Description
Past research into the development of trading companies has seldom been able to separate the economic raison detre of the firm from the political and legislative influence of the home-country government. In contrast with other countries, the evolution of international trade intermediaries in Hong Kong has not been directly influenced by government policy.
Ellis, P.D. (2001), “Adaptive strategies of trading companies,” International Business
Review, 10: 235-259.
Adaptive Strategies of Trading Companies
Paul Ellis
1
Associate Professor
Department of Business Studies, Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong
Kong, Tel: (852) 2766 7108; Fax: (852) 2765 0611
Email: [email protected]
15 December 2000
Abstract
Past research into the development of trading companies has seldom been able to separate the
economic raison d’être of the firm from the political and legislative influence of the home-country
government. In contrast with other countries, the evolution of international trade intermediaries in
Hong Kong has not been directly influenced by government policy. In the absence of legislative
inducements or protection, Hong Kong traders exist purely by their ability to respond to market
forces, suggesting that important insights regarding organizational survival can be gleaned from
studying their patterns of behavior. In this paper, an evolutionary model is proposed and then
evaluated against data collected from trading companies at five different stages of development.
Specifically, these firms’ adaptive strategies are assessed in terms of their patterns of diversification
along three dimensions: product, geographic, and function. For scholars, the findings may be
considered a stepping stone towards the articulation of a truly generic model of trading company
evolution, while the value of the study for policy-makers in search of new archetypes lies in the
identification of a variety of market-responsive organizational forms.
Key words: trading company, adaptive strategy
1
The author gratefully acknowledges the constructive advice given by Wing-chor Chu of the Hong
Kong Trade Development Council, Mimi Yeung of the Hong Kong Exporters’ Association, and two
anonymous IBR reviewers. The author also wishes to acknowledge his debt of gratitude to his two
research assistants, Wong Hok Ying and Lai Wai Man. The research reported in this paper was
funded by the Hong Kong Polytechnic University (Project No.351/384).
Adaptive Strategies of Trading Companies
2
Introduction
It is ironic that despite more than three centuries of active participation in international trade, the
existence of the ubiquitous trading company is now being threatened by the globalization of markets.
Technological developments in communication, transportation, and banking, combined with the
ongoing internationalization of manufacturers and customers, have collectively led to the reallocation
of many of the traditional distribution functions performed by international trade intermediaries. This
has prompted some scholars to label modern trading companies, and in particular large-scale general
trading companies (GTCs), as “superfluous”, “inhibiting to the export of anything but the most simple
manufactured products”, and generally having “outlived their usefulness” (e.g., Wortzel and Wortzel
1983, pp.74-5). Yet, despite the economic pressures for their elimination, trading companies, for the
moment, continue to prosper at the interstices of global markets.
The research aims of this paper concern the identification of those adaptive strategies that ensure the
survival of the modern trading company in all its many guises. This goal is somewhat frustrated,
however, by the fact that many of the world’s highly diversified trade intermediaries owe their
existence to the extra-market intervention of the state (Balabanis and Baker 1993a). One classic
example is the symbiotic relationship between the policies of the Meiji Government and the
subsequent rise of the J apanese zaibatsu and sogo shosha (Kojima and Ozawa 1984; Sarathy 1985;
Yamamura 1976). Short of direct state sponsorship, governments may influence the development of
trading companies via inducive or coercive means. Inducements range in degree of intervention from
the relaxation of antitrust legislation (e.g., the US Export Trading Companies Act (1982)) to the
payment of tax rebates and direct subsidies. In contrast, coercive measures may include the
mandatory requirement to satisfy a number of criteria in order to receive GTC status along with the
commensurate benefits. For example, in 1975 the Korean government outlined six prerequisites for
GTC designation which included maintaining ten or more overseas branch offices and exporting
goods worth upwards of $1m to at least ten countries each per year (Cho 1984). A similar scheme was
adopted by the Spanish government in 1988. In this case, registered trading companies were required
to have annual sales in excess of 1,000m pesetas which included exports of at least 600m pesetas
worth of local products (Balabanis and Baker 1993a). Indeed, whether the motives are economic or
political, the hand of government is evident throughout the entire history of the trading company,
beginning with the monopoly charters assigned by the mercantilist governments of England and
Holland, to the original trading houses in the 17
th
century (Carlos 1992; Carlos and Nicholas 1988).
An exception to this interventionist rule is to be found in the case of the world’s most laissez-faire
society, Hong Kong. Unlike that of their Asian, American, and European counterparts, the evolution
of Hong Kong trading companies has not been influenced by government policy. In Hong Kong there
is no Export Trading Companies Act, no Ministry of International Trade and Industry, and no Five-
Year Plan to direct the growth of trade intermediaries. For the vast majority of Hong Kong’s import-
export companies, long-term survival is determined purely by market forces. Hong Kong’s system of
free enterprise, combined with a local trading environment that is particularly sensitive to external
shocks, suggests that important lessons on organizational survival can be gleaned from studying the
adaptive behavior of Hong Kong trading companies.
The premise of this paper follows that of Balabanis and Baker (1993a/b) who argue that any
understanding of the mechanisms by which trading companies adapt their strategies and structure
must begin with the identification of the patterns of diversification exhibited by traders at different
stages of development. “The dominant discriminating feature of the GTC concept is their diversity.
Thus, diversification theory can provide a more rigorous explanation of the development process of
GTCs” (Balabanis and Baker 1993a, p.47). The paper is divided into four sections. In the following
section the eclectic literature relating to the trading company concept is briefly reviewed. In
particular, Kim’s (1986) life cycle model is introduced and discussed as a basis for understanding
GTC development. Next, and by way of providing a backdrop for the analysis of their adaptive
strategies, a brief historical outline of the trading environment in which Hong Kong trading
companies compete is provided. Third, the methodology used to survey 1,890 traders is described,
and finally the study’s main research findings are presented and discussed.
Adaptive Strategies of Trading Companies
3
Background
Trading companies are characterized by their specialization in market-making intermediation, an
activity which may involve brokerage (selling on behalf of another) or reselling (taking title to the
goods traded) (Casson 1998). The study of trading companies is arguably more multinational in flavor
than any other topic in the field of international business. Trading companies are chauvinistic entities
whose missions, organizational structures, and internal cultures reflect the idiosyncratic political and
economic conditions of their countries of origin. As a case in point, consider the limited success with
which the uniquely J apanese sogo shosha concept has been transplanted to other countries (Hvala,
Perry, and Boddewyn 1990). Similarly, research into the topic is as context-dependent as the trading
company phenomenon itself, with many studies limited to a particular national setting (Table 1). The
result is an eclectic body of literature that is characterized by a dearth of widely applicable conceptual
schemas and typologies, and which offers limited scope for generalization (Amine 1987).
INSERT TABLE 1 ABOUT HERE
Much of the existing research on trading companies is highly particularistic in nature, focusing on
specific trading issues such as the association between different channel structures and the marketing
practices of exporter traders (Bello and Williamson 1985a), or the effect of government policies on
the performance of intermediaries (Howard 1995; Kaikati 1984; Williams and Baliga 1983). Some
attempts have been made to classify and categorize the subject matter (e.g., Barovick and Anderson
1992; Brasch 1978), but the systematic advance of research has generally been hampered by the lack
of consensus on matters of definition and limited attempts at theory-building (Amine 1987; Cho
1987). In particular, much of the research is utilitarian in purpose, providing little more than
“snapshots” of the changing roles and functions of trading companies at particular points in time (e.g.,
Brasch 1978; Kelly and Lecraw 1985; Wichmann 1997). To date there has been little progress in
developing a general framework of trading company evolution, and of the dynamic studies which do
exist, most reflect the unique characteristics associated with a particular type of trade intermediary
and cannot be applied universally (Perry 1990). For example, there are a number of longitudinal case
studies which trace the rise of the early European trading houses (Buchan 1994; Sugiyama 1978);
several growth models of the distinctive J apanese GTC have been proposed (Cho 1984; Kojima and
Ozawa 1984), and at least two scholars have developed frameworks for analyzing the adaptive
strategies of US export trading companies (Amine 1987; Perry 1990). Very few growth models exist
which are multinational in scope. Balabanis and Baker (1993b) developed a dynamic framework
based on those external and internal factors that lead to organizational change. To test their model,
they studied the “change intentions” of 29 European trading companies in terms of their geographical
expansion, product diversification, and functional and service development. These growth measures
reflect the widely-held view that the development of a trading company is evidenced in terms of
specific patterns of diversification which are brought about in response to changes in market
conditions (Balabanis and Baker 1993a; Cho 1987; Jones 1996). Pioneering conceptual work in this
area was done by Kim (1986), who was the first to conceive that the development of the trading
company can be conceptualized in terms of a number of distinct and sequential stages that
collectively define a GTC life cycle.
The General Trading Company Life Cycle
The term “trading company” encompasses a wide variety of trade intermediaries ranging in size and
scope from the small-scale import/export company to the mammoth and prototypical sogo shosha.
Because the generic concept of the GTC is no longer a strictly J apanese phenomenon, Kim (1986)
offers a three-stage explanation for these observed differences in levels of organizational
development that is based on economic and institutional factors rather than cultural differences. The
Stage I trading company is relevant to an emerging economy characterized by an underdeveloped
infrastructure, standardized manufactured goods, and fragmented channels of distribution. In addition,
the industrial base may be dominated by small- to medium-sized enterprises (SMEs) requiring the
assistance of an intermediary to go abroad. These contextual conditions create a market opportunity
for purchasing and sales agents to generate transaction efficiencies by specializing in the trading
function and by developing a global market information network. As well as lowering the costs of
information acquisition, the trading company may also be able to provide economies of scale for local
Adaptive Strategies of Trading Companies
4
firms in terms of sorting, grading and packaging standardized products, and by routinizing some
frequently-repeated transactions.
As the trading company engages in some product and area diversification to spread its risk, it
approaches the second stage of development which is characterized by the performance of a number
of trade-supporting activities, such as the provision of credit. Trade financing may be warranted if
local manufacturers have only limited access to affordable funds and if local lending institutions take
the view that local export agents are better placed to evaluate the credit risk of their trading partners
(J ETRO 1976). Political factors may also play a role at this stage if there is a perceived need to
improve the efficiency of trading activities, possibly with a view to replicating the well-publicized
success of the sogo shosha (e.g., as in the case of the thirteen GTCs created in Korea, a developing
country, in 1975), or if there is a desire to create organizations which bridge the gap between local
SMEs and export markets (e.g., as in the case of an industrialized country like the US, with its Export
Trading Company Act of 1982). As competitive pressures continue to increase, Kim (1986) notes that
the Stage II trading company is faced with two choices: (1) it can continue to focus on standardized
products (provided the firm has preferential access to low-cost supplies or captive clients), or (2) it
can trade up to higher-technology products (which offer good growth prospects and fatter margins).
The trading company enters the third stage of its life cycle when it engages in foreign direct
investment to protect its market position from the threat of internationalizing local manufacturers or
rising protectionist forces. In addition to the growth in direct exporting among manufacturers, the
competitive position of the trading company may be further undermined by the provision of trade-
supporting services by banks and government agencies. Although trading remains the principal
activity of the firm, its relative importance within the firm decreases to make room for related
functions such as shipping, warehousing, and manufacturing. For example, Wortzel and Wortzel
(1983) observe that the J apanese sogo shosha, which exemplify Stage III GTCs, play a fairly passive
role as export marketers taking title to goods primarily to expedite their financial and physical
distribution functions. Functional diversification into complementary areas may also be prompted by
a lack of local infrastructure or entrepreneurship. Further, there may be a need for the trading
company to internalize some trade-related activities to maintain quality control, preserve information
asymmetries, or pursue scope economies (J ones 1996). At this advanced stage of development, any
government mandates are more likely to be regulatory than strategic in nature.
Although Kim’s (1986) life cycle model implies that, given the right conditions, trading companies
ultimately evolve to become analogues of the sogo shosha, trading company development is by no
means deterministic. Kim concedes that life cycle stages may be by-passed, or they may occur
simultaneously rather than sequentially, while some firms may never progress beyond the first or
second stages of development. The actual evolutionary paths taken reflect strategic choice which, in
turn, is influenced by conditions within the host environment. While the life cycle model provides a
general basis for understanding those institutional factors germane to the development of GTCs, it
makes no attempt to consider those diversification strategies which may lead the firm away from
trading altogether. The implicit assumption is that trading companies never grow to become anything
other than highly diversified trading companies. In reality, adaptive parameters are not so narrowly
defined.
If the aim is to understand the adaptive strategies or developmental paths exhibited by trading
companies over time, any comprehensive explanation must take into account those non-trading
diversification alternatives which are available to the growing trading company. For example, and
despite the implied suggestion of the life cycle model, a Stage I import/export company is not
compelled by environmental forces to “evolve” into a Stage II trading company. Other diversification
options exist and perhaps chief among these is the option to vertically integrate into manufacturing
(Amine 1987). Support for this assertion comes from Cho’s (1987) study of GTC diversification
patterns. Cho was arguably the first to recognize that small traders may diversify in one of two
directions: forwards or backwards. Forward integration implies the provision of additional services
such as transportation, warehousing and insurance (i.e., progression to the next stage of GTC
development), whereas backward integration implies diversification into manufacturing, possibly to
counter the threat of opportunism in the channel. Such a move assumes that the trading company is
Adaptive Strategies of Trading Companies
5
able to leverage its knowledge of foreign markets and distribution channels into a position of
competitive parity on the production function. It is plausible that for many small trading firms
confronted with the need to diversify, investment into manufacturing may be too costly, prompting
either a hybrid arrangement (e.g., contract manufacture) or the further expansion of the firm’s traded
product lines and market coverage (i.e., diversification into a Stage II trading company). The
manufacturing option identified by Cho (1987) is arguably more suited to Stage II traders who, by
virtue of their greater size and experience, are more likely to have access to the capital necessary for
such an investment.
Another evolutionary outcome not captured in the life cycle model is observed when the general
trading company’s principal activities begin to expand into areas other than trading. Such a change in
strategic direction might result from diminished opportunities for trade intermediation as the home
economy matures, local manufacturers become more adept at international marketing, and better
investment opportunities emerge in brokerage- or service-related industries such as construction,
finance, and property development. This has been the experience of the sogo shosha as described by
The Economist (1991):
To defend their markets they have gone upstream (into mining and manufacturing) and downstream
(into retailing) through equity investments in joint ventures and affiliates. The sogo shosha are
changing from pure traders, where high-volume low-margin commission business dominates, to
become more financially sophisticated investment holding companies.
The tendency for GTCs to diversify away from trading has been evident from the time of the earliest
trading companies, whose investments in foreign production made them forerunners of the modern
multinational enterprise (MNE) (Carlos and Nicholas 1988). Whether the trading company is
investing in foreign rubber plantations, like the English firm Harrisons & Crosfield (J ones 1996), or
car manufacturing, like Korea’s Hyundai Group (Amine 1987), MNE status is achieved when the
revenue earned from offshore investments exceeds the income derived purely from trading. Although
the switch from GTC to MNE is by no means inevitable (Cho 1987), environmental forces may
provide a compelling incentive for both Stage II and III traders to engage in a significant level of
foreign direct investment to hedge against the risks associated with trading. Kim (1986)
acknowledged as much by defining his Stage III firm as “toward a conglomerate”, but he saw no
reason to incorporate this path of diversification in his model of GTC development. Despite the
omission of GTCs-turned-MNE conglomerates from Kim’s framework, it is worth noting that the
trading history of such firms may have a residual bearing on both the corporate culture and the
direction of investment activities. This is particularly evident in the case of Hong Kong, where firms
like J ardine Matheson and Swire Pacific are still referred to as hongs despite relegating their trading
functions to fairly minor, albeit active, subsidiaries many years ago. Indeed, for many of these former
trading houses (e.g., Hutchison), the current core business activity information (in this case
telecommunications and port facilities) stems directly from the firm’s early involvement in the
international flow of goods. For others such as J ardines, investment decisions in the growing market
of the People’s Republic of China (PRC) are overshadowed by the tenor of their past trading relations
with the mainland.
In summary, trading companies must adapt to meet the changing export requirements of their host
environments. At different points, in time traders may be confronted with one of two choices: either
(1) continue to exploit their core competence as market intermediaries and specialize or diversify to
improve the efficiency and effectiveness of this service, or (2) functionally diversify to the point
where trading is no longer the core business activity of the firm. The first option, played to its logical
conclusion, will lead to the development of the diversified GTC, while the second option is more
open-ended. However, given the strategic constraints imposed by available skills and resources, it is
likely that trading companies that engage in this latter type of diversification will ultimately become
either manufacturer-exporters (if some form of backward integration is pursued) or MNEs (if offshore
investments come to dominate the business activity of the firm). The main adaptive possibilities open
to trading companies are depicted in Figure 1. In contrast with the simple linear progression of the life
cycle model (Kim 1986), the figure proposed in this paper allows for six types of adaptation leading
to four distinct kinds of organization, each characterized by a particular pattern of diversification and
each having a different trading emphasis.
Adaptive Strategies of Trading Companies
6
INSERT FIGURE 1 ABOUT HERE
In terms of the nature and pace of their organizational development, trading companies, perhaps more
than any other international business entity, are influenced by the particular economic and regulatory
conditions of their home environments. Consequently, before the empirical relevance of Figure 1 is
assessed, it is first necessary to provide some background detail regarding the context in which Hong
Kong trading companies have evolved.
The Hong Kong Trading Environment
Since the earliest collection of trade statistics, it has been evident that Hong Kong’s trading
environment is characterized by its volatility. As the world’s tenth largest trading territory, Hong
Kong has not suffered from a protracted recession, but shocks and disruptions have been frequent and
disruptive, as evidenced by the magnitude of annual changes in trade growth (Figure 2). In terms of
the proportion of trade to gross domestic product, Hong Kong is the world’s second most open
economy after Singapore (Ellis 1998b), which means that the local trading environment is especially
sensitive to external shocks such as China’s Great Leap Forward (1959), the OPEC price hikes
(1973), and the Asian currency crisis (1997). On occasion the nature of an external disruption is such
that organizational survival is threatened compelling trading companies to adapt. This can be seen in
the case of at least four external jolts (Ellis 1998a):
1. the invasion of J apanese forces in 1941 (which abruptly ended Hong Kong’s entrepÔt role,
the basis for much of its economic activity at the time),
2. the J apanese surrender in 1945 (which precipitated the swift resumption of the re-export trade
with southern China),
3. the Korean War and the resulting United Nations embargo on trade with the PRC in 1951
(which led to the cessation of all entrepÔt trade virtually overnight and triggered the
industrialization of Hong Kong), and
4. the opening of China’s markets in 1979 (which stimulated the mass migration of Hong Kong
industry across the border).
INSERT FIGURE 2 ABOUT HERE
The effect of these environmental disruptions on Hong Kong’s population of trading enterprises can
be summarized as follows: the thousands of import/export businesses that had ceased trading after the
J apanese invasion in 1941 flourished again in the late-1940s on the back of the colony’s resurgent
entrepÔt trade. These firms were then obliged in 1951 to replace their mainland Chinese suppliers
with local suppliers or go out of business (Woronoff 1980). Subsequent competition for international
markets from local manufacturers of clothes, toys, and watches then compelled trading companies to
reduce commissions, engage in further product and geographic diversification, and begin investing
directly abroad. Many traders simply decided to capitalize on the swelling supply of cheap local labor
and go into manufacturing for themselves, with the result that the value of Hong Kong-made exports
surpassed re-exports for the first time in the colony’s history in 1959 (Wilson 1990). In the 1980s,
economic factors (rising rents and wages) combined with a political thaw in the PRC stimulated the
resumption of Hong Kong’s entrepÔt role, and once again re-exports replaced domestic exports as the
main form of international trade (Ellis 1999). By the mid-1990s, more than one fifth of the local
workforce, around half a million people, was employed in import/export enterprises
(http://www.tdc.org.hk/).
In the past six decades, environmental disruptions have qualitatively altered the rules of the game for
Hong Kong traders on several occasions. The net result is a trading landscape populated by a variety
of organizational forms, each of which represents an entrepreneurial response to changing market
signals rather than an induced response to political directives. To date very little research has been
done on this unique group of 100,000 or so trading companies. Little is known about their patterns of
diversification, the nature and direction of their trading activities, or their performance as middlemen
in international distribution channels. The methodology used to address these questions is described
in the following section.
Adaptive Strategies of Trading Companies
7
Methodology
Questionnaire Design
In order to gain some insight into the adaptive strategies enacted by Hong Kong traders, a four-page
questionnaire was administered to a sample of 1,890 firms. The questionnaire consisted of several
items relating to the personal characteristics of the respondent, 19 descriptive items regarding the
extent of diversification and the nature of the firm’s trading activities, and 11 objective and subjective
measures of the firm’s performance over the past three years. To ascertain the level of product, area,
and functional diversification, operational measures were adapted from the indices used in Cho’s
(1987) analysis of the GTC. An area diversification index (ADI) was calculated in two ways: (1) the
number of countries actively traded with at the time of data collection, and (2) the proportion of sales
income earned outside of Hong Kong and the PRC. Functional diversification (FDI) was ascertained
by identifying the number of support services offered as selected from a list of 26 alternatives.
Product diversification (PDI) proved to be the most troublesome variable to measure. Following
Cho’s (1987) procedure, respondents were initially asked to identify the number of two-digit Standard
International Trade Classification (SITC) categories in their product ranges and then provide
information regarding the proportion of sales earned in each category. However, several iterative
pretests of the questionnaire among a group of six traders revealed that this item was hopelessly
intractable and difficult for respondents to answer. In the end, a list of 24 product categories was
provided that approximated the two-digit level of aggregation with a question asking the respondent
to indicate the percentage of the firm’s sales, if any, earned from each category. An index was then
calculated by multiplying the number of product lines by the proportion of sales generated outside the
principle product. For example, if a respondent checked three product categories indicating relative
income shares of 80, 15, and 5 per cent, then the relevant PDI is 0.60 (or three times 0.20).
Sample Selection
The sample was defined as the population of indigenous trading companies listed in two recently
published directories (Registry of Hong Kong Traders 1997; Members’ Directory 1997). The
questionnaire was mailed twice and then faxed personally to the most senior executive identified in
the directories. Although the use of self-report data is not without its limitations (see for example,
Ganster, Hennessy and Luthans, 1993), the nature of the inquiry justified the reliance on subjective
measures. First, it can be assumed that traders' perceptions of their environment will affect their
strategy-making, irrespective of whether those perceptions are valid or not (Anderson and Paine
1975; Miller 1988). Traders who are not "clued in" to their multiple operating environments will not
make the appropriate adjustments when those environments change. Thus, trader experience or
survival should be correlated with a track record of accurate "sense-making" of environmental
change. Second, past research has indicated that relying on a knowledgeable "informant's perceptions
of an organisation's strategic orientation is a valid approach to measuring strategy" (Shortell and
Zajac 1990:829). Finally, subjective performance assessments are suitable in situations where "(1)
accurate objective measures are unavailable and (2) the alternative is to remove the consideration of
performance from the research design" (Dess and Robinson 1984:271). Traders are universally
secretive by nature and are indeed particularly so when it comes to measuring their performance (see
for example, Guex 1998). This is no less true in Hong Kong, where owner-operators are the sole
gatekeepers to such information, and the vast majority (more than 95 per cent) of firms are privately-
held entities.
At the close of data collection, 47 questionnaires had been returned uncompleted because the
respondent did not consider their firm to be significantly involved in the trading (i.e., import/export)
business. A further 35 questionnaires were discarded from the database because the respondents
reported that their headquarters were located outside of Hong Kong. The large number of ineligible
responses indicated that the criteria for inclusion in the published directories were somewhat broader
than the sampling frame used in this study. In addition to the ineligibles, 300 questionnaires were
returned undelivered, reflecting both the attrition rate of the recent Asian economic crisis and the
footloose nature of small import/export companies. In all, 211 useable questionnaires were received.
The effective response rate was calculated based on the formula advocated by the Council of
American Survey Research Organizations (CASRO) for surveys with single-stage sampling and an
Adaptive Strategies of Trading Companies
8
eligibility requirement (Churchill 1995; Burns and Bush 2000). The critical variable in determining
the response rate in a study of this nature is estimating the number of eligible firms receiving the
questionnaire. There were two methods for calculating this figure. First, of those who made the effort
to return the questionnaire, only 72 per cent were eligible for inclusion in the study (i.e., they were
significantly involved in trading and had their head office in Hong Kong). Second, follow-up phone
calls made to a random subsample of 150 listed traders revealed that 24 per cent of the surveyed firms
were ineligible for inclusion for one of the two reasons mentioned. Using the more conservative
estimate of 24 per cent indicates that 454 firms receiving the questionnaire did not meet the sampling
requirements of the study. The resulting response rate of 35.0 per cent was considered satisfactory
given the performance of similar surveys done in Hong Kong where response rates can be less than
ten per cent (e.g., Chan and Ellis 1998). Nevertheless, to test for possible non-response bias,
descriptive data collected during the follow-up phone-calls were compared with information provided
by respondents. No significant differences between the two samples were found in terms of the
number of years trading (t=.31, p=.75), the number of local (t=1.18, p=.24), mainland Chinese
(t=1.02, p=.31), and foreign employees (t=1.07, p=.28), and the proportion of products sourced from
Hong Kong (t=.77, p=.44), China (t=1.36, p=.17), or elsewhere (t=-1.65, p=.10), suggesting that non-
response was not a source of bias in this study.
Analysis and Findings
Table 2 provides a summary of the general descriptive characteristics of the trading companies and
the respondents. More than 90 per cent of the firms in the sample reported more than five years’
trading activity. Similarly, more than four-fifths of the respondents themselves have been actively
involved in trading for more than five years, indicating their experience and suitability for completing
the questionnaire. A wide range of industries is represented by the traders, as indicated by the
assortment of product lines carried (Table 3). Chief among these are exports of clothing and electrical
goods, two product categories which have been the mainstay of Hong Kong’s economic development.
INSERT TABLES 2 & 3 ABOUT HERE
Classifying the Traders
As a starting point for the analysis, respondents were classified according to the five trader-types
identified in Figure 1. Although all the firms in the survey were defined as trading companies by
virtue of their inclusion in a commercially-produced directory of traders, to qualify as a trading
company for the purposes of this study, the products traded had to come primarily from sources
external to the firm. Consequently, any firm manufacturing 50 per cent or more of their traded
products was deemed to be an exporter, and any exporter with significant foreign direct investment,
as proxied by the employment of overseas workers (outside Hong Kong and China), was labeled an
MNE. This classification choice led to the identification of 35 manufacturer-exporters and 16 MNEs
in the sample. The surrogate indicator of foreign employees was also used to determine those traders
with significant foreign investments, resulting in the identification of 28 Stage III trading companies.
The remaining traders were split into two groups by looking at the specific trade-supporting services
used to calculate the degree of functional diversification. According to Kim (1986), Stage II traders
can be distinguished from their Stage I counterparts by their provision of financial resources to
trading partners. In the survey, 41 respondents indicated that they offer trade financing to clients, thus
deeming themselves to be Stage II traders and leaving 91 Stage I traders.
Having identified the five types of trading company, ANOVAs were conducted to ascertain
differences among the main descriptive variables (Table 4). The findings generally support the view
that, as trading companies evolve, they tend to exhibit increasing levels of product, geographic and
functional diversification (Balabanis and Baker 1993a). As expected, Stage I traders were the
youngest trading companies in the sample and reported lower levels of diversification on all measures
except the number of foreign markets, which was similar across all three stages of development. As a
group, the traders, and particularly the more experienced traders, evidenced higher levels of product
diversification than the exporters (i.e., the manufacturers and MNEs), perhaps indicating their greater
need to safeguard against the threat of opportunism. For example, manufacturer-exporters run little
risk of being supplanted by suppliers if they are responsible for their own production. In contrast,
Adaptive Strategies of Trading Companies
9
traders need to carry a wider variety of product lines to minimize their dependence on any one
supplier. Interestingly, the exporters compensate for their relatively low levels of product
diversification by leveraging their product expertise across a larger number of foreign markets. The
traders in the sample, irrespective of their stage of development, were on average involved in fewer
than seven foreign markets. In contrast, the manufacturer-exporters were active in almost ten foreign
markets, while the MNEs surveyed were active in twice as many markets as the traders, as befits their
multinational designation.
INSERT TABLE 4 ABOUT HERE
Other significant differences observed include firm size, with the advanced levels of development
being associated with an increase in both the scale and dispersion of operations. In general, Stage II
traders are significantly larger and employ more workers in China than Stage I traders, while Stage III
traders have the largest and, by definition, most regionally-dispersed workforce of all trading groups.
Unsurprisingly, the largest companies in the sample are those responsible for their own production
(manufacturer-exporters and MNEs). This group of firms has evidently capitalized on the supply of
cheap labor in the mainland, as indicated by their very large numbers of PRC employees.
Respondents were asked to estimate the proportion of total income derived from trading activities.
Again consistent with the literature (Kim 1986), the findings show that as a trader becomes more
diversified and develops into a GTC (i.e., Stage II ? Stage III), the relative importance of trading
activities within the firm’s overall service offering diminishes (Table 4). Nevertheless, trading
remains the dominant activity of the firm accounting for more than four-fifths of the income earned
by all trading groups. The findings also reveal that the two exporter groups have significantly lower
trading shares than the trading companies (75% for manufacturer-exporters and 53% for the MNEs),
further indicating that the classification schemas used were sufficiently discriminating.
The Nature and Direction of Trading Activities
Hong Kong’s entrepÔt role as the preeminent service provider to Guangdong province is clearly
evident in the pattern of trade reported by respondent firms (Table 5). Although respondents were
able to select from six distinct types of trading activity (imports, re-exports, exports, transshipments,
third-country trade, and wholesale/retail trade), Stage I and II traders were generally biased towards
imports/re-exports, with China being the dominant supplier for the latter group. Stage I traders were
also heavily involved in local distribution (wholesale and retail trade), more so in fact than any other
group in the sample. At the other end of the scale, the Stage III traders reported the lowest levels of
domestic trade (6%) and the highest levels of third-country trade (24%) where goods are shipped
direct to the market without stopping in Hong Kong. This finding is significant as functional
expansion into third-country trade is a hallmark of a mature trading company and is seldom
undertaken by novice traders (Balabanis and Baker 1993b). With their regionally-dispersed
workforces, Stage III traders source nearly three quarters of their products from outside of Hong
Kong and China. Of the two self-manufacturing groups, the MNEs also recorded high levels of
offshore trade (15%) in comparison with the more localized manufacturer-exporters (1%). Consistent
with their high levels of employment in China, the manufacturer-exporters had by far the highest
levels of re-exports in the sample.
Finally, there was a readily-observable association between the level of development of the trading
company and the overall size and regional dispersion of their supplier networks. Mature Stage III
traders were the least likely group to source from local suppliers and were reliant on foreign factories
for 70 per cent of their business. As expected, MNEs also had larger and more dispersed networks of
suppliers than manufacturer-exporters. This finding suggests that the social capital inherent within a
trader’s list of contacts takes time to develop. Indeed, organizational development may itself be
constrained by the boundary horizon of the trading network (Balabanis and Baker 1993a; Rauch
1996).
INSERT TABLE 5 ABOUT HERE
Adaptive Strategies of Trading Companies
10
Performance Comparisons
Finally, the five groups of firms were compared in terms of their trading margins and operating
performance (Table 6). MANOVA was used to assess differences across three composite
performance measures; sales growth and profitability over each of the last three years, and the
respondents’ own subjective appraisal of four performance measures. Although the Stage III traders
and MNEs generally enjoyed superior performance across the board, none of the observed differences
are significant at the p
Past research into the development of trading companies has seldom been able to separate the economic raison detre of the firm from the political and legislative influence of the home-country government. In contrast with other countries, the evolution of international trade intermediaries in Hong Kong has not been directly influenced by government policy.
Ellis, P.D. (2001), “Adaptive strategies of trading companies,” International Business
Review, 10: 235-259.
Adaptive Strategies of Trading Companies
Paul Ellis
1
Associate Professor
Department of Business Studies, Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong
Kong, Tel: (852) 2766 7108; Fax: (852) 2765 0611
Email: [email protected]
15 December 2000
Abstract
Past research into the development of trading companies has seldom been able to separate the
economic raison d’être of the firm from the political and legislative influence of the home-country
government. In contrast with other countries, the evolution of international trade intermediaries in
Hong Kong has not been directly influenced by government policy. In the absence of legislative
inducements or protection, Hong Kong traders exist purely by their ability to respond to market
forces, suggesting that important insights regarding organizational survival can be gleaned from
studying their patterns of behavior. In this paper, an evolutionary model is proposed and then
evaluated against data collected from trading companies at five different stages of development.
Specifically, these firms’ adaptive strategies are assessed in terms of their patterns of diversification
along three dimensions: product, geographic, and function. For scholars, the findings may be
considered a stepping stone towards the articulation of a truly generic model of trading company
evolution, while the value of the study for policy-makers in search of new archetypes lies in the
identification of a variety of market-responsive organizational forms.
Key words: trading company, adaptive strategy
1
The author gratefully acknowledges the constructive advice given by Wing-chor Chu of the Hong
Kong Trade Development Council, Mimi Yeung of the Hong Kong Exporters’ Association, and two
anonymous IBR reviewers. The author also wishes to acknowledge his debt of gratitude to his two
research assistants, Wong Hok Ying and Lai Wai Man. The research reported in this paper was
funded by the Hong Kong Polytechnic University (Project No.351/384).
Adaptive Strategies of Trading Companies
2
Introduction
It is ironic that despite more than three centuries of active participation in international trade, the
existence of the ubiquitous trading company is now being threatened by the globalization of markets.
Technological developments in communication, transportation, and banking, combined with the
ongoing internationalization of manufacturers and customers, have collectively led to the reallocation
of many of the traditional distribution functions performed by international trade intermediaries. This
has prompted some scholars to label modern trading companies, and in particular large-scale general
trading companies (GTCs), as “superfluous”, “inhibiting to the export of anything but the most simple
manufactured products”, and generally having “outlived their usefulness” (e.g., Wortzel and Wortzel
1983, pp.74-5). Yet, despite the economic pressures for their elimination, trading companies, for the
moment, continue to prosper at the interstices of global markets.
The research aims of this paper concern the identification of those adaptive strategies that ensure the
survival of the modern trading company in all its many guises. This goal is somewhat frustrated,
however, by the fact that many of the world’s highly diversified trade intermediaries owe their
existence to the extra-market intervention of the state (Balabanis and Baker 1993a). One classic
example is the symbiotic relationship between the policies of the Meiji Government and the
subsequent rise of the J apanese zaibatsu and sogo shosha (Kojima and Ozawa 1984; Sarathy 1985;
Yamamura 1976). Short of direct state sponsorship, governments may influence the development of
trading companies via inducive or coercive means. Inducements range in degree of intervention from
the relaxation of antitrust legislation (e.g., the US Export Trading Companies Act (1982)) to the
payment of tax rebates and direct subsidies. In contrast, coercive measures may include the
mandatory requirement to satisfy a number of criteria in order to receive GTC status along with the
commensurate benefits. For example, in 1975 the Korean government outlined six prerequisites for
GTC designation which included maintaining ten or more overseas branch offices and exporting
goods worth upwards of $1m to at least ten countries each per year (Cho 1984). A similar scheme was
adopted by the Spanish government in 1988. In this case, registered trading companies were required
to have annual sales in excess of 1,000m pesetas which included exports of at least 600m pesetas
worth of local products (Balabanis and Baker 1993a). Indeed, whether the motives are economic or
political, the hand of government is evident throughout the entire history of the trading company,
beginning with the monopoly charters assigned by the mercantilist governments of England and
Holland, to the original trading houses in the 17
th
century (Carlos 1992; Carlos and Nicholas 1988).
An exception to this interventionist rule is to be found in the case of the world’s most laissez-faire
society, Hong Kong. Unlike that of their Asian, American, and European counterparts, the evolution
of Hong Kong trading companies has not been influenced by government policy. In Hong Kong there
is no Export Trading Companies Act, no Ministry of International Trade and Industry, and no Five-
Year Plan to direct the growth of trade intermediaries. For the vast majority of Hong Kong’s import-
export companies, long-term survival is determined purely by market forces. Hong Kong’s system of
free enterprise, combined with a local trading environment that is particularly sensitive to external
shocks, suggests that important lessons on organizational survival can be gleaned from studying the
adaptive behavior of Hong Kong trading companies.
The premise of this paper follows that of Balabanis and Baker (1993a/b) who argue that any
understanding of the mechanisms by which trading companies adapt their strategies and structure
must begin with the identification of the patterns of diversification exhibited by traders at different
stages of development. “The dominant discriminating feature of the GTC concept is their diversity.
Thus, diversification theory can provide a more rigorous explanation of the development process of
GTCs” (Balabanis and Baker 1993a, p.47). The paper is divided into four sections. In the following
section the eclectic literature relating to the trading company concept is briefly reviewed. In
particular, Kim’s (1986) life cycle model is introduced and discussed as a basis for understanding
GTC development. Next, and by way of providing a backdrop for the analysis of their adaptive
strategies, a brief historical outline of the trading environment in which Hong Kong trading
companies compete is provided. Third, the methodology used to survey 1,890 traders is described,
and finally the study’s main research findings are presented and discussed.
Adaptive Strategies of Trading Companies
3
Background
Trading companies are characterized by their specialization in market-making intermediation, an
activity which may involve brokerage (selling on behalf of another) or reselling (taking title to the
goods traded) (Casson 1998). The study of trading companies is arguably more multinational in flavor
than any other topic in the field of international business. Trading companies are chauvinistic entities
whose missions, organizational structures, and internal cultures reflect the idiosyncratic political and
economic conditions of their countries of origin. As a case in point, consider the limited success with
which the uniquely J apanese sogo shosha concept has been transplanted to other countries (Hvala,
Perry, and Boddewyn 1990). Similarly, research into the topic is as context-dependent as the trading
company phenomenon itself, with many studies limited to a particular national setting (Table 1). The
result is an eclectic body of literature that is characterized by a dearth of widely applicable conceptual
schemas and typologies, and which offers limited scope for generalization (Amine 1987).
INSERT TABLE 1 ABOUT HERE
Much of the existing research on trading companies is highly particularistic in nature, focusing on
specific trading issues such as the association between different channel structures and the marketing
practices of exporter traders (Bello and Williamson 1985a), or the effect of government policies on
the performance of intermediaries (Howard 1995; Kaikati 1984; Williams and Baliga 1983). Some
attempts have been made to classify and categorize the subject matter (e.g., Barovick and Anderson
1992; Brasch 1978), but the systematic advance of research has generally been hampered by the lack
of consensus on matters of definition and limited attempts at theory-building (Amine 1987; Cho
1987). In particular, much of the research is utilitarian in purpose, providing little more than
“snapshots” of the changing roles and functions of trading companies at particular points in time (e.g.,
Brasch 1978; Kelly and Lecraw 1985; Wichmann 1997). To date there has been little progress in
developing a general framework of trading company evolution, and of the dynamic studies which do
exist, most reflect the unique characteristics associated with a particular type of trade intermediary
and cannot be applied universally (Perry 1990). For example, there are a number of longitudinal case
studies which trace the rise of the early European trading houses (Buchan 1994; Sugiyama 1978);
several growth models of the distinctive J apanese GTC have been proposed (Cho 1984; Kojima and
Ozawa 1984), and at least two scholars have developed frameworks for analyzing the adaptive
strategies of US export trading companies (Amine 1987; Perry 1990). Very few growth models exist
which are multinational in scope. Balabanis and Baker (1993b) developed a dynamic framework
based on those external and internal factors that lead to organizational change. To test their model,
they studied the “change intentions” of 29 European trading companies in terms of their geographical
expansion, product diversification, and functional and service development. These growth measures
reflect the widely-held view that the development of a trading company is evidenced in terms of
specific patterns of diversification which are brought about in response to changes in market
conditions (Balabanis and Baker 1993a; Cho 1987; Jones 1996). Pioneering conceptual work in this
area was done by Kim (1986), who was the first to conceive that the development of the trading
company can be conceptualized in terms of a number of distinct and sequential stages that
collectively define a GTC life cycle.
The General Trading Company Life Cycle
The term “trading company” encompasses a wide variety of trade intermediaries ranging in size and
scope from the small-scale import/export company to the mammoth and prototypical sogo shosha.
Because the generic concept of the GTC is no longer a strictly J apanese phenomenon, Kim (1986)
offers a three-stage explanation for these observed differences in levels of organizational
development that is based on economic and institutional factors rather than cultural differences. The
Stage I trading company is relevant to an emerging economy characterized by an underdeveloped
infrastructure, standardized manufactured goods, and fragmented channels of distribution. In addition,
the industrial base may be dominated by small- to medium-sized enterprises (SMEs) requiring the
assistance of an intermediary to go abroad. These contextual conditions create a market opportunity
for purchasing and sales agents to generate transaction efficiencies by specializing in the trading
function and by developing a global market information network. As well as lowering the costs of
information acquisition, the trading company may also be able to provide economies of scale for local
Adaptive Strategies of Trading Companies
4
firms in terms of sorting, grading and packaging standardized products, and by routinizing some
frequently-repeated transactions.
As the trading company engages in some product and area diversification to spread its risk, it
approaches the second stage of development which is characterized by the performance of a number
of trade-supporting activities, such as the provision of credit. Trade financing may be warranted if
local manufacturers have only limited access to affordable funds and if local lending institutions take
the view that local export agents are better placed to evaluate the credit risk of their trading partners
(J ETRO 1976). Political factors may also play a role at this stage if there is a perceived need to
improve the efficiency of trading activities, possibly with a view to replicating the well-publicized
success of the sogo shosha (e.g., as in the case of the thirteen GTCs created in Korea, a developing
country, in 1975), or if there is a desire to create organizations which bridge the gap between local
SMEs and export markets (e.g., as in the case of an industrialized country like the US, with its Export
Trading Company Act of 1982). As competitive pressures continue to increase, Kim (1986) notes that
the Stage II trading company is faced with two choices: (1) it can continue to focus on standardized
products (provided the firm has preferential access to low-cost supplies or captive clients), or (2) it
can trade up to higher-technology products (which offer good growth prospects and fatter margins).
The trading company enters the third stage of its life cycle when it engages in foreign direct
investment to protect its market position from the threat of internationalizing local manufacturers or
rising protectionist forces. In addition to the growth in direct exporting among manufacturers, the
competitive position of the trading company may be further undermined by the provision of trade-
supporting services by banks and government agencies. Although trading remains the principal
activity of the firm, its relative importance within the firm decreases to make room for related
functions such as shipping, warehousing, and manufacturing. For example, Wortzel and Wortzel
(1983) observe that the J apanese sogo shosha, which exemplify Stage III GTCs, play a fairly passive
role as export marketers taking title to goods primarily to expedite their financial and physical
distribution functions. Functional diversification into complementary areas may also be prompted by
a lack of local infrastructure or entrepreneurship. Further, there may be a need for the trading
company to internalize some trade-related activities to maintain quality control, preserve information
asymmetries, or pursue scope economies (J ones 1996). At this advanced stage of development, any
government mandates are more likely to be regulatory than strategic in nature.
Although Kim’s (1986) life cycle model implies that, given the right conditions, trading companies
ultimately evolve to become analogues of the sogo shosha, trading company development is by no
means deterministic. Kim concedes that life cycle stages may be by-passed, or they may occur
simultaneously rather than sequentially, while some firms may never progress beyond the first or
second stages of development. The actual evolutionary paths taken reflect strategic choice which, in
turn, is influenced by conditions within the host environment. While the life cycle model provides a
general basis for understanding those institutional factors germane to the development of GTCs, it
makes no attempt to consider those diversification strategies which may lead the firm away from
trading altogether. The implicit assumption is that trading companies never grow to become anything
other than highly diversified trading companies. In reality, adaptive parameters are not so narrowly
defined.
If the aim is to understand the adaptive strategies or developmental paths exhibited by trading
companies over time, any comprehensive explanation must take into account those non-trading
diversification alternatives which are available to the growing trading company. For example, and
despite the implied suggestion of the life cycle model, a Stage I import/export company is not
compelled by environmental forces to “evolve” into a Stage II trading company. Other diversification
options exist and perhaps chief among these is the option to vertically integrate into manufacturing
(Amine 1987). Support for this assertion comes from Cho’s (1987) study of GTC diversification
patterns. Cho was arguably the first to recognize that small traders may diversify in one of two
directions: forwards or backwards. Forward integration implies the provision of additional services
such as transportation, warehousing and insurance (i.e., progression to the next stage of GTC
development), whereas backward integration implies diversification into manufacturing, possibly to
counter the threat of opportunism in the channel. Such a move assumes that the trading company is
Adaptive Strategies of Trading Companies
5
able to leverage its knowledge of foreign markets and distribution channels into a position of
competitive parity on the production function. It is plausible that for many small trading firms
confronted with the need to diversify, investment into manufacturing may be too costly, prompting
either a hybrid arrangement (e.g., contract manufacture) or the further expansion of the firm’s traded
product lines and market coverage (i.e., diversification into a Stage II trading company). The
manufacturing option identified by Cho (1987) is arguably more suited to Stage II traders who, by
virtue of their greater size and experience, are more likely to have access to the capital necessary for
such an investment.
Another evolutionary outcome not captured in the life cycle model is observed when the general
trading company’s principal activities begin to expand into areas other than trading. Such a change in
strategic direction might result from diminished opportunities for trade intermediation as the home
economy matures, local manufacturers become more adept at international marketing, and better
investment opportunities emerge in brokerage- or service-related industries such as construction,
finance, and property development. This has been the experience of the sogo shosha as described by
The Economist (1991):
To defend their markets they have gone upstream (into mining and manufacturing) and downstream
(into retailing) through equity investments in joint ventures and affiliates. The sogo shosha are
changing from pure traders, where high-volume low-margin commission business dominates, to
become more financially sophisticated investment holding companies.
The tendency for GTCs to diversify away from trading has been evident from the time of the earliest
trading companies, whose investments in foreign production made them forerunners of the modern
multinational enterprise (MNE) (Carlos and Nicholas 1988). Whether the trading company is
investing in foreign rubber plantations, like the English firm Harrisons & Crosfield (J ones 1996), or
car manufacturing, like Korea’s Hyundai Group (Amine 1987), MNE status is achieved when the
revenue earned from offshore investments exceeds the income derived purely from trading. Although
the switch from GTC to MNE is by no means inevitable (Cho 1987), environmental forces may
provide a compelling incentive for both Stage II and III traders to engage in a significant level of
foreign direct investment to hedge against the risks associated with trading. Kim (1986)
acknowledged as much by defining his Stage III firm as “toward a conglomerate”, but he saw no
reason to incorporate this path of diversification in his model of GTC development. Despite the
omission of GTCs-turned-MNE conglomerates from Kim’s framework, it is worth noting that the
trading history of such firms may have a residual bearing on both the corporate culture and the
direction of investment activities. This is particularly evident in the case of Hong Kong, where firms
like J ardine Matheson and Swire Pacific are still referred to as hongs despite relegating their trading
functions to fairly minor, albeit active, subsidiaries many years ago. Indeed, for many of these former
trading houses (e.g., Hutchison), the current core business activity information (in this case
telecommunications and port facilities) stems directly from the firm’s early involvement in the
international flow of goods. For others such as J ardines, investment decisions in the growing market
of the People’s Republic of China (PRC) are overshadowed by the tenor of their past trading relations
with the mainland.
In summary, trading companies must adapt to meet the changing export requirements of their host
environments. At different points, in time traders may be confronted with one of two choices: either
(1) continue to exploit their core competence as market intermediaries and specialize or diversify to
improve the efficiency and effectiveness of this service, or (2) functionally diversify to the point
where trading is no longer the core business activity of the firm. The first option, played to its logical
conclusion, will lead to the development of the diversified GTC, while the second option is more
open-ended. However, given the strategic constraints imposed by available skills and resources, it is
likely that trading companies that engage in this latter type of diversification will ultimately become
either manufacturer-exporters (if some form of backward integration is pursued) or MNEs (if offshore
investments come to dominate the business activity of the firm). The main adaptive possibilities open
to trading companies are depicted in Figure 1. In contrast with the simple linear progression of the life
cycle model (Kim 1986), the figure proposed in this paper allows for six types of adaptation leading
to four distinct kinds of organization, each characterized by a particular pattern of diversification and
each having a different trading emphasis.
Adaptive Strategies of Trading Companies
6
INSERT FIGURE 1 ABOUT HERE
In terms of the nature and pace of their organizational development, trading companies, perhaps more
than any other international business entity, are influenced by the particular economic and regulatory
conditions of their home environments. Consequently, before the empirical relevance of Figure 1 is
assessed, it is first necessary to provide some background detail regarding the context in which Hong
Kong trading companies have evolved.
The Hong Kong Trading Environment
Since the earliest collection of trade statistics, it has been evident that Hong Kong’s trading
environment is characterized by its volatility. As the world’s tenth largest trading territory, Hong
Kong has not suffered from a protracted recession, but shocks and disruptions have been frequent and
disruptive, as evidenced by the magnitude of annual changes in trade growth (Figure 2). In terms of
the proportion of trade to gross domestic product, Hong Kong is the world’s second most open
economy after Singapore (Ellis 1998b), which means that the local trading environment is especially
sensitive to external shocks such as China’s Great Leap Forward (1959), the OPEC price hikes
(1973), and the Asian currency crisis (1997). On occasion the nature of an external disruption is such
that organizational survival is threatened compelling trading companies to adapt. This can be seen in
the case of at least four external jolts (Ellis 1998a):
1. the invasion of J apanese forces in 1941 (which abruptly ended Hong Kong’s entrepÔt role,
the basis for much of its economic activity at the time),
2. the J apanese surrender in 1945 (which precipitated the swift resumption of the re-export trade
with southern China),
3. the Korean War and the resulting United Nations embargo on trade with the PRC in 1951
(which led to the cessation of all entrepÔt trade virtually overnight and triggered the
industrialization of Hong Kong), and
4. the opening of China’s markets in 1979 (which stimulated the mass migration of Hong Kong
industry across the border).
INSERT FIGURE 2 ABOUT HERE
The effect of these environmental disruptions on Hong Kong’s population of trading enterprises can
be summarized as follows: the thousands of import/export businesses that had ceased trading after the
J apanese invasion in 1941 flourished again in the late-1940s on the back of the colony’s resurgent
entrepÔt trade. These firms were then obliged in 1951 to replace their mainland Chinese suppliers
with local suppliers or go out of business (Woronoff 1980). Subsequent competition for international
markets from local manufacturers of clothes, toys, and watches then compelled trading companies to
reduce commissions, engage in further product and geographic diversification, and begin investing
directly abroad. Many traders simply decided to capitalize on the swelling supply of cheap local labor
and go into manufacturing for themselves, with the result that the value of Hong Kong-made exports
surpassed re-exports for the first time in the colony’s history in 1959 (Wilson 1990). In the 1980s,
economic factors (rising rents and wages) combined with a political thaw in the PRC stimulated the
resumption of Hong Kong’s entrepÔt role, and once again re-exports replaced domestic exports as the
main form of international trade (Ellis 1999). By the mid-1990s, more than one fifth of the local
workforce, around half a million people, was employed in import/export enterprises
(http://www.tdc.org.hk/).
In the past six decades, environmental disruptions have qualitatively altered the rules of the game for
Hong Kong traders on several occasions. The net result is a trading landscape populated by a variety
of organizational forms, each of which represents an entrepreneurial response to changing market
signals rather than an induced response to political directives. To date very little research has been
done on this unique group of 100,000 or so trading companies. Little is known about their patterns of
diversification, the nature and direction of their trading activities, or their performance as middlemen
in international distribution channels. The methodology used to address these questions is described
in the following section.
Adaptive Strategies of Trading Companies
7
Methodology
Questionnaire Design
In order to gain some insight into the adaptive strategies enacted by Hong Kong traders, a four-page
questionnaire was administered to a sample of 1,890 firms. The questionnaire consisted of several
items relating to the personal characteristics of the respondent, 19 descriptive items regarding the
extent of diversification and the nature of the firm’s trading activities, and 11 objective and subjective
measures of the firm’s performance over the past three years. To ascertain the level of product, area,
and functional diversification, operational measures were adapted from the indices used in Cho’s
(1987) analysis of the GTC. An area diversification index (ADI) was calculated in two ways: (1) the
number of countries actively traded with at the time of data collection, and (2) the proportion of sales
income earned outside of Hong Kong and the PRC. Functional diversification (FDI) was ascertained
by identifying the number of support services offered as selected from a list of 26 alternatives.
Product diversification (PDI) proved to be the most troublesome variable to measure. Following
Cho’s (1987) procedure, respondents were initially asked to identify the number of two-digit Standard
International Trade Classification (SITC) categories in their product ranges and then provide
information regarding the proportion of sales earned in each category. However, several iterative
pretests of the questionnaire among a group of six traders revealed that this item was hopelessly
intractable and difficult for respondents to answer. In the end, a list of 24 product categories was
provided that approximated the two-digit level of aggregation with a question asking the respondent
to indicate the percentage of the firm’s sales, if any, earned from each category. An index was then
calculated by multiplying the number of product lines by the proportion of sales generated outside the
principle product. For example, if a respondent checked three product categories indicating relative
income shares of 80, 15, and 5 per cent, then the relevant PDI is 0.60 (or three times 0.20).
Sample Selection
The sample was defined as the population of indigenous trading companies listed in two recently
published directories (Registry of Hong Kong Traders 1997; Members’ Directory 1997). The
questionnaire was mailed twice and then faxed personally to the most senior executive identified in
the directories. Although the use of self-report data is not without its limitations (see for example,
Ganster, Hennessy and Luthans, 1993), the nature of the inquiry justified the reliance on subjective
measures. First, it can be assumed that traders' perceptions of their environment will affect their
strategy-making, irrespective of whether those perceptions are valid or not (Anderson and Paine
1975; Miller 1988). Traders who are not "clued in" to their multiple operating environments will not
make the appropriate adjustments when those environments change. Thus, trader experience or
survival should be correlated with a track record of accurate "sense-making" of environmental
change. Second, past research has indicated that relying on a knowledgeable "informant's perceptions
of an organisation's strategic orientation is a valid approach to measuring strategy" (Shortell and
Zajac 1990:829). Finally, subjective performance assessments are suitable in situations where "(1)
accurate objective measures are unavailable and (2) the alternative is to remove the consideration of
performance from the research design" (Dess and Robinson 1984:271). Traders are universally
secretive by nature and are indeed particularly so when it comes to measuring their performance (see
for example, Guex 1998). This is no less true in Hong Kong, where owner-operators are the sole
gatekeepers to such information, and the vast majority (more than 95 per cent) of firms are privately-
held entities.
At the close of data collection, 47 questionnaires had been returned uncompleted because the
respondent did not consider their firm to be significantly involved in the trading (i.e., import/export)
business. A further 35 questionnaires were discarded from the database because the respondents
reported that their headquarters were located outside of Hong Kong. The large number of ineligible
responses indicated that the criteria for inclusion in the published directories were somewhat broader
than the sampling frame used in this study. In addition to the ineligibles, 300 questionnaires were
returned undelivered, reflecting both the attrition rate of the recent Asian economic crisis and the
footloose nature of small import/export companies. In all, 211 useable questionnaires were received.
The effective response rate was calculated based on the formula advocated by the Council of
American Survey Research Organizations (CASRO) for surveys with single-stage sampling and an
Adaptive Strategies of Trading Companies
8
eligibility requirement (Churchill 1995; Burns and Bush 2000). The critical variable in determining
the response rate in a study of this nature is estimating the number of eligible firms receiving the
questionnaire. There were two methods for calculating this figure. First, of those who made the effort
to return the questionnaire, only 72 per cent were eligible for inclusion in the study (i.e., they were
significantly involved in trading and had their head office in Hong Kong). Second, follow-up phone
calls made to a random subsample of 150 listed traders revealed that 24 per cent of the surveyed firms
were ineligible for inclusion for one of the two reasons mentioned. Using the more conservative
estimate of 24 per cent indicates that 454 firms receiving the questionnaire did not meet the sampling
requirements of the study. The resulting response rate of 35.0 per cent was considered satisfactory
given the performance of similar surveys done in Hong Kong where response rates can be less than
ten per cent (e.g., Chan and Ellis 1998). Nevertheless, to test for possible non-response bias,
descriptive data collected during the follow-up phone-calls were compared with information provided
by respondents. No significant differences between the two samples were found in terms of the
number of years trading (t=.31, p=.75), the number of local (t=1.18, p=.24), mainland Chinese
(t=1.02, p=.31), and foreign employees (t=1.07, p=.28), and the proportion of products sourced from
Hong Kong (t=.77, p=.44), China (t=1.36, p=.17), or elsewhere (t=-1.65, p=.10), suggesting that non-
response was not a source of bias in this study.
Analysis and Findings
Table 2 provides a summary of the general descriptive characteristics of the trading companies and
the respondents. More than 90 per cent of the firms in the sample reported more than five years’
trading activity. Similarly, more than four-fifths of the respondents themselves have been actively
involved in trading for more than five years, indicating their experience and suitability for completing
the questionnaire. A wide range of industries is represented by the traders, as indicated by the
assortment of product lines carried (Table 3). Chief among these are exports of clothing and electrical
goods, two product categories which have been the mainstay of Hong Kong’s economic development.
INSERT TABLES 2 & 3 ABOUT HERE
Classifying the Traders
As a starting point for the analysis, respondents were classified according to the five trader-types
identified in Figure 1. Although all the firms in the survey were defined as trading companies by
virtue of their inclusion in a commercially-produced directory of traders, to qualify as a trading
company for the purposes of this study, the products traded had to come primarily from sources
external to the firm. Consequently, any firm manufacturing 50 per cent or more of their traded
products was deemed to be an exporter, and any exporter with significant foreign direct investment,
as proxied by the employment of overseas workers (outside Hong Kong and China), was labeled an
MNE. This classification choice led to the identification of 35 manufacturer-exporters and 16 MNEs
in the sample. The surrogate indicator of foreign employees was also used to determine those traders
with significant foreign investments, resulting in the identification of 28 Stage III trading companies.
The remaining traders were split into two groups by looking at the specific trade-supporting services
used to calculate the degree of functional diversification. According to Kim (1986), Stage II traders
can be distinguished from their Stage I counterparts by their provision of financial resources to
trading partners. In the survey, 41 respondents indicated that they offer trade financing to clients, thus
deeming themselves to be Stage II traders and leaving 91 Stage I traders.
Having identified the five types of trading company, ANOVAs were conducted to ascertain
differences among the main descriptive variables (Table 4). The findings generally support the view
that, as trading companies evolve, they tend to exhibit increasing levels of product, geographic and
functional diversification (Balabanis and Baker 1993a). As expected, Stage I traders were the
youngest trading companies in the sample and reported lower levels of diversification on all measures
except the number of foreign markets, which was similar across all three stages of development. As a
group, the traders, and particularly the more experienced traders, evidenced higher levels of product
diversification than the exporters (i.e., the manufacturers and MNEs), perhaps indicating their greater
need to safeguard against the threat of opportunism. For example, manufacturer-exporters run little
risk of being supplanted by suppliers if they are responsible for their own production. In contrast,
Adaptive Strategies of Trading Companies
9
traders need to carry a wider variety of product lines to minimize their dependence on any one
supplier. Interestingly, the exporters compensate for their relatively low levels of product
diversification by leveraging their product expertise across a larger number of foreign markets. The
traders in the sample, irrespective of their stage of development, were on average involved in fewer
than seven foreign markets. In contrast, the manufacturer-exporters were active in almost ten foreign
markets, while the MNEs surveyed were active in twice as many markets as the traders, as befits their
multinational designation.
INSERT TABLE 4 ABOUT HERE
Other significant differences observed include firm size, with the advanced levels of development
being associated with an increase in both the scale and dispersion of operations. In general, Stage II
traders are significantly larger and employ more workers in China than Stage I traders, while Stage III
traders have the largest and, by definition, most regionally-dispersed workforce of all trading groups.
Unsurprisingly, the largest companies in the sample are those responsible for their own production
(manufacturer-exporters and MNEs). This group of firms has evidently capitalized on the supply of
cheap labor in the mainland, as indicated by their very large numbers of PRC employees.
Respondents were asked to estimate the proportion of total income derived from trading activities.
Again consistent with the literature (Kim 1986), the findings show that as a trader becomes more
diversified and develops into a GTC (i.e., Stage II ? Stage III), the relative importance of trading
activities within the firm’s overall service offering diminishes (Table 4). Nevertheless, trading
remains the dominant activity of the firm accounting for more than four-fifths of the income earned
by all trading groups. The findings also reveal that the two exporter groups have significantly lower
trading shares than the trading companies (75% for manufacturer-exporters and 53% for the MNEs),
further indicating that the classification schemas used were sufficiently discriminating.
The Nature and Direction of Trading Activities
Hong Kong’s entrepÔt role as the preeminent service provider to Guangdong province is clearly
evident in the pattern of trade reported by respondent firms (Table 5). Although respondents were
able to select from six distinct types of trading activity (imports, re-exports, exports, transshipments,
third-country trade, and wholesale/retail trade), Stage I and II traders were generally biased towards
imports/re-exports, with China being the dominant supplier for the latter group. Stage I traders were
also heavily involved in local distribution (wholesale and retail trade), more so in fact than any other
group in the sample. At the other end of the scale, the Stage III traders reported the lowest levels of
domestic trade (6%) and the highest levels of third-country trade (24%) where goods are shipped
direct to the market without stopping in Hong Kong. This finding is significant as functional
expansion into third-country trade is a hallmark of a mature trading company and is seldom
undertaken by novice traders (Balabanis and Baker 1993b). With their regionally-dispersed
workforces, Stage III traders source nearly three quarters of their products from outside of Hong
Kong and China. Of the two self-manufacturing groups, the MNEs also recorded high levels of
offshore trade (15%) in comparison with the more localized manufacturer-exporters (1%). Consistent
with their high levels of employment in China, the manufacturer-exporters had by far the highest
levels of re-exports in the sample.
Finally, there was a readily-observable association between the level of development of the trading
company and the overall size and regional dispersion of their supplier networks. Mature Stage III
traders were the least likely group to source from local suppliers and were reliant on foreign factories
for 70 per cent of their business. As expected, MNEs also had larger and more dispersed networks of
suppliers than manufacturer-exporters. This finding suggests that the social capital inherent within a
trader’s list of contacts takes time to develop. Indeed, organizational development may itself be
constrained by the boundary horizon of the trading network (Balabanis and Baker 1993a; Rauch
1996).
INSERT TABLE 5 ABOUT HERE
Adaptive Strategies of Trading Companies
10
Performance Comparisons
Finally, the five groups of firms were compared in terms of their trading margins and operating
performance (Table 6). MANOVA was used to assess differences across three composite
performance measures; sales growth and profitability over each of the last three years, and the
respondents’ own subjective appraisal of four performance measures. Although the Stage III traders
and MNEs generally enjoyed superior performance across the board, none of the observed differences
are significant at the p