1、 - 1 - 英文原文： FAIR VALUE ACCOUNTING IN THE BANKING SECTOR Mary Pacaperscu The Financial Instruments Joint Working Group (JWG) of Standard Setters issued in December 2000 the consultative document entitled “Draft Standard and Basis for Conclusions Financial Instruments and Similar Items”. The Draft St
2、andard reviews and assesses an extensive use of fair value accounting (FVA) as the basis for the valuation of all financial instruments in a banks balance sheet. The work of the JWG is linked to the long-term strategy of the International Accounting Standards Committee (IASC) recently replaced by th
3、e International Accounting Standards Board (IASB) to introduce a comprehensive FVA framework for the recognition and measurement of financial instruments. The JWG invited comments on the Draft Standard from all interested parties by 30 September 2001. The IASB will evaluate the long-term prospects o
4、f FVA in the light of the comments received. This note conveys the comments of the European Central Bank (ECB) on an important dimension of the proposal put forward by the JWG, notably the application of FVA to the banking sector. After reviewing the main innovations of the Draft Standard, the note
5、focuses on the critical aspects associated with the application of a full FVA regime to the banking sector and presents a possible way forward. I. The main innovations of the Draft Standard for the banking sector The present accounting rules for banks in the European Union distinguish between financ
6、ial instruments held for trading purposes (in the trading book) and those intended to be held to maturity (in the banking book). Instruments held in the trading book are valued at market prices. A profit and/or loss arising from the revaluation of trading book instruments is recognised in the profit
7、 and loss account. The accounting rules for the trading book thereby take all market risks (i.e. price risk, interest rate risk, foreign exchange risk and liquidity risk) into account. Banking book instruments, by contrast, are carried in the balance sheet at the lower of historical cost and market
8、value. Whereas a loss on a banking book instrument is transferred to the profit and loss account, unrealised gains are not recognised and can therefore become hidden reserves in the balance sheet. Therefore, the accounting rules for the banking book do not take market risks into account (except for
9、the foreign exchange risk, where the end-period value is usually applied to almost all balance sheet items). The Draft Standard proposes a uniform rule for all financial instruments. The assets and liabilities are carried in the balance sheet at market values, if they are available, or at fair value
10、s calculated as an approximation of the market value by using a present value model for discounting the expected future cash flow. For banks, this would imply that the trading and banking books would receive equal accounting treatment, whereby all changes in value would be recognised in the balance
11、sheet and transferred to the profit and loss account. The foreseen revaluation applies irrespective of whether a profit or loss has been realised or remains unrealised because all instruments are either marked to market or the fair value is estimated. The hidden reserves that may arise under the exi
12、sting accounting rules thus disappear. Market - 2 - risks would be taken into account when calculating the value of financial instruments in both the trading and the banking book. II. Critical aspects According to its proponents, an FVA regime may constitute, from a conceptual point of view, an alte
13、rnative approach to reporting financial performance in order to avoid some of the problems associated with the current historical cost accounting. One of its main advantages would be to enhance the degree of transparency of financial statements. However, this point of view remains theoretical due to
14、 the absence of homogeneity and therefore comparability in FVA methodologies. Furthermore, the possible concrete application of a full FVA regime (applying to all assets and liabilities) to the banking sector gives rise to some serious problems and concerns. The application of FVA may be suitable fo
15、r the trading book of banks, which refers to transactions (buying and selling) of marketable securities and related instruments with the objective of making a profit from short-term price variations. The use of fair value for these transactions is consistent with the availability of market prices an
16、d the short-term horizon. However, the application of FVA to the banking book of banks, i.e. to non-negotiable instruments such as loans, appears to be inappropriate for at least three main reasons. First, the issue of relevance. FVA principles do not reflect properly the way in which banks manage t
17、heir core business, namely the granting of loans. The essence of bank management in this area lies in taking long-term decisions about credit quality and concentration and fostering customer relationships over the life of the contracts. It is less concerned about short-term variations that represent
18、 the basis for the use of FVA principles. Therefore, there is the possibility that the introduction of FVA for the banking book might in principle create incentives for banks to alter their core business. This would be the case if banks decided to reduce their exposure to increased volatility of inc
19、ome (stemming from the accounting recognition of interest rate risk in the banking book) by shortening the average maturity of loans. Other ways to achieve the same goal would be the recourse to hedging techniques and the increased use of variable interest rates. The decision to reduce the average m
20、aturity of loans would depend also on other factors, including the nature of customer demand and the specific cost structure of individual banks. Second, the issue of feasibility. There are serious doubts that an adequate fair value can be determined for bank loans, which are non-negotiable instrume
21、nts precisely because they embody elements that cannot be easily quantified in a standardised manner. First, there are, by definition, no secondary markets for these instruments. This is particularly true where credit risk markets do not appear to be sufficiently deep and liquid for the purpose conc
22、erned. Second, some relevant information for the determination of the fair value of loans (i.e. that stemming from the bilateral relationship between the borrower and the lender) would never be priced in a market. Third, the estimation techniques currently available (including the one proposed in th
23、e Draft Standard) suffer from methodological problems (e.g. modelling of non-interest income, appropriate discount rate, etc.), which increase the risks of error. Accordingly, they do not represent an effective benchmark for obtaining reliable fair values for loans. Therefore, the application of FVA
24、 to bank loans would give rise to many uncertainties hindering and working against the transparency and comparability of financial - 3 - statements. It is acknowledged, however, that the current and future developments in banks credit risk management systems recognised also in the new capital adequa
25、cy regime proposed by the Basel Committee on Banking Supervision may provide accounting standard-setters with useful elements to refine their methodologies, in particular regarding the measurement of credit risk. Doubts are also raised with regard to the application of FVA to the liability side of b
26、anks. For instance, the suggested methodology (the so-called “own credit risk”) to determine the fair value of debt instruments issued by banks entails that, if the rating of a bank deteriorates, the value of its equity will ultimately increase (since the difference in revaluation of debt instrument
27、s is accounted in the profit and loss account). This outcome is counter-intuitive and can be misleading for shareholders and creditors. Third, the issue of prudence. The use of FVA in the banking book would entail that potential profits and losses would be treated in the same way, by being recognise
28、d as soon as they emerge. This goes against the principle of prudence according to which losses stemming from the banking book should be recognised as soon as they are known, even if only potential, whereas profits should be recognised only if they are actually realised. Potential profits should be
29、recognised only for marketable instruments. Therefore, there is the possibility that the application of FVA to the banking book might induce banks to adopt an imprudent behaviour. This is a crucial aspect also from the viewpoint of the banking supervisory function. III. Possible way forward In light
30、 of the critical aspects mentioned above, the ECB has a negative stance towards the possibility of applying an FVA regime to the banking book of banks. Against this background, the following developments could be considered in order to make a constructive use of the valid arguments that lie behind F
31、VA. A first development would entail that, whereas FVA would not be recognised as an accounting standard for the banking book of banks, supervisory authorities might use it as a supplementary instrument to complement their assessment of the situation of individual credit institutions. A second devel
32、opment involves the adoption by banks of the so-called “dynamic provisioning”. This entails recognising that a proportion of the loan portfolio can deteriorate in the future and that this proportion can be measured ex ante on the basis of a specific statistical analysis. It would also involve the di
33、sclosure by banks of the results of stress-test analyses conducted on the interest rate sensitivity of the banking book. This approach would allow two criticisms associated with the current accounting standards to be overcome, notably that potential credit losses remain hidden until signs of deterioration are evident and that market participants have insufficient information about the interest rate risk profile of banks.
