Description
This paper examines the issues surrounding the aborted attempt by the International Accounting Standards Board
in early 2005 to regulate the accounting for the European Union’s new Emissions Trading Scheme under the Kyoto
Protocol. The paper argues that the features that made the trading scheme attractive to governments were precisely
the ones that created difficulties for accountants to capture under existing standards. After showing why the challenge
has to be faced, the paper suggests a possible way forward that the IASB might consider when it revisits the subject, as it
is expected to do in the near future.
Emission rights: From costless activity to market operations
Allan Cook
*
,1
c/o International Accounting Standards Board, 30 Cannon Street, London EC4M 6XH, United Kingdom
Abstract
This paper examines the issues surrounding the aborted attempt by the International Accounting Standards Board
in early 2005 to regulate the accounting for the European Union’s new Emissions Trading Scheme under the Kyoto
Protocol. The paper argues that the features that made the trading scheme attractive to governments were precisely
the ones that created di?culties for accountants to capture under existing standards. After showing why the challenge
has to be faced, the paper suggests a possible way forward that the IASB might consider when it revisits the subject, as it
is expected to do in the near future.
Ó 2008 Elsevier Ltd. All rights reserved.
Background – an aborted interpretation
While Europe was preparing for the adoption
of International Accounting Standards (IASs) by
all listed companies from January 2005, the
International Accounting Standards Board
(IASB) was urged to develop mandatory guid-
ance for the ?nancial reporting of Emission
Rights, the basis for the European Union’s
(EU’s) new scheme designed to reduce emissions
by major polluters.
In view of the specialised nature of the subject,
the IASB asked its International Financial Report-
ing Interpretations Committee (IFRIC) to under-
0361-3682/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2007.12.001
*
Tel.: +44 20 8399 4368.
E-mail address: [email protected]
1
Allan Cook was the co-ordinator for the International Financial Reporting Interpretations Committee (IFRIC) of the
International Accounting Standards Board (IASB) at the time that the issues described in the paper came to a head. He is particularly
grateful to Henry Rees, the IASB senior project manager responsible for the topic, for permission to draw on his work for the IFRIC
and the IASB and to Professor Donald Mackenzie for his advice on the relevance of existing papers dealing with the economic
implications of various possible mechanisms for reducing emissions. However, the views expressed are the author’s own and do not
necessarily re?ect those of either of these gentlemen or the members of the IASB or the IFRIC.
Available online at www.sciencedirect.com
Accounting, Organizations and Society 34 (2009) 456–468
www.elsevier.com/locate/aos
take the task. The IFRICpublished a draft interpre-
tation in May 2003 and, after considering public
comments, developed an interpretation, which the
IASB issued in December 2004, just within the
European Commission’s deadline for the adoption
of IASs.
The result was a public outcry. Companies com-
plained that application of the interpretation
would force them into showing a completely dis-
torted picture of their performance in their annual
and interim ?nancial statements. The IASB, while
recognising that the IFRIC had made a valid inter-
pretation of the relevant IASs, accepted that the
end result was confusing in certain respects. Per-
haps fortunately, the expected market for Emis-
sion Rights was slow to develop and the IASB
took the opportunity to withdraw the interpreta-
tion in June 2005, only six months after it had been
issued.
What was the cause of this reverse?
The problem
The Emission Rights a?air is worth study-
ing because it illustrates the problems faced by
standard setters as they explore the frontiers of
accounting. Just as the science of medicine some-
times advances by examination of its failures to
deal with newly presenting diseases, so, in other
?elds, an understanding of the roots of failures
can lead to a fuller understanding of the discipline
involved, its limitations and its possibilities.
Three features in particular lay at the heart of
the challenge posed to standard setters as they
sought to deal with Emission Rights:
A previously costless activity had become
costly.
Governments mitigated the cost.
By means of marketable allowances.
It is the combination of these features that has
so far proved an insuperable problem for account-
ing to re?ect. Before examining the accountants’
problem, we need to understand what led govern-
ments to adopt such a scheme in their attempts to
reduce emissions.
Controlling emissions – alternative approaches
There is a voluminous literature, dating back
over 30 years, discussing the relative merits of dif-
ferent forms of economic instrument for achieving
government policies with respect to the use of a
scarce resource or service. The choice in a given
situation depends on many factors, including the
degree of uncertainty regarding the marginal costs
and bene?ts of a particular course.
Economic instruments divide broadly into those
that set a price for the good or service, leaving the
quantity demanded to adjust accordingly and those
that set a total quantity, leaving the price to be
determined by market forces. In the ?eld of emis-
sions an example of a price instrument is a carbon
tax: a previously costless, or less costly, activity
becomes costly and the cost is borne in proportion
to the level of emissions involved. An example of a
quantitative instrument is a system under which
government issues a predetermined total of licences
to emit. If the licences are auctioned rather than
issued free and emitters are allowed to trade them,
it can be shown that under idealised conditions
either instrument will result in the same optimal
price and quantity. Note that an ability to trade
licences is necessary for a quantitative system to
achieve equivalence with a price instrument.
Of course, conditions are rarely idealised nor is
it only economic factors that determine political
decisions by governments or inter-governmental
agreements. This paper does not attempt to evalu-
ate the relative merits of price versus quantitative
instruments in the control of emissions
2
. Rather,
the paper examines the accounting implications
of the quantitative systems that have been put in
place by governments. Although price instruments
also exist, they are not currently as widespread as
quantitative systems and, in any event, do not raise
any new accounting issues.
2
An excellent, recent analysis of those issues is given by
Hepburn 2006. That article explains why some economists
believe that the relationships between the marginal costs and
bene?ts of alternative instruments to reduce emissions suggest
that, in theory, a price instrument may be more e?cient (i.e.
arrive closer to the optimum price and quantity) than a
quantitative instrument.
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 457
The principal quantitative instruments in use
for controlling emissions are ‘base line and credit’
and ‘cap and trade’.
Base line and credit
Under this system a government allocates
allowances based on some assessment of a normal
rate of emissions. For each entity a base line is set
below which no charge for emissions will be made.
Entities that manage to operate below the base line
for a period are given credits, which can be sold
through a market to those which exceed their base
line. At the end of the period any entity that has
exceeded its base line must deliver to the govern-
ment su?cient credits, bought in the market, to
cover the excess.
A weakness of this system is that allowances that
are traded may be insu?cient to sustain a market.
Moreover, the government may lay itself open to
charges of unfair discrimination in its determina-
tion of the base lines for di?erent entities.
Cap and trade
Under a cap and trade system the government
mitigates the cost for entities, not by setting a
base line, but by allocating tradable allowances
for the period. After the end of the period each
entity must pay for its emissions by surrendering
allowances granted or buying in from the market
any shortfall. Since the allowances are allocated
from the beginning of a period, in theory an entity
that perceives a market opportunity can sell more
than the amount of its expected savings in the hope
of buying back later at a more advantageous price.
This systemplaces greater reliance on the market
mechanism than does the base line and credit
method. Although initially all allowances are allo-
cated by governments for nil consideration, it is
intended, under the EU scheme, that a gradually
increasing proportion should be auctioned rather
than granted free. Some believe that an auction of
all allowances would be the most e?ective means
of reducing total emissions by ensuring that the sav-
ings that were easiest to make were addressed the
?rst. However, too rapid a move to the auction of
all allowances might be counter productive by oblig-
ing entities to pass on the cost to the consumer
instead of encouraging them to seek savings.
The EU scheme under the Kyoto protocol
The EU scheme covers all 25 member states and
was adopted as the principal means for the EU to
meet its emission reduction commitments under
the Kyoto protocol. The protocol set reduction tar-
gets in the form of ‘Allowed Amount Units’ of
emissions by country but allowed for countries to
group together to redistribute their total allowances
among themselves. Under the EU scheme, the
Union’s overall target of eight percent reduction
was redistributed by giving the UK, for example,
a target of 12.5% reduction but allowing Portugal
an increase of 27%. Under the EU scheme entities
may participate in the Kyoto Clean Development
Mechanism, which enables them to acquire further
allowances by undertaking a development to
reduce emissions in another country
3
.
The EU scheme has nearly completed an initial
period of three years, which will be succeeded by
the ?rst full ?ve year phase under the Kyoto proto-
col running from 1 January 2008 to 31 December
2012. For the initial phase, 95% of allowances
had to be allocated free to entities. During the fol-
lowing phase, only 90% must be free. The remain-
ing ten per cent may be auctioned, thus obliging
most entities to have recourse to the market, either
at the time of the auction or subsequently.
Allowances are allocated in January to cover
emissions up to the end of December; settlement
is required by the end of the following April. Enti-
ties that during the initial phase failed to deliver suf-
?cient allowances to cover their level of emissions
are ?ned €40 per tonne of shortfall. During the sec-
ond phase this is expected to rise to €100 per tonne.
Payment of the ?ne is not a substitute for delivery of
the required allowances. The shortfall must be
3
MacKenzie (2009) gives a good illustration of the scienti?c
imponderables that underlie ‘making things the same’ by using
allowances received in one industry to o?set the cost of
emissions in another. The paper stresses the important role of
accounting if market based schemes are to motivate successfully
by making carbon costs visible.
458 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
made up from the subsequent year’s allocation plus
market purchases.
Allowances are not themselves a licence to emit:
each entity covered by the scheme must obtain
from the government an emissions permit. If the
entity ceases production during the year, it must
surrender its emissions permit but may sell any
surplus allowances on the market.
All features of the scheme are focused on the
objective that somewhere in the world emissions
get reduced – even if not necessarily by the entity
whose expected emissions exceed its allocation of
allowances nor even within the jurisdiction of the
government determining that allocation. By allow-
ing an entity that ceases production to sell its allow-
ances, the scheme maintains the originally assessed
level of allowances for the economy as a whole. It is
not necessary to claw them back from the entity to
which they had been granted. Moreover, an e?-
cient producer may judge that it can reduce more
emissions – and earn more credits – by undertaking
a project to enable a less e?cient entity to reduce its
emissions than it could by investing the same level
of resources in reducing its own emissions. The
design of the scheme encourages it to take the for-
mer, more e?ective course.
The accountants’ problem rooted in government
objectives
From the above description we can see why the
three features causing the most problems for
accountants are particularly attractive from the
point of view of governments.
Emission reduction schemes are designed to
give a cost to a previously costless activity, in
order to motivate producers to regard emissions
as an input cost that must be monitored and
controlled like any other.
However, unlike licences for oil and gas explo-
ration or 3G development, emissions costs are
being introduced only slowly, by gradual reduc-
tion of the total level of allowances or the pro-
portion of the total that is allocated free. This
strategy avoids a shock to existing systems for
the production, supply and use of energy, which
might result in the new cost simply being passed
on to the consumer instead of being controlled
at the point where savings were possible.
The cap and trade system responds well to the
previous two objectives while also making most
e?ective use of the market. Use of the market
both simulates other real costs and is the most
e?cient means of passing the incentives to save
emissions to those best able to make such sav-
ings at an early date.
The remainder of this paper analyses the
accounting problems created by these very reason-
able governmental objectives.
One accounting solution: maintain the status quo
Many producers and others commenting on this
subject to the IASB asserted that the key to the
accounting solutionlay indealing only withthe mar-
ginal e?ect on cost. They argued that, as long as an
entity emitted no more than the amount covered
by its emissions allowances, no new cost emerged.
The only cost would be that of acquiring additional
allowances fromthe market: the only credit, the pro-
ceeds of a sale of surplus allowances to the market.
This solution relied on netting the bene?t of trad-
able allowances received against the newly created
cost of emissions, as though it were dealing with a
base line and credit scheme. It sought to simulate
as far as possible the status quo ante. The question
for accountants was whether it was true that the sta-
tus quo had not changed: whether in fact something
was being overlooked if income from the grant of
allowances was netted against the cost of emissions.
The circumstances in which costs may be netted
against income are, of course, severely limited both
under IASs and under the EU Fourth Directive.
The weakness of the netting solution is that it
ignores the market oriented thrust of cap and trade
schemes. Although beginning slowly, the market
for allowances seems likely to play an increasingly
important part in governments’ attempts to reduce
emissions both in the EU and internationally. The
netting solution attributes a cost to purchased
allowances but not to those granted to an entity
by government. Yet the allowances themselves
are fungible: neither the government nor the mar-
ket distinguishes allowances according to their
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 459
source. Allowances are assets in their own right, as
demonstrated by their ability to be sold for cash
and even to be transferred between di?erent
schemes under the Clean Development Mecha-
nism of the Kyoto protocol.
The netting solution also ignores the reality of
the liability for emissions, a liability that exists
and behaves quite independently from the allow-
ances held by the emitting entity. It is an under-
statement of that liability to net it against
allowances held. The netting of assets and liabili-
ties is restricted, if anything, even more severely
than the netting of income and costs.
Even as a practical expedient, the netting solu-
tion breaks down as soon as an entity begins to
trade its allowances. From that point the signi?-
cance of the initial grant disappears as granted
allowances are replaced by others bought in from
the market. At the extreme, the whole of the initial
grant may have been sold and replaced; on what
basis is the amount of o?set then to be calculated?
If the reply is that no o?set should be allowed in
such a case, there would be no comparability
between one entity that had retained its original
grant of allowances and one that had sold but
replaced them, even if there had been no signi?-
cant price movements in the interim. If the reply
is that o?set is to be allowed, some method would
have to be devised to limit the amount of the o?set
to the volume and unit price of allowances at the
time the grant was made. The reason is that any
variation of the current value of allowances from
the amount of the grant would be the result of
price movements or market transactions rather
than an adjustment of the original grant.
The issues for IFRIC
The IFRIC
4
approached the problem by analy-
sing separately the nature of the allowances
granted and the obligation to pay for emissions
by delivering allowances. It asked:
Are the allowances an asset
– if purchased?
– if allocated by government?
If so, what kind of asset
– Financial instrument? or
– intangible?
If the allowances are an asset, is the credit (the
grant)
– income or
– a liability on receipt?
Allowances
Most people would instinctively agree that pur-
chased allowances are an asset, probably because
they have a cost. However, the IASB Framework
for the Preparation and Presentation of Financial
Statements (‘the Framework’) does not require a
cost. It de?nes an asset as ‘a resource controlled by
the entity as a result of past events and from which
future economic bene?ts are expected to ?ow to
the entity’. The IFRIC found that allowances allo-
cated by government, no less than those purchased,
met this de?nition. They are controlled by the entity
from the moment of the past event – allocation – by
which they were acquired. They are expected to gen-
erate future bene?ts for the entity either when ten-
dered in settlement of an emissions liability or
through sale into the market. As noted above, the
rights attaching to them are indistinguishable from
those of purchased allowances, a point that is likely
to assume practical, as well as conceptual impor-
tance as markets develop and entities begin to trade
signi?cant portions of their allocations.
Although, as we shall see later, this conclusion
gave rise to uncomfortable consequences, it is di?-
cult to challenge it on conceptual grounds. How-
ever, in determining whether allowances were a
?nancial instrument or an intangible asset, the
IFRIC encountered a more arti?cial di?culty of
the IASB’s own making. It would have been conve-
nient if the allowances could have been classi?ed as
one of the classes of assets that IASs require to be re-
measured to fair value at each reporting date, with
gains andlosses reportedinpro?t or loss. The reason
is that under IAS 37 Provisions, Contingent Liabil-
ities and Contingent Assets the liability for emis-
sions must be measured and reported in that way.
4
This section summarises the requirements and rationale
contained in IFRIC3 Emission Rights. A fuller description of
the debate surrounding the ?nalisation of IFRIC3 is given by
Casamento (2005).
460 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
IASs require ?nancial instruments held for trading
to be on this basis but the IFRIC decided that the
allowances did not meet the, somewhat detailed,
de?nitionof a ?nancial instrument. Nor didthey fall
within the limited class of other assets that could be
measured like ?nancial instruments.
Instead, the IFRIC decided that the allowances
fell within the de?nition of an intangible asset.
Although such assets may be re-measured to fair
value if they can be traded in an active market,
IASs currently require the resulting gains and
losses to be reported outside pro?t or loss. Symme-
try of treatment with that of the liability therefore
eluded the IFRIC.
Emissions liability
The IFRIC decided that the obligation to pay for
emissions bydeliveringallowances tothe government
after the endof the periodwas a liability. It clearly fell
within the Framework’s de?nition, ‘a present obliga-
tion of the entity arising from past events, the settle-
ment of which is expected to result in an out?ow
from the entity of resources embodying economic
bene?ts’. Anobligationarose as emissions were made
– the past event – which would require the entity to
part in the future with resources – the allowances –
embodying economic bene?ts – the ability to be used
to settle further obligations or be sold in the market.
Some argued that even if the asset and liability
were distinct, they should be permitted to be o?set,
since the ultimate use of the allowances was to set-
tle the liability. However, as noted above, the
allowances would not necessarily be used to settle
that entity’s liability. Furthermore, the allowances
were not a debt owed by the government, which
might have quali?ed as o?settable against a liabil-
ity to the government.
Others argued that hedge accounting should be
permitted between the allowances and the liability.
Certainly, the existing rules for hedge accounting
would not permit that but, since hedge accounting
is essentially anarti?cial convention, this couldhave
been a route worth exploring. As with all forms of
hedge accounting, however, such a solution would
enable one entity to portray itself di?erently from
another entity in the same or a very similar eco-
nomic position. Some of the practical complexities
that would be involved in applying rules for such
hedge accounting have been referred to above.
Grant of allowances
If the allowances are recognised as an asset, it is
necessarytodetermine howthe correspondingcredit
for free allocation of allowances (the grant) should
be treated. The IFRIC decided that any allocation
below fair value fell within the de?nition of govern-
ment grants, essentially a government’s ‘transfers of
resources to an entity in return for past or future
compliance with certain conditions relating to the
operating activities of the entity’. It argued that the
obligation under an Emission Rights scheme to
reduce emissions or deliver allowances was a condi-
tion‘relating tothe operating activities of the entity’.
At the outset, the grant and the allowances are
measured at the fair value of the allowances at the
time the grant is made. As noted above, under
IASs the allowances may, but do not have to be
re-measured to fair value; the grant, however, is
not so re-measured.
Under the existing version of IAS 20 Govern-
ment Grants, income from the grant is recognised
gradually in line with recognition of the activity
to which the grant relates. For example, a grant
towards the purchase of a machine is recognised
in the income statement over the same period as
depreciation of the machine. In a paragraph of
the draft interpretation that attracted virtually no
comment, the IFRIC noted the IASB’s long-stand-
ing intention to propose amendment of IAS 20: the
IASB had already discussed publicly its proposal to
require income from a government grant to be
recognised over the period to the point at which
the grant could no longer be recalled. The rationale
for the change was that beyond that point no liabil-
ity existed: the whole grant therefore needed by
then to be recognised in income. The e?ect of such
a change would be that a grant of emissions allow-
ances would have to be recognised in income imme-
diately on its receipt at the start of the annual cycle.
E?ects of the IFRIC decisions
While each of the IFRIC’s decisions outlined
above was soundly based on existing IASs, which
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 461
in their turn, had stood the test of time, the com-
bined e?ect when applied to Emission Rights pro-
duced some very strange results. The various
measurement bases and recognition points in the
income statement are summarised in Table 1 and
the e?ect on the balance sheet is illustrated graphi-
cally in Fig. 1. For simplicity, the analysis that fol-
lows ignores the e?ects of market transactions.
Moreover, no attempt has been made to portray
the complications that arise when an entity’s ?nan-
cial year does not correspond with the annual cycle
for the allocation and settlement of allowances.
When allowances are received at the start of the
annual cycle, they are recognised at fair value,
which is the market price for allowances at that
date. Since the entity does not pay for allowances
received from the government, it recognises the
grant at the same amount as the allowances.
Unless and until IAS 20 is amended, the grant is
recognised initially as a liability. Pro?t or loss at
that point is therefore not a?ected.
As emissions are produced, a new liability
emerges for the amount that will have to be paid
to the government in the form of allowances after
the end of the annual cycle. In common with most
other liabilities, this emissions liability is re-mea-
sured to ‘current settlement value’, in this case
essentially fair value at each reporting date. The
grant is gradually transferred from liability status
to income to compensate for the emissions cost
incurred. However, unlike most other liabilities,
the liability for the grant is not re-measured for
price movements. The reason is that, as explained
above, it is not a true liability but in reality no
more than deferred income, which is being recog-
nised for convenience sake in line with the cost it
is designed to subsidise.
Various imbalances nowbegin to emerge. Even if
the entity continues throughout the annual cycle to
expect that its cumulative emissions will be exactly
covered by the grant of allowances, the e?ect of
price changes will generally prevent the diminishing
liability for the grant fromexactly compensating for
the growing liability for the cost of emissions. Fur-
thermore, the e?ect of price changes on the asset
side of the balance sheet di?ers from that on the lia-
bilities side. If the entity chooses not to fair value the
allowances, as it is permitted to do, there will natu-
rally be no equivalent price movements on the assets
side. However, even if the allowances are re-mea-
sured, the e?ect of price changes during the year dif-
fers from that a?ecting the liability. The reason is
that a full year’s allowances are held by the entity
fromthe start of the annual cycle, whereas the liabil-
ity for emissions builds up only gradually. It is
therefore only at the end of the annual cycle that
an entity that fair values its allowances will ?nd that
the price movements on the allowances broadly
compensate for those on the liability. If the entity
measures its allowances at historical cost, they will
only be adjusted for cumulative price changes at
the time of settlement, four months into the subse-
quent annual cycle.
Table 1
IFRIC
requirements
Measurement basis Recognition point in income statement Repriced
Grant Fair value of allowances at start
of annual cycle
In line with emissions but could
change to start of annual cycle
Not repriced
Allowances Historical cost or fair value If historical cost, on disposal.
If fair value, continuously in equity
Emissions Fair value of allowances As emitted Cumulative continuously in
pro?t or loss
Liabilities Assets
Grant
Emissions
Allowances
1 Jan 31 Dec 1 Jan 31 Dec
Fig. 1. Price change e?ects on balance sheet.
462 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
In the graph in Fig. 1 the dotted line curves
show the e?ects of one upward and one downward
price movement during the year. Note that the
price change that occurs early in the annual cycle
has nearly twice the e?ect on the allowances as
on the liability for emissions. When the second
price change occurs, towards the end of the annual
cycle, the two e?ects are nearly equal.
The third column of Table 1 summarises the
e?ects of price changes on the income statement.
As in the balance sheet, the e?ect of price changes
on the cost of emissions increases as the emissions
liability builds up. However, the imbalances noted
in relation to the balance sheet are aggravated by
the requirement under existing IASs that, if anintan-
gible asset is revalued, the gain or loss should gener-
ally be recognised in equity rather than in pro?t or
loss. Even by the end of the annual cycle, therefore,
an entity with allowances that broadly match the
total of its annual emissions and which revalues
those allowances will report price change e?ects in
pro?t or loss without the compensating e?ects of
those same price changes on its allowances.
Evaluation
Howfar was the widespread criticismof IFRIC3
justi?ed and how far was it essentially a rejection of
the real implications of a new phenomenon? The
critics concentrated on the volatility generated in
the income statement over the course of the annual
cycle, at the end and even in some cases into the next
period. To what extent was that volatility arti?cial
and what could reasonably be done to reduce the
arti?cial element?
The least justi?able aspect of the interpretation
and probably the easiest, at least in principle, to
remedy was the mismatch arising from recogni-
tion of gains and losses on the emissions liability
in pro?t or loss and on the allowances in equity.
IASs apply the same revaluation rules to intangi-
ble assets as to property, plant and equipment,
e?ectively treating them all as long-term assets.
Emission allowances, however, are current rather
than long-term in nature and it would be appro-
priate for gains and losses on them to be recogni-
sed in pro?t or loss along with other trading gains
and losses.
A simple, though unchanged, answer can also be
given to the fundamental question identi?ed at the
start of this paper, whether allowances granted by
government should be recognised separately from
the emissions to which they relate. For the reasons
discussed already, these two items must be recogni-
sed separately in order to re?ect the various ways in
which the allowances can be used.
There are then two questions that need to be
considered together:
when should the grant be recognised in inc-
ome? and
shouldany balance of the grant that is recogni-
sed as a liability be re-measured for price
changes?
We have seen that under existing IASs a govern-
ment grant is recognised in income in line with the
costs it is designed to subsidise. However, the IASB
has made no secret of its desire to revise this require-
ment to one that would recognise a grant as a liabil-
ity only so long as it is repayable if its conditions
cease to be met.
Emission allowances are in principle current
rather than long-term, depreciable assets. Unlike
other assets that are the subject of grants, emission
allowances are designed to be ultimately payable to
the government in settlement of the newly created
environmental liability. Some would say that such
payment is not the same as repayment of a grant
because its conditions cease to be met. They point
out that, if the entity reduces its emissions below
the level of its allowances for whatever reason –
even that it ceases that line of business – it does
not have to pay back the allowances to the govern-
ment but may sell them on the market. To the
extent that emissions have occurred, IFRIC 3 rec-
ognises a liability and re-measures it continuously
at current market prices. The question is whether
the same recognition can be given to the unused
portion of the grant, which will be available to
cover future emissions. If it can, that portion of
the grant would be recognised as a liability and in
principle should be re-measured continuously to
current prices; if it cannot, the whole grant should
be recognised in income on receipt. IFRIC 3’s
treatment, which is to recognise a liability for the
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 463
grant but not re-measure it, is unlikely to survive a
revision of the government grants standard.
Although treating the unused portion of the
grant in the same way as the emerging liability
for past emissions would satisfy most critics by
removing what they perceive as arti?cial volatility
from the income statement, there is at present no
obvious principle within the Framework by which
it could be justi?ed. If one is to be found or an
exception made, it is necessary to examine the rea-
sons for the disquiet shown by so many people for
IFRIC 3’s strict application of existing principles.
The reason is essentially that the government is
giving at the start of the year what it will in all prob-
ability take away at the end. Managers resent rec-
ognising a pro?t early in their reporting period on
an asset that will have to be surrendered at the end
of the period in settlement of a liability that the asset
was designed to meet. The same logic is at work as
leads managers toseekhedge accountingfor transac-
tions deemed to be incomplete until a further trans-
action is completed in a later period. An important
di?erence from hedge accounting, however, is that
emission allowances are given by and repaid to the
same party. In that sense, the initial gain is more clo-
sely related to the eventual cost than are the gains
and losses on hedged and hedging items, which are
unrelated except in the mind of the manager.
It is true that for the duration of the annual
cycle the allowances have an independent existence
from the emissions cost that they will be required
to settle. That requires their separate recognition
from the moment of receipt. However, it does
not necessarily require immediate recognition in
income of the grant.
Although one can view the grant as without con-
ditions once it has been made, that view ignores one
of the two basic purposes of the grant. The ?rst pur-
pose is to provide the entity with the resource to set-
tle a predetermined portion of its expected level of
emissions. The second is to create a market for the
allowances that are used in settlement. Those who
view the grant as without conditions focus on the
second purpose, noting that the allowances granted
may be sold for the entity’s own account. However,
the ?rst purpose is the more fundamental. The ?rst
purpose suggests that the grant of allowances is akin
to the granting of a loan, which will be called in at
the end of a period – except insofar as the entity is
able to reduce its emissions below the level of the
allowances. At intermediate reporting dates the
cumulative level of emissions is unlikely to have
reached the level of the allowances. The unused por-
tion of the grant is therefore a kind of conditional
liability, from which the entity can only escape if it
can reduce its total emissions for the annual cycle
below the level of its allowances. It is conditional,
not because the event to which it relates has not
yet occurred (which would not justify recognition
as a liability), but because the allowances received
have to be repaid unless savings in emissions can
be made. In other words, the past event required
by the Framework for recognition of a liability is
the receipt of the allowances rather than the produc-
tion of emissions, which still lies in the future.
If the grant can be recognised as a conditional
liability, the question arises how to avoid double
counting in respect of both the emissions liability
itself and the liability to repay the allowances.
The previous paragraphs refer to the ‘unused’ por-
tion of the grant, i.e. only that portion that has not
yet been credited to income to cover a proportion
of the emissions produced to date. But, if the con-
dition attached to the grant is that annual emis-
sions should be reduced below the level of
allowances, why should not the whole grant remain
a liability unless or until such an event occurs?
The answer lies in the explicit linkage of the grant
to the expected future liability. We have earlier lik-
ened the grant to a loan because it is in most cases
repayable to the grantor after the end of the annual
cycle. However, it is instructive tocontrast the di?er-
ent e?ects of a simple loan of cash and a grant of
emissions allowances. The emissions liability itself
is assumed to be the same in both cases, building
up day by day as emissions are produced. In the ?rst
scenario, a loan, the accounting entries would be:
Dr Cash
Cr Loan liability
Dr Income statement
Cr Emissions liability
Dr Emissions liability
Cr Cash
Dr Loan liability
Cr Cash
464 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
In other words, two cash payments are required:
one to settle the emissions liability and the other
to repay the loan.
By contrast, in the second scenario, the grant of
emissions allowances, the recipient of the grant
makes only one payment. The entries are:
Dr Allowances
Cr Grant liability
Dr Income statement
Cr Emissions liability
Dr Grant liability
Cr Income statement
Dr Emissions liability
Cr Allowances
Unlike in the ?rst scenario, the emissions liability is
extinguished by the repayment of the asset granted,
the allowances. Insofar as a proportion of the emis-
sions liability can be covered by allowances, there
are not two liabilities but only one. Accordingly,
as the emissions liability, which is the fundamental
one, builds up, the grant liability reduces.
The above analysis suggests that the conditionality
of a grant of emissions allowances should be
expressed, not simply in terms of the obligation to
repaythe allowances unless emissions canbe reduced,
but in terms of the use to which they must then be
put, i.e. settlement of the emissions liability.
Conclusion
A ?nding along the above lines would resolve
most of the di?culties encountered by IFRIC 3.
The question remains whether the IASB would
be faithful to its mission if it interpreted the
Framework in such a way. We noted at the start
of this paper that the issue of Emission Rights
illustrates the di?culties of standard setting at
the frontiers of accounting, of being able to discern
the limitations but also the possibilities of ?nancial
reporting. The IASB knows that it must resist the
temptation, so often urged on it by its critics, to
resolve every problem by a ‘quick ?x’ that relies
on a standalone rule rather than clear application
of a principle in the Framework. That way lie
increasing complexity and inconsistencies.
To assist in the search for a principled solution,
this paper has drawn attention to certain features
of the cap and trade scheme that, in combination,
are probably unique:
a grant of assets having two functions – an
immediately marketable security and a ‘cur-
rency’ whichwill be the sole means of satisfying
a highly probable future liability tothe grantor.
The Framework necessarily focuses on the
?nancial position at the reporting date. However,
the implications of expected future events are taken
into account to the extent that they throw light on
the existence of assets and liabilities at that date.
This paper suggests that a grant of assets that are
extremely likely to have to be repaid is a liability,
notwithstanding that the assets can be used inde-
pendently in the meantime in market transactions.
Appendix A. Summary of IFRIC 3 illustrative
example
The following schedules summarise the illustra-
tive example in IFRIC 3, ?rst on the assumption
that the allowances are accounted for at cost until
disposed of and, secondly, on the assumption that
they are continuously re-measured to fair value.
Other assumptions are:
The annual cycle for the allocation of allow-
ances and assessment of emissions, January–
December, coincides with the entity’s ?nan-
cial year.
The entity receives a grant of allowances cov-
ering 12,000 tonnes of emissions.
The fair value of allowances is.
o 1 January CU10 (i.e. 10 currency units) per
tonne,
o 30 June CU12 per tonne,
o 31 December till settlement CU11 per tonne.
Emissions January–June are 5500 tonnes.
o July–December 7000 tonnes.
At 30 June the entity still expects emissions
of 12,000 tonnes.
On 31 December the entity buys at CU11 per
tonne additional allowances to cover 500
tonnes.
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 465
466 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 467
Annual emissions are settled four months
after the year end.
References
Casamento, Robert (2005). Accounting for and taxation of
emission allowances and credits. In Charlotte Streck &
David Freestone (Eds.), Legal aspects of implementing the
Kyoto protocol mechanisms: Making Kyoto work
(pp. 55–70). Oxford University Press.
Hepburn, Cameron (2006). Regulation by prices, quantities or
both: A review of instrument choice. Oxford Review of
Economic Policy, 22, 226–247.
IFRIC 3 Emission Rights. (2004). Withdrawn June 2005.
International Accounting Standards Board.
MacKenzie Donald. (2009). Making things the same: Gases,
Emission Rights and the politics of carbon markets.
Accounting, Organizations and Society, 34, 440–455.
468 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
doc_600997323.pdf
This paper examines the issues surrounding the aborted attempt by the International Accounting Standards Board
in early 2005 to regulate the accounting for the European Union’s new Emissions Trading Scheme under the Kyoto
Protocol. The paper argues that the features that made the trading scheme attractive to governments were precisely
the ones that created difficulties for accountants to capture under existing standards. After showing why the challenge
has to be faced, the paper suggests a possible way forward that the IASB might consider when it revisits the subject, as it
is expected to do in the near future.
Emission rights: From costless activity to market operations
Allan Cook
*
,1
c/o International Accounting Standards Board, 30 Cannon Street, London EC4M 6XH, United Kingdom
Abstract
This paper examines the issues surrounding the aborted attempt by the International Accounting Standards Board
in early 2005 to regulate the accounting for the European Union’s new Emissions Trading Scheme under the Kyoto
Protocol. The paper argues that the features that made the trading scheme attractive to governments were precisely
the ones that created di?culties for accountants to capture under existing standards. After showing why the challenge
has to be faced, the paper suggests a possible way forward that the IASB might consider when it revisits the subject, as it
is expected to do in the near future.
Ó 2008 Elsevier Ltd. All rights reserved.
Background – an aborted interpretation
While Europe was preparing for the adoption
of International Accounting Standards (IASs) by
all listed companies from January 2005, the
International Accounting Standards Board
(IASB) was urged to develop mandatory guid-
ance for the ?nancial reporting of Emission
Rights, the basis for the European Union’s
(EU’s) new scheme designed to reduce emissions
by major polluters.
In view of the specialised nature of the subject,
the IASB asked its International Financial Report-
ing Interpretations Committee (IFRIC) to under-
0361-3682/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2007.12.001
*
Tel.: +44 20 8399 4368.
E-mail address: [email protected]
1
Allan Cook was the co-ordinator for the International Financial Reporting Interpretations Committee (IFRIC) of the
International Accounting Standards Board (IASB) at the time that the issues described in the paper came to a head. He is particularly
grateful to Henry Rees, the IASB senior project manager responsible for the topic, for permission to draw on his work for the IFRIC
and the IASB and to Professor Donald Mackenzie for his advice on the relevance of existing papers dealing with the economic
implications of various possible mechanisms for reducing emissions. However, the views expressed are the author’s own and do not
necessarily re?ect those of either of these gentlemen or the members of the IASB or the IFRIC.
Available online at www.sciencedirect.com
Accounting, Organizations and Society 34 (2009) 456–468
www.elsevier.com/locate/aos
take the task. The IFRICpublished a draft interpre-
tation in May 2003 and, after considering public
comments, developed an interpretation, which the
IASB issued in December 2004, just within the
European Commission’s deadline for the adoption
of IASs.
The result was a public outcry. Companies com-
plained that application of the interpretation
would force them into showing a completely dis-
torted picture of their performance in their annual
and interim ?nancial statements. The IASB, while
recognising that the IFRIC had made a valid inter-
pretation of the relevant IASs, accepted that the
end result was confusing in certain respects. Per-
haps fortunately, the expected market for Emis-
sion Rights was slow to develop and the IASB
took the opportunity to withdraw the interpreta-
tion in June 2005, only six months after it had been
issued.
What was the cause of this reverse?
The problem
The Emission Rights a?air is worth study-
ing because it illustrates the problems faced by
standard setters as they explore the frontiers of
accounting. Just as the science of medicine some-
times advances by examination of its failures to
deal with newly presenting diseases, so, in other
?elds, an understanding of the roots of failures
can lead to a fuller understanding of the discipline
involved, its limitations and its possibilities.
Three features in particular lay at the heart of
the challenge posed to standard setters as they
sought to deal with Emission Rights:
A previously costless activity had become
costly.
Governments mitigated the cost.
By means of marketable allowances.
It is the combination of these features that has
so far proved an insuperable problem for account-
ing to re?ect. Before examining the accountants’
problem, we need to understand what led govern-
ments to adopt such a scheme in their attempts to
reduce emissions.
Controlling emissions – alternative approaches
There is a voluminous literature, dating back
over 30 years, discussing the relative merits of dif-
ferent forms of economic instrument for achieving
government policies with respect to the use of a
scarce resource or service. The choice in a given
situation depends on many factors, including the
degree of uncertainty regarding the marginal costs
and bene?ts of a particular course.
Economic instruments divide broadly into those
that set a price for the good or service, leaving the
quantity demanded to adjust accordingly and those
that set a total quantity, leaving the price to be
determined by market forces. In the ?eld of emis-
sions an example of a price instrument is a carbon
tax: a previously costless, or less costly, activity
becomes costly and the cost is borne in proportion
to the level of emissions involved. An example of a
quantitative instrument is a system under which
government issues a predetermined total of licences
to emit. If the licences are auctioned rather than
issued free and emitters are allowed to trade them,
it can be shown that under idealised conditions
either instrument will result in the same optimal
price and quantity. Note that an ability to trade
licences is necessary for a quantitative system to
achieve equivalence with a price instrument.
Of course, conditions are rarely idealised nor is
it only economic factors that determine political
decisions by governments or inter-governmental
agreements. This paper does not attempt to evalu-
ate the relative merits of price versus quantitative
instruments in the control of emissions
2
. Rather,
the paper examines the accounting implications
of the quantitative systems that have been put in
place by governments. Although price instruments
also exist, they are not currently as widespread as
quantitative systems and, in any event, do not raise
any new accounting issues.
2
An excellent, recent analysis of those issues is given by
Hepburn 2006. That article explains why some economists
believe that the relationships between the marginal costs and
bene?ts of alternative instruments to reduce emissions suggest
that, in theory, a price instrument may be more e?cient (i.e.
arrive closer to the optimum price and quantity) than a
quantitative instrument.
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 457
The principal quantitative instruments in use
for controlling emissions are ‘base line and credit’
and ‘cap and trade’.
Base line and credit
Under this system a government allocates
allowances based on some assessment of a normal
rate of emissions. For each entity a base line is set
below which no charge for emissions will be made.
Entities that manage to operate below the base line
for a period are given credits, which can be sold
through a market to those which exceed their base
line. At the end of the period any entity that has
exceeded its base line must deliver to the govern-
ment su?cient credits, bought in the market, to
cover the excess.
A weakness of this system is that allowances that
are traded may be insu?cient to sustain a market.
Moreover, the government may lay itself open to
charges of unfair discrimination in its determina-
tion of the base lines for di?erent entities.
Cap and trade
Under a cap and trade system the government
mitigates the cost for entities, not by setting a
base line, but by allocating tradable allowances
for the period. After the end of the period each
entity must pay for its emissions by surrendering
allowances granted or buying in from the market
any shortfall. Since the allowances are allocated
from the beginning of a period, in theory an entity
that perceives a market opportunity can sell more
than the amount of its expected savings in the hope
of buying back later at a more advantageous price.
This systemplaces greater reliance on the market
mechanism than does the base line and credit
method. Although initially all allowances are allo-
cated by governments for nil consideration, it is
intended, under the EU scheme, that a gradually
increasing proportion should be auctioned rather
than granted free. Some believe that an auction of
all allowances would be the most e?ective means
of reducing total emissions by ensuring that the sav-
ings that were easiest to make were addressed the
?rst. However, too rapid a move to the auction of
all allowances might be counter productive by oblig-
ing entities to pass on the cost to the consumer
instead of encouraging them to seek savings.
The EU scheme under the Kyoto protocol
The EU scheme covers all 25 member states and
was adopted as the principal means for the EU to
meet its emission reduction commitments under
the Kyoto protocol. The protocol set reduction tar-
gets in the form of ‘Allowed Amount Units’ of
emissions by country but allowed for countries to
group together to redistribute their total allowances
among themselves. Under the EU scheme, the
Union’s overall target of eight percent reduction
was redistributed by giving the UK, for example,
a target of 12.5% reduction but allowing Portugal
an increase of 27%. Under the EU scheme entities
may participate in the Kyoto Clean Development
Mechanism, which enables them to acquire further
allowances by undertaking a development to
reduce emissions in another country
3
.
The EU scheme has nearly completed an initial
period of three years, which will be succeeded by
the ?rst full ?ve year phase under the Kyoto proto-
col running from 1 January 2008 to 31 December
2012. For the initial phase, 95% of allowances
had to be allocated free to entities. During the fol-
lowing phase, only 90% must be free. The remain-
ing ten per cent may be auctioned, thus obliging
most entities to have recourse to the market, either
at the time of the auction or subsequently.
Allowances are allocated in January to cover
emissions up to the end of December; settlement
is required by the end of the following April. Enti-
ties that during the initial phase failed to deliver suf-
?cient allowances to cover their level of emissions
are ?ned €40 per tonne of shortfall. During the sec-
ond phase this is expected to rise to €100 per tonne.
Payment of the ?ne is not a substitute for delivery of
the required allowances. The shortfall must be
3
MacKenzie (2009) gives a good illustration of the scienti?c
imponderables that underlie ‘making things the same’ by using
allowances received in one industry to o?set the cost of
emissions in another. The paper stresses the important role of
accounting if market based schemes are to motivate successfully
by making carbon costs visible.
458 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
made up from the subsequent year’s allocation plus
market purchases.
Allowances are not themselves a licence to emit:
each entity covered by the scheme must obtain
from the government an emissions permit. If the
entity ceases production during the year, it must
surrender its emissions permit but may sell any
surplus allowances on the market.
All features of the scheme are focused on the
objective that somewhere in the world emissions
get reduced – even if not necessarily by the entity
whose expected emissions exceed its allocation of
allowances nor even within the jurisdiction of the
government determining that allocation. By allow-
ing an entity that ceases production to sell its allow-
ances, the scheme maintains the originally assessed
level of allowances for the economy as a whole. It is
not necessary to claw them back from the entity to
which they had been granted. Moreover, an e?-
cient producer may judge that it can reduce more
emissions – and earn more credits – by undertaking
a project to enable a less e?cient entity to reduce its
emissions than it could by investing the same level
of resources in reducing its own emissions. The
design of the scheme encourages it to take the for-
mer, more e?ective course.
The accountants’ problem rooted in government
objectives
From the above description we can see why the
three features causing the most problems for
accountants are particularly attractive from the
point of view of governments.
Emission reduction schemes are designed to
give a cost to a previously costless activity, in
order to motivate producers to regard emissions
as an input cost that must be monitored and
controlled like any other.
However, unlike licences for oil and gas explo-
ration or 3G development, emissions costs are
being introduced only slowly, by gradual reduc-
tion of the total level of allowances or the pro-
portion of the total that is allocated free. This
strategy avoids a shock to existing systems for
the production, supply and use of energy, which
might result in the new cost simply being passed
on to the consumer instead of being controlled
at the point where savings were possible.
The cap and trade system responds well to the
previous two objectives while also making most
e?ective use of the market. Use of the market
both simulates other real costs and is the most
e?cient means of passing the incentives to save
emissions to those best able to make such sav-
ings at an early date.
The remainder of this paper analyses the
accounting problems created by these very reason-
able governmental objectives.
One accounting solution: maintain the status quo
Many producers and others commenting on this
subject to the IASB asserted that the key to the
accounting solutionlay indealing only withthe mar-
ginal e?ect on cost. They argued that, as long as an
entity emitted no more than the amount covered
by its emissions allowances, no new cost emerged.
The only cost would be that of acquiring additional
allowances fromthe market: the only credit, the pro-
ceeds of a sale of surplus allowances to the market.
This solution relied on netting the bene?t of trad-
able allowances received against the newly created
cost of emissions, as though it were dealing with a
base line and credit scheme. It sought to simulate
as far as possible the status quo ante. The question
for accountants was whether it was true that the sta-
tus quo had not changed: whether in fact something
was being overlooked if income from the grant of
allowances was netted against the cost of emissions.
The circumstances in which costs may be netted
against income are, of course, severely limited both
under IASs and under the EU Fourth Directive.
The weakness of the netting solution is that it
ignores the market oriented thrust of cap and trade
schemes. Although beginning slowly, the market
for allowances seems likely to play an increasingly
important part in governments’ attempts to reduce
emissions both in the EU and internationally. The
netting solution attributes a cost to purchased
allowances but not to those granted to an entity
by government. Yet the allowances themselves
are fungible: neither the government nor the mar-
ket distinguishes allowances according to their
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 459
source. Allowances are assets in their own right, as
demonstrated by their ability to be sold for cash
and even to be transferred between di?erent
schemes under the Clean Development Mecha-
nism of the Kyoto protocol.
The netting solution also ignores the reality of
the liability for emissions, a liability that exists
and behaves quite independently from the allow-
ances held by the emitting entity. It is an under-
statement of that liability to net it against
allowances held. The netting of assets and liabili-
ties is restricted, if anything, even more severely
than the netting of income and costs.
Even as a practical expedient, the netting solu-
tion breaks down as soon as an entity begins to
trade its allowances. From that point the signi?-
cance of the initial grant disappears as granted
allowances are replaced by others bought in from
the market. At the extreme, the whole of the initial
grant may have been sold and replaced; on what
basis is the amount of o?set then to be calculated?
If the reply is that no o?set should be allowed in
such a case, there would be no comparability
between one entity that had retained its original
grant of allowances and one that had sold but
replaced them, even if there had been no signi?-
cant price movements in the interim. If the reply
is that o?set is to be allowed, some method would
have to be devised to limit the amount of the o?set
to the volume and unit price of allowances at the
time the grant was made. The reason is that any
variation of the current value of allowances from
the amount of the grant would be the result of
price movements or market transactions rather
than an adjustment of the original grant.
The issues for IFRIC
The IFRIC
4
approached the problem by analy-
sing separately the nature of the allowances
granted and the obligation to pay for emissions
by delivering allowances. It asked:
Are the allowances an asset
– if purchased?
– if allocated by government?
If so, what kind of asset
– Financial instrument? or
– intangible?
If the allowances are an asset, is the credit (the
grant)
– income or
– a liability on receipt?
Allowances
Most people would instinctively agree that pur-
chased allowances are an asset, probably because
they have a cost. However, the IASB Framework
for the Preparation and Presentation of Financial
Statements (‘the Framework’) does not require a
cost. It de?nes an asset as ‘a resource controlled by
the entity as a result of past events and from which
future economic bene?ts are expected to ?ow to
the entity’. The IFRIC found that allowances allo-
cated by government, no less than those purchased,
met this de?nition. They are controlled by the entity
from the moment of the past event – allocation – by
which they were acquired. They are expected to gen-
erate future bene?ts for the entity either when ten-
dered in settlement of an emissions liability or
through sale into the market. As noted above, the
rights attaching to them are indistinguishable from
those of purchased allowances, a point that is likely
to assume practical, as well as conceptual impor-
tance as markets develop and entities begin to trade
signi?cant portions of their allocations.
Although, as we shall see later, this conclusion
gave rise to uncomfortable consequences, it is di?-
cult to challenge it on conceptual grounds. How-
ever, in determining whether allowances were a
?nancial instrument or an intangible asset, the
IFRIC encountered a more arti?cial di?culty of
the IASB’s own making. It would have been conve-
nient if the allowances could have been classi?ed as
one of the classes of assets that IASs require to be re-
measured to fair value at each reporting date, with
gains andlosses reportedinpro?t or loss. The reason
is that under IAS 37 Provisions, Contingent Liabil-
ities and Contingent Assets the liability for emis-
sions must be measured and reported in that way.
4
This section summarises the requirements and rationale
contained in IFRIC3 Emission Rights. A fuller description of
the debate surrounding the ?nalisation of IFRIC3 is given by
Casamento (2005).
460 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
IASs require ?nancial instruments held for trading
to be on this basis but the IFRIC decided that the
allowances did not meet the, somewhat detailed,
de?nitionof a ?nancial instrument. Nor didthey fall
within the limited class of other assets that could be
measured like ?nancial instruments.
Instead, the IFRIC decided that the allowances
fell within the de?nition of an intangible asset.
Although such assets may be re-measured to fair
value if they can be traded in an active market,
IASs currently require the resulting gains and
losses to be reported outside pro?t or loss. Symme-
try of treatment with that of the liability therefore
eluded the IFRIC.
Emissions liability
The IFRIC decided that the obligation to pay for
emissions bydeliveringallowances tothe government
after the endof the periodwas a liability. It clearly fell
within the Framework’s de?nition, ‘a present obliga-
tion of the entity arising from past events, the settle-
ment of which is expected to result in an out?ow
from the entity of resources embodying economic
bene?ts’. Anobligationarose as emissions were made
– the past event – which would require the entity to
part in the future with resources – the allowances –
embodying economic bene?ts – the ability to be used
to settle further obligations or be sold in the market.
Some argued that even if the asset and liability
were distinct, they should be permitted to be o?set,
since the ultimate use of the allowances was to set-
tle the liability. However, as noted above, the
allowances would not necessarily be used to settle
that entity’s liability. Furthermore, the allowances
were not a debt owed by the government, which
might have quali?ed as o?settable against a liabil-
ity to the government.
Others argued that hedge accounting should be
permitted between the allowances and the liability.
Certainly, the existing rules for hedge accounting
would not permit that but, since hedge accounting
is essentially anarti?cial convention, this couldhave
been a route worth exploring. As with all forms of
hedge accounting, however, such a solution would
enable one entity to portray itself di?erently from
another entity in the same or a very similar eco-
nomic position. Some of the practical complexities
that would be involved in applying rules for such
hedge accounting have been referred to above.
Grant of allowances
If the allowances are recognised as an asset, it is
necessarytodetermine howthe correspondingcredit
for free allocation of allowances (the grant) should
be treated. The IFRIC decided that any allocation
below fair value fell within the de?nition of govern-
ment grants, essentially a government’s ‘transfers of
resources to an entity in return for past or future
compliance with certain conditions relating to the
operating activities of the entity’. It argued that the
obligation under an Emission Rights scheme to
reduce emissions or deliver allowances was a condi-
tion‘relating tothe operating activities of the entity’.
At the outset, the grant and the allowances are
measured at the fair value of the allowances at the
time the grant is made. As noted above, under
IASs the allowances may, but do not have to be
re-measured to fair value; the grant, however, is
not so re-measured.
Under the existing version of IAS 20 Govern-
ment Grants, income from the grant is recognised
gradually in line with recognition of the activity
to which the grant relates. For example, a grant
towards the purchase of a machine is recognised
in the income statement over the same period as
depreciation of the machine. In a paragraph of
the draft interpretation that attracted virtually no
comment, the IFRIC noted the IASB’s long-stand-
ing intention to propose amendment of IAS 20: the
IASB had already discussed publicly its proposal to
require income from a government grant to be
recognised over the period to the point at which
the grant could no longer be recalled. The rationale
for the change was that beyond that point no liabil-
ity existed: the whole grant therefore needed by
then to be recognised in income. The e?ect of such
a change would be that a grant of emissions allow-
ances would have to be recognised in income imme-
diately on its receipt at the start of the annual cycle.
E?ects of the IFRIC decisions
While each of the IFRIC’s decisions outlined
above was soundly based on existing IASs, which
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 461
in their turn, had stood the test of time, the com-
bined e?ect when applied to Emission Rights pro-
duced some very strange results. The various
measurement bases and recognition points in the
income statement are summarised in Table 1 and
the e?ect on the balance sheet is illustrated graphi-
cally in Fig. 1. For simplicity, the analysis that fol-
lows ignores the e?ects of market transactions.
Moreover, no attempt has been made to portray
the complications that arise when an entity’s ?nan-
cial year does not correspond with the annual cycle
for the allocation and settlement of allowances.
When allowances are received at the start of the
annual cycle, they are recognised at fair value,
which is the market price for allowances at that
date. Since the entity does not pay for allowances
received from the government, it recognises the
grant at the same amount as the allowances.
Unless and until IAS 20 is amended, the grant is
recognised initially as a liability. Pro?t or loss at
that point is therefore not a?ected.
As emissions are produced, a new liability
emerges for the amount that will have to be paid
to the government in the form of allowances after
the end of the annual cycle. In common with most
other liabilities, this emissions liability is re-mea-
sured to ‘current settlement value’, in this case
essentially fair value at each reporting date. The
grant is gradually transferred from liability status
to income to compensate for the emissions cost
incurred. However, unlike most other liabilities,
the liability for the grant is not re-measured for
price movements. The reason is that, as explained
above, it is not a true liability but in reality no
more than deferred income, which is being recog-
nised for convenience sake in line with the cost it
is designed to subsidise.
Various imbalances nowbegin to emerge. Even if
the entity continues throughout the annual cycle to
expect that its cumulative emissions will be exactly
covered by the grant of allowances, the e?ect of
price changes will generally prevent the diminishing
liability for the grant fromexactly compensating for
the growing liability for the cost of emissions. Fur-
thermore, the e?ect of price changes on the asset
side of the balance sheet di?ers from that on the lia-
bilities side. If the entity chooses not to fair value the
allowances, as it is permitted to do, there will natu-
rally be no equivalent price movements on the assets
side. However, even if the allowances are re-mea-
sured, the e?ect of price changes during the year dif-
fers from that a?ecting the liability. The reason is
that a full year’s allowances are held by the entity
fromthe start of the annual cycle, whereas the liabil-
ity for emissions builds up only gradually. It is
therefore only at the end of the annual cycle that
an entity that fair values its allowances will ?nd that
the price movements on the allowances broadly
compensate for those on the liability. If the entity
measures its allowances at historical cost, they will
only be adjusted for cumulative price changes at
the time of settlement, four months into the subse-
quent annual cycle.
Table 1
IFRIC
requirements
Measurement basis Recognition point in income statement Repriced
Grant Fair value of allowances at start
of annual cycle
In line with emissions but could
change to start of annual cycle
Not repriced
Allowances Historical cost or fair value If historical cost, on disposal.
If fair value, continuously in equity
Emissions Fair value of allowances As emitted Cumulative continuously in
pro?t or loss
Liabilities Assets
Grant
Emissions
Allowances
1 Jan 31 Dec 1 Jan 31 Dec
Fig. 1. Price change e?ects on balance sheet.
462 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
In the graph in Fig. 1 the dotted line curves
show the e?ects of one upward and one downward
price movement during the year. Note that the
price change that occurs early in the annual cycle
has nearly twice the e?ect on the allowances as
on the liability for emissions. When the second
price change occurs, towards the end of the annual
cycle, the two e?ects are nearly equal.
The third column of Table 1 summarises the
e?ects of price changes on the income statement.
As in the balance sheet, the e?ect of price changes
on the cost of emissions increases as the emissions
liability builds up. However, the imbalances noted
in relation to the balance sheet are aggravated by
the requirement under existing IASs that, if anintan-
gible asset is revalued, the gain or loss should gener-
ally be recognised in equity rather than in pro?t or
loss. Even by the end of the annual cycle, therefore,
an entity with allowances that broadly match the
total of its annual emissions and which revalues
those allowances will report price change e?ects in
pro?t or loss without the compensating e?ects of
those same price changes on its allowances.
Evaluation
Howfar was the widespread criticismof IFRIC3
justi?ed and how far was it essentially a rejection of
the real implications of a new phenomenon? The
critics concentrated on the volatility generated in
the income statement over the course of the annual
cycle, at the end and even in some cases into the next
period. To what extent was that volatility arti?cial
and what could reasonably be done to reduce the
arti?cial element?
The least justi?able aspect of the interpretation
and probably the easiest, at least in principle, to
remedy was the mismatch arising from recogni-
tion of gains and losses on the emissions liability
in pro?t or loss and on the allowances in equity.
IASs apply the same revaluation rules to intangi-
ble assets as to property, plant and equipment,
e?ectively treating them all as long-term assets.
Emission allowances, however, are current rather
than long-term in nature and it would be appro-
priate for gains and losses on them to be recogni-
sed in pro?t or loss along with other trading gains
and losses.
A simple, though unchanged, answer can also be
given to the fundamental question identi?ed at the
start of this paper, whether allowances granted by
government should be recognised separately from
the emissions to which they relate. For the reasons
discussed already, these two items must be recogni-
sed separately in order to re?ect the various ways in
which the allowances can be used.
There are then two questions that need to be
considered together:
when should the grant be recognised in inc-
ome? and
shouldany balance of the grant that is recogni-
sed as a liability be re-measured for price
changes?
We have seen that under existing IASs a govern-
ment grant is recognised in income in line with the
costs it is designed to subsidise. However, the IASB
has made no secret of its desire to revise this require-
ment to one that would recognise a grant as a liabil-
ity only so long as it is repayable if its conditions
cease to be met.
Emission allowances are in principle current
rather than long-term, depreciable assets. Unlike
other assets that are the subject of grants, emission
allowances are designed to be ultimately payable to
the government in settlement of the newly created
environmental liability. Some would say that such
payment is not the same as repayment of a grant
because its conditions cease to be met. They point
out that, if the entity reduces its emissions below
the level of its allowances for whatever reason –
even that it ceases that line of business – it does
not have to pay back the allowances to the govern-
ment but may sell them on the market. To the
extent that emissions have occurred, IFRIC 3 rec-
ognises a liability and re-measures it continuously
at current market prices. The question is whether
the same recognition can be given to the unused
portion of the grant, which will be available to
cover future emissions. If it can, that portion of
the grant would be recognised as a liability and in
principle should be re-measured continuously to
current prices; if it cannot, the whole grant should
be recognised in income on receipt. IFRIC 3’s
treatment, which is to recognise a liability for the
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 463
grant but not re-measure it, is unlikely to survive a
revision of the government grants standard.
Although treating the unused portion of the
grant in the same way as the emerging liability
for past emissions would satisfy most critics by
removing what they perceive as arti?cial volatility
from the income statement, there is at present no
obvious principle within the Framework by which
it could be justi?ed. If one is to be found or an
exception made, it is necessary to examine the rea-
sons for the disquiet shown by so many people for
IFRIC 3’s strict application of existing principles.
The reason is essentially that the government is
giving at the start of the year what it will in all prob-
ability take away at the end. Managers resent rec-
ognising a pro?t early in their reporting period on
an asset that will have to be surrendered at the end
of the period in settlement of a liability that the asset
was designed to meet. The same logic is at work as
leads managers toseekhedge accountingfor transac-
tions deemed to be incomplete until a further trans-
action is completed in a later period. An important
di?erence from hedge accounting, however, is that
emission allowances are given by and repaid to the
same party. In that sense, the initial gain is more clo-
sely related to the eventual cost than are the gains
and losses on hedged and hedging items, which are
unrelated except in the mind of the manager.
It is true that for the duration of the annual
cycle the allowances have an independent existence
from the emissions cost that they will be required
to settle. That requires their separate recognition
from the moment of receipt. However, it does
not necessarily require immediate recognition in
income of the grant.
Although one can view the grant as without con-
ditions once it has been made, that view ignores one
of the two basic purposes of the grant. The ?rst pur-
pose is to provide the entity with the resource to set-
tle a predetermined portion of its expected level of
emissions. The second is to create a market for the
allowances that are used in settlement. Those who
view the grant as without conditions focus on the
second purpose, noting that the allowances granted
may be sold for the entity’s own account. However,
the ?rst purpose is the more fundamental. The ?rst
purpose suggests that the grant of allowances is akin
to the granting of a loan, which will be called in at
the end of a period – except insofar as the entity is
able to reduce its emissions below the level of the
allowances. At intermediate reporting dates the
cumulative level of emissions is unlikely to have
reached the level of the allowances. The unused por-
tion of the grant is therefore a kind of conditional
liability, from which the entity can only escape if it
can reduce its total emissions for the annual cycle
below the level of its allowances. It is conditional,
not because the event to which it relates has not
yet occurred (which would not justify recognition
as a liability), but because the allowances received
have to be repaid unless savings in emissions can
be made. In other words, the past event required
by the Framework for recognition of a liability is
the receipt of the allowances rather than the produc-
tion of emissions, which still lies in the future.
If the grant can be recognised as a conditional
liability, the question arises how to avoid double
counting in respect of both the emissions liability
itself and the liability to repay the allowances.
The previous paragraphs refer to the ‘unused’ por-
tion of the grant, i.e. only that portion that has not
yet been credited to income to cover a proportion
of the emissions produced to date. But, if the con-
dition attached to the grant is that annual emis-
sions should be reduced below the level of
allowances, why should not the whole grant remain
a liability unless or until such an event occurs?
The answer lies in the explicit linkage of the grant
to the expected future liability. We have earlier lik-
ened the grant to a loan because it is in most cases
repayable to the grantor after the end of the annual
cycle. However, it is instructive tocontrast the di?er-
ent e?ects of a simple loan of cash and a grant of
emissions allowances. The emissions liability itself
is assumed to be the same in both cases, building
up day by day as emissions are produced. In the ?rst
scenario, a loan, the accounting entries would be:
Dr Cash
Cr Loan liability
Dr Income statement
Cr Emissions liability
Dr Emissions liability
Cr Cash
Dr Loan liability
Cr Cash
464 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
In other words, two cash payments are required:
one to settle the emissions liability and the other
to repay the loan.
By contrast, in the second scenario, the grant of
emissions allowances, the recipient of the grant
makes only one payment. The entries are:
Dr Allowances
Cr Grant liability
Dr Income statement
Cr Emissions liability
Dr Grant liability
Cr Income statement
Dr Emissions liability
Cr Allowances
Unlike in the ?rst scenario, the emissions liability is
extinguished by the repayment of the asset granted,
the allowances. Insofar as a proportion of the emis-
sions liability can be covered by allowances, there
are not two liabilities but only one. Accordingly,
as the emissions liability, which is the fundamental
one, builds up, the grant liability reduces.
The above analysis suggests that the conditionality
of a grant of emissions allowances should be
expressed, not simply in terms of the obligation to
repaythe allowances unless emissions canbe reduced,
but in terms of the use to which they must then be
put, i.e. settlement of the emissions liability.
Conclusion
A ?nding along the above lines would resolve
most of the di?culties encountered by IFRIC 3.
The question remains whether the IASB would
be faithful to its mission if it interpreted the
Framework in such a way. We noted at the start
of this paper that the issue of Emission Rights
illustrates the di?culties of standard setting at
the frontiers of accounting, of being able to discern
the limitations but also the possibilities of ?nancial
reporting. The IASB knows that it must resist the
temptation, so often urged on it by its critics, to
resolve every problem by a ‘quick ?x’ that relies
on a standalone rule rather than clear application
of a principle in the Framework. That way lie
increasing complexity and inconsistencies.
To assist in the search for a principled solution,
this paper has drawn attention to certain features
of the cap and trade scheme that, in combination,
are probably unique:
a grant of assets having two functions – an
immediately marketable security and a ‘cur-
rency’ whichwill be the sole means of satisfying
a highly probable future liability tothe grantor.
The Framework necessarily focuses on the
?nancial position at the reporting date. However,
the implications of expected future events are taken
into account to the extent that they throw light on
the existence of assets and liabilities at that date.
This paper suggests that a grant of assets that are
extremely likely to have to be repaid is a liability,
notwithstanding that the assets can be used inde-
pendently in the meantime in market transactions.
Appendix A. Summary of IFRIC 3 illustrative
example
The following schedules summarise the illustra-
tive example in IFRIC 3, ?rst on the assumption
that the allowances are accounted for at cost until
disposed of and, secondly, on the assumption that
they are continuously re-measured to fair value.
Other assumptions are:
The annual cycle for the allocation of allow-
ances and assessment of emissions, January–
December, coincides with the entity’s ?nan-
cial year.
The entity receives a grant of allowances cov-
ering 12,000 tonnes of emissions.
The fair value of allowances is.
o 1 January CU10 (i.e. 10 currency units) per
tonne,
o 30 June CU12 per tonne,
o 31 December till settlement CU11 per tonne.
Emissions January–June are 5500 tonnes.
o July–December 7000 tonnes.
At 30 June the entity still expects emissions
of 12,000 tonnes.
On 31 December the entity buys at CU11 per
tonne additional allowances to cover 500
tonnes.
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 465
466 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
A. Cook/ Accounting, Organizations and Society 34 (2009) 456–468 467
Annual emissions are settled four months
after the year end.
References
Casamento, Robert (2005). Accounting for and taxation of
emission allowances and credits. In Charlotte Streck &
David Freestone (Eds.), Legal aspects of implementing the
Kyoto protocol mechanisms: Making Kyoto work
(pp. 55–70). Oxford University Press.
Hepburn, Cameron (2006). Regulation by prices, quantities or
both: A review of instrument choice. Oxford Review of
Economic Policy, 22, 226–247.
IFRIC 3 Emission Rights. (2004). Withdrawn June 2005.
International Accounting Standards Board.
MacKenzie Donald. (2009). Making things the same: Gases,
Emission Rights and the politics of carbon markets.
Accounting, Organizations and Society, 34, 440–455.
468 A. Cook / Accounting, Organizations and Society 34 (2009) 456–468
doc_600997323.pdf