International Affairs INTESA SANPAOLO COMMENTS TO THE DRAFT CESR STATEMENT ON FAIR VALUE MEASUREMENT AND RELATED DISCLOSURES OF FINANCIAL INSTRUMENTS ON ILLIQUID MARKETS September 2008 Intesa Sanpaolo, created as from 1 January 2007 as a result of the merger of Banca Intesa and the Sanpaolo IMI Group, is the largest banking group in Italy and one of the major players in the European market. Intesa Sanpaolo welcomes CESR s open consultation process and would like to submit the following comments. General Remarks Under turbulent market conditions, such as those we are currently experiencing, Intesa Sanpaolo appreciates CESR s effort aimed at providing market participants with guidelines on the preparation of the financial statements and on the application of the financial reporting standards. From this standpoint, we support the general principles set out in the consultation paper. We would nonetheless stress the fact that, as far as the valuation of financial instruments is concerned (and in particular, the use of quoted prices on active markets rather than the use of valuation techniques), the current turbulence and volatility call for a clarification from the IASB (which is the only body that should respond to the accounting issues in general and in particular those arising as a result of the current market turmoil). In particular, in illiquid markets the use of quoted prices for the valuation of many financial instruments becomes problematic while the use of valuation techniques, namely the comparable approach or the mark-to-model approach, assumes a prominent role. 1
These valuation techniques involve significant uncertainties (first of all the measurement of the illiquidity risk); therefore a clearer interpretation from the IASB on their use would contribute to the comparability of the financial statements. Question 1: Do you agree with CESR s views above regarding the distinction between active and non active markets for fair value measurement? The statement reported in CESR document is in line with IAS 39 guidance on the judgment regarding whether to consider if the market is active or not. The factors mentioned are correct and they are generally used by market practice to assess the value of financial instruments and they are a common ground both for the dealing room activity and for accounting valuation process. As to the issues covered in paragraph 28 of the draft statement, we would like to put forward the following remarks: If on a liquid market a situation of forced sales can be difficult to identify, on a market with few transactions (or even just one transaction), and particularly on a market that has shifted from liquid to illiquid, it is possible that some of those few transactions (or even the transaction) have been entered into in order to address specific problems of the seller, and therefore the observed price do not represent a reliable input. However, the final part of paragraph 28 seems to imply that, unless there is sufficient evidence to the contrary, even on a market with very few transactions (and therefore, an illiquid one), the few observed transaction prices are to be considered a better measure of fair value than that resulting from a valuation model. As IAS 39 excludes the use of prices resulting from distressed conditions, the potential inclusion among the few transactions observed of distressed sales could have an important impact on price formation. Only if we can observe a sufficient number of transactions, narrow bid-ask spreads, with prices readily and regularly available from reliable contributors, we can consider quoted prices the best evidence of fair value. If, on the contrary, the above mentioned conditions are not met, quoted prices could be easily marred by transactions carried out not on an arm s length basis. Under such circumstances, we would deem it correct to rely on a valuation technique. Observed prices should be monitored and used as a benchmark for calibrating the model. 2
Moreover, as far structured financial products are concerned, it is very difficult to identify comparable transactions, as each product has peculiar features. Of course the use of models, as well as the process of definition of active market, should be properly documented in a formal framework - generally a fair value policy - as defined by independent risk management and approved by a management committee. Such document would provide consistency across time and any change should be documented and subject to a formal process of approval. To conclude as financial instruments should be valued on the assumption of the going concern concept, and not on a forced sale or distress sale situation - we would recommend the reversal of the presumption in paragraph 28: on an illiquid market the use of sporadic market prices (instead of a valuation model) to measure fair value is acceptable only if there is sufficient evidence that the transactions were carried out on an arm s length basis (paragraph AG69 of IAS 39 refers to the concept of going concern in the valuation of financial instruments). Question 2: Do you agree with CESR s views above regarding inputs to valuation techniques for financial instruments? Intesa Sanpaolo agrees with CESR statement, which is in line with related IFRS. The risk factors outlined in the report are the main sources of fair value uncertainty, in the light of the present market conditions and the structured credit turmoil. We fully agree with the concerns raised in using indices as inputs for the valuation process: in this case judgmental analysis is required in order to assess a comparison between these indices and the financial instruments to value. For Abx we agree that differences in the underlying credit structure, geographical location, and prepayment estimation, can be relevant sources for mis-pricing. Fair value back-testing in case of a sale is a powerful instrument we have widely used in our IPV process. In order to manage these sources of uncertainties a robust process of fair valuing should encompass a formalized process of model risk assessment and valuation. In this case a financial institution identifies the sources of uncertainties (complexity of the models or unobservable market parameters) and states the rule to assess and manage these. 3
Question 3: Do you agree with CESR s views above regarding disclosures of financial instruments in illiquid markets? We share CESR s general approach regarding disclosures of financial instruments. However, as to point 5 in Box 1, it seems unclear how prices (or values, if resulting from valuation models) for classes of financial instruments should be presented, as each instrument within a class has a different value. From a quantitative point of view, we would consider reasonable to present, for the various homogeneous classes, the notional / nominal values, the fair value at the end of the financial year, the fair value at the end of the previous financial year and any possible explanatory note on the reasons for the deviations. In general, however, the disclosure on valuation models should be easy to understand, whereas too detailed a technical annex on financial engineering underlying the valuation process may turn out not to be appropriate for disclosure purposes, to the point of being misleading and adding confusion. Question 4: Do you agree that the benefits of the presentation of disclosures regarding financial instruments in illiquid markets in the example in Box 2 outweigh the costs of preparing this information? Particularly during a time of financial turbulence, a comprehensive disclosure about the policies, criteria and processes used in the fair value measurement of financial instrument (and in particular, structured financial instruments) and about their economic impact, is clearly opportune, even if it imposes an additional burden on companies However, taking into account the wide range of financial instruments and valuation models (and their constant evolution) it seems hard to identify a standard tabular form fit for all products and businesses. Moreover IFRS principles provide for strict format of disclosures either. As to Box 2, we stress the fact that so far issuers have presented their measurements of illiquid financial instruments splitting them into homogeneous classes (e.g. CDO, ABS, RMBS, etc.) rather than according to the valuation 4
model applied. This approach is also recommended in the SSG report mentioned in the draft statement. In our opinion this presentation would be clearer than that proposed in the draft statement. The valuation model can be referred to through the three levels provided for by IAS 39 for determining the fair value (quoted prices, comparable approach, and mark-to-model approach). For any further comments or questions, please contact: Alessandra Perrazzelli Head of International Affairs alessandra.perrazzelli@intesasanpaolo.com Stefano Mazzocchi Regulatory Advisor Stefano.mazzocchi@intesasanpaolo.com Francesca Passamonti Regulatory Advisor francesca.passamonti@intesasanpaolo.com Intesa Sanpaolo S.p.A. International Affairs Square de Meeûs, 35 B 1000 Brussels Brussels, 12 September 2008 5