Mitigating against pro-cyclicality in regulatory policy is one of the points from the Declaration: Summit on financial markets and the world economy for Finance ministers to come back on, early 2009.

Preliminary remark: risk-managers have long warned that regulations as they were would be pro-cyclical: making things better when they're good, making things worse when they're bad. Politics, prompt at the moment to blame bankers and to think about strengthening regulations, consistently ignored the feedback though. Was it, for good reason, because the feedback was coming from the private sector, and lobbies should not be allowed to tweak laws? Not quite so. Warnings came from authorities and international institutions such as the Bank of England (e.g. working paper 181, 2003: Procyclicality and the new Basel Accord - banks' choice of loan rating system) or the Bank for International Settlements (e.g. working paper 193, 2006: Procyclicality in the financial system: do we need a new macrofinancial stabilisation framework?)... Politics find it very convenient at the moment forgetting they did not act at the time; and now playing the saviors against the banking villain!

The key regulations one needs to know of to tackle procyclicality are: the Basel Committee on Banking Supervision's Basel II framework (for capital adequacy), and the Financial Accounting Standards Board's statement FAS 157 (on fair value measurements).

Already some decent amount of ink has been splashed regarding FAS 157. Usually, it is blamed for forcing a mark-to-market to distressed prices. The issue at hand is a vicious circle: forcing companies to mark "paper" losses may significantly endanger the system. Due to these losses, confidence in and credit rating of a company may go down, which could trigger e.g. business slowdown or collateral margin calls. Although the company might not have needed to sell its assets at the current distressed levels in the first place, it ends up in trouble because of paper losses, then forced to sell assets to cover treasury constraints. The paper losses may de facto force the company to materialize such losses. Not only the forced sale is at a bad timing, it also maintains pressure on the price of the assets themselves, potentially making things worse for other companies holding such assets and, in turn, forced to mark corresponding paper losses... The spiral in this scenario is evident but few commentators notice that this scenario is not and should not be the only possible outcome of FAS 157.

The FAS 157 norm just does not force marking at distressed levels. It requires substantial inadequate rigidity (or willingness to protect one from any potential litigation) from regulators, accountants, auditors, etc. to interpret FAS 157 as forcing mark-to-distressed. People at all levels and in all functions who try to be blameless while killing companies by blindly ticking boxes are just in practise evading their responsibilities of using the rules intelligently. Intelligently does not mean bending the rules, but one can nonetheless recognise that stupidly following guidelines is not properly following guidelines. FAS 157 has scope for interpretation of the value of assets and the responsibility of interpretation should not be abandoned.

FAS 157 is clear: the fair value of an asset or a liability is the hypothetical exit price (selling the asset or buying a cover of the liability) in an orderly transaction. The goal of such accounting standards is the fair representation of the value of a company thanks to rules ensuring that different companies report similar values for similar assets and liabilities. The simplest option to achieve such consistency is to value assets and liabilities based on observable prices on an organized and liquid market. Hence, when an organized and liquid market is available, it makes sense to rely on such a market in priority and the norm may say "fair value is a market-based measurement, not an entity-specific measurement."

The existence of such a market is not automatic though; numerous contracts (OTC, over-the-counter) are not valued in any organized (or liquid) way and their value must be established based on the loose notion of "market practise" (for consistency) and on observation of other contracts linked more or less directly and tightly to the OTC. Moreover, at a particular point in time, it is possible that observations available on regular basis temporarily disappears (the markets they're based on sees their liquidities dry out). In such a case, the norm envisions the use of market hypothesis or models, based on the best degree of information available in the circumstances (e.g. by extrapolating from variations of historical observations). One would then talk about unobservable inputs. The preamble of FAS 157 is absolutely limpid regarding these: The notion of unobservable inputs is intended to allow for situations in which there is little, if any, market activity for the asset or liability at the measurement date. In those situations, the reporting entity need not undertake all possible efforts to obtain information about market participant assumptions. (bold stress by myself).

Unfortunately, debate comes back in with the next sentence of the preamble: However, the reporting entity must not ignore information about market participant assumptions that is reasonably available without undue cost and effort. Is the observation of the price of a distressed asset in a panic sale a reliable information, reasonably available without undue cost and effort? Unreliable information may well not be information at all! If it is part of a block of similar information items, a holistic approach might allow for conclusions but, in isolation, poor information is... poor. When a big American investment bank get rid of $5bn notional of CDOs at the hugely discounted price of 22%, what is the information? What do auditors of other firms know about the quality of the assets in the mentioned CDOs, their composition, their representativeness of the market as a whole? What information does one get regarding how orderly the transaction was? If the sale was forced by treasury constraints, can one assume other big American investment banks necessarily have equivalent issues of treasury or regulatory capital?

FAS 157 offers scope for debates. It states e.g. that A fair value measurement should include an adjustment for risk if market participants would include one in pricing the related asset or liability, even if the adjustment is difficult to determine. This sounds reasonable at first sight... However, is a distressed price a price adjusted for some risk? Yes. Is the risk related to the asset? No, it is related to the potential buyer/seller who is scared of "locking" some money in an illiquid investment and having to sell at an even-bigger discount later because it cannot hold-and-wait due to regulatory or treasury reasons. Would all market participants have the same adjustment? No. Should they all have one? No, it only depends on the financial strength of the investor (no foreseeable problem of cash, no adjustment). Should fair value include such adjustment? Most accountants, regulators, auditors, would say yes. Such adjustment would not be consistent across entities though, defeating the very purpose of the norm... Another example: as a financial institution, I prefer a diversified portfolio so I'm selling financial products but I tend to adjust my price so I become less and less competitive as I pile up a specific risk (so market competition naturally helps me avoiding too much of one single risk). The market as a whole would not have the same adjustment because it does not have the same portfolio or even the same risk aversion. What adjustment should be considered, which would offer the consistency the norm is about?

As a conclusion, we have seen that the accounting norms on mark-to-market induce pro-cyclicality in the system (due to the vicious circle of distressed assets sales triggering paper losses triggering other sales): when things go bad, the regulation makes them go worse. We have however also seen that the archetypal FAS 157 would not have such pro-cyclical consequences, should accountants, regulators and auditors do their jobs instead of trying to cover themselves by ticking boxes and forgetting that the norm requires judgement, not just a spreadsheet program. Being conservative when valuing assets is not trashing them as much as possible. Mark-to-market rules do not need amendments, the G20 does not need to tackle it... The issue is about the "blame culture", about witch hunting, about lowly-paid regulators, accountants, auditors, with no willingness or skill to actually value assets and liabilities instead of just using any data available no matter its quality.

If there must be a reform related to mark-to-market, it should not be changing the accounting standards. it could be about paying regulators and accountants properly (so to attract skills and willingness). Looking for a more profound proposal? Reform could be about creating "Chinese walls" between the traders pricing products, and the traders managing them... In finance, management is about locking the value of a portfolio by buying / selling products and thus cancelling the risks of one with another. Traders managing products should be paid on their ability to lock the value of the portfolio: the closer they stay of 0 during a year, the higher their bonus... If they make a lot of money, it actually means they took some decent risk, and their bonus should be reduced... Proprietary traders should be the only traders allowed to take substantial risk. Hedgers/managers don't need to know the value of the book (let alone the price attached to the initial trade); better, they should not know (to avoid the bias of being lazy if the context was very profitable so far or the bias of having to catch-up if losses were suffered). Marketers should be paid on their profits or losses (traded price - fair price), risk managers should be paid on their consistency, proprietary traders should be paid on their profits or losses (knowing they trade at market price and exit at market price but are allowed to run risks (within limits)). Most hedgers are too concerned with profits and losses to do a good job... If new policies are about ensuring financial institutions manage risks better, let's make sure the traders managing risks are just doing that (that does require intervening in remuneration policies, but that does not prevent financial innovations, etc. Not so heavy handed as one could anticipate!)