5/2008: AICPA views on fair value accounting

The Role of Fair Value Accounting in the Subprime Mortgage Meltdown
Three opinions in the May issue of AICPA's Journal of Accountancy

First the view "Both Sides Make Good Points" by Michael R. Young

We’re all familiar with what happened. This past summer, two Bear Stearns funds ran into problems, and the result was increasing financial community uncertainty about the value of mortgage backed financial instruments, particularly collateralized debt obligations (CDOs). As investors tried to delve into the details of the value of CDO assets and the reliability of their cash flows, the extraordinary complexity of the instruments provided a significant impediment to insight into the underlying financial data.

As a result, the markets seized. In other words, everyone got so nervous that active trading in many instruments all but stopped.

The practical significance of the market seizure was all too apparent to both owners of the instruments and newspaper readers. What was largely missed behind the scenes, though, was the accounting significance under Statement no. 157, which puts in place a “fair value hierarchy” that prioritizes the inputs to valuation techniques according to their objectivity and observability (see also “Refining Fair Value Measurement,” JofA, Nov. 07, page 30). At the top of the hierarchy are “Level 1 inputs” which generally involve quoted prices in active markets. At the bottom are “Level 3 inputs” in which no active markets exist.

The accounting significance of the market seizure for subprime financial instruments was that the approach to valuation for many instruments almost overnight dropped from Level 1 to Level 3. The problem was that, because many CDOs to that point had been valued based on Level 1, established models for valuing the instruments at Level 3 were not in place.
...
[W]hen subprime instruments were trading in active, observable markets, valuation did not pose much of a problem. But that changed all too suddenly when active markets disappeared and valuation shifted to Level 3. At that point, valuation models needed to be deployed which might potentially be influenced by such things as the future of housing prices, the future of interest rates, and how homeowners could be expected to react to such things.
...
Still, the subprime experience also demonstrates that there are two legitimate sides to this debate. For the difficulties in financial markets were not purely the consequences of an accounting system. They were, more fundamentally, the economic consequences of a market in which a bubble had burst.

And advocates of fair value can point to one aspect of fair value accounting—and Statement no. 157 in particular—that is pretty much undeniable. It has given outside investors real-time insight into market gyrations of the sort that, under old accounting regimes, only insiders could see. True, trying to deal with those gyrations can be difficult and the consequences are not always desirable. But that is just another way of saying that ignorance is bliss.

Sabaziotatos says:

A little too much fence sitting. Young nevertheless offers the useful observation that, when the markets seized up, financial institutions did not have in place robust Level 3 models. See, for instance, Citigroup's comments in its most recent conference call: "The methodology that we use has been refined and the inputs have been modified to reflect current conditions. The two principal refinements and modifications this quarter are the use of a more direct method of calculating projected HPA and a more refined method for calculating the discount rate."

Now, for the view "The Capital Markets’ Needs Will Be Served: Fair value accounting limits bubbles rather than creates them" by Paul B.W. Miller

The key to converging market and intrinsic values is understanding that more information, not less, is better. It does no good, and indeed does harm, to leave markets guessing. Reports must be informative and truthful, even if they’re not flattering.

To this end, all must grasp that financial information is favorable if it unveils truth more completely and faithfully instead of presenting an illusory better appearance. Covering up bad news isn’t possible, especially over the long run, and discovered duplicity brings catastrophe.

Sabaziotatos says:

As with most arguments that defend fair value accounting, the problem here is that Miller asserts that what is required is more information and that fair value accounting provides that information, while ignoring the fact that fair value and mark-to-market accounting by its pro-cyclical nature distorts information. As Plantin, Sapra, and Shin have written: "While the historical cost regime leads to some inefficiencies, marking to market may lead to other types of inefficiencies by injecting artificial risk that degrades the information value of prices, and induces sub-optimal real decisions. ... In this way, the mark-to-market regime generates endogenous volatility of prices that impede the resource allocation role of prices."

Finally, the view "The Need for Reliability in Accounting: Why historical cost is more reliable than fair value" by Eugene H. Flegm

HOW WE GOT HERE
The debate over the need for any standards began with the 1929 market crash and the subsequent formation of the SEC. Initially, Congress intended that the chief accountant of the SEC would establish the necessary standards. However, Carmen Blough, the first SEC chief accountant, wanted the American Institute of Accountants (a predecessor to the AICPA) to do this. In 1937 he succeeded in convincing the SEC to do just that. The AICPA did this through an ad hoc committee for 22 years but finally established a more formal committee, the Accounting Principles Board, which functioned until it was deemed inadequate and FASB was formed in 1973.

FASB’s first order of business was to establish a formal “constitution” as outlined by the report of the Trueblood Committee (Objectives of Financial Statements, AICPA, October 1973). With the influence of several academics on that committee, the thrust of the “constitution” was to move to a balance sheet view of income versus the income view which had arisen in the 1930s. Although the ultimate goal was never clarified, it was obvious to some, most notably Robert K. Mautz, who had served as a professor of accounting at the University of Illinois and partner in the accounting firm Ernst & Ernst (a predecessor to Ernst & Young) and finally a member of the Public Oversight Board and the Accounting Hall of Fame. Mautz realized then that the goal was fair value accounting and traveled the nation preaching that a revolution was being proposed. Several companies, notably General Motors and Shell Oil, led the opposition that continues to this day.

The most recent statement on the matter was FASB’s 2006 publication of a preliminary views (PV) document called Conceptual Framework for Financial Reporting: Objective of Financial Reporting and Qualitative Characteristics of Decision-Useful Financial Reporting Information. It is clear that FASB has abandoned the real daily users who apply traditional accounting to manage their businesses. The PV document refers to investors and creditors only. It mentions the need for comparability and consistency but does not attempt to explain how this would be possible under fair value accounting since each manager would be required to make his or her own value judgments, which, of course, would not be comparable to any other company’s evaluations.

Sabaziotatos says:

Excellent brief overview of the historical development of fair value accounting and the coup carried out by those who place more emphasis on the balance sheet than the income statement (or, say, the cash flow of underlying securities).

5/7/2008: Understanding Fair Value

Excerpts from Understanding the Issues: Some Facts About Fair Value
By FASB Chairman Robert Herz and FASB Director Linda A. MacDonald

However, the main purpose of this article is not to debate the pros and cons of fair value accounting. Rather, it is to provide some basic facts about fair value accounting that are important in understanding the current debate. Specifically, (1) where fair value is (and where it is not) used in financial reporting currently, (2) what fair value is (and what it is not), and (3) the approach for developing fair value estimates, including in illiquid markets.
...
For an asset, the fair value estimate is determined by reference to the price that would be received in an orderly transaction for the asset at the measurement date (an exchange price notion), not, as some have asserted, the price that would be received in a fire sale or forced liquidation transaction for the asset at the measurement date. An orderly transaction is one that involves market participants that are willing to transact and allows for adequate exposure to the market before the measurement date. In contrast, a fire sale or forced liquidation transaction is one that involves market participants that are compelled to transact (under duress) and allows for little (or no) exposure to the market before the measurement date.
...
The fair value hierarchy prioritizes observable inputs over unobservable inputs. However, the weighting of the inputs in the fair value estimate will depend on the extent to which they provide information about the value of an asset or liability and are relevant in developing a reasonable estimate of a current exchange price for the asset or liability. In making that determination, many factors need to be considered. Examples include the following:

• The extent to which observable inputs relate to transactions
  involving comparable assets or liabilities, considering both
  the nature of the transactions (orderly vs. forced) and the
  timing of the transactions (current vs. stale)


• The magnitude and subjectivity of adjustments to the inputs


• Factors specific to the market(s) in which the inputs are
  observed, such as a change in the volume of transactions and
  liquidity in a market that previously was active, a change in
  the availability of observable inputs, and a change in bid/ask
  spreads.
  

In some cases, for example, when there is little (or no) market activity for comparable assets or liabilities at the measurement date (illiquid markets) or when information about transactions involving comparable assets or liabilities is not publicly disclosed (principal-to-principal markets), the fair value estimate might rely principally on unobservable inputs (Level 3 estimates).

Sabaziotatos says:

This is an excellent article that provides much guidance from senior people at FASB on how to apply SFAS 157. This article is almost as useful as the SEC's Sample Letter Sent to Public Companies on MD&A Disclosure Regarding the Application of SFAS 157 (Fair Value Measurements) . In particular, FASB provides guidance on what to do in illiquid markets and more information on what an "orderly transaction" is.

5/11/2008: The fight over fair value

Excerpts from Gloves off on fair value
By Jeremy Woolfe

Signs of disharmony, verging on disarray, are emerging in the world of accountancy regulation over fair value reporting and its alleged potential to trigger a downward spiral in asset values.
...
The IASB’s consistent position of upholding the gold standard of fair value, as covered mainly by IAS 39, was reflected at a recent meeting with the US Financial Accounting Standards Board and the British Corporate Reporting Users’ Forum.

Reflecting the views of professional investors, the Cruf members said they “prefer fair value”, but also expressed concern about the reliability of their valuation models. Additionally, they admitted “fair value might be pro-cyclic, but that this would not be the real issue”. One example of what was “real” was need for cash flow data.

Putting pressure on the regulators from the other side of the fence are the banks. The International Banking Federation notes in a press release “that a mixed measurement model [for reporting financial instruments] is essential for the faithful representation of an entity’s business model”. Other banking institutions, including the European Banking Federation, endorses the view that fair value should not “always be an exit price”.

3/31/2008: Nouriel Roubini on mark-to-market accounting

Excerpts from Ten Fundamental Issues in Reforming Financial Regulation and Supervision in a World of Financial Innovation and Globalization
By Nouriel Roubini

Seventh, there are fundamental accounting issues on how to value securities, especially in periods of market volatility and illiquidity when the fundamental long term value of the asset differs from its market price. The current “fair value” approach to valuation stresses the use of mark-to-market valuation where, as much as possible, market prices should be used to value assets, whether they are illiquid or not.

There are two possible situations where mark-to-market accounting may distort valuations: first, when there are bubbles and market values may be above fundamental values; second, when bubbles burst and, because of market illiquidity, asset prices are potentially below fundamental values. The latter case has become a concern in the latest episode of market turmoil as mark-to-market accounting may force excessive writedowns and margin calls that may lead to further fire sales of illiquid assets; these, in turn, could cause a cascading fall in asset prices well below their long term fundamental value. However, mark-to-market accounting may also create serious distortions during bubbles when its use may lead to excessive leverage as high valuation allow[s] investors to borrow more and leverage more and feed even further the asset bubble. In either case, mark-to-market accounting leads to pro-cyclical capital bank capital requirement given the way that the Basel II capital accord is designed.

The shortcomings of mark-to-market valuation are known but the main issue is whether one can find an alternative that is not subject to gaming by financial institutions. Some have suggested the use of historical cost to value assets (where assets are booked at the price at which they were bought); others have proposed the use of a discounted cash flow (DCF) model where long run fundamentals – cash flows – would have a greater role. However, historical cost does not seem to be an appropriate way to value assets. The use of a DCF model may seem more appealing but it is not without flaws either. How to properly estimate future cash flows? Which discount rate to apply to such cash flows? How to avoid a situation where those using this model to value asset[s] subjectively game the model to achieve the valuations that they want as the value of the asset in a DCF model strongly depend on assumptions about future cash flows and the appropriate discount factor? Possibly mark-to-market may be a better approach when securities are held in a trading portfolio while DCF may be a more appropriate approach when such securities a[re] held as a long term investment, i.e. until maturity. But the risk of a DCF approach is that different firms will value very differently identical assets and that firms will use any approach different from mark-to-market to manipulate their financial results.

The other difficult problem that one has to consider is that any suspension of mark-to-market accounting in periods of volatility would reduce – rather than enhance – investors’ confidence in financial institutions. Part of the recent turmoil and increase in risk aversion can be seen as an investors’ backlash against an opaque and nontransparent financial system where investors cannot properly know what is the size of the losses experienced by financial institutions and who is holding the toxic waste. Mark-to-market accounting at least imposes some discipline and transparency; moving away from it may further reduce the confidence of investors as it would lead to even less transparency.

Some suggest that the problem is not mark-to-market accounting but the pro-cyclical capital requirements of Basel II; that is correct. But even without such pro-cyclical distortions there is a risk that financial institutions – not just banks - would retrench leverage and credit too much and too fast during periods of turmoil when they become more risk averse. Thus, the issue remains open of whether there are forms of regulatory forbearance - that are not destructive of confidence - that can be used in periods of turmoil in order to avoid a cascading and destructive fall in asset prices. But certainly solutions should be symmetric, i.e applied both during periods of rising asset prices and bubbles (when market prices are above fundamentals) and when such bubbles go bust (and asset prices may fall below fundamentals). But so far there is no clear and sensible alternative to mark-to-market accounting.

4/1/2008: CMSA calls for CMBX trading data

Excerpts from CMSA Calls for Greater Transparency in Trading Data
Commercial Mortgage Securities Association (CMSA), the leading voice of the commercial real estate capital market finance industry, last week requested that trading data on the CMBX Index, including total volume and number of daily trades, be made publicly available in order to increase market transparency. CMSA made the request in a letter to Markit, the administrator of the CMBX Index, a synthetic credit default swap derivatives index introduced in March 2006.

CMSA, an industry organization dedicated to promoting the ongoing strength, liquidity and integrity of commercial real estate capital market finance worldwide, has been instrumental in increasing the transparency in the structured credit markets.

"Public disclosure of derivatives trading data in the CMBX Index would provide an invaluable service to investors in the commercial real estate capital market finance arena," said Leonard W. Cotton, Vice Chairman of Centerline Capital Group and President of CMSA. "We believe the volatility in the CMBX index caused by momentum traders, rather than fundamental traders, distorts the true picture of the value of CMBS bonds, which are backed by the cash flows from loans on income-producing commercial real estate."

"Given the role the Index has come to play in determining the ‘mark-to-market’ value of securities held by financial institutions in the current market environment, greater transparency on CMBX trading volumes and the number of daily trades would aid investors in assessing the merit of values as indicated by the Index," Cotton added.

Dottie Cunningham, CEO of CMSA, expressed concern that the Index is not indicative of the underlying fundamentals of the investment product. "In a volatile market, this mark-to-market process becomes a self-fulfilling prophecy, driving prices down based on index trading activity rather than asset fundamentals," she said. "Some market participants may be relying on what we believe is a distorted value that perpetuates the current cycle of no issuance, erroneous spread widening and additional mark-to-market write downs."

4/10/2008: Bernanke on mark-to-market accounting

Excerpts from Bernanke: mark-to-market accounting challenging
Reuters story on question-and-answer session after Ben Bernanke's Richmond speech

Federal Reserve Chairman Ben Bernanke said on Thursday mark-to-market accounting has helped to destabilize markets for illiquid assets, but regulators need to be careful about any changes to the system.

"It's also true in the current context, that mark-to-market accounting has been sometimes destabilizing in that sales of assets into very illiquid markets had led to reductions in prices, which have caused writedowns which have sometimes caused firesales, and you get into an adverse dynamic which has caused problems in some of our markets," Bernanke said in a question-and-answer session before a business group,

On balance, he said mark-to-market accounting has been a positive influence for investors, but valuations should be determined during normally functioning, stable markets, not times when assets are illiquid.

5/5/2008: REIT Wrecks

REIT Wrecks is an interesting blog that discusses the commerical real estate market. A number of the entries discuss the impact of mark to market accounting on commercial real estate and CMBS's.

5/3/2008: Berkshire Hathaway, AIG, and fair value accounting for CDS's

It seems that even Berkshire Hathaway cannot escape the pernicious effects of fair value accounting.

The following are excerpts from Berkshire Hathaway's press release announcing Q108 earnings:

In the table at the top of the page (which, as noted, reports after-tax results), we give investment and derivatives gains (losses) a line of their own because the amount of these in any given quarter or year is usually meaningless.
...
[A]ccounting rules require that any unrealized gain or loss from derivatives contracts be regularly recorded in earnings. During the first quarter, Berkshire’s derivatives had an unrealized pre-tax loss of $1.6 billion. However, that is a “net” figure, incorporating gains and losses in several different kinds of derivatives. Fundamentally, the size of that figure reflects the fact that we recorded a $1.7 billion unrealized loss in our two major categories of derivatives.

In explanation of that loss, we will begin with the first of these major categories, credit default swaps (CDS). The ones we hold are related to various high-yield indices. As is explained in the section of our 2007 annual report that is appended to this release, at yearend Berkshire had received $3.2 billion in premiums on these CDS contracts and, to that date, had paid $472 million in losses. The contracts – with original terms of about five years – expire in years 2009 through 2013, and our yearend financial statements carried a liability of $1.84 billion for future losses. The weighted length of the contracts was then 42 months and was therefore reduced to 39 months by the end of the first quarter.

In the first three months of 2008 our paid losses were $52 million, considerably less than we expect to average on a quarterly basis. Despite this favorable experience, we raised our liability figure to $2.26 billion, an increase dictated by market changes that reflect heightened worries about corporate credit. That increase led us to record a loss of $471 million for this particular CDS group.

We also added some new five-year positions, receiving premiums of $229 million. On these contracts we paid no losses but established a quarter-end liability of $248 million, This gave us an additional loss of $19 million. Therefore our aggregate first-quarter loss on our CDS positions was $490 million (the total of $471 million and $19 million).

Overall, we now hold $2.9 billion of cash from CDS contracts, all available to us for investing as we choose. When prices are right, as they have been recently, we like the credit insurance business and believe it will be profitable, even without our taking into account investment earnings on the substantial funds we hold.

The press release also contains an excerpt from Berkshire Hathaway Inc.’s 2007 Annual Report – Chairman’s Letter, which states:

[W]e have written 54 contracts that require us to make payments if certain bonds that are included in various high-yield indices default. ... I believe that on premium revenues alone, these contracts will prove profitable ... Our derivative positions will sometimes cause large swings in reported earnings, even though Charlie and I might believe the intrinsic value of these positions has changed little. He and I will not be bothered by these swings – even though they could easily amount to $1 billion or more in a quarter – and we hope you won’t be either. You will recall that in our catastrophe insurance business, we are always ready to trade increased volatility in reported earnings in the short run for greater gains in net worth in the long run. That is our philosophy in derivatives as well.

The following are excerpts from Berkshire Hathaway's Q108 10-Q:

Under certain circumstances, including a downgrade of its credit rating below specified levels, Berkshire may be required to post collateral against derivative contract liabilities. However, Berkshire is not required to post collateral with respect to most of its long-dated credit default and equity index put option contracts. At March 31, 2008 and December 31, 2007, Berkshire had posted no collateral with counterparties as security on derivative contract liabilities.
...
Derivative gains and losses in the preceding table primarily represent the non-cash (or unrealized) changes in fair value associated with derivative contracts that do not qualify for hedge accounting treatment. As of March 31, 2008, outstanding derivative contracts primarily pertain to credit default risks of entities in the United States and equity price risks associated with major equity indexes, including three outside the United States. In the first quarter of 2008, pre-tax losses of $1.6 billion from derivative contracts were principally attributable to declines in equity indexes, declines in the value of the U.S. dollar versus the Euro and Japanese Yen, as well as a widening of credit default spreads in the United States.

The estimated fair value of credit default contracts at March 31, 2008 was approximately $2.5 billion, an increase of $667 million since December 31, 2007. The increase was due to fair value losses of $490 million, as well as $229 million in premiums from new contracts entered into in 2008 partially offset by loss payments of $52 million. The estimated fair value of the equity index put option contracts at March 31, 2008 was approximately $6.2 billion, an increase of $1.6 billion since December 31, 2007. The increase was primarily due to fair value losses of $1.2 billion as well as $383 million in premiums from new contracts entered into in 2008. There were no cash payments made under the equity index put option contracts.

The aforementioned contracts are not traded on an exchange. The contracts were entered into with the expectation that amounts ultimately paid to counterparties for actual credit defaults or declines in equity index values (measured at the expiration date of the contract) will be less than the premiums received. The contracts generally may not be terminated or fully settled before the expiration dates (up to 20 years in the future with respect to equity index put option contracts) and therefore the ultimate amount of cash basis gains or losses will not be known for years.

Berkshire does not actively trade or exchange these contracts, but rather intends to hold such contracts until expiration. Nevertheless, current accounting standards require derivative contracts to be carried at estimated fair value with the periodic changes in estimated fair value included in earnings. Fair value is estimated based on models that incorporate changes in applicable underlying credit standings, equity index values, interest rates, foreign currency exchange rates, risk and other factors. The fair values on any given reporting date and the resulting gains and losses reflected in earnings will likely be volatile, reflecting the volatility of equity and credit markets. Management does not view the periodic gains or losses from the changes in fair value as meaningful given the long term nature of the contracts and the volatile nature of equity and credit markets over short periods of time.
...
Berkshire’s estimated liabilities for credit default and equity index put option contracts were approximately $8.7 billion at March 31, 2008. Payments under the equity index put option contracts are contingent upon the future value of the related indexes at the expiration date of the contracts, the earliest of which is in 2019. Payments under credit default contracts are contingent on the occurrence of a default as defined under the contracts, and in the first quarter of 2008 were $52 million. The contract expiration dates begin in 2009.
...
The timing and amount of the payment of other obligations, such as unpaid property and casualty loss reserves and long duration credit default and equity index put option contracts are contingent upon the outcome of future events. Actual payments will likely vary, perhaps significantly, from estimates.

Note 6 of the 10-Q identifies the Berkshire's credit default swaps as Level 3 assets and describes the inputs for valuation of Level 3 assets as follows:

Inputs are unobservable inputs that are used in the measurement of assets and liabilities. Management is required to use its own assumptions regarding unobservable inputs because there is little, if any, market activity in the asset or liability or related observable inputs that can be corroborated at the measurement date. Measurements of non-exchange traded derivative contract assets and liabilities are primarily based on valuation models, discounted cash flow models or other valuation techniques that are believed to be used by market participants. Unobservable inputs require management to make certain projections and assumptions about the information that would be used by market participants in pricing an asset or liability.

Sabaziotatos says:

From the press release and the 10-Q, we may conclude:

1.) The CDS's are Level 3 assets.

2.) The increase in the liability figure for the CDS's was "dictated by market changes that reflect heightened worries about corporate credit." The phrase "market changes" refers to the "widening of credit default spreads." Thus, although the CDS's are classified as Level 3 assets whose valuation is based on "unobservable" inputs, the CDS's are being marked to market in the sense that their valuation reflects the widening of spreads in the market.

3.) The losses on the CDS's are for the most part "unrealized."

4.) Berkshire intends to hold the CDS's until expiration of the contracts. Berkshire thinks that the amount of unrealized CDS losses "in any given quarter or year is usually meaningless," In the CEO's letter, Warren states that "on premium revenues alone, these contracts will prove profitable" that "[o]ur derivative positions will sometimes cause large swings in reported earnings, even though Charlie and I might believe the intrinsic value of these positions has changed little," and that "[Charlie Munger] and I will not be bothered by these swings."

Berkshire's positions that the unrealized, temporary losses on the CDS's will likely not cause the CDS's ultimately to be unprofitable is identical to the position that AIG took on unrealized, temporary losses in its CDS portfolio at the end of Q407:

AIG continues to believe that the unrealized market valuation losses on this super senior credit default swap portfolio are not indicative of the losses AIGFP may realize over time. Under the terms of these credit derivatives, losses to AIG would result from the credit impairment of any bonds AIG would acquire in satisfying its swap obligations. Based upon its most current analyses, AIG believes that any credit impairment losses realized over time by AIGFP will not be material to AIG’s consolidated financial condition, although it is possible that realized losses could be material to AIG’s consolidated results of operations for an individual reporting period. Except to the extent of any such realized credit impairment losses, AIG expects AIGFP’s unrealized market valuation losses to reverse over the remaining life of the super senior credit default swap portfolio.

AIG's stock price was punished brutally when these valuation losses on its CDS portfolio were disclosed. The losses even moved AIG CEO Martion Sullivan to request that regulators modify fair value accounting rules:

The American International Group (NYSE:AIG) is urging regulators to change controversial accounting rules on asset valuations to stem the tide of writedowns that have wreaked havoc on Wall Street. The US insurer, which this month recorded a record quarterly loss after an $11bn mortgage-related writedown, is the first company publicly to air a proposal to move away from "fair value" accounting. … Martin Sullivan, AIG's chief executive, told the Financial Times that "mark-to-market" rules forced companies to recognise losses even when they had no intention of selling assets at the current prices. He said the practice created a vicious circle whereby companies recorded huge losses, lost investors' confidence and were then forced to raise funds at unfavourable prices. "This is not a criticism of fair value accounting. I just think it has had unintended consequences in these uncharted waters," he said.

So, even Berkshire Hathaway cannot escape the pernicious effects of fair value accounting. One wonders whether shares of Berkshire will suffer the same fate as those of AIG. I find it baffling that no one at the Omaha annual "Woodstock for Capitalists" did not ask Warren what he thinks of fair value and mark-to-market accounting.

5/2/2008: Vinod Kothari on mark to market accounting

Excerpts from Sentimental Accounting: Is it time to question mark to market?
By Vinod Kothari


First, on the pro-cyclicality. A business cycle arises when there are forces that move the market further away from equilibrium, when the market is already away from equilibrium. There are short-term self-cancelling spikes in markets but the central tendency of all markets is convergence. When losses exacerbate losses, or gains propel gains, the result is a business cycle. In normal circumstances, the beauty of the marketplace is that people hold different, and mutually opposite views on a certain thing. That is why they trade, and do so happily. In the MTM scenario, since every player in the market is an MTM player, everyone has losses to report on the P&L the moment the market takes an adverse turn and there is a broad-based fall in prices of financial assets. Though the loss is still a notional loss, the equity gets eroded. Equity is the basic premise on which all leverage is built in the world of finance – be it hedge funds, or CDOs or the banks. A dollar of equity being eroded might mean anywhere between 10 to 25 dollar of reduction of asset-holding capacity of the MTM player. This would mean, if the player was fully leveraged, it would have to shed assets or curtail exposures several times the notional loss. In a scenario where the market is already bad, the leveraged reduction in asset-holding ability would mean there are lots of sellers as compared to buyers. Obvious enough, this pushes the market further lower. This would mean more MTM losses, and more erosion of equity… Very quickly, we have the downside of the business cycle.

5/1/2008: Avinash Persaud on mark to market accounting

Excerpts from The Inappropriateness of Financial Regulation
By Avinash Persaud

My own view of banking regulation would be considered quaint next to today’s practice. I consider the primary objective of intervening in the banking market to be mitigating the substantial systemic consequences of market failure in banking. It is therefore puzzling to me that market prices are now placed at the heart of modern financial regulation, whether in the form of mark-to-market accounting or the market price of risk in risk models. It is not clear to me how we can rely on market prices to protect us from a failure of market prices.
...
Almost every economic model will tell you that if all the players have the same tastes (reduce capital adequacy requirements) and have the same information (public ratings, approved risk-models using market prices) that the system will sooner or later send the herd off the cliff edge ... And no degree of greater sophistication in the modelling of the price of risk will get around this fact. In this world, where falling prices generate more sell-orders from price-sensitive risk models, markets will not be self-stabilising but destabilising and the only way to short-circuit the systemic collapse is for a non-market actor, like some agent of the tax payer, to come in and buy up assets to put a floor under their prices.

Excerpts from Why bank risk models failed
By Avinash Persaud

When a market participant's risk model detects a rise in risk in his portfolio, perhaps because of some random rise in volatility, and he or she tries to reduce his exposure, many others are trying to do the same thing at the same time with the same assets. A vicious cycle ensues as vertical price falls, prompting further selling. Liquidity vanishes down a black hole. The degree to which this occurs has less to do with the precise financial instruments and more with the depth of diversity of investor behaviour. Paradoxically, the observation of areas of safety in risk models creates risks, and the observation of risk creates safety. Quantum physicists will note a parallel with Heisenberg's uncertainty principle.
...
This brings us to the philosophical problem of the reliance of supervisors on bank risk models. The reason we regulate markets over and above normal corporate law is that from time to time markets fail and these failings have devastating consequences. If the purpose of regulation is to avoid market failures, we cannot use, as the instruments of financial regulation, risk-models that rely on market prices, or any other instrument derived from market prices such as mark-to-market accounting. Market prices cannot save us from market failures. Yet, this is the thrust of modern financial regulation, which calls for more transparency on prices, more price-sensitive risk models and more price-sensitive prudential controls. These tools are like seat belts that stop working whenever you press hard on the accelerator.

The reliance on risk models to protect us from crisis was always foolhardy. In terms of solutions, there is only space to observe that if we rely on market prices in our risk models and in value accounting, we must do so on the understanding that in rowdy times central banks will have to become buyers of last resort of distressed assets to avoid systemic collapse.

Excerpts from Sending the herd off the cliff edge: the disturbing interaction between herding and market-sensitive risk management practices
By Avinash Persaud

While many believe that market sensitive risk management, prudential standards and transparency are probably not enough to avoid future crises, they believe these measures will probably help to provide the right discipline for governments and can surely do no harm. These measures are likely to be a positive force in the long run when markets are better at discerning between the good and bad. But in the short-run, there is growing evidence that market participants find it hard to discern between the good and the unsustainable, they often herd and contagion from one crisis to another is common. The problem is that in a world of “herding”, tighter market-sensitive risk management regulations and improved transparency can, perversely, turn events from bad to worse, creating volatility, reducing diversification and triggering contagion. How can this happen?
...
The growing fashion in risk management, supported by the Basel Committee on Banking Supervision, is a move away from discretionary judgements about risk and a move to more quantitative and market sensitive approaches ... This is well illustrated by how banks now tend to manage market risks by setting a DEAR limit - daily earnings at risk. DEAR answers the question: how much can I lose with, say, a 1% probability over the next day. It is calculated by taking the bank’s portfolio of positions and estimating the future distribution of daily returns based on past measures of market correlation and volatility. Both rising volatility and rising correlation will increase the potential loss of the portfolio, increasing DEAR. Falling volatility and correlation will do the opposite. Banks set a DEAR limit: the maximum dollar amount they are prepared to put at risk of losing with a 1% probability. When DEAR exceeds the limit, the bank reduces exposure, often by switching into less volatile and less correlated assets such as US dollar cash.
...
Imagine that over time a herd of banks have acquired stocks in two risky assets that have few fundamental connections, say, Korean property and UK technology stocks. Imagine too that some bad news causes volatility in UK technology stocks and the banks most heavily invested there find that their DEAR limits are hit. As these banks try and reduce their DEAR by selling the same stocks (Korean property and UK technology) at the very same time, there are dramatic declines in prices, rises in volatility in both markets and in the correlation between Korean and UK markets. Rising volatility and correlation triggers the DEAR limits of banks less heavily invested in these markets but invested in other markets. As they join the selling milieu, volatility, correlation and contagion rises.

5/1/2008: Henry Kaufman on mark to market accounting

Excerpts from On Money and Markets, McGraw Hill, 2000
By Henry Kaufman

That highly securitized markets are, on the whole, more volatile than less securitized markets has important implications. In volatile markets -- as in recent times, and surely during periods in the future -- liquidity may disappear suddenly, accurate pricing becomes exceptionally difficult to obtain, and marking to market may be practically impossible.

Why is this so? And what are the consequences? To begin with, the price of the last trade may be completely invalid in rapidly moving markets, particularly for illiquid securities and certainly for most options. Moreover, the price that a dealer is prepared to quote may be little more than an indication of what the security is worth, not the price at which the dealer is actually willing to trade. Another dealer may quote -- on the same “indications-only” basis -- a wildly different price. For the institution trying to mark the position to market, there is simply no reliable arbiter of “true” price.

More than that, the price quoted may be valid for trading only a very small amount, not the full amount that the investor has in his portfolio.

…

The consequences of not being able to mark to market accurately are often far from trivial. At best, dealers face embarrassment when clients call attention to inaccurate estimates of portfolio returns. At worst, they face the calamity of enormous losses and the impairment of capital. [pp. 56-7]

In years past, the procedure of reevaluating the net asset values of portfolios on a daily basis (to reflect changes in asset prices) was practiced mainly by securities firms. ... Banks and insurance companies did not mark to market, and their regulators did not want them to do so. They mainly held securities on their balance sheets at book values. They did not recognize interim (or unrealized) profits or losses. They tended to buy and hold. The great bulk of their assets was not securitized, and was not amenable to repricing in an active secondary market.

But in recent years ... changes in accounting guidelines and regulatory rules have encouraged an irreversible movement toward greater marking to market by traditional institutions. Securitization has undermined the key argument against marking to market -- namely, that valid prices cannot be determined -- while the passage of time has worn down both official and industry opposition to the practice.

Even so, it is important to recognize that marking to market is an imperfect process, especially under difficult market conditions. Generally speaking, [in times of rising asset prices?] it tends to overstate values and to offer investors a kind of false comfort. When market conditions deteriorate and liquidity seizes up, no one can really claim that the last quoted price in organized markets (such as the NYSE) or quoted by the dealers in the over-the-counter market is the real market value – at least not without taking into consideration the size of an intended trade, the credit quality of the issuer, the activity of other market participants, and related dimensions of the transaction. [pp. 316-7]

4/28/2008: BOE Financial Stability Report, Issue 23

Excerpts from Bank of England's Financial Stability Report, April 2008, Issue 23 :


While realised losses in sub-prime markets have been small to date, large complex financial institutions (LCFIs) have announced substantial write-downs ... These estimates of expected future losses have been based on a range of indicators, including internal models and market prices such as the ABX. There has been considerable variation in reported valuations across products by different financial institutions ... And disclosures have often provided only partial information on the assumptions underlying valuations and the uncertainties around reported point estimates of losses.

Falls in indices such as the ABX have led to increasing estimates of system-wide mark-to-market losses ... The gap between these estimates and reported write-downs to date by firms have fuelled expectations that further losses are in store but have yet to be revealed. This has added to the mood of uncertainty and pessimism in financial markets, which is retarding the recovery of confidence and risk-taking.

But credit losses from the turmoil are unlikely to ever rise to levels implied by current market prices unless there is a significant deterioration in fundamentals, well beyond the slowdown currently anticipated. That is because prices are likely to reflect substantial discounts for illiquidity and uncertainty that have emerged as markets have adjusted but which should ease over time. While market-based estimates and the write-downs announced by firms may be unduly pessimistic, if such concerns persist there is a risk they could become self-fulfilling.

...

[One] respect in which the loss estimates may be misleading is because they confuse true credit losses and losses implied by market prices. These two approaches can differ markedly at times when market prices deviate significantly from credit fundamentals — for example, when illiquidity and uncertainty discounts in market prices are large, as at present.

...

The above analysis suggests that using a mark-to-market approach to value illiquid securities could significantly exaggerate the scale of losses that financial institutions might ultimately incur. It will exaggerate to an even greater extent the potential damage to the real economy that these losses might inflict, since there are always winners and losers to financial contracts. This does not deny, however, the possibility of some adverse consequences for the real economy as a result of recent events — for example, due to a higher cost of capital for some borrowers.

...

ABX prices are often used to estimate mark-to-market losses on sub-prime securities. These loss estimates have also risen, to around US$380 billion. These losses have often acted as a benchmark for public commentary on possible bank write-downs. But as [our analysis] explains, estimates based on projected credit losses are considerably lower, suggesting there are large illiquidity and uncertainty premia in the ABX market. So although ultimate realised losses on sub-prime mortgage securities could be high, market prices appear to be giving an overly pessimistic impression of their eventual scale. This may be one important factor weighing on market confidence and retarding the recovery of risk appetite.

...

Under fair-value accounting (FVA) rules, firms are required to mark to market when an active market exists and a price is observable. This is the appropriate way to ensure that those parts of the balance sheet that are subject to FVA reflect current market values. But difficulties can arise when markets suffer [certain types of dislocation] — that is, when prices may be distorted by temporary factors, such as poor or no liquidity. In these circumstances, accounting standards setters need to provide authoritative guidance on the valuation of financial instruments when markets are no longer active and on what constitutes an active market. And audit standards setters need to provide robust guidance that promotes consistent auditing practices.