within the agency
theory, managerial opportunism and expropriation by large controlling shareholders is assumed
to be a primary concern. However, blockholders may become more proactive and involved as
their wealth in the firm increases, which makes Jostarndt and Sautner (2008) arguing that the top
blockholder rather than the total group of blockholders exert the strongest monitoring effect. To
develop the concept of blockholder dominance, Lai and Sudarsanam (1997) characterise the
dominant blockholder as being the (top) blockholder with the largest ownership position in the
total firm equity. Dominant shareholders are benefiting from appreciation in share price and
dividends, why a performance decline may have a considerable negative wealth-effect on the
largest blockholders (Jostarndt & Sautner, 2008). Dominant blockholders are expected to
disfavour equity-based strategies such as dividend cuts and equity issue, while having the
incentive to favour operational, strategic, managerial, and debt strategies (Lai & Sudarsanam,
1997), which do not require issue of new equity. Therefore, dominant shareholders are
ultimately expected to influence management to initiate the necessary turnaround strategies due
to their significant cash-flow rights and wealth-constraints (Lai & Sudarsanam, 1997; Jostarndt
& Sautner, 2008). Based on these perspectives, the top blockholder may have a positive impact
on the turnaround outcome.
Jostarndt and Sautner (2008) point out the fact that blockholders may exert heavy resistance
to equity-based measures as a part of a turnaround strategy, which may jeopardize firm-survival.
For example, threat of firm-existence arises from the fact that a creditor may withdraw
necessary bridge capital and support during the turnaround process. Bibeault (1999) mentions
unwillingness and ignorance among large blockholders as one among many reasons of
turnaround failure. Contrary to the previous implications, this aspect suggests that dominant
blockholders present a constraining factor in the turnaround process. Similar, a large ownership
position provide control only available to the dominant blockholder, which may encourage and
give incentive to expropriate corporate resources on the expense of smaller shareholders (Denis
& McConnell, 2003). However, the incentive to reap private benefits of control may diminish in
Page 25 of 111
a turnaround situation where firm-survival is threatened unless performance improves. The
review by Laeven and Levine (2008) supports the existence of a positive relationship between
the large cash-flow rights of the dominating blockholder and firm performance in terms of firm
value. They suggest that large and concentrated cash-flow rights in the hand of a dominant
blockholder discourage and reduce incentives to divert and expropriate firm resources. Hence,
dominant blockholders may hold an increasing incentive to use their influence on management
to initiate turnaround strategies as firm failure will discontinue the blockholders ability to divert
corporate resources.
Much related, there is a gradual convergence towards stronger legal shareholder protection
and enhanced governance systems, which continuously reduce the degree of private benefits
derived from dominant ownership (Denis & McConnell, 2003). For example, private benefits
extracted from large blockholdings are in some countries close to zero (Denis & McConnell,
2003), indicating dominant blockholdings may be either positive or insignificantly related to
turnaround outcome and performance. Findings suggest that firms with majority blockholders
are weakly more profitable than firms with few shareholders, while other find no relationship
(Denis & McConnell, 2003).
Based on both theory and empirical literature, there has been established a relationship
between dominant blockholdings and performance that provide two competing perspectives.
Therefore, having these two perspectives in mind that dominant blockholders may be either
positively or negatively related to performance and turnaround, the second hypothesis will be
divided into two hypotheses:

H2a: Turnaround firms are more likely to have a dominant blockholder than non-turnaround
firms, suggesting that a blockholder with a dominant ownership position is positively associated
with the extent of turnaround performance and outcome.

H2b: Turnaround firms are less likely to have a dominant blockholder than non-turnaround
firms, suggesting that a blockholder with a dominant ownership position is negatively
associated with the extent of turnaround performance and outcome.

Page 26 of 111
3.1.3. Change in ownership structure: Takeovers and block investments
Although stated that this thesis does not involve examining the role of external governance
mechanisms, I still consider the role of takeover activity. However, I consider takeovers as
change in ownership and to encompass several events. Takeovers cover the following three
events in my thesis: 1) the acquisition of majority block of shares, i.e. ownership position equal
to 50 pct. or above, by a new blockholder from an incumbent dominant blockholder, 2) the
acquisition of a majority block of shares by a new shareholder without a prior dominant
blockholder, or 3) the increase in equity position of an incumbent blockholder without prior
dominant blockholding status. Therefore, the term takeovers are related to large changes in the
ownership structure.
Lai and Sudarsanam (1997) argue that large changes in ownership may act as a necessary
catalyst of restructuring. They find that significant managerial, asset, and financial restructurings
are undertaken following large ownership changes. New large owner(s) introduce renewed
monitoring and control which often is associated with replacement of current management. In
the observation of boards in turnarounds, Bibeault (1999) describe that boards are not the reason
of decline, but that poor management actions are the reason of decline. Therefore, management
change may lead to firm performance improvements (Bibeault, 1999). Management change has
been accepted in the turnaround literature and has been advocated necessary to impose new
understandings of the business (Mueller & Barker, 1997; Abebe et al., 2012), suggesting
takeovers or large ownership changes may impact turnarounds positively.
Furthermore, Lai and Sudarsanam (1997) show that subsequent large block acquisitions
increase firm value and are either maintained or further enhanced in the following 3-year period.
Assuming performance follows the value increase, the findings suggest that large block
acquisitions may be positively related to turnaround performance and the ability to successfully
turn around. Bethel and Liebeskind (1993) show that block investments, i.e. ownership
investment below 50 pct., are positively related to restructuring activities and subsequent
improvement in performance. They suggest that the entry of new blockholders have a
disciplinary influence on management, pressuring the management to initiate turnaround
measures, and that new blockholder may bring resources otherwise unavailable to the firm.
However, they also summarize previous empirical evidence suggesting an insignificant relation
between acquisitions of large minority-blocks of equity and firm performance, indicating block
investments have no affect on firm performance. Additionally, it stresses the fact that there is a
Page 27 of 111
difference between majority blocks, i.e. above 50 pct. of firm equity, and minority block, i.e.
between 5 pct. and 50 pct. of total firm equity. In practice, takeovers do not seem to be an
important mechanism in Europe and are rarely taking place (Denis & McConnell, 2003).
However, my perception of takeovers, which cover several events of large ownership changes,
is helpful to examine the role of shifts in the ownership structure in distressed situations.
Due to the different size of changes in ownership, I distinguish between takeovers and block
investments to differentiate between large minority-block (block investments) and majority-
block (takeover) investments. As discussed above, the perspective on takeovers and block
investments is suggested to be positive related to turnarounds. Consequently, the last hypotheses
can be constructed as below:

H3: The extent of turnaround performance and turnaround outcome is positively associated
with takeovers, implying that turnaround firms are more likely have experienced a large change
in its ownership structure.

H4: Turnaround firms are associated with a higher degree of block investments than non-
turnaround firms, meaning that the extent of turnaround performance is positively related to
block investments.

3.1.4 Other aspects of ownership structure
In assessing the relationship between ownership structure and turnarounds, I have emphasized
examining the above aspects. It is an undisputed fact that other ownership effects exert a likely
significant impact. For example, the type of blockholder, e.g. management and private, are
likely to posses conflicting objectives and incentives, and the identity of the blockholder has
important influence on the effect on turnarounds. Laeven and Levine (2008) mention pyramids
and collaboration as factors that affect the actual control in a company. However, I have
restricted my thesis to investigate the discussed ownership aspects.

3.2. Structural differences in governance across countries
With respect to ownership structure, country differences have been found to be reflected in both
in ownership structure and firm performance (Denis & McConnell, 2003), while La Porta et al.
(2000) argue that investor protection, which may differ substantially due to its legal origin (e.g.
Page 28 of 111
common vs. civil law), has a significant influence on the incentive to hold a large ownership
position. Similar, I include several countries in my sample and these may contain possible
important national differences in their regulatory frameworks, external governance mechanisms,
financial systems, and institutions. This may introduce different interactions and substitutions
between mechanisms depending on the specific country. In response to these possible country-
specific effects, I will later elaborate and introduce variables to address such differences.

3.3. Summary of hypotheses
Table 1 summarizes the suggested relationships between ownership structure and corporate
turnarounds.
Table 1: Summary of the hypothesized relation between ownership structure and turnarounds
Hypothesis Ownership aspect Hypothesized effect on turnaround
Hypothesis 1 Ownership concentration ( + )
Hypothesis 2a and 2b Blockholder dominance ( + ) / ( - )
Hypothesis 3 Takeover ( + )
Hypothesis 4 Block investment ( + )
The hypotheses are formulated under the ceteris paribus condition.


Page 29 of 111
4. DATA AND METHODOLOGY
This section has the objective to set up the overall configuration on how to analyse the
hypothesized role of ownership structure in corporate turnaround in Western European firms. It
requires a well-considered and well-defined approach of the phenomenon turnarounds to select
representative cases for the analyses. Therefore, the first steps are to devise an overall definition
of the phenomenon corporate turnarounds based on previous research frameworks. That is the
performance measures, the cycle period, benchmark, and the required characteristics.
Afterwards, the remainder of this section includes a description of the data, the variables,
descriptive statistics and the econometric approach.

4.1. Turnaround cycle and measures
Before discussing the sample procedure, the first step in my empirical analysis is to formulate
the basic fundamentals of corporate turnaround to set up the framework for the sample
procedure. When viewing corporate turnaround as the recovery of a firm’s performance and
financial health following an existence threatening performance decline, there is two important
aspects to be determined when identifying and classifying corporate turnarounds: 1) the time
frame and cycle, that is, the period of deteriorated performance (the decline phase) which is
followed by a potential improvement in performance (the recovery phase), and 2) the definition
and measurements of firm performance and measures to indicate severe firm decline (Pandit,
2000).
In determining the fundamentals, the basic definition should possess the ability to take in all
elements in genuine turnaround cases. While this may be difficult, the best suiting definition
must be chosen based on a strong evaluation of prior research. I first discuss the time frame of
the turnaround cycle and secondly the measures used to indicate poor performing firms.

4.1.1. Turnaround cycle
Corporate performance decline may be present for several years while others experience shorter
periods of decline due to poor strategic decisions, poor managerial oversight or similar. Similar,
the time before recovery in performance and recurrence to sound financial health will vary from
situation to situation. As discussed by Bibeault (1992), there may be a time lag between
turnaround efforts are taken and a possible subsequent performance improvement. The time lag
Page 30 of 111
is affected by several factors, where especially delays in recognition by the current management
and reluctance to initiate turnaround actions, which often are related to ignorance of the severity,
heavily influence the time lag. As Bibeault (1992) observes, many companies do not initiate
turnaround actions and continue to be financially vulnerable for an extended period of time or
until end of existence, which stresses the fact that not all firms experience a turnaround. The
definition of the time frame should ideally encompass these aspects.
Previous and especially the early turnaround research use very different lengths of the
turnaround cycle. Appendix 1 provides a short review of the different time frames. As a
consequence, Pandit (2000) recommends that time cycle definitions should be in coherence with
a generally conceptualisation of the turnaround cycle characterizing turnarounds. Therefore, I
adopt the most common and predominant approaches used in turnarounds studies, which
enhance the comparability to the previous and latest empirical work.
In accordance with many of the turnaround studies (e.g. Bruton et al., 2003; Morrow et al.,
2004; Mueller & Barker, 1997), I examine turnarounds in a 6-year cycle, where the first 3 years
is the decline period and the last 3 years is the potential recovery period. As stated by Robbins
and Pearce (1993), the time required for a company to be considered as suffering from
performance decline, should take into account that the longer a firm experience declining
performance, the greater is the probability that the firm actually is in decline and not just
experiencing a temporary fluctuations in performance. Therefore, the 3-year decline period is
chosen to secure a significant period of decline and that a firm actually experienced a period of
decline, which is consistent with other researchers (e.g. Mueller & Barker, 1997; Barker &
Duhaime, 1997). Similarly, the 3-year potential recovery period ensures either that recoveries
are more than short-term improvement in performance and that declines actually are
underperformance. As noted by Bibeault (1992) and Barker and Barr (2002), in order to be
theoretical meaningful and comprise firms in a turnaround situation, the sample has to consist of
former profitable firms that encounter severe performance decline subsequent to a period of
prosperity. Therefore, as suggested by Robbins and Pearce (1992) and consistent with Francis &
Desai (2005), I use a 2-year base period prior to the turnaround cycle to ensure that the
performance decline not is a part of a longer extended period of underperformance.
The time frame is subject to both advantages and disadvantages. The major limitation is that
by establishing the characteristics a company must experience during (and prior to) the
turnaround cycle, not all companies will be included in my sample. For instance, the 6-year time
Page 31 of 111
frame excludes firms that decline for more than 3 years before undergoing a successful
turnaround subsequent to the investigated time period. Related, not all firms categorized as
unsuccessful turnarounds will continue to underperform or eventually fail. As discussed in
section 4.2.1., I attempt to overcome the disadvantages by using several approaches when
determining the characteristics for the recovery phase to ensure above disadvantages are not
exercising their full influence in my study. The multiple approaches are also an
acknowledgment of the fact that the turnaround phenomenon is highly firm-specific in nature,
owing to differences in recognition and degree of turnaround actions initiated. In connection to
this aspect of the turnaround duration, Bibeault (1992) argues that the length of the decline
period and subsequent recovery period depends on the suddenness of decline and severity of the
firm’s financial health. However, the 6-year turnaround period is generally accepted as being
sufficient to track decline and recovery performance (Morrow et al., 2004).
The advantage of the turnaround cycle time period is not only its connection to the
generally conceptualisation of the time period, but it is also close to empirically findings
(Bibeault, 1992) and connected to the theory, e.g. two-stage turnaround model and life-cycle of
firms. The specific recovery characteristics of turnaround and non-turnaround firms during the
turnaround cycle are discussed in the sampling procedure.
Figure 2: Time structure of the panel data

Source: This illustration is adopted from Jostarndt and Sautner (2008), but adjusted to the context of this thesis. The
time structure illustrates both the base (year 1-2) and turnaround cycle period (year 3-8).

4.1.2. Turnaround measures
The second significant aspect that is necessary to address is the performance and turnaround
measures, and the characteristics it must possess in the turnaround cycle in order to determine
whether a firm is in decline. Likewise, a generally accepted benchmark is necessary to
distinguish between good and poor performance.
As pointed out by Pearce and Robbins (1993) and Pandit (2000), many studies solely use
profitability measures to define performance, e.g. return on assets (ROA), return on investment
(ROI), or net income. Balcaen and Ooghe (2006) and Pandit (2000) recommend employing
additional measure(s) besides profitability measures for two reasons: 1) the possibility of
Page 32 of 111
inflated and manipulated financial figures, and 2) time lag between profitability measures and
competiveness. First, incorrect accounting-based profitability measures are often a problem in
situations with poor performance. As observed in practice by Bibeault (1992), previous
management often attempt to downgrade the severity and magnitude of decline by having
accounting figures embellished by questionable or “creative” accounting practices.
8
For
example, creative practices could be postponing crucial maintenance or investment plans to push
costs forward for later periods, avoiding restructuring plans that the current result or value asset
items at inflated values, which create manipulated and unreliable accounting-based profitability
measures (Bibeault, 1992). Second, there may be a lag between a firm’s loss of competitive
position and deterioration in profits. A weakening competitive position could be as a result of
lower market share, inability to keep up in the marketplace, organisational ineffectiveness, etc.
(Bibeault, 1992). As Bibeault (1992) describes, a bad trend do not happen overnight, but will
span over a longer period, but will only be evident in profitability measures at a later state, e.g.
ROA may still be positive despite a gradual loss of competitiveness. In connection to this
aspect, even profitability measures can differ. There is an important difference between
profitability measures by financial statement (accrual-based) figures and cash flow figures. For
instance, a fast-growing company may operate with profitable ROA while bleeding cash, i.e. a
negative cash-flow. This stresses the importance of using more than one profitability measure to
indicate the existence of a turnaround situation (Slatter & Lovett, 1999). Additional turnaround
measures are necessary in order to capture the actual financial and competitive state of a
company.
For the two above reasons described above, Pandit (2000) recommends incorporating 1)
multiple accounting-based profitability measures, 2) expert opinions, and 3) a generally
accepted benchmark of performance. Based on the recommended approach and research designs
of prior studies, I develop a broad-based framework consisting of suitable and accessible
measures to identify firms that have experienced a turnaround situation. Further, the framework
should ideally consist of measures that are complementary in such way they alleviate the
problem of using accounting-based measures only and decrease the gap between competiveness
and performance (Bibeault, 1999; Pandit, 2000).

8
As stated by Barker and Duhaime (1997), the financial ratios capture only what the income statement reflect.
There exist several examples of firms that manipulate and inflate financial statements, while others simply use less
credible financial disciplines. Examples are the classical cases of Enron, Tyco International, etc.
Page 33 of 111

Accounting-based profitability measures
Although the importance of using broad conceptualized profitability measures is widely
recognised, the topic has received little attention in prior turnaround research. The most widely
used approach by researchers is to define performance by a single profitability measure, e.g.
return on assets [ROA] (Abebe et al., 2010; Abebe, 2011; Mueller & Barker, 1997), return on
investments [ROI] (Morrow et al., 2004; Francis & Desai, 2005), and return on invested capital
[ROIC] (Barker & Duhaime, 1997), while few use measures based on both accounting and non-
accounting, i.e. market values, figures to measure performance, e.g. Morrow et al. (2004) use
Tobin’s Q. However, there do not seem to be consensus regarding the profitability measure to be
used. Instead prior researches tend to leave very little or no argumentation for their choice of
profitability measure.
The objective is to select a profitability ratio that effectively capture the overall single-
period operational performance, the company’s earnings capacity, and utilization of resources,
while allowing an assessment of the performance in each individual year. Measures such as
return on equity (ROE), return on assets (ROA) and return on invested capital (ROIC) are
possible profitability ratios to measure performance. ROE is calculated as net income divided by
common equity. However, many researchers dismiss ROE due to the fact that the measure is
affected by the degree of leverage. ROE included the impact of the accumulated decisions
regarding financing, i.e. leverage, when assessing profitability and thus performance. A more
strong measure of profitability is ROIC, which is reflecting the return on the capital invested in
its operating activities. Invested capital can be viewed as net operating assets or the funds to
finance operations. Similar, ROA provides a measure on how efficiently total assets is used to
generate operating profits. ROA is ignoring the type of financing by measuring the total return
to all providers of capital (both debt and equity). As all relative performance measures the
mentioned ratios are short-term looking based on historical financial data which is
advantageously in measuring performance for each single year in the turnaround period.
Given the above, I use profitability based on return on assets (ROA) as the performance
measure based on the premise that the measure reflects a firm’s ability to generate earnings and
utilize resources, while reflecting the ability to generate income from the total resources
invested. However, I recognize the drawbacks of using ROA to measure performance. The
measure is a very industry specific measure that varies from industry to industry. Capital
Page 34 of 111
intensive firms, e.g. telecommunication, require higher levels of fixed assets in order to operate,
while firms operating in less capital intensive industries may be able to generate higher returns
on their assets. Similar, firms with a high level of intangible assets (which cannot be accounted
for) will understate the asset level, e.g. Apple. For this reason, I employ industry dummies to
control for industry effects in the analysis. Despite the limits, I reckon ROA to be more robust
than other measures. However, I will perform tests using ROIC as the performance measure.
In addition to measuring performance by ROA, I use an additional absolute accounting-
based profitability measure in the base period. As Robbins and Pearce (1992), I employ return
on sales (ROS) defined as earnings before interest and taxes as a percentage of total sales. By
using ROS in the base period, I attempt to ensure that a positive ROA in the base period are
driven by the core business; that the profitability are generated by the operating activities.
Compared to many other studies, I attempt to move away from defining corporate turnaround on
the basis of a single profitability measure to use multiple accounting-based measures instead,
which also is recommended by Pandit (2000). I do not use cash-flow based measures due to the
fact that, as empirically demonstrated, the accrual accounting-based measures capture
performance better than the cash-flow performance measures (Plenborg & Petersen, 2011).
The absolute (ROS) and relative (ROA) accounting-based performance measures are
reflecting the single-period performance and success of utilizing invested resources. However,
the measures are reflecting competiveness and the financial state very poorly, in which case the
potential problem of a lag between performance and competitiveness remains. Additionally,
profitability is not alone a reliable measure of the existence of turnaround. The next section
attempts to provide a perspective on the last statements.

Expert opinion
Supplementing accounting-based performance measures in defining successful or non-
successful turnaround performance with expert opinions or interviews would be a clear
advantage when identifying firms that have experienced genuine corporate turnaround
situations. A disadvantage of using financial data only to identify firms for the sample is that the
approach ignores whether the individual firm acknowledge the turnaround situation. Robbins
and Pearce (1992) required agreement from at least one of the firm’s executives that the firm
had experienced a turnaround situation, while Barker and Duhaime (1997) used an expert panel
to develop a list of fundamental turnaround actions that were designed into a questionnaire
Page 35 of 111
mailed to the total sample in order to derive whether the firm actually experienced and
undertook turnaround actions or not.
The advantage of this approach is it captures the opinion of experts and firm executives
which enhance the quality of the sample. Further, turnaround cases are rather heterogeneous in
nature, that is the situation is unique to the individual firm and requires customized measures
(Kazozcu, 2011), and successful turnaround actions is linked to many contextual factors (e.g.
industry and environment context). Expert opinions can decrease and take into account the
influence of contextual variations (Pandit, 2000), which is difficult to measure and capture by
financial data. It is obvious that it is necessary to use additional indicators for sample selection.
However, due to the extensive scoop of the above approach, I instead use a measure of financial
health to ensure that the firms selected for the sample are genuine turnaround candidates.
As argued by Robbins and Pearce (1992), companies experiencing severe distress, thus
being financial unhealthy, due to a severe performance decline are to a great extent forced to
initiate turnaround activities. In this relation, the Altman’s Z-score is proven to be a strong
measure when assessing a firm’s financial condition (Barker & Duhaime, 1997; Barker et al.,
2001; Abebe et al., 2010). A lower value of the Z-score reflects deteriorating financial health
and therefore increased possibility of bankruptcy (Altman, 2000). Altman’s Z-score is based on
both accounting-based ratios and market values and the formula was build for public
manufacturing firms. The Z-score is calculated as follows (Altman, 1983):

Z = 1.2X
1
+ 1.4X
2
+ 3.3X
3
+ 0.6X
4
+ 0.999X
5

where,
X
1
= working capital / total assets
X
2
= retained earnings / total assets
X
3
= EBIT / total assets
X
4
= market value of equity / total liabilities
X
5
= sales / total assets, and
Z = overall score.
(1)

Altman (2000) consider two critical values and describes firms as no longer being in the “safe
zone” when the Z-score falls below the cut-off value 2.99, while a firm with a Z-score below the
cut-off value 1.81 has even higher probability of bankruptcy.
Page 36 of 111
Altman (2000) revisited the original Z-score model and the original model was modified to
also apply for non-manufacturing firms. The adapted model does not include X
5
, which is highly
influenced by industry effects, while the X
2
was changed to be calculated with the book value of
equity instead of the market value of equity. The modified Z-score model is as follows:

Z = 6.56X
1
+ 3.26X
2
+ 6.72X
3
+ 1.05X
4

where,
X
4
= book value of equity / total liabilities, and
Z = overall score
(2)

The upper threshold value is 2.6, while the lower cut-off value is 1.1. Appendix 2 provides a
detailed explanation of the financial measures in the Z-score models and their relation to my
thesis.
The discussed turnaround framework models suggest that the extent and the speed of
initiation and activation the overall turnaround response to the turnaround situation depends on
the severity and nature of decline (e.g. Robbins & Pearce, 1992; Pearce & Robbins, 1993;
Barker & Duhaime, 1997). In this relation, past findings suggest that a firm’s financial health
measured by the severity of decline reflect the threat of firm-survival and extent of performance
decline. In an attempt to address these attributes, I employ Altman’s Z-score to ensure that the
performance and financial condition of the individual firm during the decline period are severe
and life-threatening enough to warrant the initiation and activation of an appropriate turnaround
strategy.
9


Benchmarking
The last definition issue, benchmarking, outlines the single objective of selecting a benchmark
measure making it possible to differentiate successful and unsuccessful turnaround performance.
Mueller and Barker (1997) suggest using industry average of performance as a benchmark.
However, according to Pandit (2000), the type of industry is poorly related to profitability, why
using industry average performance as benchmark is inappropriate. Although, requiring a firm
to be benchmarked against industry average would ensure that firms represented in the final

9
However, Francis and Desai (2005) emphasize that firms experiencing severe declines may find it more difficult to
reverse decline than firms experiencing less severe declines. They suggest that fast performance decline and greater
severity of decline impact the ability to achieve turnarounds negatively.
Page 37 of 111
sample would not be industry low-performers, which often are first to decline compared to the
industry as a whole (Mueller & Barker, 1997). However, few industries include enough firms to
make this approach practical in my study, in which case assessing performance against average
industry performance is inadequate. Pandit (2000) suggests adapting the cost of capital as the
most appropriate benchmark. However, the cost of capital may not be appropriate as it is
affected by the firm’s financial health, market value, and other characteristics
10
(Fich & Slezak,
2008).
Instead, the turnaround literature has adopted the risk-free rate of return (r
f
) as the
benchmark because a firm is suggested to be failing in economic terms if it does at least earn a
return above the risk-free rate of return (Bruton et al., 2003; Abebe et al., 2010). The return of a
government zero-coupon bond is normally used as an alternative for the risk-free rate, which
holds a negligible amount of risk. The duration of the government bond should ideally match the
time horizon of the period (Plenborg & Petersen, 2011). I use the yield of each countries
respective 1-year government bond as a proxy of the risk-free rate (Appendix 3). Therefore, the
benchmark is the rate of return on the individual 1-year government bonds.
The rationale of using the rate of return is that it is an appropriate benchmark for the
minimum required performance for each individual firm. The main advantage with this
approach is that it discriminates between the origins of the firm by evaluating performance
based on different benchmark levels. Bruton et al. (2003) explains the necessarily of using the
country-specific rate of return to account for different regulatory frameworks, why there must be
a variation in the minimum performance requirement. Thus, a Danish firm in decline may have
an ROA below the benchmark, i.e. the yield on a Danish 1-year government bond for the given
year, while performing above the Swedish benchmark. Such firms are excluded from the
sample.

Summery
As recommended by Pandit (2000), I have defined the 6-year turnaround cycle period according
to the conceptualized conception. The framework relies on measures based on accounting-,
financial-, and economical-based information. I use ROA and ROS to measure performance,
while I employ Altman’s Z-score to decrease the disadvantage of using accounting-based data.

10
Other characteristics could be the beta, the capital structure, the market premium, etc. (Plenborg & Petersen,
2011).
Page 38 of 111
The overall framework is constructed in order to classify genuine turnaround candidates. The
following sample procedure builds on this approach and outlines the specific characteristics that
a firm’s performance is required to follow.

4.2. Sampling procedure
The general population of this study consists of all publicly traded Western European
companies, both manufacturing and non-manufacturing, in the period embracing the fiscal years
of 1995 to 2010. The firms in the population were identified through Standard and Poor’s
COMPUSTAT database. Financial firms are not considered and are excluded given the structure
and regulated environment of the financial industry and the companies operating herein (i.e.
banks, investment funds, private equity firms, and insurance companies). All firms are active
publicly traded companies in the turnaround period, which will increase the accessibility and
availability in terms of data collection.
By not restricting the sample to a specific industry and by considering all Western European
firms, I allow for a more heterogeneous and cross-cultural study, which also is necessary to
yield a suitable sample size in a European context. Most previous studies restrict their focus to a
single industry group, e.g. manufacturing, and do not examine the service industry. However,
this industry group should not be ignored as it may not be less vulnerable to performance
declines and should therefore be examined on equal basis with manufacturing firms (Pearce &
Robins, 1993). Heterogeneous samples in terms of industry are according to Barker and
Duhaime (1997) necessary and samples should not be restricted to single industries or sub-
groups.

4.2.1. Sampling criteria
Classification of decline
Based on the sample procedure of several turnaround studies (e.g. Mueller & Barker, 1997;
Abebe et al., 2010; Robbins & Pearce, 1992; Barker & Duhaime, 1997; Francis & Desai, 2005),
a firm is considered for the sample if it comply with the following sampling criteria:
1) Two consecutive years of ROA above the risk free rate of return (r
f
) and positive return
on sales (ROS) prior to the decline period. This is to eliminate continually poor
performing firms, while limiting the sample to consist of firms that are actually
experiencing a decline and are in a turnaround situation.
Page 39 of 111
2) Three consecutive years of ROA below the risk free rate during the 3-year decline. The
decline occurs after the base period, when the firm’s performance was above the risk-
free rate. This conservative benchmark is employed to ensure the turnaround firms are
failing in economic terms during the downturn in performance. This 3-year period is
considered as the decline period.
3) During the decline period, the firm’s performance have to become low enough to cause
negative net income (i.e. negative ROA) for at least one year. This additional criterion is
an attempt to strengthen the validity of the definition and ensure that the firms not only
have experienced a performance decline, but have additionally experienced net losses
that could have potentially threatened the viability of the business.
4) During the decline period, the firms had to experience an Altman’s Z-score of less than
2.99 for manufacturing firms and 2.60 for non-manufacturing firms for at least one year
in the downturn period. As the previous criteria, Altman’s Z-score is to ensure that the
firms are experiencing a significant performance decline of such a severe character that it
could threaten the viability of the firm and warrant a turnaround attempt. Consistent with
prior studies, Altman’s Z-score is used to express the financial soundness of the
individual firm (e.g. Abebe et al., 2011; Barker & Duhaime, 1997), and as described,
lower values generally indicate lower financial health.

Hence, all firms included in the sample have experienced declining and deteriorating
performance measured by ROA for 3 consecutive years, with ROA being below the risk-free
rate of return for 3 consecutive years, experienced an accounting loss measured by net income in
at least one year in the 3-year decline period, and have had an Altman’s Z-score below the
threshold limit in at least one of the years during the 3-year decline period.

Classification of recovery
As consequence of the variety of turnaround definitions in the literature, Pandit (2000) and
Pearce and Robbins (1993) recommend using a generally agreed conceptualisation of the
phenomenon turnaround. Therefore, I define successful recovery, and thus successful
turnaround, in three ways. The main approach is defined to maximize the sample size due to
data availability constraints, while the two subsequent definitions in greater detail will be
consistent with the approaches used in the turnaround literature. The two subsequent definitions
Page 40 of 111
will act as a robustness check and to test the relation between ownership aspects and
turnarounds outcome.
In line with the approach used by Bruton et al. (2003), the degree of turnaround success is
measured by the following definition:
1) The degree of corporate turnaround success is measured by the turnaround performance
(ROA). Thus, I am not restricting the dependent variable into discrete choices, i.e. either
turnaround or non-turnaround, but I am instead using the actual performance in the
turnaround period.

Contrary to the definition above, the two following definitions will classify the firms into
turnarounds and non-turnarounds. Consistent with Abebe et al. (2011) and Abebe (2010), a firm
is defined to have achieved a successful turnaround if the firm comply with the following
recovery characteristics:
2a) Three consecutive years of positive and increasing return on assets (ROA) above the
risk-free rate of return during the recovery period. By increase in ROA in this criterion
is meant that ROA is above the risk-free rate of return, i.e. the minimum threshold,
while it does not necessarily mean that ROA occur in an actual increasing pattern.
2b) At least three consecutive years of increasing return of assets (ROA) with performance
in the last year (year 6) at least being above the minimum threshold and profitable, i.e.
positive net income and thus positive ROA.

Similar, non-turnaround firms were identified in 2a by replacing the recovery characteristics
with the fact that ROA was decreasing and below the benchmark during the recovery period (i.e.
year 4, 5, and 6). In 2b non-turnaround firms were characterized by experiencing deteriorating
and fluctuating ROA below the risk-free rate of return during the entire recovery period (i.e.
year 4, 5, and 6). The sampling procedure allows me to follow the individual firm through the
turnaround process. The underlying ideas are illustrated in the figure below.
Page 41 of 111
Figure 3: Illustration of the turnaround process including sampling criteria

Source: The illustration is adopted from Francis and Desai (2005) and is adjusted to the sampling criteria for my thesis. The actual
turnaround- and performance-pattern depend on the definition and the figure has only an illustrative purpose.

4.2.2. Final Sample
The turnaround definition and sample selection criteria’s were applied to the COMPUSTAT
database for the period 1995 to 2010, which resulted in a general population consisting of 3.227
publicly-held firms, where 301 firms were identified as meeting the specified sample selection
criteria’s. Missing ownership information reduced the sample by 10, while another two was
restricted from the sample due to irregular values. The final sample consists of 289 firms that
have experienced severe performance decline. The two additional definitions for the robustness
analysis resulted in a sample size of respectively 152 and 199 firms.
Table 2: Summary of the number of companies in the analysis
Characteristics of the sample # number of companies
Total observations extracted from Compustat as the general population 3.227
Companies meeting the sample criteria 301
Companies eliminated due to missing ownership information 10
Companies eliminated due to irregular values 2
Total sample size for analysis (definition 1) 289
Sample size for definition 2a 152
Sample size for definition 2b 199
The table summarizes information regarding the observations and sample size for each definition. Observations are extracted from Compustat
and reduced by applying the given sample selection criteria. The companies restricted from the sample due to irregular values were as a
consequence of no operational revenues in periods of the turnaround cycle.

4.3. Data Sources and Sample Characteristics
I obtained the financial data for my analyses through two sources. I obtained the annual
financial statement and stock price data from Standard and Poor’s global fundamental files in
COMPUSTAT made available through Wharton Research Data Services (WRDS). Information
Page 42 of 111
on firm size was obtained and extracted from the Amadeus and Orbis databases published by
Bureau van Dijk, and was supplemented with information from annual reports when data was
missing or incomplete. The proxy for the risk-free rates was computed based on data from
Thomson Reuters Datastream. I obtained ownership data from annual reports either through firm
websites or Thomson Reuters Research. In cases of missing information, I reluctantly used
information from Amadeus or Orbis. Data quality is discussed in section 4.3.1.
The final sample consists of 289 firms experiencing a turnaround situation, which provide
me with a panel dataset with financial and ownership information for 6 years for each individual
firm. Although the firms are drawn without consideration to their industry group, the two top
industries represented in the final sample are manufacturing (SIC 2000-3999) with 157
companies and service (SIC 7000-8899) with 79 companies. The industries in terms of final
representation in the sample are distributed as follows:
Table 3: Sample description of industry group representation
Division code SIC Code Industry name # number of companies
B 1000 < 1500 Mineral 3
D 2000 < 4000 Manufacturing 157
E 4000 < 5000 Transportation, Communication, Utilities 23
F 5000 < 5200 Wholesale Trade 10
G 5200 < 6000 Retail Trade 17
I 7000 < 8900 Services 79
This table provide information regarding the industries in this study. The four industries A) Agriculture, Forestry and Fishing (SIC <1000),
Construction (SIC 1500 < 1800), H) Finance, Insurance, and Real Estate (6000 < 6800) and J) Public administration (SIC 9100 < 10.000) are
not included in the table since no companies from the respective industry groups are represented or the industry group is restricted from the
sample. Industry representation in the sample for definition 2a and 2b is presented in Table 15 and Table 16 (Appendix 4Appendix ).

The firms included in the final sample are drawn from 15 different countries. The countries
that make up a large part of the final sample in terms of representation are Germany with 53
companies, Great Britain with 76 companies and France with 59 companies. It is in the scope of
this thesis to draw the sample from a wide area of countries, making a heterogeneous and
diverse sample. This will inevitably lead to some countries being more represented than others
due to their economic size. The firm distribution in terms of country is represented below.
Table 4: Distribution of firms by country
Country Abbr. # number of firms Country Abbr. # number of firms
Denmark DNK 8 Spain ESP 5
Sweden SWE 20 Finland FIN 8
Norway NOR 5 France FRA 59
Germany DEU 53 Ireland IRL 3
Great Britain GBR 76 Italy ITA 10
Austria AUT 5 Holland NLD 17
Belgium BEL 6 Portugal PRT 2
This table reports the number of firms represented in the final sample for each country.
Page 43 of 111
As a validation of the rationale behind the sample procedure, Figure 4 illustrates the average
performance of the participating firms in the sample. As the figure illustrates, the average
participating firm in the sample experienced a severe decline in performance leading to financial
losses, while every firm had at least 1 year of negative net income.
Figure 4: Performance of sample firms during the turnaround cycle


Using the sample characteristics as a starting point, Appendix 5 shortly elaborate on the
performance measure, difference in performance between turnaround and non-turnarounds
firms, the development of the Z-score among the firms, and sampling window by referring to the
actual dataset.

4.3.1. Validity and reliability of data
Although I build the thesis on several reliable sources, the thesis is subject to noise in the
measurement in the dataset, which can create two potential problems. First, there is a possibility
that information contain errors, e.g. due to data accessibility, and second there will be a
possibility that firms are incorrectly classified, i.e. that turnaround firms are characterised as
non-turnaround and vice versa. Thus, this is a question of validity and reliability of data.
First, the most possible problem is concerning the reliability of data and the potential issue
of measurement errors. I have extracted all financial accounting data from Compustat, which is
a database widely and often used in empirical studies. The advantage of Compustat is that
financial statements and market information are standardized by specific data item definitions,
making information more comparable across companies, industries, countries, and time periods
(Standard & Poor’s, 2003). Thus, Compustat data may differ from those reported in company
annual financial reports. However, the Compustat approach mitigate the fact that companies
often present their annual financial data in different formats, thereby allowing for a more
-50%
-30%
-10%
10%
1 2 3 4 5 6 7 8
A
v
e
r
a
g
e

o
f

R
O
I
C

a
n
d

R
O
A
,

%

Year in the turnaround cycle period
Average of ROA Average of ROIC
Page 44 of 111
meaningful and reliable analysis by removing reporting biases and data discrepancies. In
addition, I have performed random tests on financial data and compared it information with
other databases, e.g. Amadeus. I did not encounter any divergence.
Datastream was used to gain access to historical financial time series of government interest
rates. In two cases, information was not available in the first two years of the sampling period
for a specific country. The lacking information was mitigated by not considering the two
respective periods. Balanced against this problem, the Datastream database is in conclusion a
reliable source of information.
Initially, I used Amadeus to gather ownership information. The Amadeus database contains
historical shareholder ownership data from 2002 until now, while information prior to 2002 is
accessible at CD’s through CBS Library. Amadeus rely on several information providers of
ownership information and publish data at the date of transmission, e.g. information for one year
can be collected at different dates. This results in different information conditional on the source
and date of information. Further, Amadeus attempts to track relationships of control, i.e.
reporting any pyramidal structures, which often results in total ownership exceeding 100 pct.,
making it difficult to determine the true ownership structure.
As a consequence of these obvious weaknesses, I have gathered ownership data through the
annual reports of each firm for the years of interest. Despite this being a rather time-consuming
task, I found it necessary to maintain a satisfactory level of reliability. However, I used
Amadeus in cases of missing information on ownership in firms’ annual reports. Employee
information is also gathered from Amadeus and validated by checking the annual reports of the
firm. Therefore, this information is very unlikely to be influenced by measurement errors.
Second, there is a risk of diagnosing turnaround firms as non-turnaround firms and
opposite. This is a problem connected to validity, which is whether the data and approach
represent the actual phenomenon of corporate turnaround. However, as discussed in previous
section and earlier, I construct a comprehensive sample procedure in order to ensure that firms
are classified correctly.

4.4. Variables and measure definitions
After having established the sampling procedure and identified the companies that fulfil the
turnaround characteristics, I have to address several measurement issues for the empirical
analysis. First, I need to define and construct a measure of the performance for the different
Page 45 of 111
approaches. Second, I need to create the most appropriate measures of ownership structure to
explain turnarounds. Finally, I need control measures that capture turnaround-specific
characteristics.

4.4.1. Performance measures: The dependent variables
In this thesis, I employ a clear distinction between turnaround outcome and turnaround
performance, which is a consequence of the different definitions. Firm turnaround performance
is measured by return on assets (ROA), while I use return on invested capital (ROIC) as an
alternative performance measure in additional tests.
Turnaround performance is the dependent variable in the main model, which describes the
level of performance in the turnaround process. Based on the explanations in the sample
selection section and consistent with Bruton et al. (2003) and Morrow et al. (2004), the
turnaround performance is measured by ROA, which are adjusted by the risk-free rate for the
given year to ensure performance are benchmarked against the minimum required return.
In testing the definition 2a and 2b, I adopt elements from the research framework employed
by Robbins and Pearce (1992), Mueller and Barker (1997), and Barker and Duhaime (1997), and
combine these in an attempt to answer my research objective. This approach makes me able to
distinguish between performance and outcome. I present two models of turnaround outcome,
which is the prediction of whether a firm achieves a successful turnaround or not. The
turnaround outcome is explained by the dependent variable TURNa and TURNb respectively,
which is a discrete (dummy) variable and takes on the value 1 if the firm achieves a successful
turnaround, and takes on the value 0 if otherwise.

4.4.2. I ndependent variables
In the empirical testing, I use two variables to describe ownership concentration. First, as
motivated by Jostarndt and Sautner (2008) and Laeven and Levine (2008), I use an
approximation to the Herfindahl index that measures the level of ownership concentration in a
firm. The measure is defined as follows:



(3)
where s
i
is the percentage of common stock owned by blockholder i.
Page 46 of 111
An increase in the Herfindahl ownership index is the results of the entry of new
blockholders or an increase in the holding by an incumbent blockholder, or both. The Herfindahl
ownership index has the advantage that it gives more weight to larger blockholders in measuring
the ownership concentration. The variable ranges from 0 to 10.000. As stressed by Jostarndt and
Sautner (2008) and applied in this thesis, the index measure is based on equity ownership rights,
which is equal to the cash-flow rights
11
. As discussed, I identify all shareholders who own at
least 5 pct. of the firm’s outstanding shares as blockholders.
Secondly, I identify the total concentration of shareholders blockholders and aggregate their
ownership percentage to measure their combined stake in the firm.
12
In contrast to the
Herfindahl ownership index variable, the aggregated ownership concentration does not assign
weight to the size of shareholder and may, therefore, be viewed as a pure concentration ratio
ranging from 0 to 100 pct. In the empirical testing, I will construct two groups of model
specifications, where I switch between using the two measures of ownership concentration to
test their applicability and robustness.
Jostarndt & Sautner (2008) argue that the top blockholder exercise the strongest influence,
while Lai and Sudarsanam (1997) advocate to distinguish between top and dominant
blockholders. In this perspective, I incorporate a dummy variable to indicate the presence of a
dominant blockholders, which takes the value 1 if there is a dominant blockholder with an
ownership position above 50 pct. and 0 otherwise.
In addition, I measure changes in ownership between each year by 1) takeover and 2) block
investment. Block investment is measured by a binary variable taking the value 1 when there is
an entry of a new blockholder, otherwise 0. Similar, takeover is measured by a dummy variable
taking the value 1 in the case of an acquisition of a majority block of shares or if a blockholder
increases its holdings to have a majority ownership position, i.e. ownership of more than 50 pct.
of the shares in a firm, as discussed in the hypothesis building.


11
Ownership of equity can be defined in terms of either cash-flow rights or voting rights. I use cash-flow rights to
measure ownership rights. Voting rights reflect control rights, which may differ due to difference in classes of
shares. In the cases with divergence between cash-flow rights and voting rights, the difference was rarely
significant. Only a few cases presented a significant difference between voting and cash-flow rights that had an
influence on the actual control of the given firm, i.e. where a party’s voting rights greatly exceeded the cash-flow
rights. I used the voting rights as proxy of control rights in three cases, where cash-flow rights were not disclosed.
12
Restricted data availability, due to different law requirement, prevents me from combining the stake of the top
three or top five of the shareholders, which would be an alternative measure of the ownership concentration.
Page 47 of 111
4.4.3. Control variables
This thesis focus at ownership structure as determinant of corporate turnaround, but I do not
want to ignore other potential important firm-specific factors influencing turnaround. Some
firms may require certain turnaround actions more than others firms, both across industries and
countries. Therefore, I use several control variables to account for firm-specific factors. I draw
on existing literature in order to choose the individual firm-specific variables, and these are
chosen based both on their theoretical and empirical relationship with the models in this thesis.
I include the following firm-specific control variables: 1) firms size, 2) asset retrenchment,
and 3) cost retrenchment. All variables are used to account for firm-specific turnaround
characteristics during the turnaround cycle period, which potentially could have an important
explanatory effect. Furthermore, I include variables to control for 1) country-, 2) industry-, and
3) time-effects to reduce the concerns regarding differences across countries, industries, and
time.
Past research has found size to positively affect firms to undertake the necessary
adjustments during adversity and changing environment, and thus achieve greater turnaround
success (Barker et al., 2001; Abebe, 2010). However, Bruton et al. (2003) show that the size of
East Asian firms is negatively associated with turnaround performance. Firm size is measured
by the natural logarithm of the total number of employees employed by the firm in each year
(e.g. Mueller & Barker, 1997; Morrow et al., 2004; Abebe, 2011). Some researchers use the
natural logarithm of total assets or total market capitalization to measure firm size (e.g. Bruton
et al., 2003). These alternative definitions were discarded due to currency-differences between
the firms considered in the sample.
As a consequence of retrenchment being deeply rooted in the turnaround literature and the
arguments presented in the hypothesis building, I control for retrenchment by constructing the
two variables cost and asset retrenchment. The variables are calculated as follows:









(4)








(5)

Page 48 of 111
Consistent with previous research (e.g. Bruton et al., 2003; Morrow et al., 2004), the cost
base includes costs of goods sold, and total administrative and general expenses.
13
Cost
retrenchment was initiated by the given firm if the measure is negative, i.e. the cost base was
reduced in the given year. The asset base of the individual firm is measured by the total assets in
the firm. A negative measure will indicate that the firm initiated asset retrenchment and reduced
their total assets compared to the previous period. The variables describe the percent change in
the cost and asset base respectively between the two points of time, which also mitigate any
currency differences.
Industry and country effects may also impact turnaround performance and outcome. I use
industry and country dummies to control to which extent a firm’s ability to complete a
turnaround are influenced by their national context and industry affiliation. Such effects may
display significant impact, and as I take the advantage of using data from a large number of
countries, individual country and industry characteristics should not be ignored. For example,
there is likely to be variance in market conditions, country policies, formal institutions (e.g. law
enforcement), and regulatory environment (accountability policies, shareholder rights,
ownership protection) across countries. Similar, industry differences are likely to be present due
to difference in industry conditions, which for example may arise from differences in intensity
of knowledge-capital, capital requirements, and product and service offerings.
Therefore, the final sample is divided into five industry groups based on industry
classification codes (SIC) to control for specific industry-related effects. The used industry
groupings are “mineral”, “manufacturing”, “transportation, communication, and utilities”,
“trade”, and “service”. The five industries are measured by dummy variables taking the value 1
if the firm belongs to the given industry and 0 otherwise. Nationality is measured by country
dummy variables taking the value 1 if the firm is based in the given country and 0 otherwise.
Time dummies are also introduced to control for possible year fixed effects. I initially
considered treating possible time effects unfixed because the turnaround process often are
viewed as an independent and time-isolated event. This is despite the fact that the sample period
often includes a wide-ranging time period. Treating time unfixed imply that performance are
considered to be unaffected by time effects. Some researchers attempt to mitigate time effects by

13
The item “Total administrative and general expenses” is not being compiled by Compustat for the firms within
the Global category. Instead, as suggested by Morrow et al., (2004), the cost base is measured by a proxy, which
may be calculated as sales minus cost of goods sold minus operating income.
Page 49 of 111
paring turnaround and non-turnaround firms within the approximately same time periods (e.g.
Mueller & Barker, 1997). However, as noted by Bibeault (1999), “a boom covers many sins,
and a bust uncovers many weaknesses”. Bibeault refers to the fact that macroeconomic events,
economic change, and business cyclic behaviour often reveal unsound corporations, which are
reflected by a larger number of firms experiencing severe performance declines at the onset of
economic downturns (Bibeault, 1999). Based on this conception, I introduce time dummies to
capture potential fixed year effects for the firms in the sample.

4.4.4. Summary of variable definitions and data sources
Table 5 summarizes the variables used in the econometric analyses.
Table 5: Summary of explanatory and control variables
Variable Variable explanation Definitions and description Expected sign
HHI Herfindahl index The sum of individual squared ownership share by all
blockholders.
+
OCR Ownership concen-
tration ratio
The variable is defined as the percentage of the total
ownership share of all blockholders
+
DOMI Blockholder dominance Takes on the value 1 if the firm is dominated by a single
blockholder, otherwise 0
+ / -
BI Block investment Takes on the value 1 if there is a block investment in the
given year, otherwise 0.
+
TO Takeover Takes on the value 1 if the firm experience a takeover or a
blockholder increases its share to above 50 pct., otherwise 0.
+
COSTRY Cost retrenchment Change in cost base defined as (Cost base
t
– Cost base
t-1
)/Cost
base
t-1

-
ASSETRy Asset retrenchment Change in asset base defined as (Asset base
t
– Asset base
t-
1
)/Asset base
t-1

-
SIZE Firm size Natural logarithm of the number of employees + / -
The table summarizes the independent variables applied in this thesis except dummies to control for industry, country and time specific effects. The
hypotheses and expected signs are formulated under the ceteris paribus condition. The dependent variables are given the following abbreviations:
Turnaround performance (AdjROA), turnaround performance measured by ROIC (AdjROIC), turnaround outcome depending on the definition;
TURNa and TURNb.

4.5. Descriptive statistics
The descriptive statistic of the dependent variable, explanatory variables and additional
measures are presented in Table 6, which provides descriptive data for the full sample of firms
in the turnaround cycle period, i.e. year 3 to 8.




Page 50 of 111
Table 6: Sample descriptive statistics
Variable

Mean Std.dev. Min Max Median
Turnaround performance -0.1009 0.2494 -3.4390 0.3779 -0.0385
Herfindahl ownership index 1859.07 2104.22 0.0000 1000.00 1005.08
Ownership concentration ratio 0.4927 0.2559 0.0000 1.0000 0.5139
Blockholder dominance 0.2468 0.4213 0.0000 1.0000 0.0000
Takeover 0.0156 0.1238 0.0000 1.0000 0.0000
Block investment 0.3126 0.4637 0.0000 1.0000 0.0000
Cost retrenchment 0.0868 1.6170 -16.3764 32.7089 -0.0151
Asset retrenchment 0.0574 0.6257 -1.0000 8.1238 -0.0253
Size 6.9489 1.8473 1.6094 13.0470 6.8101
N=1734 (equal to 289 cases in each year). The sample is restricted to the years in the turnaround cycle period, i.e. year 3-8. This table reports
descriptive statistics for all variables (both dependent and independent) and measures used in my estimations related to the main definition.

It is evident that the mean (average) of turnaround performance among the firms in the period is
negative 10.19 pct., indicating that the average firm averagely has a non-viable and poor
performance during the turnaround cycle period. The average firm has an HHI of 1859, which is
the alternative definition to ownership concentration that also is measured by the ownership
concentration ratio, taking the average of 49.27 pct. The average of the dominant blockholder
variable is 24.69 pct., indicating the variable takes the value 1 in approximately one fourth of the
observations. The descriptive statistics reveals that on average 1.56 pct., and in absolute values
amounting to 27, of the firms experienced a takeover activity during the turnaround process.
Similar, the block investment is on average 31.26 pct., meaning that a block investment
occurred in approximately one third of the observations, which confirms the perception of that
acquisitions and/or increases in holdings of shares increases regularly. The mean value of size is
6.94, suggesting the average firms has approximately 1042 employees. The mean value of cost
and asset retrenchment is 8.68 pct. and 5.74 respectively, while the median is -1.51 and -2.53 for
the average firm. This suggests that outliers
14
in the sample affect the mean value, which is
sensitive to extreme observations, why the median value is also reported to describe the middle
observation. The standard deviations are reported to describe the spread of the data, indicating
that the values for some variables are wide spread from the mean.

14
The descriptive statistics reflect that the panel dataset is likely to be subject to (extreme) outliers, which is also
confirmed by assessing the distribution of the variables. This ignites the considerations to remove some of these
observations. Two options are possible: 1) restrict the sample to not include the (most extreme) outliers, or 2) keep
outliers in the sample. As the first option implies removing data from the analysis, which additionally would reduce
the number of firms in the sample, I do not remove any outliers for two reasons. First, I have checked the most
extremes outliers for miscalculations and validated the data, which did not lead to any incorrect measures and, thus,
no exclusion of outliers. Second, I follow the mindset that altering the dataset is constructing the reality as wished
for. The outliers present the fact that some turnaround measures yield extreme values, and removing outlying
observations may change the relation among variables. Therefore, outliers are not restricted from the sample despite
the fact that outlying observations may affect the panel data estimations. To mitigate the issue with outliers, I take
corrective actions in the SAS procedures when possible.
Page 51 of 111
In Table 7 the descriptive statistics are classified by the year in the turnaround cycle period,
which illustrates the development of the variables year-wise during this process on an
aggregated level.
Table 7: Sample descriptive data represented for each year in the turnaround cycle period

Year in the turnaround process
Variable Year 3 Year 4 Year 5 Year 6 Year 7 Year 8
Turnaround performance -0.0862
(0.2080)
-0.1433
(0.2214)
-0.1476
(0.2114)
-0.1025
(0.2316)
-0.0652
(0.2368)
-0.0608
(0.3471)
Herfindahl ownership index 1876.06
(2143.47)
1793.50
(2019.84)
1830.92
(2061.92)
1843.71
(2099.51)
1861.91
(2107.86)
1948.33
(2202.93)
Ownership concentration ratio 0.4798
(0.2594)
0.4798
(0.2569)
0.4914
(0.2568)
0.4898
(0.2587)
0.5036
(0.2479)
0.5116
(0.2563)
Dominant blockholder 0.2595
(0.4391)
0.2422
(0.4292)
0.2457
(0.4312)
0.2318
(0.4227)
0.2422
(0.4292)
0.2595
(0.4391)
Takeover 0.0104
(0.1015)
0.0069
(0.0830)
0.0138
(0.1170)
0.0104
(0.1015)
0.0277
(0.1643)
0.0242
(0.1540)
Block investment 0.2457
(0.4312)
0.3010
(0.4595)
0.3080
(0.4625)
0.3149
(0.4653)
0.3702
(0.4837)
0.3356
(0.4730)
Cost retrenchment 0.2908
(2.0744)
0.1915
(1.8087)
0.0297
(0.8451)
-0.0212
(0.9565)
-0.1244
(1.0360)
0.1546
(2.3106)
Asset retrenchment 0.3537
(1.1069)
-0.0526
(0.4612)
-0.0470
(0.5322)
-0.0437
(0.2561)
0.0334
(0.3455)
0.1004
(0.5710)
Firm size 7.0172
(1.8412)
7.0431
(1.8405)
6.9912
(1.8252)
6.9170
(1.8405)
6.8674
(1.8534)
6.8576
(1.8896)
N=289 in each year. The table presents means and standard deviations in parentheses for the variables each year during the turnaround process, and is
related to the main definition.

A noteworthy development is the average firm size that decreases, which indicate the average
firm reduces the amount of employees during the turnaround process. The performance of the
average firm decreases during the decline period, while increasing in the recovery period. Asset
retrenchment follows an expected pattern by being negative in turnaround year 4 to 6, while cost
retrenchment takes a pattern that is diverging to the theorized pattern.
In Table 17 and Table 18 (Appendix 6) are reported the mean and standard deviations for
the two additional definitions, which are grouped by the turnaround outcome and, thus, provides
a prelude of what to expect from these models. Generally, the descriptive statistics in Table 17
and Table 18 reveal that turnaround firms on average are larger than non-turnaround firms. The
average turnaround firm seem less likely to be dominated by a single blockholder, while there
on average are fewer cases of takeovers in turnaround firms than compared to the average non-
turnaround firms. Ownership concentration does not seem to differentiate the two groups.
Table 8 reports the correlation coefficients between the dependent variable and all
independent variables considered in the alternative model specifications.
Page 52 of 111
Table 8: Correlations between variables considered in this thesis
Variables

1 2 3 4 5 6 7 8 9
1. Turnaround performance 1
2. Herfindahl ownership index .05** 1
3. Ownership concentration ratio .07*** .79*** 1
4. Dominant shareholder .02 .81*** .59*** 1
5. Block investment -.02 -.26*** -.11*** -.23*** 1
6. Takeover .00 .17*** .14*** .21*** -.01 1
7. Cost retrenchment -.01 -.03 .00 -.03 -.01 -.02 1
8. Asset retrenchment .18*** .01 .02 -.02 .02 .00 .17*** 1
9. Firm size .19*** -.07*** -.15*** -.07*** .01 -.01 -.10*** -.04 1
N=1734 (289 cases multiplied by six years of interest). This table reports correlations between the variables used in testing the main approach. Stars
indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01, which are used to indicate the result of the null hypotheses testing for zero correlation.
If the null hypothesis is rejected, there is an indication of either positive or negative relationship between the two given variables. The sample embraces all
years in the turnaround cycle period, e.g. year 3-8.

In relation to the model general model specification, the dependent variable are significantly
correlated with Herfindahl index, ownership concentration, asset retrenchment and firm size,
although this does not necessarily imply significance in the individual model estimations.
A high degree of correlation between the explanatory variables may suggest
multicollinearity problems, which could limit the usefulness of my estimation results. For
example, the correlation coefficient is 0.81 between Herfindahl ownership index and
blockholder dominance, indicating a significant linear relationship between these two variables,
while the latter is correlated to ownership concentration with a value of 0.59 and, thus, only
moderately correlated. The pair-wise correlation between the two variables used in measuring
ownership concentration is 0.79. These two variables will be substituted by each other in the
model specification to test different ownership concentration measures and their inter-
correlation is therefore not relevant. The high correlation between Herfindahl ownership index
and blockholder dominance is strongly correlated by exceeding 0.80, suggesting there may be
severe multicollinareairty problems. However, high correlation is not a necessary condition for
multicollinearity to exist. Therefore, I apply variance inflation factors (VIF) and condition index
(CI) in order to test the presence of multicollinearity, which do not suggest any problems with
multicollinearity (Appendix 7).
Table 20 and Table 21 (Appendix 8) report correlations of the alternative sample definitions
used in the discrete response models. Furthermore, the tables report the mean and standard
deviation for the full sample for definition 2a and 2b respectively without grouping the
descriptive statistics.

Page 53 of 111
4.6. Empirical methodology and Econometric model specification
In this section, I establish the empirical approach used in testing the theoretical and
empirical arguments presented throughout the thesis. More specifically, I take advantage of the
panel dataset and concentrate on specifying regression models to test the argued relationship
between ownership structure and corporate turnaround performance and outcome. Since I have
arranged a panel dataset consisting of 289 firms for 6 years each gathered within a 15-year time-
period and with large cross-sectional dimensions, e.g. various industries and countries, then it is
relevant to use econometric panel models in testing the suggested relationships by taking both
cross-sectional and time elements into consideration. The employed dataset has been arranged in
Excel, where the variables for the econometric analysis have been constructed. The econometric
testing is performed in SAS EG.

4.6.1. Standard panel models
To estimate the relationship between governance structure and corporate turnaround the most
common options are to apply one of the following panel regression models; pooled, random
effects, and/or fixed effects. As I suspect there are unobservable characteristics specific to the
individual firm that is time-invariant (e.g. management philosophy, ability to maintain a certain
level of management quality, board of director routines, board composition, power agreements
among blockholders, and similar over time stable firm-specific effects)
15
that may affect
turnaround, I set up and apply fixed effect panel models, which can be specified through the
following equation:




(6.1.)
e.g.




(6.2.)


15
Unobserved effects and characteristics of the given firm that are time-invariant and thus do not change over time
may be difficult to justify. However, the examples are potential effects that could stay fixed (or assumed to
gradually change over a longer period than the one investigated) for a firm and account for a significant amount of
difference between firms in the sample. It is assumed unobserved factors within the individual firm are not
correlated with any of the explanatory variables as estimates otherwise would be biased.
Page 54 of 111
where the i and t subscript denotes the firm and number of time-periods respectively, and ?
i
is
the time-invariant (assumed fixed for all time) unobserved heterogeneity and reflect any
individual firm-specific effects that not are included as the explanatory variables. The y
it
is the
dependent variable. The X
it
is a vector of explanatory and control variables, while ?
k
is a matrix
of coefficient of independent variables (k equals the number of explanatory variables), and u
it
is
the idiosyncratic disturbance term, which change across time and firm (Gujarati & Porter, 2009).
The industry and country dummies are left out of the example since time-invariant effects are
absorbed in the individual intercepts. Time dummies are included in most of the estimations. In
these cases the total error term will include ?
t
that denote the year-specific effects.
The pooled OLS regression model is not explicitly considered as it ignores the panel
structure of the dataset and the firm-specific uniqueness that may exist in the individual firms,
which are not appropriate since I have constructed the sample to be rather heterogeneous.
16
The
fixed effect model allows intercept to vary across firms, while the random effects models absorb
all heterogeneity in the error term.
16
Unfortunately, in the random model the error term are
represented by a not observable and latent variable, which not allow for interpretation of the
unobserved heterogeneity (Gujarati & Porter, 2009)
17
. Unobserved firm-specific effects and
characteristics is an important issue if not controlled for since the heterogeneity would otherwise
be included in the error term, causing the error term to be potentially correlated with the
explanatory variables. If the individual firm-specific error term induces autocorrelation then the
random effects estimators will be biased, while the fixed effect would absorb such time-
invariant heterogeneity in the firm-individual intercepts, leading to unbiased estimators (Gujarati
& Porter, 2009). Related, random effects are sensitive to misspecification, while fixed effect is
assumed to absorb time-invariant effects, e.g. individual non-measurable effects that normally
are difficult to measure explicitly. Thus, fixed effect models disregard, or at least greatly reduce,
the potential of omitted variable bias (Gujarati & Porter, 2009).
A disadvantage of the fixed effects approach is that it does not allow invariant variables
among the explanatory variables. Therefore, an argument for the random effects model is that it
allows time-invariant and individual-invariant explanatory variables. Further, the fixed effects

16
Although the pooled OLS regression model and random effects model not are explicitly considered as a part of
the empirical analysis, Appendix 9 reports the pooled and random effect estimation results for comparison to the
fixed effect estimation results presented later.
17
In random effect models the error term

consists of two components, which are the individual-specific error
component

and the idiosyncratic error component

that varies over cross-section and time (Gujarati & Porter,
2009).
Page 55 of 111
model consumes a large degree of freedoms, while the random effects approach holds the
advantage of consuming less degree of freedom, making it more efficient if the underlying
assumptions not are violated (Gujarati & Porter, 2009). Last, a disadvantage of the fixed effect
models are that the individual unobserved effects may be correlated with the error term (Gujarati
& Porter, 2009).
To ensure that the expectations regarding fixed effect to be the appropriate approach, I
conduct the Hausman specification test, which compares fixed effects against random effects
under the null hypothesis that the unobserved effects are uncorrelated with the explanatory
variables (Gujarati & Porter, 2009). If the null hypothesis is rejected, the fixed effects
specification is appropriate since the random effects are likely to be inconsistent (Gujarati &
Porter, 2009). The Hausman tests are executed with different sets of variables (following the
model specifications) and are rejected in all cases. In addition, SAS EG provides an F-test for
the null hypothesis of no fixed subject effects that can be rejected every time, which also reject
poolability (SAS Institute, 2010). Therefore, I use fixed effects OLS models.
In order to check the robustness of the model specifications I stepwise introduce the
explanatory variables of interest. I mainly test the adjusted return of assets, but also test the
adjusted return on invested capital as the dependent variable to test the alternative performance
measure. I use two different definitions for ownership concentration (Herfindahl ownership
index and ownership concentration ratio) to test the robustness of the different ownership
concentration measures. To prevent issues in the presence of heteroscedasticity, e.g. issues
arising from outliers, all models are adjusted to ensure efficient estimates and unbiased standard
errors (SAS Institute, 2010). I have compared results for all models using adjusted and non-
adjusted standard errors and variations between the results are all negligible.

4.6.2. Dynamic models
Past firm turnaround performance may have an influence on future turnaround performance, i.e.
there may be persistent elements in the determination of turnaround performance (Cameron &
Trivedi, 2005). Hence, I estimate dynamic models to take the role of past turnaround
performance into account. The dynamic model specification threats the lagged dependent
variable, i.e. past turnaround performance, as an explanatory variable by introducing it into the
right-hand side of the equation. This allows me to examine turnaround performances own
Page 56 of 111
determining effect in the turnaround process. Thus, I estimate a fixed effect dynamic model to
take unobserved heterogeneity and persistency into account.
By including the lagged dependent variable together with the explanatory variables, the
dynamic model specification is expressed by the following equation:




(7)

where the i and t subscript denotes the cross sectional and time-periods dimension respectively,
y
it
is the turnaround performance variable, y
i,t-1
is the lagged turnaround performance, ? the
coefficient of adjustment, X
it
is the vector of explanatory and control variables excluding the
intercept ?
0
and y
i,t-1
, while ?
k
is a matrix of all parameters of independent variables, and error
term w
it
consisting of the two components ?
i
, which is the unobserved firm-specific effects
capturing heterogeneity, and u
it
, which is the idiosyncratic error term (Baltagi, 2005; Gujarati &
Porter, 2009).
As stressed by Baltagi (2005), the dynamic model specification complicates estimation and
introduces two highly potential econometric issues, which create inconsistent estimates in the
OLS panel models for two reasons (Gujarati & Porter, 2005). First, the lagged dependent
variable is by construction endogenous and correlated with the error term, why the presence of
the lagged dependent variable is problematic (Baltagi, 2005). In a fixed effect dynamic model
only time-invariant heterogeneity will be absorbed by the individual intercept. Hence,
correlation between an endogenous explanatory variable and error term is likely to persist.
Second, the individual effects reflecting the heterogeneity may be correlated with some or all of
the explanatory variables, which also may induce autocorrelation (Baltagi, 2005). Additionally
as briefly discussed, previous researchers (e.g. Holderness, 2003) suggest the relationship
between firm performance and ownership (and retrenchment (Barker & Mone, 1994)) could be
endogenous. Therefore, if assumed to be endogenous and actually influenced by turnaround
performance and the other way around, this may create a problem related to above. Thus, the
way of causality and assumptions of exogeneity may be questioned.
A remedy to these problems is suggested by Baltagi (2005), who suggest using the
generalized method of moment (GMM) estimation approach to obtain efficient estimates in the
dynamic models. The GMM approach transforms variables into first differences. The first
difference transformation eliminates the econometric issues with 1) individual constant and
Page 57 of 111
unobservable firm-specific effects and 2) correlation that arise from incorporating the lagged
dependent variable. The GMM application in SAS EG follows the Arellano and Bond
methodology and introduces instrumental variables to address the potential problem of
endogeneity (SAS Institute, 2010). The approach involves introducing the dependent variable as
an instrument variable, which is argued to result in consistent estimates for dynamic models
(SAS Institute, 2010)
18
. The GMM model assumes that error terms present no autocorrelation to
be consistent (Baltagi, 2005).
To test the validity of the instrumental variables, I perform the Sargan test of over-
identification of restrictions. The underlying null hypothesis is that there is no correlation
between the used instruments and the error term, i.e. the instruments are exogenous. Instruments
are valid if the null hypothesis is confirmed, that is, if I fail to reject the hypothesis. In addition,
I examine the existence of first and second order autocorrelation. Therefore, based on the above
argumentations, I estimate dynamic models by using the GMM approach suggested by SAS
Institute (2010).

4.6.3. Logit models
In addition to the standard panel models that examine turnaround performance, I have
established an alternative approach. The alternative approach involves a binary dependent
variable that takes on the value 1 if the firm is defined as a turnaround case and 0 otherwise,
which calls for binary choice models. I consider the logistic approach that predicts the
occurrence of an event, here successful turnaround, by fitting the data to a function of the
cumulative logistic distribution function (logistic CDF) that constrain the function to be between
one and zero. The logit model expresses the outcome probability function and by building on the
argumentations above, the logit fixed effects model can be expressed by the following (Baltagi,
2005):














(8)


18
The underlying estimation technique (e.g. that the dependent variable y
i,t-1
are used as instruments) and
explanation hereof is deemed beyond the scoop of this thesis.
Page 58 of 111
More conveniently, the model can be expressed by an odds ratio, describing the ratio between
the likelihood of one event occurring (p) and the likelihood of the event not occurring (1-p), and
may be expressed by the following reduced logit specification:

logit(






(9)

where p
i
is the probability of successful turnaround outcome, while the last part in the equation
build on equation 7, where X
it
is a vector of explanatory and control variables, while ?
k
is a
matrix of coefficient of independent variables, ?
i
captures the unobserved time-invariant
individual firm heterogeneity and u
it
is the idiosyncratic error term (Baltagi, 2005). The fixed
effect logit model assumes a logistic distribution of the idiosyncratic error terms in equation (8)
and (9), and uncorrelated to the variables (Baltagi, 2005).
Estimating fixed effect logit models may cause a potential problem. The inclusion of
dummy variables for each firm may introduce problems when estimating the individual
intercepts, i.e. the incidental parameters ?
1
,...,?
N
. Furthermore, the estimation of the dummies
depends on the number of time-period observations, which in short panels also are likely to
induce estimation problems (Cameron & Trivedi, 2009). Consequently, these two aspects are
likely to cause the incidental problem, which result in poor estimation of the common
coefficients ?
k
, yielding inconsistent estimates (Cameron & Trivedi, 2005; Gujarati & Porter,
2009) or convergence problems (Baltagi, 2005; Greene & Hensher, 2010). Especially, the
problem of convergence is a problem with the structure of my panel dataset. In the standard
models, estimation is made by mean-differenced transformation, i.e. based on deviations from
group means, which is the logistic models, will cause every firm with the same values for all
time-periods to be dropped (Baltagi, 2005). The solution to the incidental parameter problem is
often referred to be conditional and unconditional logit (SAS Institute, 2010; Greene & Hensher,
2010). However, these methods do not converge given the structure of my data, where there is
no variation within the firm, i.e. the dependent variable is always 1 or 0 (SAS Institute, 2010;
Greene & Hensher, 2010).
Instead, I have to threat the firm-specific intercepts as explicit random variables, which are
the most appropriate approach to simulate the fixed effects. The equation can be describes as
follows:
Page 59 of 111

logit(







(10)

where the individual firm-specific intercept ?
i
is a determined by a overall mean ? and
individual the deviation from the mean ?
i
. The standard error ?
i
is assumed to be normal
distributed (Cameron & Trivedi, 2005; Greene & Hensher, 2010). Taking this approach is
necessary to estimate the logit model specifications in SAS EG, and estimates are deemed to be
indicative only.

Page 60 of 111
5. EMPIRICAL RESULTS AND ANALYSES
This section presents the empirical analysis of the relationship between ownership structure and
corporate turnaround. The first part serve to present and structure the estimation results of
several econometric model specifications applied with various econometric methods in SAS EG.
Afterwards, this section elaborates on the potential econometric issues and their presence in the
analyses in this thesis. The last part of this section shows the estimations of my two alternative
definitions 2a and 2b, which will work as a robustness test belonging to the first definition and
to examine the potential difference between turnaround performance and outcome.

5.1. Evidence from panel regressions: Estimation results
5.1.1. Fixed effect estimation results
First, I begin by reporting the estimation results from four model specifications using the
method of fixed effect, where I control for any unobserved time-invariant firm heterogeneity. I
use Herfindahl ownership index and ownership concentration ratio respectively to represent the
measure of ownership concentration. Table 9 below presents estimation results of the model
specifications, where the first four columns report estimates when using Herfindahl ownership
index (Model 1-4) and the next four columns when using ownership concentration ratio (Model
5-8) to represent ownership concentration. In general, all factors except asset retrenchment are
found to be statistically insignificant, suggesting that ownership structure and changes herein are
not associated with turnaround performance.










Page 61 of 111
Table 9: Fixed effect estimation results
Herfindahl ownership index Ownership concentration ratio
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Herfindahl ownership
index
-9.56E-6
(6.87E-6)
-2.98E-6
(9.02E-6)
-3.53E-6
(9.03E-6)
-3.50E-6
(9.04E-6)
- - - -
Ownership concentra-
tion ratio
- - - - -0.0458
(0.0535)
-0.0134
(0.0574)
-0.0167
(0.0574)
-0.0185
(0.0351)
Dominant blockholder - -0.0470
(0.0417)
-0.0533
(0.0421)
-0.0530
(0.0422)
- -0.0530
(0.0341)
-0.0602*
(0.0347)
-0.0593*
(0.0351)
Takeover - - 0.0547
(0.0497)
0.0547
(0.0498)
- - 0.0544
(0.0497)
0.0544
(0.0497)
Block investment - - - 0.0016
(0.0139)
- - - 0.0026
(0.0141)
Cost retrenchment 0.0014
(0.0038)
0.0014
(0.0038)
0.0016
(0.0038)
0.0016
(0.0038)
0.0013
(0.0038)
0.0014
(0.0038)
0.0015
(0.0038)
0.0016
(0.0038)
Asset retrenchment 0.0886***
(0.0099)
0.0879***
(0.0099)
0.0878***
(0.0099)
0.0878***
(0.0099)
0.0882***
(0.0099)
0.0878***
(0.099)
0.0877***
(0.0099)
0.0877***
(0.0099)
Firm size -0.0129
(0.0164)
-0.0127
(0.0164)
-0.0126
(0.0164)
-0.0127
(0.0164)
-0.0116
(0.0164)
-0.0124
(0.0164)
-0.0122
(0.0164)
-0.0123
(0.0164)
Industry dummies Yes Yes Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes Yes Yes
Time dummies Yes Yes Yes Yes Yes Yes Yes Yes
F-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
R
2
-value 0.3607 0.3612 0.3618 0.3618 0.3601 0.3612 0.3618 0.3618
# cross-section/time 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 12 289 / 12
This table shows fixed effects (FE) OLS estimation results when the Herfindahl ownership index and ownership concentration ratio is the ownership
concentration variable. Standard errors are presented below the parameter estimates in parentheses and are corrected for heteroscedasticity. The sample is
restricted to the years in the turnaround process, i.e. year 3-8. The time and individual intercepts are not shown to save space. F-tests for no fixed effects are all
rejected. Hausman tests suggest fixed effects as presented in Table 24. Although included, industry and industry effects are conditioned out and their effects
absorbed by the individual intercepts. Stars indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; *** p<0.01.

The coefficients for both ownership concentration definitions are negative, suggesting
turnaround performance is negatively related to ownership structure, but the coefficients are all
statistically insignificant. Blockholder dominance negatively influences turnaround
performance, supporting the two-sided hypothesis of a negative relationship. Blockholder
dominance is weakly statistically significant in Model 7 and 8, suggesting that firms having a
dominant blockholder is weakly significantly experiencing a turnaround performance 6 pct.-
points lower than non-dominated firms. Blockholder dominance is insignificant in the remaining
models, i.e. Model 1 to 6. Takeover and block investments are having the hypothesized sign by
suggesting a positive relationship to turnaround performance, but both variables are found to be
statistically insignificant. The variable cost retrenchment is also found to both be insignificant
related to turnaround performance and to take the incorrect sign of the expected relationship. In
addition, firm size is also found to be insignificant associated with performance.
Most importantly, I find asset retrenchment to be highly significant with a negative
(positive sign) influence in the relationship with turnaround performance. The effect of asset
Page 62 of 111
retrenchment is generally very robust across the different model specifications. In terms of
effect, an increase in the asset base, which is equal to a decrease in asset retrenchment, by 1 pct.-
point is related with an increase in turnaround performance by approximately 8.8 pct.-points.
Asset retrenchment is measured by the change in asset base, where a negative coefficient
indicates that firms decreasing its asset base experience improved performance. That is asset
retrenchment results in better performance. Somewhat different, my results suggest a positive
relationship between increases in the asset base and turnaround performance, meaning asset
retrenchment is negatively related to turnaround performance.
In terms of stability, the ownership concentration variables were sensitive to the different
model specifications, which mainly stem from the inclusion of the dominant blockholder
variable. None of the other variables, as reported above, were sensitive to different
specifications. This was expected as they are strongly and highly inter-correlated. The adjusted
R
2
is reported to indicate model performance and it is approximately 36 pct. in all models.
Table 22 and Table 23 (Appendix 9) replicate the model specifications in Table 9 above and
report results for fixed effect estimations when not controlling for fixed time effects and pooled
estimations respectively. The fixed effect models in Table 22 do not differ substantially except
the variable dominant blockholder is more significant in Model 7 and 8. Asset retrenchment is
highly significant, while firm size is weakly significant when not considering time effects. The
remaining estimation of parameters is not altered in terms of sign or significance. Table 23
reports pooled estimation results. The results differ compared to the fixed effects method
estimations, which is evident from the parameters changing signs, parameters becoming
significant, and the adjusted R
2
decreasing considerably. This behaviour emphasizes the
importance of using the fixed effect approach by including firm and time fixed effects. This is
important since the variations between the two methods indicate that the independent variables
are correlated with the error term in the pooled regression models, causing the estimates to be
biased and inconsistent (Borsch-Supan & Koke, 2000)
Table 25 (Appendix 10) shows estimation results when using return on invested capital
(ROIC) as the dependent variable. Surprisingly, estimating random effects is the most
appropriate methods in this case, which is confirmed by the insignificant Hausman test for all
models. However, the adjusted R
2
is significantly low, making the estimations uninteresting.
ROIC is not considered in the forthcoming models. Similar, block investment is not reported as
Page 63 of 111
the variable has no explanatory effect in the forthcoming estimations. The inclusion of the block
investment variable did not affect the estimates of the other variables.

5.1.2. Dynamic panel estimation results
The key purpose of the dynamic models in Table 11Table 10 is to evaluate the influence of past
turnaround performance on current turnaround performance through inclusion of the lagged
dependent variable in the model specifications as shown in Equation (7). In the second column,
Table 10 shows dynamic fixed effect models results, while first column present the results
obtained by the two-step GMM estimations using the Arellano and Bond methodology build
into the SAS EG procedure.
Table 10: Results of dynamic panel regression with GMM and FE estimation
Dynamic panel models GMM Fixed effects
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Lagged turnaround
performance
0.3420***
(0.0014)
0.3827***
(0.1001)
0.4130***
(0.1038)
0.0627**
(0.0302)
0.0647**
(0.0302)
0.0651**
(0.0302)
Ownership concentration ratio -0.3346
(0.3007)
-0.4649
(0.3295)
-0.4956
(0.3515)
-0.0480
(0.0534)
-0.0135
(0.0573)
-0.0168
(0.0574)
Dominant blockholder - 0.1106
(0.1773)
0.2711
(0.2134)
- -0.0559
(0.0341)
-0.0633*
(0.0347)
Takeover - - -0.4352*
(0.2452)
- - 0.0559
(0.0497)
Cost retrenchment 0.0517**
(0.0215)
0.0564**
(0.0243)
0.0569**
(0.0248)
0.0004
(0.0038)
0.0005
(0.0038)
0.0006
(0.0038)
Asset retrenchment 0.2300***
(0.0597)
0.2193***
(0.0693)
0.2318***
(0.0708)
0.0886***
(0.0099)
0.0882***
(0.0099)
0.0881***
(0.0099)
Firm size -0.1425
(0.1123)
-0.0829
(0.1228)
-0.0741
(0.1210)
-0.0260
(0.0165)
-0.0164
(0.0165)
-0.0163
(0.0165)
Industry dummies Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes
Time dummies Yes Yes Yes Yes Yes Yes
Sargan Test (Chi
2
-statistic) 18.42 17.44 13.77
1st-order autocorrelation AR(1) - - -
2th-order autocorrelation AR(2) - - -
R
2
-value 0.3621 0.3633 0.3638
# cross-section/time length 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15
This table reports GMM and fixed effects (FE) estimation results with ownership concentration ratio as the ownership concentration variable and the
lagged dependent variable. Stars indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; *** p<0.01. Standard errors are
presented below the parameter coefficients in parentheses and the FE model are corrected for heteroscedasticity. The sample is restricted to the years in
the turnaround process, i.e. year 3-8. The time and individual intercepts are not shown to save space. The FE F-statistics for no fixed time effects are all
rejected. Joint significant tests are all significant. The Sargan statistics related to the GMM estimations all verify over-restriction of restrictions in all
models. First and second order autocorrelation tests fail to report statistics, suggesting autocorrelation in the first and/or second order regression
residuals. Five lags of the dependent variable are introduced and used as instruments.

First, the estimates from the dynamic fixed effect models (Model 5-7) reflect that past
turnaround performance has a statistically significant explanatory effect on current turnaround
performance, suggesting good turnaround performance is positive related to turnaround
Page 64 of 111
performance in the current period. An increase in the level of past turnaround performance by 1
pct.-point is generally associated with a 6.5 pct.-point increase in performance. Again, asset
retrenchment is highly significant with the incorrect predicted sign, while blockholder
dominance is weakly significant at the more generous significance level in Model 8. Model
performance in terms of adjusted R
2
does not change by the inclusion of the lagged dependent
variable. The lagged turnaround performance has not a considerable effect on the magnitude of
the estimates in Model 5-7.
The assumption of exogenous variables is violated by the inclusion of the lagged turnaround
performance in the dynamic fixed effect models and the estimations do not take the endogeneity
of this variable into account. It is noteworthy that the significant estimates remain significant,
while all estimates maintain their sign and magnitude. However, the underlying assumptions are
violated due to the endogeneity of the lagged variable (Baltagi, 2005), undermining any causal
inference based on the results. Luckily, the GMM method should be able to address these
shortcomings.
Based on the econometrics issues arising when estimating the dynamic fixed effect models,
the dynamic GMM models are estimated. The estimation results are presented in the first
column in Table 10. The dynamic GMM panel model uses lags as instruments. Hence, the
Sargan test of over-identification is reported. The reported Sargan statistics does not reject the
validity of the used instrument.
19
The results provided by the GMM estimation suggest an
extremely large and highly significant relationship lagged turnaround performance and
turnaround performance. Again, asset retrenchment is highly significant and presenting a very
large effect, while cost retrenchment also are found to be significant and takeover is found to be
weakly significant. The results depict no significant relationship between ownership
concentration and firm turnaround performance. However, the first and second order
autocorrelation tests fail to be producing any statistics, which leaves me to question the
estimation results and the choice of instruments
46
.
The extreme dynamic GMM results are likely to be caused by inappropriate instrument,
wrong instrument and the ignorance of the potential endogeneity of ownership concentration.
Assuming exogeneity of ownership concentration in the GMM estimation greatly alters the
magnitude of the estimates. Table 28 shows estimation results when holding ownership
concentration exogenous, confirming the lack of proper instruments in the GMM estimation

19
Using lower lags induce the Sargan statistic to discard the instrument.
Page 65 of 111
(Appendix 11).
20
These econometric issues are addressed in the discussion section (Section
6.1.), when discussing endogeneity issues arising from ownership concentration.

5.2. Robustness tests
This section serves to analyse the ownership and control variables effect on the turnaround
outcome. By using a binary dependent variable, I test the alternative definitions and analyse the
variables in a different approach. This alternative approach will enhance the understanding both
in terms of importance but also direction of the individual variables relation to the turnaround
outcome. Further, it allows me to check whether my above results can be considered as robust
against alternative definition of turnaround, that is, the outcome instead of actual performance
during the turnaround process.

5.2.1. Logistic estimation results
By reusing the previous model specifications (Model 5-7) and introduce the binary response
variable, the regression models can be estimated by the logistic method for the two definitions.
Table 11 reports estimation results. In most cases, the computation of the estimates failed when
including dummy variables. Therefore, before addressing the results, it should be noted that
industry and country dummies explicitly are omitted in the logistic regression models due to
convergence difficulties. However, the time-invariant effects are included by the individual
intercept. Similar, I had to relax convergence criteria to ensure variables in the model
specifications were estimated.









20
Table 26 shows estimation results without considering time effects in the dynamic model specifications, while
Table 27 shows estimation results of dynamic pooled regression (Appendix 11). These tables emphasize the use of
firm-specific and time effects.
Page 66 of 111
Table 11: Estimation results from logit fixed effects models of turnaround outcome
Logistic panel models Definition 2a Definition 2b
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Ownership concentration
ratio
-4.7080*
(2.5856)
-1.3828
(2.3909)
-1.5857
(2.4731)
1.2067
(2.2772)
4.7775*
(2.7994)
-0.5156
(1.4649)
Dominant blockholder - -5.6229***
(1.7136)
-5.6630
(1.9325)
- -4.5065**
(1.7936)
-3.7338***
(1.4908)
Takeover - - -0.7396***
(3.9945)
- - -0.6856
(1.9042)
Cost retrenchment -0.4770
(0.4859)
-0.6342
(0.4570)
-0.7167
(0.4611)
-0.1524
(0.6338)
-0.5451
(0.5326)
-0.8015*
(0.3513)
Asset retrenchment 0.4708
(0.5189)
0.8490*
(0.4678)
0.8576*
(0.3452)
0.0895
(0.6513)
0.3784
(0.0000)
0.2722
(0.3305)
Firm size 1.0531***
(0.3691)
0.7839**
(0.4000)
0.8372**
(0.4245)
0.0868
(0.3129)
0.9039**
(0.4231)
0.4467**
(0.2174)
Industry dummies No No No No No No
Country dummies No No No No No No
Time dummies Yes Yes Yes Yes Yes Yes
-2 Log Likelihood 265.80 270.41 271.53 302.89 376.72 451.84
# of observations 912 912 912 1194 1194 1194
This table reports fixed effect logit regression results with the dependent variable being turnaround outcome for the two alternative definitions 2a and
2b. Due to low fit statistic output in SAS EG, the -2 Log Likelihood is the only reported fit statistics. The sample is restricted to the years in the
turnaround process, i.e. year 3-8. Standard errors are presented below the estimation results in parentheses. Intercepts are not reported to save space.
Model estimation is by Maximum Likelihood (ML). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. Table 31 (Appendix
14) report the estimation results with the industry and country dummies included.

As indicated, I had to relax convergence and estimation criteria of almost all variables. Hence, it
is reasonable to question the estimated variables and their effects. Eliminating problematic
variables in the models specification is sometimes recommended. However, this could also
result in biased estimates of the remaining variables in the models (Baltagi, 2005). Along with
other potential issues affecting the estimates, for instance such as endogeneity of ownership
concentration and potential omitted variable bias, the results are only indicative and should be
evaluated with care.
The results are with regard to ownership concentration conflicting and, aside from the
weakly significant estimates in Model 5 (Definition 2a) and Model 6 (Definition 2b), not
significant, indicating insignificant relationship between ownership concentration and
turnaround outcome. The blockholder dominance variable is highly significant and with the
opposite sign than expected. Although blockholder dominance is insignificant in Model 7
(Definition 2a), the sign has the same direction as in the other models. Takeover is significant in
Model 7 (Definition 2a), while insignificant in Model 7 (Definition 2b). Furthermore, the
relationship is opposite the one expected, and thus suggesting a negative association with
turnaround outcome. Cost and asset retrenchment is generally either insignificant or weakly
significant. Only cost retrenchment has the expected sign. Apart from previous models, firm size
Page 67 of 111
is found to be statistically significant (except Model 5, Definition 2a) and positively associated
with successful turnarounds, which was expected based on the descriptive statistics in Table 17
and Table 18 (Appendix 6).
In general, the results in Table 11 are changing both in terms of estimated direction,
magnitude and significance, suggesting the robustness of the models can be questioned.
Although, the interpretation of the estimates is possible in terms of direction, the marginal effect
of each dependent variable is difficult given the logistic structure of the models. It would be
convenient to be able to interpret the partial change in the probability of turnaround given a
change in one of the explanatory variables. Unfortunately, SAS EG fail to estimate the marginal
effects. Therefore, odds ratios are provided in Table 30 (Appendix 13). Pooled logistic
regression is reported in Table 33 and Table 34 for comparison (Appendix 15). Appendix 16
provides logit analysis for each year in the turnaround process as in Mueller and Barker (1997).
Conclusively, the robustness of the models is questionable and results are indicative.

Page 68 of 111
6. DISCUSSION
According to the estimation results, which are concisely summarized in Table 12, ownership
concentration is found to be statistically insignificantly related to turnaround. This is
discouraging for the perhaps too simplified hypothesis suggesting that higher ownership
concentration would induce better monitoring and controlling in turnaround situations, and
therefore lead to improved performance and successful turnaround outcome. Therefore, the
initial expectations are invalid and there is no evidence supporting Hypothesis 1. However,
ownership concentration is discussed further below in Section 6.1.
Table 12: Summary of estimation results
Standard fixed effect Dynamic Logit
Variables FE FE GMM FE FE FE
Herfindahl ownership index -3.53E-6
(9.03E-6)
- - - - -
Ownership concentration ratio - -0.0167
(0.0574)
-0.4956
(0.3515)
-0.0168
(0.0574)
-1.5857
(2.4731)
-0.5156
(1.4649)
Dominant blockholder -0.0533
(0.0421)
-0.0602*
(0.0347)
0.2711
(0.2134)
-0.0633*
(0.0347)
-5.6630
(1.9325)
-3.7338***
(1.4908)
Takeover 0.0547
(0.0497)
0.0544
(0.0497)
-0.4352*
(0.2452)
0.0559
(0.0497)
-0.7396***
(3.9945)
-0.6856
(1.9042)
Cost retrenchment 0.0016
(0.0038)
0.0015
(0.0038)
0.0569**
(0.0248)
0.0006
(0.0038)
-0.7167
(0.4611)
-0.8015*
(0.3513)
Asset retrenchment 0.0878***
(0.0099)
0.0877***
(0.0099)
0.2318***
(0.0708)
0.0881***
(0.0099)
0.8576*
(0.3452)
0.2722
(0.3305)
Firm size -0.0127
(0.0164)
-0.0122
(0.0164)
-0.0741
(0.1210)
-0.0163
(0.0165)
0.8372**
(0.4245)
0.4467**
(0.2174)
Lagged turnaround perform - - 0.4130***
(0.1038)
0.0651**
(0.0302)
- -
Industry dummies Yes Yes Yes Yes No No
Country dummies Yes Yes Yes Yes No No
Time dummies Yes Yes Yes Yes Yes Yes
Total observations 1734 1734 1734 1734 912 1194
This table summarized the main estimation results for the standard panel models, dynamic panel models, and logistic panel models. The summary is
based on Model 7 estimation results. Individual firm intercepts are not reported to save space. The fixed effect and dynamic models are estimated by
OLS, while the logistic model estimation is by Maximum Likelihood. In the logit column, the first column report results for definition 2a and the
latter definition 2b. Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01.

Dominant blockholdings, which are mainly weakly significant, indicate a negative effect on
corporate turnarounds. This confirms Hypothesis 2b, stating that dominant blockholders are
negatively related with turnaround performance and outcome, while Hypothesis 2a is rejected.
An explanation of this relationship may be that when firms are dominated by a dominant
blockholder, other blockholders have a less pronounced influence, which may undermine their
ability to monitor and exert control of management. Further, the dominant blockholder may
advocate strategies not favoured by minority blockholders and other stakeholders. Blockholders
Page 69 of 111
with only moderate power will have little influence on both initiation and extent of turnaround
measures taken, which is consistent with Lai and Sudarsanam (1997) and Bethel and Liebeskind
(1993), who suggest dominant blockholders may exert a constraining effect in the turnaround
process. Hence, dominant blockholders seem a poor element of the governance mechanism of
ownership in the context of corporate turnarounds.
In the empirical testing, I used takeover as an overall measure for large changes in
ownership to test the hypothesis of takeovers as a remedy against poor performance, which
increases the likelihood of successful turnaround. The measure is changing in terms of sign and
significance, making it difficult to interpret the effect of takeover. The measure covers several
events and make up rather few observations of the total observations, which is two potential
explanations for the insignificance. First, the variable covers three types of large changes in
ownership, which potentially could have contradictory effects. Second, Denis and McConnell
(2003) report that large ownership changes rarely happens in the European context, which
mainly is owing to the relative high ownership concentration and ownership stability in most of
the European countries. In this perspective, the few large ownership changes and the
insignificant effect might not be surprising. Table 39 in Appendix 17 shows the average size of
the largest top blockholder and the average of ownership concentration in the respective
countries in the sample.
I now turn to the second aspect of ownership change, which is reported in Table 9 only. The
effects of block investments, which reflect the entry of new blockholder, were found to be
insignificant in all estimations. This is inconsistent with Bethel and Liebeskind (1993), who
notice that block investments are associated with initiation of turnaround actions, which may be
a result of new blockholders having a disciplinary and pressuring effect on management to
undertake corrective measures. The control variables, which are included to control for
turnaround characteristics, are discussed in Section 6.2.

6.1. Firm turnaround performance and ownership structure – A question of endogeneity?
Ownership concentration has been treated as an exogenous variable in the empirical analysis.
This is despite the fact that ownership concentration has been advocated as being endogenous
especially since the paper by Demsetz & Lehn (1985). However, ownership concentration may
actually be endogenous, also introducing the problem of reverse causality. Ownership
concentration and changes herein may be results of turnaround performance and vice versa. For
Page 70 of 111
example, deteriorating and declining performance may cause a reduction in ownership
concentration as incumbent blockholders seek the option to lower their holding or to opt out. On
the other hand, other blockholders may be attracted by poor performance as they hold the
expectation of revived performance in the near future or confidence to the business model given
few changes and, thus, their entry as new blockholders may in turn be associated with
performance recovery. Therefore, a relationship between ownership concentration and
performance may not exist. This is consistent with the view of Borsch-Supan and Koke (2000)
and Demsetz and Lehn (1985) who suggest ownership structure and concentration is balanced
to reflect the nature of the firm, in which case the structure should have an enhancing and
maximising effect on performance irrespectively of the ownership concentration.
If the suggested endogeneity of ownership - and potentially also other explanatory variables
-is valid and present in the model specifications, the variables will be correlated with the error
term, causing standard fixed effect models to produce biased and inconsistent estimates.
Therefore, the GMM regression methodology is applied. As discussed, the applied GMM
methodology only addresses endogeneity of turnaround performance variable in the dynamic
models, which produce extreme estimation results. Comparing GMM estimation results in Table
12 with Table 28, where the ownership concentration ratio is assumed exogenous, the variation
in the size of coefficients is remarkable. In particular, the coefficients of lagged turnaround
performance, ownership concentration ratio, and firm size change significantly. This provides
support to the expectation that ownership concentration and potentially other variables are
endogenous and, hence, endogeneity issues are likely to be present.
A remedy would be to introduce other instruments and focus on ownership concentration
ratio. According to Cameron and Trivedi (2005), a good instrument should be correlated with
the endogenous variables, which in the dynamic setting include at least the lagged dependent
variable and ownership concentration. Furthermore, the instrument should be uncorrelated with
turnaround performance. However, as noted by Bennedsen and Nielsen (2010), very few strong
instruments exist, making the GMM approach inappropriate. Bennedsen and Nielsen (2010)
state that biases resulting from poor and unqualified instruments to control for endogeneity are
much more severe than the biases arising in standard panel models. Therefore, they suggest
addressing underlying issues related to endogeneity, e.g. measurement errors, omitted variable
bias and reverse causality. Some of these issues are discussed in Section 6.3., while others have
already been addressed. I have attempted various model configurations, specifications and
Page 71 of 111
instruments using my limited data, but it did not yield any satisfying results. Finding appropriate
instruments and other useful econometric techniques are deemed outside the scope of this thesis.
To end this section, I basically have difficulties in ensuring that my model specifications are
not plagued by endogenous variables – especially with regard to the ownership variables.
Another issue is that the remaining explanatory variables are also likely to be endogenous. For
now, my findings suggest that ownership concentration has no significant relationship with
turnaround, which is a confirmation of the implicit null hypothesis. However, it may not be
possible with the current model specifications to make meaningful comments regarding the
relationship as the data and model specifications are possibly affected by endogeneity issues. In
this case it makes me unable to answer the hypothesis and whether ownership concentrations in
turnaround situations is consistent with the theoretical and empirical predictions about
concentrated ownership as a good governance mechanism in a Western European context.

6.2. Corporate turnarounds: Too complex a phenomenon?
Given a large number of variables are found to be insignificant with turnaround and sometimes
showing opposite behaviour than expected, I generally have to reflect on the question whether
corporate turnarounds simply are too complex a phenomenon and much more than ownership
and governance arrangements. One may ask, can corporate turnaround be modelled and
conceptualized through equations? The easy answer is no. As a pleasant contrast to the easy
answer, a corporate turnaround is indeed a complex phenomenon and variables, where firm-
specific solutions are required to reverse performance declines. Nevertheless, common factors
are present in such situations. The empirical studies are not able to examine exact processes and
mechanisms, but they provide evidence for the existence of common relationships. For example,
whenever statistically significant, I find takeovers to exert a negative relationship with
turnaround, which is contrary to my hypothesis stating that takeovers may lead to change and
access to new resources. Employing empirical methodologies is a way forward to understand
such associations to turnaround. Although, a deeper understanding is likely only to be reached
through studies beyond the econometric perspective, e.g. case studies, where it is possible to
take into account aspects that are hard to measure and transform into a variable, e.g. strategic
process, leadership, communication, board interference and response to decline, access to
external resources, management change, sense of urgency, acceptance of crisis, etc.
Page 72 of 111
Additionally, the evolving models in the turnaround literature are a good indication of the fact
that the understanding of the phenomenon is constantly being challenged and expanded.
An implication affecting the understanding of governance in turnarounds is the framework
supporting the development of hypotheses. Despite employing life-cycle and resource-
dependency theory, I mainly use agency theory as the theoretical foundation in building the
hypotheses related to ownership, which largely builds on two assumptions that may be
questioned in the context of turnarounds. First, owners and management are likely not to have
misaligned objectives in a turnaround situation. For example, as indicated earlier, Mueller and
Barker (1997) argue that both owners and management are likely to suffer monetarily from
turnaround failure, giving both parties the incentive to collaborate for a common goal and
positive turnaround outcome. In this perspective there are no conflicting interests, which makes
Mueller and Barker (1997) conclude that it may be difficult to accept agency theory as an
appropriate theoretical approach and, thus, undermine the use of agency theory. Here, I think
such a conclusion is too one-sided and only partly true. Agency theory is indeed based on an
underlying assumption of conflict between principal and agent interests, but it is a quite stark
conclusion that all parties will unite in a common effort in a turnaround situation.
In a turnaround situation with declining and life-threatening performance, all parties will
probably work towards stabilization and recovery of the firm, e.g. by retrenchment, but
everybody within the firm will attempt to do so from an individual perspective, where their own
interests will be present. For example, every individual and department within the firm will
likely have the perception that retrenchment, head-count cuts, shutdown of activities should be
implemented in departments other than their own. Hence, there is likely to be internal power
struggles across organisation layers and departments and shifts in power structures, despite the
fact that the firm as a whole is working together towards the common goal of recovery. Based
on this example, conflicting interests exist in turnaround situations. Hence, I think that agency
theory is applicable and relevant.
Second, I reckon that assuming top management only seek to maximize personal wealth,
which necessitates concentrated ownership as a necessary governance mechanism to discipline
top management, is likely not to always be true in the context of turnarounds. In discussing poor
top management, Bibeault (1999) cite Emerson: “An institution is the lengthened shadow of one
man”. A firm is often said to be shaped and created by the top management and in particular the
CEO, who will be strongly motivated to ensure firm-survival to maintain personal status and
Page 73 of 111
career aspects. This perception conflicts with the normal perception in agency theory. However,
as indicated when discussion conflicting objectives, individual interests are likely to persist in
the turnaround situation.
Based on the two perspectives addressed above, the role of concentrated ownership may be
more blurred in turnarounds and in practice than suggested by the agency theory. Governance
functions change between value-protecting and value-creating activities, which not necessarily
are reflected in the agency theory, and the ownership structure is not uniformly effective during
the firm’s life-cycle. Therefore, it is important to incorporate other theoretical approaches when
discussing turnarounds and address other governance mechanisms in the turnaround process, but
this does not imply that agency theory does not have its usefulness when investigating
turnaround situations.
Another limitation of this thesis is the fundamental assumption that blockholder exercise
their power and engage actively in the turnaround and non-turnaround firms included in the
sample. Although shareholder activism and blockholder engagement is prevalent in most
European countries, especially some type of blockholders have been criticised for their passivity
(Nielsen, 2012). Hence, some types of blockholders may be having more disciplinary effect on
management than others.

6.2.1. Retrenchment
Besides trying to explain turnaround performance and outcome from the alternative perspective
of ownership structure and variations herein, I have focused particularly on the role of
retrenchment. Aside from the potential issues arising from the model specifications, I believe it
is an important finding that asset retrenchment seems significantly but oppositely related to
turnaround than expected.
The actual association in my sample contradicts the role of retrenchment as otherwise warmly
advocated by Pearce and Robbins (1992) and, thus, the results do not support asset retrenchment
as an essential strategic action in the turnaround process as otherwise normally proven (e.g.
Pearce & Robbins, 1992; Robbins & Pearce, 1994; Bruton et al., 2003; Francis & Desai, 2005).
Even Barker & Mone (1994), who has discussed the question of generalizability and causality of
retrenchment, support that asset retrenchment only among firms that experience severe and life-
threatening performance declines leads to performance improvement. Thus, my findings
contradict previous empirical findings by suggesting that firms suffering from decline should
Page 74 of 111
increase their asset base to improve performance, meaning that asset retrenchment appears to be
negatively related to firm turnaround performance.
Much of the literature that advocate asset retrenchment as a fundamental turnaround
strategy has been conducted with samples restricted to specific types of industry, e.g.
manufacturing, growth-intensive, competitive environment and similar (e.g. Pearce & Robbins,
1992; Morrow et al., 2004). It is further argued that since basic industry characteristics are
individual from industry to industry, e.g. between manufacturing and service, the content of
turnaround strategies diverge significantly (Barker & Duhaime, 1997). For example, Morrow et
al. (2004) note that both cost and asset retrenchment are insignificantly related to firm value for
firms attempting turnaround in declining industries, while Francis and Desai (2005) find cost
retrenchment negatively related to turnaround performance in growth industries. I have,
compared to other studies, constructed a rather heterogeneous sample combining several
industries, which may have created a comparability issue between the industries of interest.
When looking at the average asset retrenchment across industries, the difference is highly
significant (Table 40, Appendix 19). Hence, I suspect my conflicting results arise from
differences between industries. For example, firms operating in the service industry may
respond differently to performance decline than manufacturing firms. An explanation to the
difference may also stem from variation in turnaround measures taken within industries. For
example, manufacturing firms are stated to normally initiate restructuring activities to correct for
overexpansion and over-diversification, thereby shrinking back to the viable core business
(Bethel and Liebeskind, 1993). Similar, manufacturing firms are noticed to often improve their
competitive position by decreasing expenses and improving asset utilisation (Bibeault, 1999;
Francis & Desai 2005). Retrenchment may not be appropriate to regain competitiveness in
different industries. Table 40 (Appendix 19) shows that the average manufacturing firm
increased the asset base by 0.35 pct. in the turnaround process, while the average increase
amounted to a total of 14.17 pct. in asset base by firms in the transportation, communication,
and utilities industry. Firm in the latter industry may need larger asset investment to overcome
decline and regain its competitive position. These considerations are only examples, but
illustrate the possible explanations.
Retrenchment is subject to two limitations. First, the definition and measurement of
retrenchment do not take the phases into account. Firms are expected to be retrenching more
during the decline phase, while retrenching less or actually increasing the asset base during the
Page 75 of 111
recovery period. Second, it does not address the question regarding causality of retrenchment
raised by Barker & Mone (1994). They argue retrenchment activities to be a consequence of
severe performance decline while it is not causing improvements in turnaround performance.
This is contrary to the perception of Pearce and Robbins (1994), who advocate retrenchment as
an essential means for improved turnaround performance. Therefore, further investigations
should note these diverging views.
Overall, the results indicate that the nature of the industry in which the firm operates
influences the level and effect of retrenchment, and challenge the perspective that retrenchment
is a fundamental turnaround action among declining firms seeking to reverse performance
decline.

6.2.2. Firm size – I s size of importance?
Firm size is found to be insignificantly related to turnaround performance, while being
significant and positively related to turnaround outcome. This finding is consistent with Barker
et al. (2010) and Abebe et al. (2011). Thus, firm size significantly set non-turnaround firms
aside from turnaround firms, but does not exert impact on the level of performance. A possible
explanation is that larger firms may have larger slack resources, superior resources, are better
able to change strategy and seize business opportunities or are more prone to replacing poor top
management.

6.3. Econometric erosions, limitations and considerations
Overall, the methodological approach is subject to weaknesses and strengths. I have throughout
the thesis attempted to avoid potential caveats and several aspects have already been pinpointed
and addressed, e.g. unobserved heterogeneity must be assumed time-invariant and constant to
produce unbiased and consistent estimations in the fixed effect models, endogenous impact not
captured by the individual intercepts will create biased estimations, potential bias due to omitted
time-varying characteristics and so on.
Further, two other issues that generally plague empirical research by the use of
econometrics is sample selectivity and measurement errors. Limitations arising from
measurement errors are already discussed in Section 4.3.1., where I addressed constraints
regarding data accessibility and availability. Data availability has an influence on the
investigated aspects and is therefore greatly connected to potential omitted variables, e.g. type of
Page 76 of 111
shareholder and pyramids. Further, I do not examine share buybacks and share issues due to lack
of data. Similar, Section 4.1 addresses issues related to sampling selection, and the sampling
procedure is defined with considerations to minimize pitfalls and selectivity bias by following
recommendations provided in the literature on improved turnaround and governance studies
(e.g. Pandit, 2000; Borsch-Supan & Koke; 2000; Balcaen & Ooghe, 2006). I have constructed
the sampling procedure to ensure comparability of the results, while focusing on the sample
size. It is complicated to identify turnaround firms due to the nature of causes, variation in
strategic actions and multiple understandings of decline and recovery, making the perfect
identification impossible. Hence, sample selectivity is always a potential issue when examining
corporate turnarounds empirically.

Page 77 of 111
7. CONCLUSION
My thesis has empirically explored the relationship between ownership structure and corporate
turnaround performance and outcome. Based on prior literature, I have suggested that ownership
structure and governance arrangements must be fundamentally aligned during the turnaround
process to ensure successful recovery. I employ a detailed sample procedure to construct a
sample consisting of firms that have experienced severe and life-threatening performance
declines, thus experiencing genuine turnaround situations. I estimate models by specifying fixed
effect, dynamic, and logistic models based on a heterogeneous panel dataset consisting of
Western European firms gathered within the period 1995 to 2010.
In general, the findings of my analyses suggest that ownership concentration has no
relationship to corporate turnaround performance and outcome. I find dominant blockholding to
be weakly related to turnaround and having a negative influence, which confirms my competing
hypothesis, suggesting that dominant blockholdings have a destroying effect on ownership
concentration as a governance mechanism in turnaround situations. The entry of new
blockholders and large changes in ownership contrast my expectations and are not related to
turnaround performance or outcome. Furthermore, I find firm size to exercise no influence on
turnaround performance, while being positively related to turnaround outcome.
Cost and asset retrenchment are advocated to have an impressive and positive impact on
turnaround performance. However, I find cost retrenchment, although having the predicted
direction, to generally have no effect. Opposite, I find asset retrenchment to be negatively
related to turnaround performance and outcome. This is in partial contrast to the general
conception of retrenchment as an essential element in the turnaround strategy. Lastly, past
turnaround performance has a large explanatory effect on current turnaround performance.
Unfortunately, my findings reveal that that econometric issues and particularly endogeneity
complicate the investigation of a relationship between ownership concentration and turnaround,
confirming that the results should be treated with care and considered indicative.
The findings of my thesis provide a first step towards understanding the relation between
ownership structure and turnaround performance and outcome in Western European firms. I add
to the area by integrating and investigating the effect of governance arrangements on the
turnaround process. More importantly, I show that the governance mechanism of ownership
concentration does not affect corporate turnarounds, suggesting the effect of governance
Page 78 of 111
mechanism may shift as the firm moves through the life-cycle stages and other mechanisms may
be more effective in the turnaround process. In light of the latest theory, the alignment of
governance functions may be so individually rooted that no common pattern exits.

7.1. Future research
In general, there are several unanswered questions with regard to ownership and turnarounds,
and in hindsight, there are many variables and extensions of the study framework that could
potentially enhance the understanding and ability to distinguish turnaround firms from non-
turnaround firms. In addition to already suggested extensions, future studies are specifically
recommended to divide ownership by the type of blockholder, e.g. institutional or private, due to
their potentially different objectives and commitment. The effect of governance mechanisms
(both internal and external) may shift between stages in the life cycle and, for this reason, other
studies are encouraged to investigate other governance aspects, e.g. board of directors. Future
studies are encouraged to employ more advanced econometric techniques to address issues that
arise beyond the area of basic econometrics in order to gain a refined and deeper empirical
understanding of the subject.



Page 79 of 111
LIST OF REFERENCES
Abebe, M. A. (2010). Top team composition and corporate turnaround under environmental
stability and turbulence. Leadership & Organization Development Journal, 31(3), 196-212.
Abebe, M. A., Angriawan, A., & Ruth, D. (2012). Founder-CEOs, external board appointments,
and the likelihood of corporate turnaround in declining firms. Journal of Leadership &
Organizational Studies, 19(3), 273-283.
Abebe, M. A., Angriawan, A., & Yanxin Liu, . (2011). CEO power and organizational
turnaround in declining firms: Does environment play a role? Journal of Leadership &
Organizational Studies, 18(2), 260-273.
Altman, E. I. (1983). Corporate financial distress: A complete guide to predicting and avoiding
bankruptcy. New York: Wiley,
Altman, E. I. (2000). Predicting financial distress of companies: Revisiting the Z-score and
ZETA models. Stern School of Business, New York University,
Arogyaswamy, K., Barker, V. L., & Yasai-Ardekani, M. (1995). Firm turnarounds: An
integrative two-stage model*. Journal of Management Studies, 32(4), 493-525.
Balcaen, S., & Ooghe, H. (2006). 35 years of studies on business failure: An overview of the
classic statistical methodologies and their related problems. The British Accounting
Review, 38(1), 63-93.
Baltagi, B. H. (2005). Econometric analysis of panel data (3rd ed.) John Wiley & Sons Ltd.
Barker III, V. L., & Barr, P. S. (2002). Linking top manager attributions to strategic
reorientation in declining firms attempting turnarounds. Journal of Business Research,
55(12), 963-979.
Barker III, V. L., & Duhaime, I. M. (1997). Strategic change in the turnaround process: Theory
and empirical evidence. Strategic Management Journal, 18(1), 13-38.
Barker III, V. L., Patterson Jr, P. W., & Mueller, G. C. (2001). Organizational causes and
strategic consequences of the extent of top management team replacement during
turnaround attempts. Journal of Management Studies, 38(2), 235-270.
Barker, V. L., & Mone, M. A. (1994). Retrenchment: Cause of turnaround or consequence of
decline? Strategic Management Journal, 15(5), 395-405.
Becht, M., Bolton, P., & Röell, A. (2003). Chapter 1: Corporate governance and control. In G.
Constantinides, M. Harris & R. Stulz (Eds.), Handbook of the economics of finance (pp. 1-
109) Elsevier.
Page 80 of 111
Bethel, J. E., & Liebeskind, J. (1993). The effects of ownership structure on corporate
restructuring. Strategic Management Journal, 14(S1), 15-31.
Bibeault, D. B. (1999). In Originally published: New York, McGraw-Hill, 1982 (Ed.),
Corporate turnarounds: How managers turn losers into winners! (2nd ed.). U.S.: Beard
Books.
Borsch-Supan, A. H., & Koke, J. F. (2000). An applied econometricians' view of empirical
corporate governance studies. ZEW Discussion Paper, 00(17)
Bruton, G. D., Ahlstrom, D., & Wan, J. C. C. (2003). Turnaround in east asian firms: Evidence
from ethnic overseas chinese communities. Strategic Management Journal, 24(6), 519-
540.
Cameron, A. C., & Trivedi, P. K. (2005). Microeconometrics: Methods and applications. New
York, USA: Cambridge University Press.
Chakrabarti, A., Vidal, E., & Mitchell, W. (2011). Business transformation in heterogeneous
environments: The impact of market development and firm strength on retrenchment and
growth reconfiguration. Global Strategy Journal, 1(1-2), 6-26.
Demsetz, H., & Lehn, K. (1985). The structure of corporate ownership: Causes and
consequences. Journal of Political Economy, 93(6), 1155-1177.
Denis, D. K., & McConnell, J. J. (2003). International corporate governance. Journal of
Financial and Quantitative Analysis, 38(01), 1.
Fich, E. M., & Slezak, S. L. (2008). Can corporate governance save distressed firms from
bankruptcy? an empirical analysis. Review of Quantitative Finance and Accounting, 30(2),
225-251.
Filatotchev, I., & Toms, S. (2006). Corporate governance and financial constraints on strategic
turnarounds. Journal of Management Studies, 43(3), 407-431.
Filatotchev, I., Toms, S., & Wright, M. (2006). The firm's strategic dynamics and corporate
governance life-cycle. International Journal of Managerial Finance, 2(4), 256-279.
Francis, J. D., & Desai, A. B. (2005). Situational and organizational determinants of turnaround.
Management Decision, 43(9), 1203-1224.
Fuglsang, L., Hagedorn-Rasmussen, P., & Olsen, P. B. (2007). In Olsen P. B. (Ed.), Teknikker i
samfundsvidenskaberne (1st ed.) Roskilde Universitetsforlag.
Furman, J. L., & McGahan, A. M. (2002). Turnarounds. Managerial and Decision Economics,
23(4-5), 283-300.
Page 81 of 111
Greene, W. H., & Hensher, D. A. (2010). Modeling ordered choices: A primer (1st ed.)
Cambridge University Press.
Gujarati, D. N., & Porter, D. C. (2009). Basic econometrics (5th ed.) McGraw-Hill/Irwin.
Hofer, C. W. (1980). Turnaround strategies. Journal of Business Strategy, 1(1), 19-31.
Holderness, C. G. (2003). A survey of blockholders and corporate control. Economic Policy
Review, 9(1)
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency
costs and ownership structure. Journal of Financial Economics, Printed in "Reader for
Firm Theory and Corporate Governance", 3(4), 305-360.
Jostarndt, P., & Sautner, Z. (2008). Financial distress, corporate control, and management
turnover. Journal of Banking & Finance, 32(10), 2188-2204.
Kazozcu, S. B. (2011). Role of strategic flexibility in the choice of turnaround strategies: A
resource based approach. Procedia - Social and Behavioral Sciences, 24(0), 444-459.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor protection and
corporate governance. Journal of Financial Economics, 58(1–2), 3-27.
Laeven, L., & Levine, R. (2008). Complex ownership structures and corporate valuation. Review
of Financial Studies, Oxford University Press for Society for Financial Studies, 21(2), 579-
604.
Lai, J., & Sudarsanam, S. (1997). Corporate restructuring in response to performance decline:
Impact of ownership, governance and lenders. European Finance Review, 1(2), 197-233.
Loui, D., & Smith, M. (2006). Financial distress and corporate turnaround: A review of the
literature and agenda for research. Unpublished School of Accounting, Finance &
Economics, Edith Cowan University.
Morrow Jr., J. L., Johnson, R. A., & Busenitz, L. W. (2004). The effects of cost and asset
retrenchment on firm performance: The overlooked role of a Firm’s competitive
environment. Journal of Management, 30(2), 189-208.
Mueller, G. C., & Barker III, V. L. (1997). Upper echelons and board characteristics of
turnaround and nonturnaround declining firms. Journal of Business Research, 39(2), 119-
134.
Nielsen, H. B. (2012, 20/09). Danskes storaktionær raser over dovne investorer. Message posted
to http://finanswatch.dk.esc-
web.lib.cbs.dk/secure/Finansnyt/Pengeinstitutter/article4846415.ece
Page 82 of 111
Paint, L. W. (1991). In investigation of industry and firm structural characteristics in corporate
turnarounds. Journal of Management Studies, 28(6), 623-643.
Pandit, N. R. (2000). Some recommendations for improved research on corporate turnarounds.
Management, 3(2), 31-56.
Pearce, J. A., & Robbins, K. D. (1993). Toward improved theory and research on business
turnaround. Journal of Management, 19(3), 613-636.
Pearce, J. A., & Robbins, K. D. (1994). Retrenchment remains the foundation of business
turnaround. Strategic Management Journal, 15(5), 407-417.
Pearce, J. A., & Robbins, K. D. (2008). Strategic transformation as the essential last step in the
process of corporate turnaround. Business Horizons, 51, 121-130.
Plenborg, T., & Petersen, C. (2011). Financial statement analysis: Valuation - credit analysis -
executive compensation (1st ed.) Financial Times Press.
Robbins, K. D., & Pearce, J. A. (1992). Turnaround: Retrenchment and recovery. Strategic
Management Journal, 13(4), 287-309.
SAS Institute. (2010). SAS/ETS 9.22 user's guide (1st ed.). Cary, NC: SAS Institute Inc.
Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. The Journal of
Finance, 52(2), 737-783.
Slatter, S., & Lovett, D. (1999). Corporate turnaround: Managing companies in distress (9th
ed.) Penguin Books.
Smith, M., & Graves, C. (2005). Corporate turnaround and financial distress. Managerial
Auditing Journal, 20(3), 304-320.
Standard & Poors. (2003). Compustat user's guide, The McGraw-Hill Companies, Inc.
Sudarsanam, S., & Lai, J. (2001). Corporate financial distress and turnaround strategies: An
empirical analysis. British Journal of Management, 12(3), 183-199.


Page 83 of 111
APPENDIXES
Appendix 1
If deemed important, the following appendix provides a short review of the different
considerations and time frames of the turnaround cycle employed in the empirical literature.
Bibeault (1992) attempted to examine the average length of the turnaround cycle and
demonstrated the average length of the turnaround cycle to be 7.8 years, resulting from the
average time period of the decline phase being 3.7 years and the average time period of the
recovery phase being 4.1 years. However, the firms included in the sample all had to have at
least 3 years of decline and were all major U.S. based companies, which bias the length of the
recovery period and lessens the comparability to a Western European context. A connection to
prior research is therefore needed.
As Slatter and Lovett (1999) explain the typical turnaround cycle, the typical length is
several years with successively lower performance and severe distress, which they describe as a
situation with significant financial losses and negative cash-flows, before the firm either
continue to perform poorly or return to prosperity. The difficulties in determining the time frame
is evident from the different definitions in prior empirical studies (e.g. Barker & Duhaime, 1997;
Paint, 1991; Sudarsanam & Lai, 2001; Furman & McGahan, 2002; Smith & Graves, 2005),
where the turnaround cycle time period span from a few years to almost a decade. A range of
definitions has been used to define the turnaround cycle time period, e.g. Smith and Graves
(2005) uses a turnaround cycle of 4 years divided into 2 years of decline followed by two years
of potential recovery period. They argue that it is sufficient time to observe a successful
turnaround in this time period. Secondly, they explain that extending the turnaround cycle time
period beyond 4 years will significantly reduce sample size, thus reducing the reliability of their
findings. Similar, Robbins and Pearce (1992) construct their turnaround cycle to consist of a
decline phase of at least 2 years and at least 2 years of increased performance, resulting in a
turnaround cycle time period of at least 4 years. Also, the study performed by Sudarsanam and
Lai (2001) also belongs to the category of short turnaround time periods. They operate with a
base period prior to the actual turnaround cycle consisting of two years, while the actual decline
or distress year consists of one year. Opposite both Mueller and Barker (1997), Francis and
Desai (2005), and Abebe et al. (2011) all use a turnaround cycle consisting of 6 years, while
Paint (1991) uses a turnaround cycle consisting of 8 years. More extensively, Barker and
Page 84 of 111
Duhaime (1997) use a time frame of potentially 9 years by allowing up to 3 years of fluctuating
performance between the decline and recovery period. In total, it is evident that the definitions
are widespread and the length is often more or less arbitrarily chosen.

Appendix 2
The following appendix provides a throughout explanation of Altman’s Z-score model and how
the individual ratios capture important turnaround aspects. Especially the last part of the
appendix addresses financial slack, which are indirectly reflected by the score. Altman (1983)
developed the Z-score using financial measures to predict bankruptcy for publicly traded
manufacturing firms, why the Z-score value is a powerful measure of firms’ financial condition.
The model in a linear model based on five firm-level financial measures which are weighted by
five estimated coefficients and then summed up to an overall score. As stated, the Z-score model
is specified as follows:
Z = 1.2X
1
+ 1.4X
2
+ 3.3X
3
+ 0.6X
4
+ 0.999X
5

where,
X
1
= working capital / total assets
X
2
= retained earnings / total assets
X
3
= EBIT / total assets
X
4
= market value of equity / total liabilities
X
5
= sales / total assets,
Z = overall score.
(11)

The five financial ratios weighted and included in the Z-score are individually described below
based on the paper by Altman (2000) and the individual ratio is linked to the context of my
study:

X
1
, Working Capital / Total Assets (WC/TA):
The WC/TA ratio, frequently used in studies of corporate financial performance, expresses the
liquidity position of the company towards the total amount of assets. Working capital is defined
as the difference between current assets (e.g. inventories, receivables, prepayments)
21
and

21
Current assets are characterized as being part of the normal operating cycle and are intended for sale, trade or use,
2) expected to be realized within 12 months, or 3) be either cash or cash equivalents (Plenborg & Petersen, 2011).
Page 85 of 111
current liabilities (e.g. payables)
22
. The ratio explicitly considers the liquidity and size by
measuring the level of liquid assets in relation to the size of the company, i.e. the total amount of
resources. Normally, a firm encountering ongoing operational losses will have shrinking current
assets in relation to total assets, thus the measure will over time indicate if the firm is seeing a
cash outflow from the business or not.

X
2
, Retained Earnings / Total Assets (RE/TA):
The RE/TA measure indicates the amount reinvested earnings and/or losses, which reflects the
degree of corporate leverage. In other words, to which extent assets have been financed by
company net earnings. Those firms with low retained earnings relative to total assets have been
financing capital expenditures and resources through debt rather than through retained earnings.
Thus, firms utilizing less debt will have high retained earnings relative to total assets due to
retention of net earnings. This measure also highlights either the use of internal generated funds
for growth versus externally raised funds for growth. The more a company retains, the greater
the ability to finance capital expenditures internal generated resources. Companies taking a big
bath, i.e. making large write-offs, will reduce their retained earnings and thus reduce the total Z-
score value.
The age of a firm is implicitly considered in this measure. Relative younger firms will
probably show a weaker RE/TA ratio due to lower accumulated retained earnings. Therefore,
younger firms may be discriminated by this measure and will be classified as potential bankrupt
more often compared to older companies, everything equal. However, this is the actual situation
of younger firms (Altman, 2000).

X
3
, Earnings Before Interest and Taxes / Total Assets (EBIT/TA):
The ratios EBIT/TA is a difference version of return on assets (ROA), measuring the firm’s
operating performance and it also indicate the earning capacity of the firm. In addition, the
measure is an effective way off assessing the productivity of the firm’s assets, independent of
any tax or leverage factors. The ratio is particular appropriate for measuring return on assets
without the affect of firm borrowings, cash, or the tax regime it operates under.

22
Current liabilities are characterized by 1) being part of the normal operating cycle, 2) is to be settled within 12
months, 3) purpose of being traded, and 4) cannot be deferred for at least 12 months after the reporting date
(Plenborg & Petersen, 2011).
Page 86 of 111

X
4
, Market Value of Equity / Total Liabilities (MVE/TL):
The MVE/TL measure is the measure of the long term solvency of the firm, which is the
reciprocal of the debt-to-equity ratio. Equity is measures by the combined market value of all
outstanding shares. This ratio shows how much the assets of a firm can decline in value
(measured by market value of equity plus debt) can decline before the liabilities exceed the
assets and the firm would become insolvent. This measure adds a market value dimension to the
model that not is based on pure accounting-based measures. Clearly, the measure incorporated
the market’s confidence in the company’s position.
The measure attempts to alleviate the time lag between competiveness and company profits.
A related point on the time lags can be illustrated by the fact that a firm can be successful in that
its market capitalisation is rising rapidly, while the firm is making losses at the same time. This
situation can arise when investors and the market believe the company to be viable in the future
and expect positive performance in the longer term. For example, firms operating in the
biotechnology industry, where a considerable amount of resources and time are needed to
develop profitable products, often enjoy confidence from shareholders demonstrated by rising
market capitalisation despite experiencing heavy losses. Thus, summed up the MVE/TL
measure assists to eliminate firms being deemed viable by the financial markets.

X
5
, Sales / Total Assets (S/TA):
The S/TA ratio, also known as the capital-turnover ratio, is a standard financial ratio that
measures the sale generating capacity of the firm’s assets, i.e. how effectively assets in the
operation are used by the firm. All things equal, it is attractive to have a high turnover rate on
invested capital, but unfortunately the turnover rate varies significantly from industry to another.
In addition, the measure is a measure of the firm’s capacity to deal with competitive conditions,
thus a further attempt to reduce the lags between competitiveness and firm profits.

Non-manufacturing
Since the first formulation of the original Z-score model, the original model was successful
modifies and adapted to apply for non-manufacturing firms (Altman, 2000), which is the
original model without the X
5
(sales / total assets) in order to minimize potential industry
Page 87 of 111
effects. In addition, the book value of equity was used for X
4
instead of the market value of
equity. The modified Z-score model is as follows:

Z = 6.56X
1
+ 3.26X
2
+ 6.72X
3
+ 1.05X
4

where,
Z
4
= book value of equity / total liabilities
X
1
= working capital / total assets
X
2
= retained earnings / total assets
X
3
= EBIT / total assets
X
4
= book value of equity / total liabilities, and
Z = overall score.
(12)

All the coefficients for the variables X
1
to X
4
are changed as well as the cut-off values. The zone
of ignores and threshold values for both Z-score models are summarised below:
Table 13: Illustration of the threshold levels for the Z-score models
Zones of discriminations
Modified Z-score model for
manufacturing firms
Original Z-score model for non-
manufacturing firms
Safe zone: Low probability for
bankruptcy
Z > 2.99 Z > 2.6
Grey zone: zone of ignorance;
uncertain future
1.81 < Z < 2.99 1.1 < Z < 2.6
Red zone: High probability of
bankruptcy
Z < 1.81 Z < 1.1

Altman (2000) revised the original Z-score model using updated data (for the periods 1969-
1975, 1976-1995, and 1997-1999) on more companies due to potential biases in the original
sample. The updated model dismissed the appearance of any significant biases. However, the
updated model suggested lower cut-off values for both boundaries in the manufacturing model;
respectively 1.23 for the lower boundary and 2.67 for the safe-zone. However, the revised model
appeared slightly less reliable why the original threshold values are followed. The difference
between the two models is deemed not to exercise any significant influence when being used for
sample selection.
The Z-score include a firm’s slack resources as a factor, which have been identified as an
important feature in turnarounds (e.g. Barker & Duhaime, 1997; Barker & Barr, 2002; Abebe et
al., 2010). The level of available firm resources in the turnaround situation reflects the ability to
Page 88 of 111
initiate the necessary turnaround elements (Abebe et al., 2010). Similar, substantial slack
resources may provide the firm with the necessary flexibility when formulating the overall
turnaround strategy, meaning the available resources provides the given firm with more options
to choose from than firms with less available resources. Firms with less slack resources may as a
consequence be more constrained in their response to the turnaround situation (Barker &
Duhaime, 1997). According to Abebe et al. (2010), most researchers measure firm slack
resources by the debt-to-equity ratio, which reveals the financial leverage. The ratio reflects the
potential access and ability to raise necessary (bridge) capital, which may be a perquisite to
initiate the relevant turnaround response (Barker & Barr, 2002). High financial leverage is
associated with higher financial long-term risk. In calculating the Z-score, the debt-to-equity is
represented by the reciprocal value, i.e. equity-to-debt value
23
. Thus, the slack resources, which
have been widely identified as an important factor, are, therefore, indirectly considered when
assessing the financial health and severity of turnaround situation of the respective firm.

Appendix 3
Proxy for the risk-free rate for each respective country as computed based on daily yield data
extracted from Datastream available through Thomson Research. The annual rate is the average
rate of return for each individual government bond for each yearly period. The individual bond
indices are available in Datastream with the track codes: Denmark GVDK05(CM01), Sweden
GVSD05(CM01), Norway GVNK05(CM01), Germany GVBD03(CM01), United Kingdom
GVUK05(CM01), Austria GCOE05(CM01), Belgium GVBG05(CM01), Finland
GVFN05(CM01), France GVFR05(CM01), Ireland GVIR05(CM01), Portugal
GVPT05(CM01), Spain GVES05(CM01), Switzerland SW05(CM01), Holland
GVNL05(CM01). Rates are in percentage.
Table 14: Annual risk-free rates for each country
Denmark Sweden Norway Germany United Kingdom
DNK01Y SWE01Y NOR01Y GER01Y UK01Y
Year Rate Year Rate Year Rate Year Rate Year Rate
1995 6.24 1995 9.40 1995 5.64 1995 4.70 1995 6.88

23
The leverage ratio is calculated using market value of equity for manufacturing firms, while being calculated
using book value of equity for non-manufacturing firms. Plenborg and Petersen (2011) stress that leverage
measures based on book value opposite market value may provide very different results, leading to incorrect
conclusions about the leverage depending on the value used. It would be preferred to be consistent and use the same
type of value for the two groups of firms. However, I follow the models suggested by Altman (2008) and do not
modify the variables in his models.
Page 89 of 111
1996 4.12 1996 5.97 1996 4.91 1996 3.37 1996 6.13
1997 4.01 1997 4.63 1997 3.96 1997 3.51 1997 6.77
1998 4.22 1998 4.41 1998 5.44 1998 3.62 1998 6.47
1999 3.60 1999 3.74 1999 5.77 1999 3.13 1999 5.38
2000 5.32 2000 4.66 2000 6.87 2000 4.71 2000 6.04
2001 4.50 2001 4.21 2001 6.92 2001 3.94 2001 4.82
2002 3.76 2002 4.42 2002 6.71 2002 3.45 2002 4.19
2003 2.38 2003 3.10 2003 3.81 2003 2.25 2003 3.58
2004 2.37 2004 2.30 2004 2.00 2004 2.23 2004 4.50
2005 2.34 2005 2.09 2005 2.38 2005 2.27 2005 4.35
2006 3.34 2006 2.83 2006 4.10 2006 3.35 2006 4.73
2007 4.25 2007 3.87 2007 4.59 2007 4.10 2007 5.31
2008 4.08 2008 3.79 2008 4.88 2008 3.51 2008 3.94
2009 1.80 2009 0.47 2009 2.18 2009 0.96 2009 0.78
2010 0.86 2010 0.83 2010 2.36 2010 0.49 2010 0.62

Austria Belgium Finland France Ireland
AUS01Y BEL01Y FIN01Y FRA01Y IRE01Y
Year Rate Year Rate Year Rate Year Rate Year Rate
1995 5.14 1995 5.19 1995 1995 6.29 1995 7.05
1996 3.69 1996 3.29 1996 4.71 1996 3.92 1996 5.74
1997 3.79 1997 3.48 1997 3.72 1997 3.50 1997 5.67
1998 3.87 1998 3.66 1998 3.75 1998 3.61 1998 4.50
1999 3.20 1999 3.09 1999 3.18 1999 3.20 1999 3.25
2000 4.91 2000 4.79 2000 4.76 2000 4.69 2000 4.80
2001 4.16 2001 4.13 2001 4.36 2001 3.98 2001 4.04
2002 3.60 2002 3.48 2002 3.47 2002 3.44 2002 3.19
2003 2.30 2003 2.39 2003 2.49 2003 2.25 2003 1.78
2004 2.41 2004 2.20 2004 2.46 2004 2.23 2004 2.26
2005 2.40 2005 2.27 2005 2.33 2005 2.25 2005 2.44
2006 3.38 2006 3.36 2006 3.33 2006 3.34 2006 3.29
2007 4.10 2007 4.14 2007 4.12 2007 4.13 2007 3.87
2008 3.75 2008 3.65 2008 3.62 2008 3.68 2008 3.91
2009 1.11 2009 1.12 2009 1.01 2009 0.99 2009 1.63
2010 0.12 2010 1.09 2010 0.75 2010 0.59 2010 2.37

Italy Portugal Spain Switzerland Holland
IT01Y POT01Y SP01Y SW01Y NET01Y
Year Rate Year Rate Year Rate Year Rate Year Rate
1995 11.31 1995 10.38 1995 10.15 1995 4.28 1995 4.66
1996 9.00 1996 7.30 1996 7.27 1996 2.11 1996 3.19
1997 6.70 1997 5.32 1997 5.21 1997 1.54 1997 3.64
1998 4.45 1998 3.84 1998 3.93 1998 1.54 1998 3.73
1999 3.31 1999 2.84 1999 3.15 1999 1.73 1999 3.20
2000 4.91 2000 4.69 2000 4.74 2000 3.36 2000 4.78
2001 4.14 2001 4.20 2001 3.98 2001 2.78 2001 4.08
2002 3.50 2002 3.62 2002 3.41 2002 1.44 2002 3.44
2003 2.36 2003 2.26 2003 2.08 2003 0.43 2003 2.24
2004 2.25 2004 2.11 2004 2.24 2004 0.91 2004 2.28
2005 2.29 2005 2.24 2005 2.23 2005 0.94 2005 2.33
2006 3.35 2006 3.30 2006 3.31 2006 1.84 2006 3.33
2007 4.14 2007 4.14 2007 4.14 2007 2.58 2007 4.12
2008 3.84 2008 3.89 2008 3.71 2008 1.86 2008 3.61
2009 1.21 2009 1.23 2009 0.93 2009 0.30 2009 1.03
Page 90 of 111
2010 1.43 2010 2.50 2010 0.47 2010 0.20 2010 0.68

Appendix 4
Table 15: Sample description of industry group representation in definition 2a
Division code SIC Code Industry name # number of companies
B 1000 < 1500 Mineral 2
D 2000 < 4000 Manufacturing 114
E 4000 < 5000 Transportation, Communication, Utilities 10
F 5000 < 5200 Wholesale Trade 5
G 5200 < 6000 Retail Trade 10
I 7000 < 8900 Services 58
This table provide information regarding the industries in this study for definition 2a. The four industries A) Agriculture, Forestry and Fishing
(SIC <1000), Construction (SIC 1500 < 1800), H) Finance, Insurance, and Real Estate (6000 < 6800) and J) Public administration (SIC 9100
< 10.000) are not included in the table since no companies from the respective industry groups are represented or the industry group is
restricted from the sample.

Table 16: Sample description of industry group representation in definition 2b
Division code SIC Code Industry name # number of companies
B 1000 < 1500 Mineral 1
D 2000 < 4000 Manufacturing 84
E 4000 < 5000 Transportation, Communication, Utilities 6
F 5000 < 5200 Wholesale Trade 5
G 5200 < 6000 Retail Trade 8
I 7000 < 8900 Services 48
This table provide information regarding the industries in this study for definition 2b. The four industries A) Agriculture, Forestry and Fishing
(SIC <1000), Construction (SIC 1500 < 1800), H) Finance, Insurance, and Real Estate (6000 < 6800) and J) Public administration (SIC 9100
< 10.000) are not included in the table since no companies from the respective industry groups are represented or the industry group is
restricted from the sample.

Appendix 5
This appendix serves to elaborate on the chosen performance measure, show how the Z-score
vary over time for turnaround and non-turnaround firms, verify the sampling procedure and
provide example of firms in the sample. Figure 4 presented in the thesis add to the discussion of
performance measure. Evidently, return on assets (ROA) is displaying less volatility in
performance for sample firms than return on invested capital (ROIC). The difference in
volatility is due to total invested capital is smaller than total assets for every firm in the sample.
As a consequence, ROA will always be smaller than ROI when a firm reports positive net
income, i.e. in the base period before decline and after successful turnaround. Opposite, ROA is
always larger than ROI when a firm experience negative net income, i.e. in period 2 to period 4
in the turnaround cycle period. The illustration supports using ROA as the performance measure
since it is more conservative by being less volatile to changes in financial performance. Thus,
ROA will compared to ROIC be less likely to be above the benchmark in the recovery period,
Page 91 of 111
making this measure better in discriminating between actual performance turnarounds and
insufficient performance improvements.
The performance of the firms over the base period and the turnaround cycle period for the
alternative definitions is presented in Figure 5 and Figure 6, which compares the performance
between successful turnaround and non-turnaround firms. The performance of turnaround and
non-turnaround firms is quite similar in the base and the following decline period. However, the
performance of the two groups begin to diverge in the first year of the recovery period, i.e. year
4, with the average turnaround firm recovering from the initial performance decline in both
definitions, while the average non-turnaround firm in both cases continue the overall decline
despite a small improvement in performance in year 5.
Figure 5: ROA of firms for definition 2a

Figure 6: ROA of firms for definition 2b


The presented figures shows that the additional sampling criteria are successful in classifying
firms as either turnaround or non-turnaround firms, thus being suitable as a supplement for the
additional approach.
The above performance patterns of the participating firms reflect the use of Altman’s Z-
score as the score ensures that firms not only experience a performance downturn but that the
performance decline actually posses a severe threat in terms of firm survival. Thus, the score
acts as a tool to separate the participating firms in the sample from firms in stagnation, which do
not pose the same threat to firm survival (Barker & Duhaime, 1997). A graphical presentation of
the Altman’s Z-score for the turnaround and non-turnaround firms divided by the type of firm,
manufacturing or non-manufacturing, mirrors the pattern of ROA. The Z-score for the two
definitions are presented below:
-18%
-8%
2%
12%
1 2 3 4 5 6 7 8
R
e
t
u
r
n

o
n

A
s
s
e
t
s
,

%

Year in turnaround cycle period
Non-turnaround Turnaround
-20%
-10%
0%
10%
1 2 3 4 5 6 7 8
R
e
t
u
r
n

o
n

A
s
s
e
t
s
,

%

Year in turnaround cycle period
Non-turnaround Turnaround
Page 92 of 111
Figure 7: Z-score manufacturing firms, def. 2a

Figure 8: Z-score non-manufacturing firms, def. 2a


Figure 9: Z-score manufacturing firms, def. 2b

Figure 10: Z-score non-manufacturing firms, def. 2b


Before the decline, both turnaround and non-turnaround firms, irrespectively of industry group,
had a Z-score above the threshold, which Altman describes as the “green zone” (Altman, 2000),
where the situation of bankruptcy is unlikely. However, as the performance pattern above, both
turnaround and non-turnaround firms experience declining Z-values and slide out of the “safe
zone” in the decline period. This confirms that the average firm in the final sample faced a
severe threat of firm survival during the decline period. In the first year of the recovery period,
i.e. year 4, the turnaround firms start to improve financially, which are indicated by the
increasing Z-scores and they begin to shift back towards the “safe zone”. Opposite, the non-
turnaround firms continue to pose deteriorating Z-scores, increasing their chance of financial
distress.
1.0
2.0
3.0
4.0
5.0
6.0
7.0
1 2 3 4 5 6 7 8
Z
-
s
c
o
r
e

v
a
l
u
e

Year in the turnaround cycle period
Non-Turnaround Turnaround
-5.0
-3.0
-1.0
1.0
3.0
5.0
1 2 3 4 5 6 7 8
Z
-
s
c
o
r
e

v
a
l
u
e

Year in the turnaround cycle period
Non-Turnaround Turnaround
1.0
2.0
3.0
4.0
5.0
6.0
7.0
1 2 3 4 5 6 7 8
Z
-
s
c
o
r
e

v
a
l
u
e
s

Year in the turnaround cycle period
Non-Turnaround Turnaround
-5.0
-3.0
-1.0
1.0
3.0
1 2 3 4 5 6 7 8
Z
-
s
c
o
r
e

v
a
l
u
e

Year in the turnaround cycle period
Non-Turnaround Turnaround
Page 93 of 111
As I have screened and examined each firm individually, the process has revealed strong
indications for that the sample selection process has captured the individual outcomes
successfully. For example, companies such as Anovo, Belvedere, Richard Ginori, ID Future and
Jarvis were all classified as being non-turnarounds. The classification as being non-turnaround
firms are supported by the fact that these firms all have insolvency proceeding, are in
liquidation, or are being dissolved.
The sampling window, i.e. turnaround cycle period, requires every firm to recognize the
need of turnaround actions within the decline period. For example, Realdolmen initiated
turnaround actions in 2001, which was the last year of its decline period. Even though the firm
recognized the need of a turnaround by undertaking turnaround measures such as disposal of
non-core activities, negotiation of bridge capital, replacement of several directors on the board,
then they did not achieve a turnaround within the recovery period, which may call for a
redefinition of the turnaround cycle period. However, prolonging the turnaround cycle period to
8 years (i.e. 4 years decline period followed by 4 years of potential recovery) or working with a
period between the decline and recovery period to allow for turnaround actions to be
incorporated would have resulted in a sample size to small relative to the initial general
population. Opposite, reducing the turnaround cycle period to 4 years (i.e. 2 years decline period
followed by a 2 years recovery period) would have increased the sample significantly but
increased the probability of including firms in the sample that experienced fluctuations in the
performance and did not actually undergo a severe performance decline threatening firm
survival.









Page 94 of 111
Appendix 6
Table 17: Definition 2a: Descriptive statistics presented per year
Definition 2a Turnaround (TURNa=1)
Variable Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 3-8
Turnaround performance (ROA) -0.0648
(0.1140)
-0.1764
(0.2156)
-0.1364
(0.2053)
0.0475
(0.0521)
0.0605
(0.0554)
0.0675
(0.0545)
-0.0337
(0.1633)
Ownership concentration (OC) 0.4568
(0.2642)
0.4844
(0.2243)
0.4701
(0.2410)
0.4770
(0.2513)
0.4667
(0.2443)
0.4348
(0.2375)
0.4650
(0.2421)
Dominant blockholder 0.1951
(0.4012)
0.1707
(0.3809)
0.1463
(0.3578)
0.1220
(0.3313)
0.0976
(0.3004)
0.0976
(0.3004)
0.1382
(0.3458)
Takeover (TO) 0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0244
(0.1562)
0.0041
(0.0638)
Cost retrenchment (COSTRy) 0.2131
(1.0125)
0.2465
(0.9277)
-0.1519
(0.3204)
-0.0694
(0.2930)
0.0390
(0.3545)
-0.1404
(0.8906)
0.0228
(0.7170)
Asset retrenchment (ASSETRy) 0.1868
(0.6012)
-0.2285
(0.2400)
-0.1116
(0.2164)
0.0584
(0.1736)
0.2180
(0.3805)
0.3627
(1.1281)
0.0810
(0.5951)
Firm size (SIZE) 7.1068
(1.9081)
7.1433
(1.8544)
7.0332
(1.8950)
7.0157
(1.8557)
7.1007
(1.8099)
7.1859
(1.7898)
7.0976
(2.8346)
Non-Turnaround (TURNa=0)
Variable Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 3-8
Turnaround performance (ROA) -0.0767
(0.1465)
-0.1450
(0.2684)
-0.1605
(0.2244)
-0.1713
(0.2770)
-0.1654
(0.3390)
-0.1882
(0.5250)
-0.1512
(0.3199)
Ownership concentration (OC) 0.4664
(0.2463)
0.4514
(0.2562)
0.4687
(0.2639)
0.4627
(0.2593)
0.4964
(0.2512)
0.5234
(0.2612)
0.4782
(0.2566)
Dominant blockholder 0.2793
(0.4507)
0.2703
(0.4461)
0.3153
(0.4667)
0.2793
(0.4507)
0.3063
(0.4630)
0.3423
(0.4766)
0.2988
(0.4581)
Takeover (TO) 0.0090
(0.0949)
0.0180
(0.1336)
0.0270
(0.1629)
0.0090
(0.0949)
0.0360
(0.1872)
0.0450
(0.2083)
0.0240
(0.1532)
Cost retrenchment (COSTRy) 0.4966
(3.2381)
0.0886
(0.5155)
-0.0406
(0.5180)
0.0683
(1.4420)
-0.1438
(0.4006)
0.2598
(2.7058)
0.1215
(1.8568)
Asset retrenchment (ASSETRy) 0.5224
(1.4049)
-0.0148
(0.4139)
0.0009
(0.7184)
-0.0826
(0.2856)
-0.0475
(0.3988)
0.0292
(0.5152)
0.0679
(0.7527)
Firm size (SIZE) 6.8703
(1.8606)
6.8804
(1.9093)
6.8435
(1.9131)
6.7426
(1.9580)
6.6497
(1.9647)
6.6151
(1.9582)
6.7669
(1.9233)
For the yearly statistics the number of turnaround cases is nt=44 in each year, while the number of non-turnaround cases is nnt=111 in each year. For the
total mean, the number of turnaround cases is nt=246 (equal to 44 cases in each year). The number of non-turnaround cases is nnt=666 (equal to 111 cases
in each year). The table reports means and standard deviations in parentheses for the variables each year individually in the turnaround process, i.e. year
3-8, and for the full process. In this alternative approach, the dependent variable TURNa describes the turnaround outcome and takes the value 1 if a firm
is characterized as a successful turnaround, otherwise 0. Therefore, the descriptive statistics (means and standard deviations) are grouped by being either a
turnaround and non-turnaround cases. The alternative sample derived from definition 2a is restricted to the years in the turnaround period, e.g. year 3-8. In
contrast to the main definition, a number of variables are not reported. The measure of adjusted firm performance (ROA) is only reported to illustrate the
performance difference between the two groups.







Page 95 of 111
Table 18: Definition 2b: Descriptive statistics presented per year
Definition 2b Turnaround (TURNb=1)
Variable Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 3-8
Turnaround performance (ROA) -0.0795
(0.1334)
-0.1257
(0.1439)
-0.1531
(0.2071)
-0.0821
(0.2251)
-0.0053
(0.1088)
0.0578
(0.0639)
-0.0645
(0.1718)
Ownership concentration (OC) 0.4372
(0.2484)
0.4429
(0.2430)
0.4468
(0.2347)
0.4574
(0.2385)
0.4661
(0.2363)
0.4478
(0.2452)
0.4650
(0.2401)
Dominant blockholder 0.1705
(0.3782)
0.1477
(0.3569)
0.1250
(0.3326)
0.1136
(0.3192)
0.1364
(0.3451)
0.1364
(0.3451)
0.1382
(0.3455)
Takeover (TO) 0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0114
(0.1066)
0.0227
(0.1499)
0.0000
(0.0000)
0.0057
(0.0752)
Cost retrenchment (COSTRy) 0.2039
(0.7870)
0.0562
(0.5759)
-0.0556
(0.3202)
-0.0443
(0.4011)
-0.0138
(0.3527)
-0.0285
(0.5556)
0.0196
(0.5292)
Asset retrenchment (ASSETRy) 0.2198
(0.6846)
-0.0483
(0.4328)
-0.1124
(0.2440)
-0.0476
(0.2224)
0.0540
(0.2616)
0.0935
(0.2146)
0.0265
(0.3967)
Firm size (SIZE) 7.1171
(1.7268)
7.1709
(1.6900)
7.1293
(1.6973)
7.0639
(1.6754)
7.0375
(1.6792)
7.0301
(1.8057)
7.0915
(1.7056)
Non-Turnaround (TURNb=0)
Variable Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 3-8
Turnaround performance (ROA) -0.0767
(0.1465)
-0.1450
(0.2684)
-0.1605
(0.2244)
-0.1713
(0.2770)
-0.1654
(0.3390)
-0.1882
(0.5250)
-0.1512
(0.3199)
Ownership concentration (OC) 0.4664
(0.2463)
0.4514
(0.2562)
0.4687
(0.2639)
0.4627
(0.2593)
0.4964
(0.2512)
0.5234
(0.2612)
0.4782
(0.2566)
Dominant blockholder 0.2793
(0.4507)
0.2703
(0.4461)
0.3153
(0.4667)
0.2793
(0.4507)
0.3063
(0.4630)
0.3423
(0.4766)
0.2988
(0.4581)
Takeover (TO) 0.0090
(0.0949)
0.0180
(0.1336)
0.0270
(0.1629)
0.0090
(0.0949)
0.0360
(0.1872)
0.0450
(0.2083)
0.0240
(0.1532)
Cost retrenchment (COSTRy) 0.4966
(3.2381)
0.0886
(0.5155)
-0.0406
(0.5180)
0.0683
(1.4420)
-0.1438
(0.4006)
0.2598
(2.7058)
0.1215
(1.8568)
Asset retrenchment (ASSETRy) 0.5224
(1.4049)
-0.0148
(0.4139)
0.0009
(0.7184)
-0.0826
(0.2856)
-0.0475
(0.3988)
0.0292
(0.5152)
0.0679
(0.7527)
Firm size (SIZE) 6.8703
(1.8606)
6.8804
(1.9093)
6.8435
(1.9131)
6.7426
(1.9580)
6.6497
(1.9647)
6.6151
(1.9582)
6.7669
(1.9233)
For the yearly statistics the number of turnaround cases is nt=88 in each year, while the number of non-turnaround cases is nnt=111 in each year. For the
total mean the number of turnaround cases is nt=528 (equal to 88 cases in each year). The number of non-turnaround cases is nnt=666 (equal to 111 cases
in each year). The table reports means and standard deviations in parentheses for the variables each year individually in the turnaround process, i.e. year 3-
8, and for the full process. In this alternative approach, the dependent variable TURNb describes the turnaround outcome and takes the value 1 if a firm is
characterized as a successful turnaround, otherwise 0. Therefore, the descriptive statistics (means and standard deviations) are grouped by being either a
turnaround and non-turnaround firms. In contrast to the main definition, a number of variables are not reported. The measure of firm performance
(AdjROA) is only reported to illustrate the performance difference between the two groups.

Appendix 7
In order to check for potential multicollinearity problems, I have executed various tests in SAS
EG
24
. The tests have been performed with the model specification used throughout the thesis. I
apply variance inflation factors (VIF) and condition index (CI) to test for multicollinearity.
None of the results suggest severe problems with multicollinearity. Table 19 show VIF and CI
results for the main model specifications estimated in the thesis.

24
As SAS EG does not present an option for calculating VIF and similar measures from other than the command
“PROC REG”, I have used this procedure to assess the problems of multicollinearity and assume these apply as an
indicator of any problems with multicollinearity in the other econometric methods and model specifications.
Page 96 of 111
Table 19: Variance inflated factors (VIF) tests
Variables

VIF 1 VIF 2 VIF 3 VIF 4
1. Herfindahl ownership index 2.953 2.900 - -
2. Ownership concentration - - 1.552 1.550
3. Dominant shareholder 2.953 2.949 1.639 1.564
4. Takeover 1.047 1.046 1.048 1.046
5. Block investment 1.079 - 1.061 -
6. Cost retrenchment 1.042 1.042 1.041 1.040
7. Asset retrenchment 1.033 1.032 1.031 1.031
8. Firm size 1.018 1.018 1.034 1.034
Condition index (CI) 3.267 3.168 2.134 2.056
The condition index is calculated as the square root of the maximum eigenvalue divided by the minimum eigenvalue. The eigenvalues
are not discussed in this thesis, but applied to test for multicollinearity.

The VIF 4 corresponds to the model specification selected as the main model specification,
which also are used in the alternative tests, i.e. binary response models. The results are assumed
to be consistent despite being estimated by different methods.
The larger the value of VIF, the greater is the chance that the variable could induce
multicollinearity problems. Gujarati and Porter (2009) explain that multicollinearity is a
problem if the VIF of a variable exceeds 10, while CI value above 10 indicate moderate to
strong multicollinearity, while a CI above 30 reflect severe problems with multicollinearity. As
presented in Table 19 for VIF 1 and VIF 2, none of the VIF of the variables exceeds 3, while the
VIF values for the variables in VIF 3 and VIF 4 all are below 1.6. Therefore, none of these
variables seems to contribute to multicollinearity. All CI values are well below 10, which
suggest no issues with multicollinearity. Summarized and conclusively, both the VIF and CI
suggest that none of the model specifications suffer from moderate, strong nor severe
multicollinearity, why this is not deemed to present any important influence throughout the
thesis.

Appendix 8
Table 20: Means, standard deviations, and correlations for variables used in definition 2a
Variables

Mean S.D. 1 2 3 4 5 6 7
1. Turnaround 0.2697 0.4441 1
2. Ownership concentration 0.4746 0.2527 -.02 1
3. Dominant shareholder 0.2555 0.4364 -.16*** .61*** 1
4. Takeover 0.0186 0.1353 -.07** .17*** .22*** 1
5. Cost retrenchment 0.0949 1.6300 -.03 .00 .01 -.04 1
6. Asset retrenchment 0.0715 0.7131 .01 .02 -.02 -.02 .23*** 1
7. Firm size 6.8561 1.9045 .08** -.21*** -.16*** -.06* -.09*** -.05 1
N=912 (152 cases multiplied by six years of interest). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. This table reports
descriptive statistics for the variables used in my estimations related to the alternative turnaround definition 2a and values are pooled for this purpose.
All years in the turnaround cycle period are included, i.e. year 3 to 8.
Page 97 of 111
Table 21: Means, standard deviations and correlation for variables used in definition 2b
Variables

Mean S.D. 1 2 3 4 5 6 7
1. Turnaround 0.4422 0.4969 1
2. Ownership concentration 0.4656 0.2497 -.06* 1
3. Dominant shareholder 0.2278 0.4196 -.19*** .59*** 1
4. Takeover 0.0159 0.1252 -.07** .16*** .22*** 1
5. Cost retrenchment 0.0765 1.4311 -.04 -.01 .00 -.03 1
6. Asset retrenchment 0.0496 0.6211 -.03 .00 -.04 -.01 .23*** 1
7. Firm size 6.9104 1.8366 .09*** -.21*** -.12*** -.05 -.09*** -.05 1
N=1194 (199 cases multiplied by six years of interest). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. This table reports
descriptive statistics for the sample used in my estimations related to the alternative turnaround definition 2b and values are pooled for this purpose.
All years in the turnaround cycle period are included, i.e. year 3 to 8.

Appendix 9
Table 22: Fixed effects panel estimation results without time dummies
Herfindahl ownership index Ownership concentration
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Herfindahl ownership -1.00E-5
(6.72E-6)
-4.06E-6
(8.97E-6)
-4.87E-6
(8.89E-6)
-4.78E-6
(8.90E-6)
- - - -
Ownership concen-
tration ratio
- - - - -0.0377
(0.0528)
0.0021
(0.0568)
-0.0024
(0.0569)
-0.0078
(0.0578)
Dominant blockholder - -0.0520
(0.0417)
-0.0593
(0.0421)
-0.0582
(0.0422)
- -0.0649*
(0.0340)
-0.0735**
(0.0347)
-0.0709**
(0.0351)
Takeover - - 0.0644
(0.0496)
0.0640
(0.0496)
- - 0.0626
(0.0496)
0.0625
(0.0496)
Block investment - - - 0.0068
(0.0139)
- - - 0.0073
(0.0142)
Cost retrenchment 0.0001
(0.0038)
0.0002
(0.0038)
0.0004
(0.0038)
0.0005
(0.0038)
0.0001
(0.0038)
0.0003
(0.0038)
0.0004
(0.0038)
0.0005
(0.0038)
Asset retrenchment 0.0932***
(0.0094)
0.0926***
(0.0095)
0.0924***
(0.0095)
0.0924***
(0.0095)
0.0928***
(0.0095)
0.0922***
(0.0094)
0.0921***
(0.0094)
0.0921***
(0.0095)
Firm size -0.0293*
(0.0161)
-0.0288*
(0.0161)
-0.0286*
(0.0161)
-0.0285*
(0.0161)
-0.0280*
(0.0161)
-0.0285*
(0.0160)
-0.0281*
(0.0160)
-0.0281*
(0.0160)
Industry dummies Yes Yes Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes Yes Yes
Time dummies No No No No No No No No
F-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
R
2
0.3454 0.3461 0.3469 0.3470 0.3444 0.3460 0.3468 0.3469
#cross-section/time
length
289 / 15
289 / 15
289 / 15 289 / 15
289 / 15 289 / 15 289 / 12 289 / 12
This table shows fixed effects (FE) estimation results when the Herfindahl ownership index and ownership concentration ratio is the ownership concentration
variable. Time dummies are not included in these estimations, i.e. the estimations are a fixed effect one-way. Standard errors are presented below the parameter
estimates in parentheses and are corrected for potential heteroscedasticity. The sample is restricted to the years in the turnaround process, i.e. year 3-8. The time
and individual intercepts are not shown to save space. F-tests for no fixed effects are all rejected. Stars indicate statistically significance at the respective levels:
* p<0.10; ** p<0.05; *** p<0.01.






Page 98 of 111
Table 23: Pooled regression estimation results
Herfindahl ownership index Ownership concentration
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Herfindahl ownership 9.58E-7
(2.49E-6)
5.39E-6
(3.92E-6)
5.39E-6
(3.92E-6)
5.56E-6
(3.93E-6)
- - - -
Ownership concen-
tration ratio
- - - - 0.0375
(0.0291)
0.0621*
(0.0337)
0.0623*
(0.0338)
0.0621*
(0.0339)
Dominant blockholder - -0.0273
(0.0232)
-0.0269
(0.0237)
-0.0266
(0.0240)
- -0.0265
(0.0168)
-0.0260
(0.0171)
-0.0257
(0.0190)
Takeover - - -0.0066
(0.0244)
-0.0071
(0.0248)
- - -0.0094
(0.0243)
-0.0095
(0.0247)
Block investment - - - 0.0051
(0.0150)
- - - 0.0011
(0.0151)
Cost retrenchment -0.0033
(0.0038)
-0.0033
(0.0037)
-0.0033
(0.0037)
-0.0032
(0.0037)
-0.0033
(0.0038)
-0.0034
(0.0038)
-0.0034
(0.0038)
-0.0034
(0.0038)
Asset retrenchment 0.0844***
(0.0146)
0.0840***
(0.0145)
0.0841***
(0.0145)
0.0840***
(0.0145)
0.0839***
(0.0146)
0.0838***
(0.0145)
0.0838***
(0.0145)
0.0838***
(0.0145)
Firm size 0.0242***
(0.0039)
0.0240***
(0.0038)
0.0240***
(0.0038)
0.0241***
(0.0038)
0.0251***
(0.0039)
0.0251***
(0.0039)
0.0251***
(0.0039)
0.0251***
(0.0039)
Industry dummies Yes Yes Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes Yes Yes
Time dummies No No No No No No No No
F-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
R
2
adjusted 0.1108 0.1115 0.1115 0.1115 0.1109 0.1132 0.1132 0.1132
# observations 1734 1734 1734 1734 1734 1734 1734 1734
This table reports pooled OLS estimation results, thus ignoring any cross-section and time effects, changing between the two ownership concentration
variables. Standard errors are presented below the coefficients in parentheses and are adjusted for heteroscedasticity. To ensure valid estimation results, the
error term is assumed to be homoscedastic and uncorrelated within and between firms in order to produce consistent and efficient estimates (BLUE). The
sample is restricted to the years in the turnaround process, i.e. year 3-8. F-tests for global null coefficient are all rejected. Stars indicate statistically significance
at the respective levels: * p<0.10; ** p<0.05; *** p<0.01.

As Hausman test only is easily available through the random effect procedure in SAS EG Proc
Panel RANONE and RANTWO (SAS Institute, 2010), Table 24 below shows the estimation
results from the random effect panel method, which also produce the Hausman tests. The
probability of observing the given Hausman statistics is all below 0.0001. Thus, the underlying
hypothesis is rejected in all cases, suggesting the fixed effect estimation is most appropriate.









Page 99 of 111
Table 24: Random effects estimation results
Herfindahl ownership index Ownership concentration
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Herfindahl ownership
index
-4.98E-0
(2.49E-6)
-3.80E-6
(5.27E-6)
-3.89E-6
(5.30E-6)
-8.84E-7
(5.30E-6)
- - - -
Ownership concen-
tration ratio
- - - - -0.0744***
(0.0239)
-0.0693**
(0.0288)
-0.0691**
(0.0290)
-0.0614**
(0.0295)
Dominant
blockholder
- -0.0286
(0.0232)
-0.0286
(0.0260)
-0.0306
(0.0260)
- -0.0057
(0.0187)
-0.0063
(0.0190)
-0.0118
(0.0194)
Takeover - - 0.0032
(0.0478)
0.0055
(0.0478)
- - 0.0049
(0.0479)
0.0065
(0.0479)
Block investment - - - 0.0849*
(0.0130)
- - - -0.0178
(0.0132)
Cost retrenchment -0.0055
(0.0036)
-0.0055
(0.0036)
-0.0054
(0.0036)
-0.0055
(0.0036)
-0.0056
(0.0037)
-0.0054
(0.0036)
-0.0053
(0.0036)
-0.0054
(0.0036)
Asset retrenchment 0.0848***
(0.0094)
0.0845***
(0.0094)
0.0846***
(0.0094)
0.0849***
(0.0094)
0.0849***
(0.0094)
0.0853***
(0.0094)
0.0854***
(0.0094)
0.0854***
(0.0094)
Firm size -0.0060***
(0.0019)
-0.0062***
(0.0019)
-0.0062***
(0.0019)
-0.0050**
(0.0020)
-0.0039**
(0.0020)
0.0040**
(0.0020)
-0.0041*
(0.0021)
-0.0033
(0.0021)
Industry dummies Yes Yes Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes Yes Yes
Time dummies Yes Yes Yes Yes Yes Yes Yes Yes
Hausman Test <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
R
2
adjusted 0.1633 0.1640 0.1626 0.1642 0.1738 0.1694 0.1681 0.1688
# cross-section/time
length
289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15
This table shows random effects (RE) estimation results when the Herfindahl ownership index and ownership concentration ratio is the ownership concentration
variable. Standard errors are presented below the parameter estimates in parentheses and are corrected for heteroscedasticity. The sample is restricted to the years in
the turnaround process, i.e. year 3-8. Hausman test (only available through the random effect estimation in SAS EG, Proc Panel) statistics suggest fixed effects are
present in all models. Stars indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; *** p<0.01.













Page 100 of 111
Appendix 10
Table 25: Random effect estimation results with ROIC as dependent variable
Herfindahl ownership index Ownership concentration
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Herfindahl ownership -2.00E-5
(2.00E-5)
-1.60E-6
(2.8E-5)
-1.60E-6
(2.8E-5)
2.76E-6
(2.28E-5)
- - - -
Ownership concen-
tration ratio
- - - - -0.1245
(0.1408)
-0.0020
(0.1679)
-0.0019
(0.1680)
-0.0089
(0.1688)
Dominant blockholder - -0.1386
(0.1343)
-0.1385
(0.1352)
-0.1363
(0.1353)
- -0.1318
(0.0985)
-0.1317
(0.0995)
-0.1230
(0.1015)
Takeover - - -0.0020
(0.2711)
-0.0057
(0.2713)
- - -0.0019
(0.2712)
-0.0052
(0.2713)
Block investment - - - 0.0333
(0.0756)
- - - 0.0330
(0.0756)
Cost retrenchment -0.0055
(0.0208)
-0.0054
(0.0207)
-0.0054
(0.0207)
-0.0051
(0.0207)
-0.0048
(0.0207)
-0.0055
(0.0207)
-0.0055
(0.0207)
-0.0052
(0.0207)
Asset retrenchment 0.0855
(0.0555)
0.1058**
(0.0535)
0.1058**
(0.0535)
0.1055**
(0.0536)
0.1065**
(0.0535)
0.1060**
(0.0535)
0.1060**
(0.0535)
0.1059**
(0.0535)
Firm size 0.0245
(0.0163)
0.0477**
(0.0188)
0.0477**
(0.0189)
0.0479**
(0.0189)
0.0476**
(0.0191)
0.0477**
(0.0191)
0.0477**
(0.0191)
0.0477**
(0.0191)
Industry dummies Yes Yes Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes Yes Yes
Time dummies Yes Yes Yes Yes Yes Yes Yes Yes
Hausman Test statistic 1.69 4.62 4.66 4.54 1.96 2.56 2.58 2.50
R
2
adjusted 0.0057 0.0108 0.0108 0.0110 0.0098 0.0108 0.0108 0.0109
#cross-section/time
length
289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 12 289 / 12
This table shows random effects (RE) estimation results when the Herfindahl ownership index and ownership concentration ratio is the ownership
concentration variable. The dependent variable is return on invested capital. Standard errors are presented below the parameter estimates in parentheses and are
corrected for potential heteroscedasticity. The sample is restricted to the years in the turnaround process, i.e. year 3-8. I fail to reject the F-tests for no fixed
effects, why RE are shown. Hausman confirm that RE are appropriate. Stars indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; ***
p<0.01.













Page 101 of 111
Appendix 11
Table 26: Results of dynamic panel regression with GMM and FE estimation without time
dummies
GMM Fixed effects
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Lagged turnaround perform. 0.1264***
(0.0255)
0.1019***
(0.0299)
0.1123***
(0.0299)
0.0591**
(0.0295)
0.0615**
(0.0294)
0.0621**
(0.0294)
Ownership concentration ratio -0.0540
(0.1214)
-0.0937
(0.1189)
-0.0407
(0.1220)
-0.0365
(0.0528)
0.0052
(0.0567)
0.0006
(0.0568)
Dominant blockholder - 0.1352*
(0.0798)
0.1765*
(0.1021)
- -0.0678**
(0.0340)
-0.0767**
(0.0347)
Takeover - - -0.1217
(0.1255)
- - 0.0642**
(0.0495)
Cost retrenchment -0.0081
(0.0106)
-0.0026
(0.0115)
-0.0074
(0.0119)
-0.0008
(0.0038)
-0.0007
(0.0038)
-0.0005
(0.0038)
Asset retrenchment 0.1994***
(0.0311)
0.1940***
(0.0324)
0.1972***
(0.0397)
0.0919***
(0.0095)
0.0913***
(0.0094)
0.0911***
(0.0094)
Firm size -0.2257***
(0.0566)
-0.2833***
(0.0615)
-0.2580***
(0.0682)
-0.0327**
(0.0162)
-0.0333**
(0.0162)
-0.0330**
(0.0162)
Industry dummies Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes
Time dummies No No No No No No
Sargan Test (Chi
2
-statistics) 37.00** 38.33** 38.15**
R
2
adjusted 0.3462 0.3480 0.3488
# cross-section/time length 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15
This table reports GMM and fixed effects (FE) estimation results with ownership concentration ratio as the ownership concentration variable and the
lagged dependent variable without considering time effects. Stars indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; ***
p<0.01. Standard errors are presented below the parameter coefficients in parentheses and are heteroscedasticity corrected. The sample is restricted to
the years in the turnaround process, i.e. year 3-8. The FE F-statistics for no fixed time effects are all rejected. The Sargan statistics related to the
GMM twostep estimations all fail to verify over-identifying of restrictions for the GMM estimations. First and second order autocorrelation tests fail
to report statistics, suggesting autocorrelation in the first and/or second order regression residuals.














Page 102 of 111
Table 27: Results of dynamic OLS regression
OLS with time dummies OLS with no time dummies
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Lagged turnaround performance 0.2830***
(0.0275)
0.2834***
(0.0275)
0.2833***
(0.0276)
0.3232***
(0.0267)
0.3234***
(0.0268)
0.3244***
(0.0268)
Ownership concentration ratio 0.0225
(0.0239)
0.0429
(0.0286)
0.0430
(0.0286)
-0.0546**
(0.0217)
-0.0439*
(0.0259)
-0.0440*
(0.0259)
Dominant blockholder - -0.0217
(0.0167)
-0.0215
(0.0169)
- -0.0127
(0.0168)
-0.0130
(0.0170)
Takeover - - -0.0034
(0.0459)
- - 0.0062
(0.0467)
Cost retrenchment -0.0053
(0.0035)
-0.0054
(0.0035)
-0.0054
(0.0035)
-0.0091**
(0.0035)
-0.0093***
(0.0036)
-0.0092***
(0.0036)
Asset retrenchment 0.0783***
(0.0094)
0.0781***
(0.0094)
0.0781***
(0.0094)
0.0776***
(0.0092)
0.0775***
(0.0092)
0.0775***
(0.0092)
Firm size -0.0210***
(0.0033)
-0.0209***
(0.0033)
-0.0209***
(0.0033)
-0.0031*
(0.0018)
-0.0033*
(0.0018)
-0.0033*
(0.0018)
Industry dummies Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes
Time dummies Yes Yes Yes No No No
R
2
-value 0.2955 0.2962 0.2962 0.2538 0.2540 0.2540
# cross-section/time length 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15 289 / 15
This table reports pooled OLS estimation results for dynamic models with ownership concentration ratio as the ownership concentration variable and
the lagged dependent variable. Results with both time effects and no time effects are reported. Stars indicate statistically significance at the respective
levels: * p<0.10; ** p<0.05; *** p<0.01. Standard errors are presented below the parameter coefficients in parentheses and are corrected for
heteroscedasticity. The sample is restricted to the years in the turnaround process, i.e. year 3-8. F-tests for global null coefficient are all rejected. Stars
indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; *** p<0.01.

Table 28: Dynamic panel GMM regression with ownership held exogenous
GMM
Variables Model 5 Model 6 Model 7
Lagged turnaround performance 0.0767***
(0.0130)
0.0848***
(0.0123)
0.0812***
(0.0140)
Ownership concentration ratio -0.0304
(0.0262)
0.0493
(0.0328)
0.0438
(0.0365)
Dominant blockholder - -0.1438***
(0.0340)
-0.1684***
(0.0381)
Takeover - - 0.1011**
(0.0469)
Cost retrenchment 0.0087**
(0.0040)
0.0030
(0.0047)
0.0034
(0.0050)
Asset retrenchment 0.1330***
(0.0117)
0.1344***
(0.0123)
0.1301***
(0.0118)
Firm size -0.0362*
(0.0219)
-0.0603**
(0.0244)
-0.0721***
(0.0246)
Industry dummies Yes Yes Yes
Country dummies Yes Yes Yes
Time dummies Yes Yes Yes
Sargan Test (Chi
2
-statistics) 101.70 96.02 92.75
# cross-section/time length 289 / 15 289 / 15 289 / 15
This table reports GMM and fixed effects (FE) estimation results with ownership concentration ratio as the ownership concentration variable
and the lagged dependent variable without considering time effects. Stars indicate statistically significance at the respective levels: * p<0.10;
** p<0.05; *** p<0.01. Standard errors are presented below the parameter coefficients in parentheses and are heteroscedasticity corrected.
Sample is restricted to the years in the turnaround process, i.e. year 3-8. The FE F-statistics for no fixed time effects are all rejected. The
Sargan statistics related to the GMM twostep estimations verify over-identifying of restrictions for the GMM estimations. First and second
order autocorrelation tests fail to report statistics, suggesting autocorrelation in the first and/or second order regression residuals.
Page 103 of 111
Appendix 12
This table is reported to create a basis of comparison for the GMM estimation results in Table
10 and Table 26. The methodology of two-stage least squares (2SLS) is performed to evaluate
the magnitude of the GMM estimations. The underlying methodology and theory of 2SLS is
deemed beyond the scope of this thesis. The 2SLS procedure is performed as suggested by SAS
Institute (2010) and Baltagi (2005), where the idea is to place instruments for the endogenous
explanatory variables. The model makes the assumption that only the lagged dependent variable
is endogenous among the explanatory variables. Hence, the estimation results are similar to the
GMM estimation results. However, the estimates are likely to be biased as ownership also may
be endogenous. As addressed in the thesis, there is a lack of appropriate instruments and
modelling the instruments differently does not produce any significant results. When
introducing ownership as an endogenous variable in the models produce significantly different
results, which reduce the magnitude of the lagged turnaround performance estimates
considerably. The analysis and method of 2SLS are not considered any further.
Table 29: Two-stage least squares estimation results of dynamic models
2SLS
Variables Model 5 Model 6 Model 7
Lagged turnaround performance 0.2830***
(0.0275)
0.2834***
(0.0275)
0.2833***
(0.0276)
Ownership concentration ratio 0.0225
(0.0239)
0.0429
(0.0286)
0.0430
(0.0286)
Dominant blockholder - -0.0217
(0.0167)
-0.0215
(0.0169)
Takeover - - -0.0034
(0.0459)
Cost retrenchment -0.0053
(0.0035)
-0.0054
(0.0035)
-0.0054
(0.0035)
Asset retrenchment 0.0783***
(0.0094)
0.0781***
(0.0094)
0.0781***
(0.0094)
Firm size -0.0210***
(0.0033)
-0.0209***
(0.0033)
-0.0209***
(0.0033)
Industry dummies Yes Yes Yes
Country dummies Yes Yes Yes
Time dummies Yes Yes Yes
R
2
-value (adjusted) 0.1622 0.1625 0.1620
# cross-section/time length 289 / 15 289 / 15 289 / 15
This table reports two-stage least square (2SLS) estimation results for model 5-7 where ownership concentration ratio is the ownership
concentration explanatory variable and the lagged independent variable is included. Firm dummies are not included as Proc Syslin in
SAS EG does not yield any results in this case. The F-statistics testing the null hypothesis that all parameter estimates are zero are all
rejected. Stars indicate statistically significance at the respective levels: * p<0.10; ** p<0.05; *** p<0.01. Only the lagged dependent
variable is assumed endogenous.

Page 104 of 111
Appendix 13
The odds ratios are obtained by exponentiating the estimated parameters. For example, the
estimated coefficient for firm size in Model 7 (Definition 2b) is 0.4467, which equal an odds
ratio of 1.7141. The odds ratio indicates that for a 1 pct-point increase in firm size, the predicted
odds ratio increases by 71.4 pct. In other words, holding the other variables constant, a 1 pct.-
increase in firm size will increase the odds of successfully turnaround outcome by 71.4 pct.
Similar, for a 1 pct.-point increase in the asset base (Model 5, Definition 2a), the odds of
turnaround increases by a factor of exp(0.0895)=1.0936. The opposite is the case for odds ratios
below one. For example, the odds of turnaround is exp(-3.7338)=0.0804 times smaller for firms
dominated by a single large blockholder than non-dominated firms.
Table 30: Odds ratios from logit analysis of turnaround outcome
Odds ratios Definition 2a Definition 2b
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Ownership concentration 0.0090 0.2509 0.2048 3.3424 118.81 4.4141
Dominant blockholder - 0.0036 0.0035 - 0.0110 0.0804
Takeover - - 0.4773 - - 0.7579
Cost retrenchment 0.6206 0.5304 0.4884 0.8586 0.5798 1.0954
Asset retrenchment 1.6013 2.3373 2.3575 1.0936 1.4599 0.7711
Firm size 2.8665 2.1900 2.3099 1.0906 2.4692 1.7141
This table shows the odds ratios for the fixed effect logit estimation results in Table 11. The odds ratios are directly derived from the fixed effect
logit coefficient by exponentiation of the estimated coefficient.

Appendix 14
Table 31: Estimation results from logit models of turnaround outcome with dummies
Non-linear panel models Definition 2a Definition 2b
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Ownership concentration -1.6031
(0.0000)
-0.08970
(2.1567)
-1.7758
(2.7406)
-1.2311
(2.3281)
1.4581
(2.6895)
1.4848
(2.3533)
Dominant blockholder - -4.2974**
(1.7369)
-5.5550
(3.5603)
- -2.2889
(3.1722)
-2.5203
(1.9066)
Takeover - - -1.0666
(4.6536)
- - -0.2772
(2.5483)
Cost retrenchment -0.8561**
(0.4071)
-0.7327*
(0.4386)
-0.0075
(0.0000)
-0.0611
(0.3417)
-0.2215
(0.0000)
-0.2600
(0.0000)
Asset retrenchment 0.0008
(0.3950)
-0.2462
(0.4434)
0.1672
(0.0000)
-0.0779
(0.5325)
-0.0517
(0.0000)
0.0911
(0.0000)
Firm size 0.1750
(0.3105)
0.3672
(0.3152)
0.5404
(0.4370)
0.5687
(0.4009)
0.5519
(0.6879)
0.5389
(0.4385)
Industry dummies Yes Yes Yes Yes Yes Yes
Country dummies Yes Yes Yes Yes Yes Yes
Time dummies Yes Yes Yes Yes Yes Yes
-2 Log Likelihood 256.58 251.41 237.66 356.01 360.12 360.03
# of observations 912 912 912 1194 1194 1194
This table reports fixed effect logit regression results with the dependent variable being turnaround outcome for the two alternative definitions 2a and
2b. Due to poor fit statistic output in SAS EG, the -2 Log Likelihood is the only reported fit statistics. The sample is restricted to the years in the
turnaround process, i.e. year 3-8. Standard errors are presented below the estimation results in parentheses. Model estimation is by Maximum
Likelihood (ML). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. The estimations are performed with industry and country
dummies included.
Page 105 of 111
Table 32: Odds ratios from logit analysis of turnaround outcome with dummies
Odds ratios Definition 2a Definition 2b
Variables Model 5 Model 6 Model 7 Model 5 Model 6 Model 7
Ownership concentration 0.2013 0.4078 0.1693 0.2929 4.2978 0.3553
Dominant blockholder - 0.0136 0.0039 - 0.0917 0.0069
Takeover - - 0.3442 - - 0.5874
Cost retrenchment 0.4248 0.4806 0.9926 0.9250 0.8013 0.4125
Asset retrenchment 1.0008 0.7818 1.1820 0.9407 0.9496 1.3982
Firm size 1.1912 1.4437 1.7167 1.7660 1.7365 1.4966
This table shows the odds ratios for the fixed effect logit estimation results without time dummies above in Table 31. The odds ratios are directly
derived from the fixed effect logit coefficient by exponentiation of the estimated coefficient.

Appendix 15
Table 33: Definition 2a: Pooled logistic regression
Non-linear panel models Definition 2a
Variables Model 5 Odds ratio Model 6 Odds ratio Model 7 Odds ratio
Ownership concentration 0.0925
(0.3548)
1.0969 1.1427***
(0.4311)
3.1352 1.1444***
(0.4310)
3.1406
Dominant blockholder - - -1.1745***
(0.2721)
0.3090 -1.1409***
(0.2750)
0.3195
Takeover - - - - -0.3713
(0.8990)
0.6898
Cost retrenchment -0.0160
(0.0623)
1.0048 -0.0125
(0.1187)
0.9902 -0.0132
(0.0644)
0.9869
Asset retrenchment 0.0048
(0.1160)
0.9841 -0.0098
(0.1187)
0.9876 -0.0096
(0.1186)
0.9905
Firm size 0.0859*
(0.0470)
1.0897 0.0840*
(0.0481)
1.0876 0.0831*
(0.0481)
1.0867
Industry dummies Yes Yes Yes
Country dummies Yes Yes Yes
Time dummies No No No
-2 Log Likelihood 822.92 800.99 800.36
Global null test (p-value) 0.0000 0.0000 0.0000
Somers’ D 0.558 0.612 0.611
# of observations 912 912 912
This table shows pooled logit regression results with the dependent variable being turnaround outcome for the alternative definition 2a. The sample
is restricted to the years in the turnaround process, i.e. year 3-8 and the sample for definition 2a. Standard errors are presented below the estimation
results in parentheses. Odds ratios are reported to the right of the estimates and are directly derived from the estimates by exponentiation of the
coefficient. Model estimation is by Maximum Likelihood (ML). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. The
estimations are performed with industry and country dummies included.








Page 106 of 111
Table 34: Definition 2b: Pooled logistic regression
Non-linear panel models Definition 2b
Variables Model 5 Odds ratio Model 6 Odds ratio Model 7 Odds ratio
Ownership concentration -0.2583
(0.2727)
0.7724 0.8599***
(0.3252)
2.3629 0.8799***
(0.3261)
2.4107
Dominant blockholder - -1.3349***
(0.2081)
0.2632 -1.2998***
(0.2100)
0.2726
Takeover - - -0.7296
(0.6584)
0.4821
Cost retrenchment -0.0429
(0.0593)
0.9060 -0.0402
(0.0622)
0.8857 -0.0422
(0.0632)
0.9587
Asset retrenchment -0.0987
(0.1064)
0.9580 -0.1214
(0.1100)
0.9606 -0.1205
(0.1100)
0.8865
Firm size 0.0736**
(0.0370)
1.0764 0.0787**
(0.0378)
1.0819 0.0791
(0.0379)
1.0823
Industry dummies Yes Yes Yes
Country dummies Yes Yes Yes
Time dummies No No No
-2 Log Likelihood 1505.23 1461.62 1460.22
Global null test (p-value) 0.0000 0.0000 0.0000
Somers’ D 0.313 0.398 0.400
# of observations 1194 1194 1194
This table shows pooled logit regression results with the dependent variable being turnaround outcome for the alternative definition 2b. The sample
is restricted to the years in the turnaround process, i.e. year 3-8 and the sample for definition 2b. Standard errors are presented below the estimation
results in parentheses. Odds ratios are reported to the right of the estimates and are directly derived from the estimates by exponentiation of the
coefficient. Model estimation is by Maximum Likelihood (ML). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. The
estimations are performed with industry and country dummies included.

Appendix 16
Based on the paper by Mueller & Barker (1997), a logit analysis is performed each year in order
to test the ability to predict the turnaround outcome the given year. That is logit analysis is
performed for each year in the turnaround process and the year before the decline.
Table 35: Definition 2a: Logit estimation results presented yearly
Logit, Definition 2a Year in the turnaround process
Variables
Year before
decline
Year 3 Year 4 Year 5 Year 6 Year 7 Year 8
Ownership concentration 1.3057
(0.9332)
0.5391
(0.9497)
1.8925*
(1.0436)
1.8930**
(1.0120)
1.6051
(0.9351)
0.4783
(1.0052)
0.1247
(0.9662)
Dominant blockholder -0.8593
(0.5718)
-0.6607
(0.5609)
-1.1798*
(0.6117)
-1.6724***
(0.6175)
-1.5789**
(0.6607)
-1.2700*
(0.6744)
-1.8716**
(0.8291)
Takeover 0.0000
(0.0000)
-13.1407
(1414.00)
-12.97
(999.8)
-14.0198
(691.70)
-9.7756
(1414.00)
-14.05
(572.10)
0.7248
(1.5064)
Cost retrenchment 0.0443
(0.2672)
-0.0194
(0.0809)
0.7110**
(0.3059)
-0.5373
(0.5507)
-0.5235
(0.5594)
0.9922*
(0.5838)
-0.4208
(0.2700)
Asset retrenchment 0.5610*
(0.3165)
-0.3046
(0.2192)
-2.5377***
(0.7379)
-0.4943
(0.5842)
2.2248**
(0.8474)
2.4603**
(0.8923)
1.1234**
(0.4575)
Firm size 0.0937
(0.1107)
0.0472
(0.1027)
0.1171
(0.116)
0.0510
(0.1093)
0.0760
(0.1038)
0.0947
(0.1081)
0.1473
(0.1061)
R
2
McF
0.04 0.0291 0.1333 0.0775 0.0997 0.1615 0.1348
LR statistic 6.62 5.15 23.62*** 13.89** 17.68*** 28.62*** 23.90***
Observations 137 152 152 152 152 152 152
Estimation is by Maximum Likelihood (ML). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. Standard errors are presented below
the logit estimation results in parentheses. The dependent variable TURNa = 1 if a firm turns around, otherwise 0.
Page 107 of 111
As in the thesis, the interpretation of the direction of the variables is possible. However, it is
necessary to calculate the marginal effects to understand the actual magnitude and effect. The
marginal effects allow interpretation of the partial change in the probability a firm successfully
return to prosperity and turn around for a change in an explanatory variable. I have calculated
the average marginal effects (AME), which for the first definition is presented below.
Table 36: Definition 2a: Yearly marginal effects
Average marginal effects

Year in the turnaround process
Variables
Year before
decline
Year 3 Year 4 Year 5 Year 6 Year 7 Year 8
Ownership concentration 0.2565 0.1030 0.3163 0.3425 0.2808 0.0773 0.0210
Dominant blockholder -0.0105 -0.0101 -0.0522 -0.0535 -0.0804 -0.0902 -0.0900
Takeover 0.0000 -0.3530 -1.1195 -0.7957 -0.9958 -1.8365 0.0349
Cost retrenchment 0.0087 -0.0037 0.1188 -0.0972 -0.0916 0.1604 -0.0710
Asset retrenchment 0.1103 -0.0582 -0.4241 -0.0894 0.3892 0.3878 0.1895
Firm size 0.0184 0.0090 0.0196 0.0092 0.0133 0.0153 0.0248
Correct predictions 99 111 115 112 111 116 112
Incorrect predictions 38 41 37 40 41 36 40
Count R
2
72.26 % 73.03 % 75.66 % 73.68 % 73.03 % 76.32 % 73.68 %
This table shows average marginal effects for each year to the model estimations in definition 2a. The table also present the Count R
2
statistics giving the
percentage of correct predictions by the model.

Even though the logit estimate is significant it does not imply the marginal effect is statistically
at the average. However, it is beyond the scope of this appendix to consider the significance
testing of marginal effects in SAS EG. Therefore, a marginal effect is considered significant
when the estimate is significant. The yearly logit estimation results for the firms included in the
sample derived from definition 2b are reported below.
Table 37: Definition 2b: Logit estimation results presented yearly
Logit, Definition 2b

Year in the turnaround process
Variables
Year before
decline
Year 3 Year 4 Year 5 Year 6 Year 7 Year 8
Ownership concentration 0.5228
(0.7715)
0.3222
(0.7491)
0.9624
(0.7239)
1.7243**
(0.8028)
1.5309**
(0.7521)
0.6435
(0.7710)
0.0329
(0.7298)
Dominant blockholder -0.4720
(0.4559)
-0.7761*
(0.4439)
-0.9890**
(0.4522)
-1.7783***
(0.4908)
-1.6871***
(0.5010)
-1.1155**
(0.4819)
-0.9101**
(0.4620)
Takeover -14.7417
(1224.9)
-13.4739
(1226.1)
-13.6251
(865.1)
-14.13
(631.5)
1.8851
(1.5582)
-0.0862
(1.0114)
-13.3629
(537.80)
Cost retrenchment -0.0549
(0.2454)
-0.0351
(0.0717)
-0.0639
(0.2919)
-0.0891
(0.3586)
-0.1752
(0.1979)
0.7652
(0.4658)
-0.1873
(0.1812)
Asset retrenchment 0.4598*
(0.2539)
-0.2844**
(0.1623)
-0.1704
(0.3774)
-0.6083
(0.4736)
0.3037
(0.6083)
0.8696*
(0.5021)
0.6732
(0.4782)
Firm size 0.0836
(0.0945)
0.0449
(0.0843)
0.0930
(0.0831)
0.0992
(0.0860)
0.0972
(0.0844)
0.0854
(0.0854)
0.0757
(0.0816)
R
2
McF
0.0301 0.0343 0.0333 0.0770 0.0576 0.0660 0.0671
LR statistic 7.29 9.36 9.10 21.03*** 15.74** 18.04*** 18.35***
Observations 176 199 199 199 199 199 199
Estimation is by Maximum Likelihood (ML). Stars indicate statistically significance: * p<0.10; ** p<0.05; *** p<0.01. Standard errors are presented below
the logit estimation results in parentheses. The dependent variable TURNb = 1 if a firm turns around, otherwise 0.
Page 108 of 111
The average marginal effects for each year in definition 2b are reported below.
Table 38: Definition 2b: Yearly marginal effects.
Average marginal effects

Year in the turnaround process
Variables
Year before
decline
Year 3 Year 4 Year 5 Year 6 Year 7 Year 8
Ownership concentration 0.1244 0.0761 0.2276 0.38848 0.3492 0.1451 0.0075
Dominant blockholder -0.0026 -0.0097 -0.0095 -0.0331 -0.0207 -0.0186 -0.0223
Takeover -0.1365 -0.2630 -0.2299 -0.4199 0.0231 -0.0014 -0.6070
Cost retrenchment -0.0131 -0.0083 -0.0151 -0.0199 -0.0400 0.1725 -0.0426
Asset retrenchment 0.1094 -0.0671 -0.0403 -0.1358 0.0693 0.1961 0.1530
Firm size 0.0199 0.0106 0.0220 0.0221 0.0222 0.0192 0.0172
Correct predictions 109 108 114 119 121 131 130
Incorrect predictions 90 91 85 80 78 68 69
Count R
2
53.98 54.27 % 57.29 % 59.80 % 60.80 % 65.83 % 65.33 %
This table shows average marginal effects for each year to the model estimations in definition 2a. The table also present the Count R
2
statistics giving the
percentage of correct predictions by the model.

Appendix 17
Table 39: Table of the average top blockholders and ownership concentration ratio by country
Country Number of countries Average of top blockholders Average ownership conc. ratio
AUT 5 47.8 (28.9) 53.9 (27.5)
BEL 6 39.5 (21.6) 51.9 (25.0)
CHE 12 30.1 (26.6) 38.3 (26.7)
DEU 53 41.0 (26.1) 54.9 (28.5)
DNK 8 30.6 (13.6) 56.8 (23.7)
ESP 5 29.6 (16.9) 53.0 (27.0)
FIN 8 23.8 (17.9) 39.1 (15.0)
FRA 59 39.7 (23.5) 57.9 (24.4)
GBR 76 19.9 (17.8) 36.0 (22.7)
IRL 3 40.7 (38.0) 64.2 (22.8)
ITA 10 53.1 (19.4) 66.6 (16.3)
NLD 17 39.8 (24.4) 57.7 (23.9)
NOR 5 31.6 (17.3) 44.1 (19.3)
PRT 2 49.2 (25.9) 50.8 (27.4)
SWE 20 26.9 (20.4) 46.3 (21.1)
This table provide information regarding the average size of the top blockholder in each country. Standard deviation is reported next to the
average value in parentheses. Irrespectively of country, the total average of top blockholder is 33.01 pct., while the total average of ownership
concentration ratio is 49.11 pct. All average measures are in percentages (%).

Appendix 18
Ensuring ownership concentration in relation to firm performance does not appear to have a
curve-linear shape (e.g. V-shape), eroding the potential association due to trends being non-
linear, the figures below provide evidence for these aspects during the sample period.
Page 109 of 111
Figure 11: ROA plotted against ownership concentration ratio


Figure 12: Distribution of ownership concentration in the sampling period

Examining the above figure with ownership concentration ratio distributed by the turnaround
cycle years, the figure show a straight line for almost every year, indicating no change in
average ownership concentrations during the turnaround process.

Appendix 19
The F-statistic reveals that the null hypothesis, testing that the means are equal, can be rejected.
However, average cost retrenchment is only weakly significant different from each other across
industries.
Table 40: Comparing average retrenchment across industries
Asset retrenchment Cost retrenchment
Industry group N Mean Std.dev.
Mineral 18 0.0406 0.2916 18 0.3434 1.1688
Manufacturing 942 0.0035 0.4149 942 -0.0064 1.1520
Transportation, Communication, Utilities 138 0.1417 0.6449 138 0.1947 1.1285
Trade 162 0.0281 0.5086 162 0.0752 0.8322
Service 474 0.1504 0.9294 474 0.2348 2.4981
F-test statistic 5.13*** 2.05*
This table reports testing of the whether asset retrenchment differ across industries. Stars indicate statistically significance: * p<0.10; ** p<0.05;
*** p<0.01.
-1.00
-0.80
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.00 0.20 0.40 0.60 0.80 1.00
R
e
t
r
u
n

o
n

a
s
s
e
t
s

(
R
O
A
)

Ownership concentration ratio
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1999 2001 2003 2005 2007 2009
O
w
n
e
r
s
h
i
p

c
o
n
c
e
n
t
r
a
t
i
o
n

r
a
t
i
o

Year
Page 110 of 111
Appendix 20
It is evident from looking at average cost and asset retrenchment that both display a non-linear
behaviour over the turnaround cycle period (Year 3-8). Firm’s are on average decreasing the
growth in their cost base, while cost retrenchment only appears in the two first years in the
recovery period. On average, asset retrenchment starts, i.e. the asset base is decreasing, in
turnaround year 4, which are in the middle of the decline period, while being negative in year 7
and 8, meaning the firms on average increased their asset base in the last two years of the
turnaround cycle period.
Figure 13: Level of asset and cost retrenchment during the turnaround period


Appendix 21
The following show formulae for the individual performance ratios used in the sample selection
process and regression. Further, this appendix addresses the profitability and performance ratios
addressed and mentioned in the discussion of appropriate performance measures and
benchmarks:

Return on sales (ROS) is defined as follows:


(13)
where sales also are known as net turnover.

Return on assets (ROA):




(14)
-20%
-10%
0%
10%
20%
30%
40%
2 3 4 5 6 7 8
A
s
s
e
t

a
n
d

c
o
s
t

r
e
t
r
e
n
c
h
m
e
n
t
,

%

Year in the turnaround cycle period
Average Asset retrenchment Average Cost retrenchment
Page 111 of 111
where the item IB in Compustat is employed. This item represents the income of a company
after all expenses, including special items, income taxes, and minority interest.

Return on invested capital (ROIC):



(15)

In extension to the equation above, return on investment [ROI] is often defined as measuring net
income against either invested capital or total assets in the literature. In this thesis, ROI is
defined as net income divided by net investments during the year and is therefore not to be
mistaken with ROIC, where net income are considered in relation to total capital invested.
Additionally, it is not always clear which definition the researchers are employing.
Some studies employ Tobin’s Q. I do not use Tobin’s Q as it is characterized as an
additional measure due to the fact that the metric not is based on profitability figures from the
income statement. Opposite, Tobin Q’s is based on the market value of equity and balance sheet
figures (Plenborg & Petersen, 2011). As I will not use market value measures in defining
performance because the changes in the financial market valuations do not necessarily reflect
changes in the utilization of the firm’s resources (Chakrabarti, 2012), I have not defined this
measure in the formulae despite the term has been used in the thesis.
In the text, I mention the term cost of capital. There are three different methods in
calculating cost of capital, which are deemed appropriate in benchmarking profitability, but not
applied in the thesis: 1) Required rate of return on assets (r
a
), 2) Required rate of return on
equity (r
e
), and 3) Weighted average cost of capital (WACC) (Plenborg & Petersen, 2011).


doc_480513681.pdf