American Enterprise Institute

April 8, 2008

[Edited transcript from audio tapes]

2:45 p.m.	Registration
3:00	Introduction:	Peter J. Wallison, AEI
3:15	Panelists:	Craig Gifford, Guaranty Financial Group, Inc.
		Harvey L. Pitt, Kalorama Partners LLC
		Vincent R. Reinhart, AEI
		Mark Scoles, Grant Thornton LLP
		Leslie Seidman, Financial Accounting Standards Board
		Gerald I. White, Grace & White, Inc.
	Moderator:	Peter J. Wallison, AEI
5:30	Adjournment

Proceedings:

 Peter Wallison:  Okay, I think we’ll get started if everyone can take his or her seat.  I’m Peter Wallison.  I’m a Senior Fellow here at the American Enterprise Institute, and I want to welcome all of you to what I believe will be a terrifically interesting and informative conference today on fair value accounting. 

A few weeks ago, the SEC’s Division of Corporation Finance sent an unusual letter to public companies about the applicability and interpretation of Statement of Financial Accounting Standards 157.  This is the FASB statement that is at the heart of the fair value accounting controversy.  You’ll find a copy of that in the blue folders.  This is the SEC’s letter, and it deserves reading because you have to parse it somewhat carefully in order to understand what the SEC was trying to say, but it seemed to me the key paragraph was this:  “Fair value accounting,” the SEC’s Division wrote, “assumes the exchange of assets or liabilities in orderly transactions.  Under SFAS 157, it is appropriate for you to consider that here they’re talking to the financial and accounting staffs at the public companies.  It is appropriate for you to consider actual market prices or observable inputs, even when the market is less liquid than historical market volumes, unless, still quoting here, those prices are the result of a forced liquidation or a distress sale.  Only when actual market prices, or relevant observable inputs, are not available is it appropriate for you to use unobservable inputs which reflect your assumptions of what market participants would use in pricing the asset or liability.” 

 This is a nice straddle, it seems to me, and probably left recipients with as many questions as they had before.  Preparers of financial statements, said the Division, are supposed to use market prices, even if the market is “less liquid” than normal, unless – and this is a big unless - those prices are the result of a forced liquidation or distress sale.  The SEC has, thus, left companies with the key decision: Whether to use the prices the market is generating.

 There is no question that markets for asset-backed securities, and even interbank and secured lending are not functioning normally right now.  This is because of uncertainty about losses – where they are and their ultimate size.  According to some participants, the credit markets are now virtually shut down.  Bid-asked spreads are wider than most market veterans can remember.  But some sales are going on anyway because some companies must sell in order to raise cash.  Are these sales what the SEC or the auditors would define as “distress prices” or is this just a market with lower volumes, as the SEC described it, than usual?  There are also derivative markets of various kinds that provide some information about how the underlying securities might be priced if sold.  Are these observable market inputs as described in SFAS 157?

 Despite all the heat that has been generated about fair value accounting, there has been little light shed on what companies are actually doing.  Yet this is the key question.  Under SFAS 157, they must use market prices, unless these are the direct result of a distress sale.  If they conclude, somehow, that what they are seeing in the market is the result of a distress sale, they may use other unobservable valuation methods that are more obscure – models, estimates of cash flows, and other internal means.

 How are they deciding whether the prices in the market are the result of distress sales?  And if they so decide, what methods are they using to calculate the value of their assets?  What role, in this connection, do independent auditors play in this process, and how do they interpret SFAS 157 and the SEC’s message?

 A few weeks ago, S&P analysts wrote, “when we dissect the percentage of writedowns taken against various types of exposures, the magnitude is greater than any reasonable estimate of ultimate losses.”  This is a very serious charge.  If asset values have been written down excessively, financial intermediaries like banks and securities firms are financially stronger than they look.  Consequently, the credit crunch – in which financial markets have been virtually shut down for months – is worse than it needed to be.  Investors have taken unnecessary losses, and the U.S. economy may slide into a recession that may be more serious than it otherwise would be.

 On the other hand, what is the alternative?  If asset values are falling, is it realistic to treat companies that are holding these assets as though there was no change in their financial condition?  Investors and lenders certainly don’t think so.  They have been calling on companies to back their borrowings with more collateral, or withdrawing their financing entirely.  In part, this is responsible for the distress sales, since companies have been compelled to liquidate assets that are no longer acceptable collateral. 

In addition, if market prices do not reflect reality, what does?  If companies are permitted to use unobservable methods to value their assets, how are we sure that they are not overstating those values?  Could things be worse than they actually look?  Again, what role are the auditors playing in this decision?

 Josh Rosner, a shrewd observer of the markets, noted in Friday’s Wall Street Journal, “A failure of institutions to rectify opacity will propel us further down the road toward a Japanese-style lost decade.”  This is the other side of the problem.  If assets are not written down enough, or at least to a level that reflects a rough reality, banks and other lenders will be slow to lend.  This is true because their capital condition is weak and they must allocate capital to new loans.  This reduced level of credit will restrain growth and recovery from any recession we might experience. 

This is exactly what happened in Japan.  The Japanese banks refused either to recapitalize or to recognize immediately losses that they had already incurred.  As a result, they recognized their losses slowly, writing them off against profits annually, and it took more than a decade to get the Japanese economy back on track.

 So there are two sides, it seems to me, to the fair value accounting story, and both reflect the seriousness of the issue we will be discussing today.  If assets are being written down too deeply, it could be producing and reinforcing an unnecessary downturn, but if companies are not recognizing the full extent of their losses, it could produce a long period of recession and slow growth.

 There are not likely to be clear answers here.  It is surely correct that assets in the normal case should be valued at what they are worth in the market.  The market is the only valid standard of value.  On the other hand, we don’t know quite what to do when the market isn’t functioning normally.  What standard is applicable then?

 What we will try to learn at this conference is what FASB had in mind when it wrote SFAS 157, how FASB thinks the rule should function in the current market, how the audit profession is enforcing the rule at the level of companies, and what companies, analysts, and people with regulatory experience think about all of this.

 What we’ll do is we’ll start with FASB.  Our FASB specialist will give her presentation, and I’ll introduce each of the people in our panel in turn.  We have an extremely distinguished and learned panel today, and so we hope to learn a lot from all of them.

 Leslie Seidman is the person from FASB who will be making their presentation.  She was appointed to the Financial Accounting Standards Board in July of 2003.  Prior to joining the Board, she managed her own firm, providing consulting services to major corporations, accounting firms and other businesses.  Previously she was Vice President of Accounting Policy at JP Morgan and Company, where she was responsible for establishing accounting policies for new financial products and analyzing and implementing new accounting standards. 

Prior to launching her consulting practice, Leslie served the FASB in various capacities, most recently as Assistant Director of Implementation and Practice Issues, but also as Industry Fellow and Project Manager.  She is the author of the first three editions of Financial Instruments, which is published by CCH, a comprehensive practice manual for accountants and other professionals.  Incidentally, as to all the people I will be introducing, there is more biographical data than I will be using in your blue folders.  Leslie.

 Leslie Seidman:  Thank you, Peter.  I’m delighted to be here today to try and provide you some background information about the accounting requirements for fair value accounting, as well as some other relevant accounting standards, the Board’s rationale behind those accounting standards, and provide some commentary on their relevance in today’s market situations.  As a member of the Board, I am required to provide a disclaimer and just let you know that the comments that I make today are my own.  It’s very important if you have the expectation that I’m going to answer your specific accounting questions, I am not allowed to do so.  And this presentation by no means represents a comprehensive analysis of the standards.

 But with that said, let me start with the background behind a lot of the buzz in the press and other circles relating to fair value accounting.  We issued a standard in September of ’06 that set forth a definition of fair value for use in financial reporting.  Now many accounting standards, I think over thirty-five, had previously used the term fair value, but it was addressed in an ad-hoc fashion, and so each standard had a slightly different definition.  The purpose of Statement 157, or I’ll call it FAS 157, that’s our lingo, was to conform the definition and sort of use a “best of” approach to come up with a clear objective for future use in financial reporting so that when the board says carry something at fair value, this is the definition that you will use.  It is a long-standing concepting gap. 

I’ll try and highlight a couple of the key areas where fair value has been used in the past.  It is not new, and furthermore, 157 did not expend any of the current requirements for using fair value.  Rather, if there was an existing standard, it simply clarified how to calculate fair value, and it also provided some new disclosure requirements relating to the use of fair value in the financial statements.

 So here is the definition just so we’re all on the same page with respect to what we mean by fair value:  It’s the price that would be received to sell an asset or pay to transfer a liability in an orderly transaction between market participants at the measurement date.  So in maybe more simple terms, it’s intended to represent an exchange price that reflects current expectations about the cash flows for the item that you’re valuing from the perspective of a willing, informed and unrelated potential counterparty.

 So what does this rule out?  Fair value is not mark to make-believe, mark to management or unfair value.  And these are a little bit snide, I would say, but I picked them out of the press because people are using these terms, so let me just try and debunk some of the misunderstanding that’s out there.  What I mean by mark to make-believe is you cannot just use a model in the most favorable assumptions that you would like to use to substantiate a value that no one would actually transact at.  Mark to management means that it is not what management paid for the item.  It is not what management hopes will be the value in the future if they hold it to maturity.  Rather, it is intended to reflect the value they would get if they were to sell in a willing transaction, an arm’s length transaction today.

 Importantly, fair value is not a liquidation value negotiated in a fire sale or another distressed situation.  This is a principle that has been in GAAP for a very long time.  It is not new to say that it does not include distressed sales because that would violate the objective of trying to reflect a willing party and an arm’s length transaction.

 So what I’d like to start with is a discussion of where fair value is required to be used in financial reporting standards today.

 The first category I’d like to talk about is ongoing fair value measurements or what a lot of people call mark to market accounting.  And what that means is that you’re going to be marking the position to fair value every reporting period and then taking the changes in value through earnings.  So it is part of a company’s net income when this model is employed.  You might take a step back and say, what are the cases where the FASB has required the use of mark to market accounting? 

And here, again, I’m just going to emphasize that I’m trying to simplify this and provide this discussion in general terms so don’t tick and tie this to the literature.  But generally speaking, fair value is required to be used to reflect a business model that reflects either trading of positions or speculating.  So in other words, the business is being managed on a fair value basis.  And it’s not so relevant whether you actually have sold.  The accounting is intended to reflect what the P & L would be or the results of the entity would be if you did sell.  So whether you did or you didn’t, the P & L reflects your decision based on the current fair value of the position. 

This model has been employed for broker dealers and mutual funds for decades, and I think its source is – Harvey, I don’t want to be wrong here, but I think it’s either the 34 Act or the 40 Act or both – the 40 Act, okay – so, again, it’s not a new concept and in fact, that regulation uses the term fair value.  And we actually cross-checked frequently with the regulation to make sure that we weren’t inventing a new approach to fair valuation.

 The second general reason that the Board would require fair value to be used in an accounting standard is that the item being measured has highly variable potential cash flows.  So a classic example of this would be derivatives.  Derivatives often have a transaction price of zero, but as market conditions change, zero is completely irrelevant for assessing the risk position in a derivative.  So we require that all derivatives be carried at fair value, and the changes in value are generally recognized in income unless it’s some sort of a hedging situation.

 The last category is what we call the fair value option, which is an election that an entity may make to carry any financial instrument at fair value.  And the primary reason someone would do this is to make their overall risk position more coherent if there isn’t already a requirement to carry things at fair value.  So if you had part of the balance sheet being mark to market but the other part not, you can elect fair value so that your whole balance sheet is at fair value.  We receive feedback from investors frequently, and the CFA Institute, for example, has repeatedly asked us to move to carry the entire balance sheet at fair value.  So that is not the current state of affairs, but we did just issue an invitation to comment that raises the question of whether all financial instruments should be carried at fair value, so I’d encourage you to take a look at that if that’s something you’d like to provide comments on.

 Fair value is also used in financial reporting to measure impairments.  Now that’s different from mark to market accounting in that it only recognizes losses, and at that, in some cases, the losses are only recognized when a trigger is hit.  The first category that I think is relevant to today’s discussion is the accounting for loans held for sale.  For example, if you have a loan portfolio that you’re intending to securitize, those loans are required to be carried at the lower of cost or market.  So, again, if in the aggregate there is a loss in the portfolio, you would write them down and only take any recoveries back up to cost. 

There is also a requirement to recognize other than temporary impairment on securities, but here you have a trigger.  To the extent that it’s probable that you don’t expect to collect all cash flows during your intended holding period, you have to take an other than temporary impairment through earnings.  So these are not the same as mark to market accounting.  They only recognize losses, not gains.

 I want to briefly touch on current accounting requirements that are not based on fair value because I’m afraid that we paint with a broad brush sometimes when we’re talking about the subprime situation and other current issues relating to the credit markets, and not everything is carried at fair value.  Guarantees is one case.  If you look at, for example the mono-line insurance companies, they are not carrying their guarantees at fair value.  And loan impairments often are not recognized at fair value. 

We have two different approaches to accounting for loans, for example, at a commercial bank or a mortgage bank where, generally speaking, the entity has to first conclude that it’s probable that they won’t collect all the case flows and then either account for the loss based on the cash flows they don’t expect to collect or, if it’s a larger loan or a loan that’s been restructured, they discount those new, lower cash flows at the original affective rate in the loan, which is not the same as fair value where you’d use a current interest rate.  So, just to emphasize, not everything that you may want to talk about today is a fair value accounting issue.

 Let me just spend a few minutes talking about the actual requirements of Statement 157.  One of the key objectives that we set out to achieve was to try and bring a little more consistency to fair value measurements that are occurring when an accounting standard requires fair value.  Generally speaking, the standard sets forth a hierarchy or a framework for accountants to use when they’re trying to value something at fair value. 

The first principle underlying a fair value measurement is that it’s intended to reflect an exchange price between market participants.  And I’m not going to repeat all of the elements of that definition, but the first place you would look, of course, is the markets to the extent that you have a quoted price and an active market for the exact item that you’re trying to measure.  That’s what we call Level 1.  And it would be really nice if everything had a Level 1 measurement attribute to use.  It would be very easy to just go through and carry everything at fair value if that were the case, but we all know that’s not the case.  So the improvement that we made here was to say, what do you do when that’s not the case?

 The next category is what we call Level 2, and it’s still predominantly observable data whether you have a price for something similar to the item that you’re measuring or you have other observable data in the markets that you can look to that are representative of the item being measured.  For example, you might have a yield curve you can look at, you might have rates, indices, et cetera, that are reflective of the position that you are trying to measure or other inputs that are derived from what we could all consider to be market sources.

With respect to the markets today, a couple of points I’d like to emphasize are that not everything that’s observable is a price.  Not every price necessarily relates to the item that’s being measured.  And management must use judgment and make adjustments, if necessary, to take the information that is available in the market and adjust it for the item that’s being measured.  So I frequently read things that say, you must take any price that you can find in the market and value your portfolio using that.  We do not believe that that’s an appropriate interpretation of the standard.  There are many steps you need to take before you would conclude that that was an appropriate valuation for a position that is not traded in an active market.

 Now when you have some data – I’m going to make up a scenario – there is some sort of a corporate bond trading that has puts and calls and some other whistles on it, and the thing that you’re trying to measure is a simple bond instrument.  To the extent that the adjustments that you need to make to take the data that’s available in the market and relate it to the position that you’re measuring, to the extent that those adjustments are significant and subjective, the nature of the adjustments may say, you no longer have something that we would consider Level 2, and instead you would move into Level 3, which is an estimate that is primarily based on unobservable inputs.

 The purpose of establishing a Level 3 category is to say, even in cases where you do not have observable inputs for every aspect of a fair value measurement, you can still try to replicate the exchange price between hypothetical market participants for the position that you’re trying to value.  So you try and put yourself in the shoes of a market participant and say, what assumptions would he or she use to value this position? 

The objective remains the same, and the key thing about this is, again, remember that slide that said here’s what fair value is not, as you approach a Level 3 measurement, none of that changes.  And the key thing to keep in mind is to the extent that a market participant would impose a haircut, for example, because there is not observable date for the position that’s being measured, then you should also include a haircut to the extent that you do not believe that the date is necessarily reflective of robust current information.

Now people used to think that Level 3 is an area you want to stay out of, but not necessarily.  To me this standard is not hardwired, and there may be cases where you’ve got a pretty good Level 2 price, but you have to make some adjustments.  But there may be cases where there’s a lot wrong with that Level 2 input and you have to make major adjustments.  It’s not clear to me that you would all automatically default to the Level 2.  You would, perhaps, work it both ways – Level 3 using your own assumptions and then adjust it for market participant assumptions or maybe start with Level 2 and make some adjustments. 

But the key here is that regardless, you need to be disclosing what you’re doing and, to the extent that you’re in Level 3, that invokes a whole different set of disclosures that are very, very informative to users.  So we try and dissuade people from suggesting that Level 3 are bad estimates or something that caused unnecessary concern because, again, they come with a very good set of disclosures.

Points that are particularly relevant to valuing subprime positions, we’ve had a change in the markets here.  These are positions that used to be, many of them, easily valued using Level 1 or Level 2 approaches.  The markets are not the same anymore, and rather than prescribe how you approach valuations in these situations, I will simply say that 157 does not require companies to doggedly mark to prices and data that may not reflect active markets, they might represent distress sales, and they might be based on quotes that are not based on actual trades. 

For example, we hear about broker quotes where you can’t actually get somebody to trade at that price.  We’re not talking about that as a price that would be used.  Now you can’t ignore those types of data points, but it is not the end of the story.  You need to start with them, consider them, and then explain and analyze why they are not representative of the value of the position that you’re trying to measure.

I made this point before, but let me emphasize it – any estimate of fair value has to include a risk premium.  It is not simply management’s optimistic assumptions about the future value if they hold something to maturity.  It’s what would it be worth if I tried to sell it to a disinterested party in an arm’s length transaction today.

So I admit that valuing items at fair value in current market conditions can be very judgmental in these circumstances, but investors have overwhelmingly told us that they still support fair value estimates even in these market conditions.  And I’ll let Gerry elaborate on that.  But when significant assumptions are used, disclosure of the approach that you’re using and your key inputs to the valuation is critical.  Other information can be provided voluntarily to the extent that you’re not thrilled about the fair value estimate that you came up with or you tend to think, I would never trade at this price. 

Go ahead and tell investors that that’s the case.  You can supplement it with a sensitivity analysis that would show the affect of changing your assumptions by various degrees, et cetera.  In my view, this was the purpose of the SEC letter.  It was simply just to encourage people to tell the story if there is extra information that might amplify the estimate that was being made.

I’d just like to take a couple more minutes if it’s okay, Peter, to really just provide my perspective on the notion that the losses are being exaggerated and the exaggeration has to do with fair value accounting.

In my view, it all starts with an impaired loan, and the idea that you would take an impairment loss on a loan that was troubled in some way is as old as the hills.  Whether you’re using historical cost accounting or any other accounting method, you always have to recognize impairments when an asset is not performing.  Now I think the fact that we’ve had complex securitizations that create synthetic or magnified exposures to these same loans creates leverage, if you will, and so you can have an amplifying effect on that original bad loan based on, for example, CDO’s or guarantees that are provided by a third party. 

We also have credit default swaps and other types of derivatives that, again, expose counterparties to those same losses on the same loans.  But because it’s a contract involving two different counterparties, you can have gains and losses on those as well.  I’m probably overstepping my area of expertise because I’m not a reporter and I’m not an economist, but I would submit that it’s possible that the tallies of losses we see reported in the media include all of these forms of exposure to the same loans.  They probably don’t include exposures that are held by private companies because they have no incentive really to report those losses, so that would tend to add to the numbers that you see reported. 

And then on the other side, I would say that for any guarantee that’s written or any credit default swap, for example, there are two parties to that trade.  So somebody is in a loss position potentially, and that’s probably included in the tally.  But the other party is in a gain position, and I wonder if the gains are being netted out against those losses in the tallies?  But it’s possible that to the extent that they are not being included, it would be overstated to that extent.  So from my perspective, the exaggeration issue really is not related to fair value accounting, but it’s related more to the magnification of the risk relating to the first bad loan.

In summary, I’d like to just emphasize that in difficult market conditions, the accounting judgment certainly can be challenging, but that is not to say that we want the accounting to ignore what’s happening in the markets.  Rather, the accounting has to reflect volatility and actual losses that are occurring in the marketplace.  The key, I think, is robust disclosure of the accounting methods that are being used, the key methods that are being employed, and assumptions so that investors can make their own judgments about whether the losses are real or not real and factor them into their analyses accordingly.  And with that, I’m interested in hearing what other participants have to think, and I’d be happy to answer any questions.

Peter Wallison:  Thanks very much, Leslie.  Before we go on to Mark, I’d just like to try to get a little bit of clarification on the issue of distressed sales because the implication of the SEC’s rule, and I take it you don’t differ in any way from that, is that the SEC’s letter, is that if what you are looking at as a price in the market is a result of a distressed sale, then you would not use that for the purpose of deciding how to value your own assets.  What did the FASB intend by using that kind of language?  How would one interpret what a distressed sale is?  How would you know, for example, whether a transaction occurred because someone was forced to liquidate or sell at a distressed price?  How is anyone supposed to know that?

Leslie Seidman:  This is an extremely judgmental question, and I’m not going to be able to provide a black and white answer to it, however, a distressed sale would be a very unusual occurrence in the marketplace.  One transaction that I’ve heard of that people seem to readily agree is a distressed sale is the price of Bear Stearns, for example, two dollars in a weekend agreement, et cetera.  People tend to think that was a distressed sale. 

Now when you look at other situations where people are having to sell assets to meet collateral calls, et cetera, et cetera, you need to ask a lot more questions.  For example, what type of asset are we talking about that they’re selling?  Are there many willing buyers of those assets?  That might suggest that that is not a distressed sale even though the seller is in distress.  So it’s a very judgmental question, and just to sort of complete the loop, even in those rare circumstances where we would say, that’s a distressed sale, it does not mean you ignore it. 

You start there and you say, that is an unusual circumstance, and I’m not going to stop my analysis there, okay?  But it does not mean you revert to cost accounting.  It doesn’t revert to mark to management, et cetera.  It’s simply a data point you would consider and look elsewhere to amplify your estimate of fair value.

Peter Wallison:  One more question and that is I just want to be sure I understand this and the audience does.  You really are looking at the circumstances of that particular sale, a particular sale or a set of particular sales to determine whether they were distressed sales?

Leslie Seidman:  Yes.  You would have to conclude that the parties were not willing buyers and willing sellers.

Peter Wallison:  You’d have to know a lot about that, wouldn’t you?

Leslie Seidman:  Yes, and you’d have to say that the normal process you would go through to sell those assets was not employed.  In other words, there wasn’t adequate exposure to the market depending on what type of asset we’re talking about, et cetera.  So it’s an extremely judgmental situation, and I’d say it’s pretty uncommon.

Peter Wallison:  Harvey?

Harvey Pitt:  If I hear you correctly, you’re saying that if a seller is in distress and feels it has to sell, that’s not a distressed sale, which is very reminiscent of Alice in Wonderland because, effectively, you are requiring parties or nonparties to a transaction to understand all of the terms of a transaction to which they are not privy.  I’m not criticizing it because I come out in favor of fair value accounting, but I just point out that there is a very vague quality to this that’s quite troubling.

Leslie Seidman:  I guess the question I would ask would be then for those who would want to consider a transaction distressed, what criteria are they using?

Harvey Pitt:  First of all, one could make the argument if you go back to the antecedents of FVA, that if somebody is willing to sell a hundred shares of IBM stock at five dollars a share, that may be the value depending on how much stock is available, et cetera, and the fact that it’s a distressed sale, the fact that it may even be a one-off transaction doesn’t necessarily make that any less a value.  By the same token, the fact that somebody might be willing to pay a premium to get control of something doesn’t make that the appropriate price either.  That’s the concern I have.  You use the word judgmental, and I’ll wait because I’m hogging in on time, but I think there are other words that could be used to describe that.

Peter Wallison:  Actually it’s a perfect segue to Mark.  Mark Scoles is a partner in the Principles Group of Grant Thornton and has over twenty years of experience in public accounting.  He’s responsible for technical matters relating to accounting, auditing and SEC work, with an emphasis on transfers of financial assets, financial instruments and derivatives, fair value measurements, and debt and equity arrangements.  He’s been involved with the AICPA in a number of capacities, both as a member of the Auditing Standards Board and as a member of the Technical Standards Subcommittee in the Division of Professional Ethics.  Mark.

Mark Scoles:  Thank you, Peter, and thanks everyone.  I’m going to kind of take off on some of the things Leslie has said, and some other areas as well.

The first thing I wanted to talk about was the notion of is fair value the right measurement attribute?  And I’m going to talk about it in the context of debt securities, for instance.  What is the best measurement attribute for debt security?  Well, for any asset, the best measurement attribute is based upon how relevant it is to users of the financial statements, and how reliable it is.  So it’s the intersect of that relevance and reliability that determines what is the best measurement attribute. 

In a perfect world you might say, well I’m going to look at every single asset I have, and I’m going to come to that conclusion based upon looking at all of them, but that doesn’t provide much comparability or transparency out there.  So we tend to put things in buckets.  For instance, debt securities — is historical cost, or more likely amortized cost, a better measurement attribute, or is some notion of valuing use or value to the entity a better measurement attribute, or is fair value a better measurement attribute?  And when you think about relevance, well, historical cost might be relevant at the point you bought the asset, but thereafter it has little relevance.  The reliability is high because you know what you paid for it.  But, again, the relevance deteriorates significantly over time.

Some entity-specific valuation, valuing use or an economic value to the entity requires the use of considerable judgment, in both determining the reliability – how reliable is it – well, it’s based on an entity’s own assumptions – and its relevance, which is based upon an entity’s own assumptions and intentions.  And overall when you look at the relevance, it may not have a whole lot of relevance because it’s difficult to compare entities to their peers or even to other companies in different industries. 

So then you look at fair value.  Well, fair value has significant relevance to looking at any asset, and in particular at securities or financial instruments.  Reliability, well, if that debt security is traded in an active market, the reliability is extremely high.  If it’s traded in a market that’s not active, or less active, it’s less reliable.  But how less reliable?

And you think about the fact that the debt security might be traded in a market that is not quite as active.  Would that mean that historical costs, for instance, would be a better measurement attribute?  Well, again, the relevance is very low.  How about an entity’s own specific value?  Again, the relevance to the users of financial statements is going to be very low.  So even though the reliability may suffer somewhat when you’re talking about assets in markets that aren’t active, I believe that it’s still, by far, the most relevant measurement attribute.  And providing disclosures about how you come up with the value in those situations is probably the best way to provide information to the marketplace and users about that fair value. 

Now thinking about Statement 157, and I think Leslie talked about this as well, did Statement 157 really change much about fair value?  Many of the principles in Statement 157 have been around for a long time.  They were in concept Statement 7 that talked about an exchange transaction – the price to sell an asset.  And when you’re talking about mortgage-backed securities, for instance, many of those are subject to an EITF, EITF 9920, that’s also been around for quite a while. 

EITF 9920 talked about an entity using the best estimate of cash flows that a market participant would use in determining the current fair value of that beneficial interest.  So it’s been around for many years.  And it sounds eerily like some of the concepts in Statement 157.  Now Statement 157 has been very beneficial from the standpoint of putting some parameters around how people determine fair value, and in particular, how you think about inputs when you’re in something other than an active market – how to prioritize an input and lead an entity down the path of determining what is the best estimate of fair value.  It also provides for significantly better disclosures than were out there before.

So that kind of leads us to the question of fair value and liquid markets.  When markets are less liquid, estimating fair value is certainly significantly more challenging for management.  It’s more challenging for auditors as well.  Fewer or even a lack of any transactions in a debt security or similar debt security makes it much more challenging to determine fair value.  But, again, even with fewer transactions, less liquidity in the marketplace, I don’t think that would lead you down the path that fair value is still not the most relevant measurement to users of the financial statements.

And it brings us to the SEC’s letter that Peter talked about.  Peter paraphrased the letter, and I won’t do that as well, but I think it’s important to understand what they brought out in that paragraph that Peter talked about was, I believe, directly out of Statement 157, paragraph 7.  And part of paragraph 7 says, “an orderly transaction is one that assumes exposure to the market for a period of time prior to the measurement date to allow for marketing activities that are usual and customary for transactions involving such assets.” 

It is not a forced transaction.  For example, a forced liquidation or distressed sale.  So it’s important to understand the context of which even forced liquidation, forced sale, distressed sale is even thought of.  It’s in the context of determining was the asset exposed to the market for a reasonable period of time that is usual and customary to sell the asset.  So what is that reasonable period of time that is customary and ordinary, for instance, for a debt security, for a mortgage-backed security or a CDO? 

Is it a day?  Days?  Weeks?  Months?  Years?  I dare say that it’s very dependant upon the type of asset you’re talking about, but it’s more in the shorter end of that scale than the longer end of the scale.  And it think some entities might like to think, well, how long I would look at to expose it to the market would be a period for which I would recover what I believe it’s worth.  That’s kind of the notion that fair value is what I would sell an asset for.  And fair value is not what I would sell an asset for, but what I could sell the asset for. 

And if you think of it, for those non-accountants out there, think of it as your own house.  If your neighborhood is like my neighborhood, yes, there are fewer transactions than there were a year ago or two years ago.  And the prices are lower.  But that doesn’t mean that my house, the fair value of my house, is somehow worth more than what I could sell if for today.  It’s not what I would sell it for, but what I could sell it for.  So with that, I’ll turn it over to Peter.

Peter Wallison:  Thanks, Mark.  Let me ask you this question though, and that is you’re much closer to what companies are doing than the rest of us, so what is your experience about what companies are doing faced with the kind of issue that the Fed was highlighting?  How are companies doing these values?  Are they using Level 3 because they think that the market is made up of distressed sales, or are they doing a combination?  What are they doing?

Mark Scoles:  I think entities are using, as Leslie said, all the data that’s available.  And I think that’s one of the things in Statement 157 that is very clear.  You don’t ignore anything.  Don’t ignore any relevant data.  The fact that there are fewer transactions, in some cases no transactions for many of these debt securities, or even for debt securities that are similar, generally would tell you that I’m going to have to adjust from those transactions that are out there or other information that’s out there in the marketplace, and that those adjustments, because maybe my particular asset is not quite so similar to those that are being sold, or that the transactions are fewer and further in between, those adjustments often times are going to lead you down a path that is a Level 3. 

But that doesn’t mean that there aren’t inputs based upon transaction data out there.  But because of the lack of liquidity, I believe many of the adjustments are bringing things down to Level 3.  And I think that will be with us for a while.

Peter Wallison:  Okay, so if companies are using Level 3 and they are using, although they are looking at what prices are in the market, they are using Level 3, what are the accountants doing?  What’s the role of accountants in that situation, or auditors?

Mark Scoles:  The auditors, the role of auditors is to determine how did they come up with the value?  What type of information did they use?  A number of entities are using various pricing services out there, they’re using broker quotes, a number of ways in which they’re trying to determine fair value, and I think the role of the auditors is to understand the methodology that management went through, how they’ve come about determining fair value, what information they used and what information they didn’t use, and coming to a conclusion, do we believe that they’ve come to the right measurement of fair value based upon that and based upon their understanding of Statement 157 and our understanding of Statement 157?

Peter Wallison:  So what you’re saying is that companies are not being forced by their auditors to adopt measurements that the companies themselves would not adopt?

Mark Scoles:  I don’t believe so.

Peter Wallison:  So who’s complaining about the values?  Who are these people who are complaining?  There’s certainly enough complaint in the market so if companies are not being forced to do this, and they are using their own methods, mostly Level 3, but making some reference to the market, why is there such concern about it in your view?

Mark Scoles:  I think probably because what I said – it’s extremely challenging, it’s very difficult when there are fewer transactions out there to look at to come up with this.  And there could be wider ranges of potential values that might be relevant, and when you’re taking losses, nobody is generally happy when they’re taking losses.  When those losses are based upon the best information available but it leaves you wanting, it’s probably more troubling.  But that doesn’t mean that that’s not the best information to provide to the marketplace and to users.

Peter Wallison:  Okay, thank you.  Harvey.  Harvey actually needs no introduction, so actually I’m not going to provide much.  But we all know Harvey’s background – lawyer, a former Chairman of the SEC, a skilled business advisor.  I won’t go into any more.  He’s now the head of, the CEO of Kalorama Partners, which is a consulting firm on global business matters.  Harvey Pitt.

Harvey Pitt:  Thank you, Peter.  I’m also pleased to be part of this panel, although, unlike the other speakers here, I have absolutely no idea what I’m doing here.  I once was a regulator, I’m not now.  I once was a lawyer, I’m not now.  I’m a consultant.  But, in any event, that has never stopped me before from having opinions or as you’ll probably determine, making certain that everyone else on this panel becomes infuriated with my comments.  I assume that was actually what my purpose was.

Peter Wallison:  And now you know.

Harvey Pitt:  Let me just say that I have been in the securities regulation and securities business field for over forty years, and something strange has happened because when I first got started, what accountants did was tell you what your assets and liabilities were, they told you what your balance sheet looked like, they audited your transactions, but they didn’t make national policy.  When I say making national policy, I mean deciding whether we go into a deep recession or not go into a deep recession, whether companies should show more in the way of losses than otherwise. 

And my own view, even though some of my very best friends and one of my partners is an accountant, is that policy shouldn’t be made by accountants any more than it should be made by lawyers.  I think it should be made by people who have a more global or at least national point of view.  So from one perspective, do we really want some of these policy issues addressed by the FASB or the IASB or, even worse, by the SEC?  And I’m very fond of the SEC. 

But my view after forty-plus years in Washington is that if you’ve got a problem, at the end of the day, the government is almost always going to make it worse.  And as a result, I think the real issues lie with the business sector and why the business sector hasn’t done more.  But I do think that one of the ways you have to look at this is to suggest that this is not really simply an accounting problem even though you’ve now heard from accountants who’ve told you how to fit everything into the rubric of FAS 157. 

The fact is this is a much bigger problem.  Now that’s on the macro level.  On the micro level, Peter asked before who was complaining.  And the people who are complaining are, by and large, the businesses because it’s not that their accountants have forced them to do anything in particular.  It’s that they think that the existence of these principles will operate to force them to do something, and those are things they don’t want to do. 

Why?  Well, first of all, many businesses, and we’ve seen this now, I happen to be very active in the hedge fund area in particular, but in a number of areas many companies have incredibly valuable assets, but what they don’t have is cash flow.  And the problem that that creates is a very simple one.  If they can’t maintain cash flow in some way, the value of the assets is going to become depleted.  So one thing that many companies are worried about is creating a self-fulfilling prophecy because if I report that it’s possible that these assets are really worth one tenth or one one hundredth of what I originally thought they are worth and what I think they will be worth if I can successfully hold them to maturity, then that’s exactly what I’m going to wind up with.  So one thing that the companies are upset about is they don’t want to create a self-fulfilling prophesy.

And lest we forget, there is another fear.  And that is, as we move, and I think the FASB is to be commended for the efforts it’s undertaken, and I think it’s to be commended for trying to move us toward a more principles-based approach.  This is what many of us have been arguing about – getting rid of prescriptive rules and moving more to a principles-based approach.  But there is a price one pays for a principles-based approach, and the judgmental factor, which Leslie alluded to several times in her comments, has a double entendre meaning because there will be judgments that get entered against companies that are criticized for the way in which they value. 

So what I think people are most concerned about is presenting themselves in a way that provides comparability across the board.  I don’t want to run the risk if I’m an entrepreneur that I’m going to do this in a way that makes me look worse than my five major competitors.  By the same token, I also don’t want to do this in a way that exposes me to litigation, and I thought it was rather telling when Peter read from the SEC’s letter because I always thought that government was a service business and that effectively what you are supposed to be doing is telling people what they need to know. 

And if it takes a panel of experts like this, myself excluded, to figure out what the SEC really meant, and it wasn’t even the SEC, it was just this division of corporation finance, but if you want to figure out what the SEC means, you shouldn’t have to struggle to figure that out.  You should be able to read the document and say, okay, I know what’s involved, and that is where the real problem lies. 

I guess my starting point is to try to figure out where we want to get to and how we want to get there, and in that sense, I frequently quote the individual who I have always thought of as the greatest sage of the twentieth century, Yogi Berra, who said that if we don’t know where we’re headed, we’re apt to wind up someplace else.  And I think if you look at what we are dealing with, we are in serious jeopardy of winding up someplace else. 

If a sale is made under distress, but it’s not a distressed sale, as a businessman what I want to understand is, what can I look to?  The problems here don’t arise because of anything FASB did.  That’s one thing I want to expressly agree with Leslie about.  What they really arise from are two things.  One is that the business community outsmarted itself.  It came up with and is continuing to come up with phenomenal synthetic securities. 

The difficulty is that those securities are not like the mutual funds to which Leslie alluded, which can mark their portfolios on the basis of net asset value just by looking at what was traded on the New York Stock Exchange or looking at some of the other markets or NASDAQ and reach a conclusion as to what they own.  We now have instruments that don’t work that way.  What’s worse is that the geniuses, and I use that in a straightforward way, a serious way, the geniuses that created a lot of the synthetic securities were outsmarting themselves because they themselves didn’t realize all of the ways in which you could hedge your investment. 

So if you look at what caused the subprime problem, what you’ll see is that the markets wound up becoming leveraged.  We did surveys on this for some of our clients and found that the markets are leveraged nearly a hundred to one in some instances.  And that’s why you saw the heads of financial services firms announcing on day one that they were going to suffer a four billion dollar write down, on day eight announcing that it was going to be eight billion dollars, and on day twelve that it was going to be twelve billion dollars. 

Companies should not be run this way.  People ought to know what their risk exposure is, and to my way of thinking, that’s not a problem that any accountant, no matter how good he or she is, is ever going to be able to solve.  So the question for us is how do we work with fair value accounting but try to fix the system?  And the answer, I think, is several-fold.  When you have markets that arise there is an absolutely critical need for transparency, and the one thing that’s been lacking, particularly in the subprime markets, is transparency. 

The SEC is investigating what went on at Bear Stearns and if normal practice applies, in about four years, we’ll figure out, or they will figure out, what happened and they’ll tell us.  The problem is we need to know today, and one of the things you’ll find is that the same instruments, the same packages, were carried on the books of various brokerage and financial services firms at multiple prices. 

Now I don’t want to suggest to you that there’s anything, per se, illegal about having multiple marks for the same securities.  But there is a problem if you can’t justify which marks are which, and it will turn out to be, I predict, that the marks that were the highest were the ones that we used to calculate what the management fees were that various people were getting.  And the marks that were the highest were also the marks that were calculated to figure out when margin calls would be made.  So if institutions are going to have multiple marks for the same instruments, they had better think about it.

So what are the solutions?  One is transparency.  I think the government has an obligation, if it is a service industry, before it thinks about regulation to assist in making these markets transparent.  That is a function of government concern and it shouldn’t get anyone upset because government wouldn’t necessarily be telling people how to run their lives.  It would just simply be trying to assess what values there are and, in fact, not even do that, just make the information available so everyone is playing from the same set of standards without having common information about what is going on in the market, whether it is a distressed sale or not.  I think you have an enormous problem.

I think the second thing that is needed is that government should provide meaningful guidance, and with all due respect to the letter, it’s not meaningful to write something that’s like a Delphic Oracle.  What you really want is advice to everyone that says these are the types of methodologies people can use, these are the types of formulae that people can apply, et cetera.  Not as a rule, but just giving people the kind of information they need and then come out with something that says, and here are the disclosures that ought to be made.  You ought to talk about what assumptions you’re making, you ought to talk about how you’re doing this so we can determine whether there’s comparability.

So I start from the same point, I guess, that the last two speakers did.  I think fair value accounting as a concept is fine.  I have no problem with that.  I just think it has no meaning in the current world, and what really needs to be done is to give it a pragmatic shape, to let people know how they can do things, how they avoid liability, and how we can foster comparability amongst all companies.

Peter Wallison:  Harvey, thanks very much.  Actually you’ve raised a fascinating philosophical question, as you know, about the role of accounting, whether we are talking as Mark suggested about the users of the financial statements, what’s useful for them, which is certainly the traditional role that public accountants had.  Or when we were talking about the effect of accounting on the economy as a whole, because of the way it works with companies having to sell their assets and downgrade their assets and so forth, whether it isn’t affected with the public interest in some way that changes the nature of the beast, in a way.  But that’s probably beyond the focus of this meeting.  If you want to ask questions about it, you’re welcome to do it, but I’m glad you raised that question because it is a very interesting one.

Okay, our next speaker is Craig Gifford.  Craig is a member of the American Bankers Association accounting administrative committee and executive vice president and chief accounting officer of Guaranty Financial Group.  He joined the Guaranty Bank in 2003 after thirteen years with Ernst and Young.  At Ernst and Young, he specialized in financial institutions and financial instruments working with a number of the largest banks in the United States.  And he also served as Ernst and Young’s derivative accounting subject matter expert.  Craig, thank you.

Craig Gifford:  Thank you, Peter.  Good afternoon.  I’m here today representing the American Bankers Association.  The American Bankers Association represents banks and thrifts of all sizes including community, regional and large money center institutions.  As Peter said, I’m the controller, I’m the accountant for a midsized thrift in Texas, Guaranty Bank, but my remarks today don’t necessarily reflect on the experiences or views of Guaranty, rather they reflect the collective experiences and views of financial officers at representative ABA member institutions gathered through committee meetings and other discussions over recent months. 

Before the session today, I was talking with Peter and he said, “Well, it’s good to have somebody here from outside the Beltway to share views.”  I think what he was really trying to say is the audience ought to get a kick out of hearing somebody with a Texas accent [laughter].

I’ve thought a lot about what I can add to the ideas of these distinguished panelists today since I don’t get the exposure to Washington in the standard setting process that they bring to the table.  What I do bring is a good deal of experience working with complex financial instruments while with a large public accounting firm, Ernst and Young, and more importantly the perspective of a registrant and financial statement preparer.  It’s preparers who deal daily with trying to ensure that financial statement information is both relevant and reliable, as Mark said, to investors and that it’s both of those things even in the face of extremely turbulent and, in some cases, nonexistent observable information about fair values. 

I’ll talk about three areas in my prepared remarks. 

First, ABA’s views about financial instrument presentation in financial statements; then the American Bankers Association’s members’ views and experiences with the current challenges in determining fair value and more importantly rapidly changing and inconsistent information about fair value from any financial instruments; and finally why ABA believes the resulting financial information reported to investors could result in degradation of market liquidity.

ABA has been involved in the consideration for financial accounting for financial instruments for many years, a fair value of accounting for financial instruments for many years.  As Leslie can attest, the ABA has been working on this for over eighteen years with the first fair value paper from the ABA written in 1990, and the points in that paper are still valid today.  We’ve actively provided our input and views through meetings and comment letters on FASB proposals.  The current accounting rules for financial instruments reflect many years of debate and crafting.  ABA does not believe that the present market conditions require changing the current accounting rules.  That ought to be a relief to you.

Leslie Seidman:  Thank you.

Craig Gifford:  Or providing forbearance from them to financial institutions.  Like any other business enterprise banks and thrifts accept financial risks in their business.  When those risks result in real losses to an institution, when the institution cannot reasonably expect to get its money back on investments, financial statement recognition of those losses is absolutely appropriate. 

The ABA also believes transparence is important and has generally supported proposals that add clarity about how financial instruments are presented and valued in financial statements.  Throughout the years, ABA’s views about financial instrument presentation have been consistent.  ABA believes investors, regulators and financial statement users are best served when assets and liabilities are presented at valuations that represent the relationship of the institution’s business model to the investment. 

For investment securities, which financial institutions frequently invest, the current accounting rules allow a business model approach to financial statement presentation.  If an institution invests in assets with the intent to sell opportunistically a trading model, the accounting rules require those assets be reporting at fair value with changes in that fair value being recorded in the current period income statement.  And ABA believes that is appropriate. 

ABA has long believed, however, that when an institution intends to hold an asset for a significant length of time, investors are best served with income statement reporting that associates the current period cash flows with the institution’s other cash flows rather than focusing on the short term vagaries of the trading market.  The vast majority of our members invest in financial instruments principally as long-term investors in those assets.  The current accounting rules do recognize the importance of financial statement consistency with the long-term investment nature of those assets and, in general, don’t require recognition in income of short-term market value fluctuations. 

The difficulty comes when trying to apply the accounting rules for determining fair value and distinguishing between credit discounts and other market considerations.  Some have suggested that Statement 157 seems to be the culprit in this market.  157 requires the use of exit prices and it delineates, as Leslie mentioned, three categories of fair value determination methods, establishing a hierarchy that reflects the observability of the information. 

When the FASB initially proposed Statement 157, the ABA did not oppose it.  In the light of the economic and market environment at the time, ABA interpreted it in a common sense manner and believed it to be very similar to existing practice.  Unfortunately when implemented, what we face today, not only did 157 not address the reliability of market pricing information, but it in fact may have validated, at least as it’s been applied, the use of low quality price reference sources over fundamental economic valuation methods that have proven effective and reliable at determining long-term investment value for many years. 

Our member are currently obliged, Mark may not agree with this, but currently obliged to use price sources supported by little market evidence or prices that are set by actual transactions or expectations of transactions with truly distressed sellers.  In effect, banks and thrifts are currently beholden to reporting many fair value measurements using bad Level 2 over good Level 3 information.  Whether 157 itself is the culprit or whether it’s because of incorrect interpretations by auditors and preparers, this is clearly a real problem. 

A year ago when mortgage loans and mortgage-backed securities were readily trading, it was possible to find many transactions of the same or similar assets in order to determine fair value of a financial instrument.  How times have changed.  And I can tell you as a preparer, they have.  Each of the last three quarters has found price discovery increasingly difficult.  It’s worse today than it was in January.  It was worse in January than it was in October.  And it was certainly worse in October than it was six months prior to that. 

Today, real world experience shows that outside of the government and government agency securities markets, which are still very liquid, there is increasingly less price transparency.  Dealers may provide quotes, though for some instruments even that’s becoming quite difficult to obtain.  But even when they do, at times the pricing information appears questionable. 

For example, our members have heard dealers acknowledge that there is very little activity to reference from any of the security classes that they’re quoting.  And that’s not just subprime assets, but even assets that are prime in almost every respect except that they were pulled into non-agency securities.  Or dealers suggest that pricing expectations by potential buyers reflect the buyer’s belief that distressed institutions, I’m not saying distressed transactions, but distressed institutions will have plenty of similar product to sell because they may have margin calls.  And if one wants to sell right now, that the price will reflect the expectations for transactions with those distressed sellers. 

Many price quotes are viewed as being based on market chatter rather than actual observable transactions.  Only distressed sellers would trade at the market chatter evaluations, many think, because when viewed from a long-term investment perspective, there often doesn’t appear to be a lack of recoverability of the investment justifying those low prices.  And even dealers, our members have even seen situations where dealers give price quotes of the same values on different securities that have different characteristics making it clear that the values that the dealers are providing were determined by reference to indices rather than actual transactions. 

These indices, such as the so-called ABX, are principally used for credit derivative transactions.  In fact, they themselves are comprised of derivative instruments rather than the underlying cash instruments.  Because of their construction, the indices can deviate greatly from the cash markets that they’re alleged to represent.  On top of the fact that they’re on the interest of investors who want to short credit risk or sell credit risk short at a macro level without any interest in or understanding or knowledge of the cash market or construction of the specific instruments to which they relate. 

What makes this most perplexing is not just that the financial instrument users are being given information about fair value that may be not of great quality, and, of course, what relevance is low quality information, but that interpretations of these valuations can trigger other accounting rules leading to tremendous impairment charges.  As I mentioned before, certainly ABA believes if there is responsible expectation that an investment will not be recovered, such as is the case with subprime and subordinate securities, the amount that will not be recovered should be charged off. 

We’re not talking about a Japanese lost decade.  But the accounting rules require that if any portion at all of an investment is not expected to be recovered, the charge, at least with respect to debt securities and secure ties to the financial instruments, the charge must include the full amount of the market discount, which may reflect many market considerations other than the most likely cash flows of the instrument, for example, lack of market depth and illiquidity. 

Our members question whether that is appropriate in the current environment where modeling approaches with cash flow inputs from published data by leading dealers suggests full recoverability, for example, for senior priority instruments, and yet quotes by those same dealers for those same securities may be sixty to eighty cents on the dollars.  Leslie mentioned that everything ultimately boils down to the underlying loan impairments, and with respect to securitizations, that’s certainly true, but for many securities, securitization actually reduces the exposure to the underlying loan impairments because the way the structures are worked, as Harvey mentioned, the business people may have outsmarted themselves. 

I’m not sure they’ve outsmarted themselves, but rather they’ve positioned risk with businesses and investors who have accepted different characteristics associated with the underlying loans.  Independent audit firms also seem to be placing excessive reliance on third party price quotes without regard to whether those come from markets with sufficient market depth, and the independent auditing firms in many cases have also established arbitrary guidelines about how much market discount there can be before a preparer must assume that there is a recoverability impairment.  And there’s a piece of literature by the emerging issues task force, a standard setting body, or a standard interpreting body, addressing securities that are not high credit quality as it defines them, which could force impairment charges far in excess of loss expectations.

ABA’s members are experiencing these price transparency difficulties for a broad range of loans and securities – mortgages, autos, student loans and auction rate securities.  When you add these challenges up, you have the potential for institutions to recognize charges far beyond what the data would suggest are actual recoverability concerns, and it’s unclear whether that may have already occurred with the size of the charges that we see.  When that happens, those institutions could find investors and counterparties looking at them as if they were in a liquidation mode, even though their operations may be doing just fine and they may not have changed their business model or their investment strategy at all. 

As we’ve seen numerous times recently, and has Harvey mentioned, it’s not surprising for the appearance of liquidation to become a self-fulfilling prophecy.  Say that one five times quickly.  ABA believes it’s not good policy to allow application of the accounting rules to drive good businesses into liquidation mode.  And the fear of being forced to apply those accounting rules this way has dramatically reduced the number of participants in the market - ABA’s members are participants in the market for many of these financial instruments – which has, in turn, reduced the credit availability for consumers and businesses alike. 

Of course the broad economy implications of this are significant.  Banking, for many of us, is a relationship business.  Presenting that business in a liquidation context commoditizes the value of the business, reporting short-term value at best, and it creates incentives for banks that have traditionally focused on originating higher credit quality assets to originate down the credit curve because high credit quality assets aren’t presented any more favorably.

So what do we do about it?  Certainly forums such as this are helpful.  There are several other courses of action that could be taken to improve the current situation.  The SEC may have started this.  Standard setters could clarify that FAS 157’s reference that pricing in distress situations does not reflect fair value.  It applies to market transactions if one of the participants is in a distress situation.  Of course determining that, for many preparers, is extremely difficult because they don’t have the insight to what the drivers behind the market transactions may be. 

Regulators could clarify for auditors and preparers that if distress transactions are the only transactions, that there’s not sufficient market depth to look at those prices for fair value evidence.  I think Leslie has clearly said that today, but it’s not clear to me and to many of our member institutions that that’s the way that auditors and, perhaps, regulators are applying the rules. 

Standard setters could clarify that a good Level 3 may be better for financial reporting than a bad Level 2.  And standard setters and regulators could provide interpretive guidance to assist preparers and auditors in recognizing that market value discounts in turbulent times may reflect many market considerations besides other than temporary impairment.

In your folders is a copy of the International Banking Federation’s conceptual paper on accounting for financial instruments.  It looks something like this.  ABA is a founding member of that global group and is releasing the paper today in order to coincide with today’s panel discussion.  I hope you will read it and let ABA know if you have any questions or thoughts.  ABA believes these actions can help to alleviate the current situation.  Thank you for allowing me to present ABA’s views and positions on the current challenges.

Peter Wallison:  Thank you very much, Craig.  Just one question just to try to clarify this a little bit - the ABA believes, we’re leaving your particular views out of this, but the ABA believes that their members are being required to use valuations that they wouldn’t use.

Craig Gifford:  I think that, as Mark put it, it’s very challenging and difficult to determine fair value today.  Many institutions have a lot of experience at analyzing instruments and understanding the ultimate cash flow considerations associated with those instruments, and, yet, in many situations auditors and regulators are requiring institutions to get price quotes in spite of the fact that the price quotes may not reflect quarterly transactions behind those quotes.  And in most situations brokers really won’t, even if they could, provide any kind of information about the underlying data behind those price quotes.  And so auditors and examiners are left with, you have a price quote and yet there’s a difficult time in reconciling that to the actual modeling that would suggest what that investment may produce in the way of cash flow in the future.

Peter Wallison:  So this is your point about Level 3 may be better in some circumstances than Level 2 and some may not make that clear.

Craig Gifford:  It very well may be.  I think that it’s important that if you’re using a Level 3 modeling approach that it be based on good data and a reasoned approach using, perhaps, published information.  But in many situations you may not be able to verify the underlying data to actual transactions or actual examples.  I certainly don’t believe our members look for a mark to make-believe approach or a mark to what-you-want-to-be approach.  Certainly in talking to my peers here, the perspective that they genuinely believe that they model in a good faith and best attempted approach and yet they come out with very different answers than what they may get from a broker price quote that they’re being held to apply.

Peter Wallison:  Okay.  Our next speaker is Gerry White.  Gerry is a chartered financial analyst and president of Grace and White, which is an investment council.  He’s a member of the CFA Institute where he serves as chair of the Corporate Policy Disclosure Council.  He was also a member - incidentally CFA is Chartered Financial Accountants... Analysts, I’m sorry, Chartered Financial Analysts.  I knew that actually.  It was a slip of the tongue.  He was also a member of CFA Institute’s financial accounting policy committee from 1970 to 2000.  Previously, he was an adjunct professor of accounting from 1978 to 1995 at the Stern School in New York at New York University, and he is the co-author of The Analysis and Use of Financial Statements.  Gerry.

Gerald White:  Thank you, Peter.  Being fifth in line has its hazards, and one of them is that a lot of what you intended to say has already been said.  There is an outline of my comments in your handout.  I’m going to be doing some cutting and pasting so as not to repeat things that have been already said.  Let me start by saying that I’m not an accountant.  I’m an investor and my view of this subject is purely based on my almost forty years of experience reading financial statements and using them or trying to use them to make investment decisions. 

Based on my experience, I believe, and I should say that I’m dealing with this on a conceptual level and on a broader level than just financial instruments, but I believe that fair value is the answer, ultimately, for the entire balance sheet.  Please hold your arrows and bullets.  The reason I hold this view is, simply, relevance.  The historical cost of an asset, as - or a liability, as has already been said, is increasingly irrelevant over the passage of time.  And I recognize there may sometimes be measurement problems, but fair values are more relevant than historical costs. 

Management decisions are, or at least ought to be, based on fair values, not historical costs.  Investor decisions are made based on fair values.  Lenders and regulators make their decisions based on fair value.  I don’t think there are many members of the ABA who would say, well we’ll ignore the fact that your collateral has gone down in value, we’ll just accept it at what you paid for it.  And I believe that management should be evaluated based on the returns that they earn on the assets and liabilities that they control with those assets and liabilities measured at fair value.  Otherwise, if they can’t earn a reasonable return on those assets, they ought to sell them and give the cash back to shareholders.

Let’s talk next about comparability.  Fair values are, in fact, more comparable, in my view, than historical costs.  Historical costs reflect transactions that were entered into at different times, perhaps in different currencies and where their carrying amount has been accounted for using different accounting methods and different assumptions.  So that over time, even if two companies acquired identical assets, had identical prices, over time their carrying amounts would diverge based on their accounting methods.  And fair values are, I believe, more objective despite the measurement problems. 

There is a market out there.  We live by markets.  I certainly do.  I’m in the investment management business, have been for thirty years.  Every three months we send reports to our clients which show the market value of their investments.  There have been quarters when I would have liked to say, well instead of using the market value, I’ll just put down what I think the stock ought to sell at and maybe charge fees based on those values.  But I’ve managed to resist that temptation, and I think my clients appreciate that. 

So much has been written on fair value in the last six months, and some of if has even been accurate.  One of the problems with a lot that has been written is that it fails to distinguish between economic events and the reporting of those events.  Yes, there has been unusual volatility in the financial markets in the last few years, the equity markets, as well as the fixed income and derivative markets. 

It seems to me that the role of financial reporting is to tell shareholders what actually happened, not what management wishes had happened or would have liked to have happened, and I think that Harvey’s comment about companies deluding themselves, or outsmarting I guess was his word, but I think there was an element of delusion as well, in that many of them believed their own hype.  And I’m talking about the investment bankers and others who trafficked in these exotic instruments.  They originated them, they told their customers that they were worth such and such an amount, and they held them on their balance sheet, and they said, “Oh, we don’t have to worry, they’re good investments, they’re worth par.” 

One of the things, those of you who want to get a better understanding of what’s happened over the past year, ought to read the report issued about a month ago by something called the Senior Supervisor’s Group, which was a group consisting of regulators from the SEC, the Fed, I think they controlled the currency, and four foreign regulatory bodies from major financial powers.  And they went back and looked to see how could they distinguish those financial institutions that were badly hurt by the crisis from those who came through it perhaps a little damaged, but fundamentally intact. 

And the answer turned out to be that those who came through it very well were those who didn’t simply accept what people told them, where management didn’t just talk risk management, but they actually performed it.  They actually looked at their risks.  They understood their risks.  They didn’t just accept their traders telling them, “Oh these bonds are worth par.”  They, in some cases, made the traders sell part of their position in order to validate those prices.  And I think there’s a real lesson in this.  I think one of the lessons here is that so many companies have talked risk management and a lot fewer actually practiced it.

Why do I believe that fair value should be used initially for all financial instruments and ultimately for all assets and liabilities?  It is because the so-called mixed attribute model we have now simply doesn’t work.  You have banks and insurance companies and other financial intermediaries whose balance sheet is a mix of some assets and liabilities at fair value, some at historical cost, and Statement 159, the so-called fair value option, allows them to pick and choose: well, we have ten billion dollars of deposits, we’re going to show two point three billion of those deposits at fair value and the rest at historical cost. 

And if you think I’m making that up, look at, read financial statements of the largest banks in this country.  I think the mixed attribute model does not convey useful information about the financial health and about the risks that companies run, and about the effects of changes in interest rates, changes in credit conditions, changes in foreign currency rates and so forth on those companies.

Hedge accounting was designed to allow companies to mitigate the effects of changes in interest rates and so forth on their balance sheets.  The problem with hedge accounting is that it’s very hard to do.  Some of the largest companies in this country have found it very hard to do and have had to go back and redo it, in some cases twice.  It’s totally opaque and there’s a whole set of arbitrary rules.  Not your fault, Leslie, but that’s the reality.

Let me finally move on, my last item, to point out that reporting fair value is not enough.  What’s on the face of the balance sheet and the income statement and the cash flow statement is not sufficient if there is no explanation as to how those numbers are arrived at.  And if there’s no explanation of what the business is, what risks are being run, what models and assumptions are being used to value assets and liabilities, and without disclosing how those fair values changed over time - in other words the causes of those change over time, how much was due to interest rate changes, for example, and which was due to changes in credit conditions.  I think probably I ought to stop there, but I’d be happy to answer any questions.

Peter Wallison:  Thank you very much, Gerry.  Just let me clarify one thing about what you said.  If there isn’t a market or if there’s only a distressed market, what do you say about that?  What say you, as Bill O’Reilly would say?

Gerald White:  I think that this distressed market is probably not an operational concept.  The market is the market, and I think there’s generally no choice but to accept what the market tells us at any given time.  There are times when I don’t like it, and I’ve just lived with it.  One exception might be, I suppose, if on December 31st there’s a block trade of some instrument at a twenty percent discount from where the price was the day before, and on January 2nd it was back to the December 30th price, that I suppose might be an argument for saying that December 31st price really wasn’t valid and that’s not what should be used.  But it seems to me, to be operational this concept of a distress sale would have to be extremely rare.

Peter Wallison:  Okay.  Vincent Reinhart, he’s our clean-up hitter today, and he’s my colleague here at AEI.  He’s a resident scholar, former director of the Federal Reserve’s Division of Monetary Affairs and has spent more than two decades working on domestic and international aspects of U.S. monetary policy.  He held a number of senior positions at the Fed and for the last six years of his Federal Reserve career, he was secretary and economist for the Federal Open Market Committee.  Vincent.

Vincent Reinhart:  Thank you, Peter.  Actually, I’m batting sixth, which usually is your much lighter hitter.  I also want to thank you --

Peter Wallison:  Not when Roy Campanella was batting sixth.

Vincent Reinhart:  I want to thank you for this opportunity to talk about accounting issues late in the afternoon after a heavy lunch.

Peter Wallison:  Their eyes are still open.

Vincent Reinhart:  Now as the [audio skips] title says, I’m supposed to provide an economist’s view about fair value accounting.  I think the real reason I’m here is to make Harvey Pitt more secure in his self-image.  By the time I’m done, he’s going to feel like a subject matter expert in 157.

We are in the midst of the worst financial crisis since the 1930’s.  Some things are familiar, including the failure of market participants to protect themselves against the withdrawal from risk taking.  Some things are new, in particular the specificity with which we report balance sheet items.  We’ve begun a discussion now, and it will be a long one actually, how much the latter has influenced the contours of financial market over the last eight months. 

Now FASB is in the unenviable positions of actually clarifying what has been going on for a long time.  That has a couple consequences that makes it an easy target for criticizing mechanisms that have been at work in previous episodes as well.  But, as I’ll bring up later, one also has to wonder if the clarification also provided a coordinating device that led to more people or more firms behaving in a similar manner, so therefore the accounting actually drove behavior rather than merely reported behavior. 

Suitable to a discussion about accounting, I’m going by the numbers.  I’m going to first make two points of review, then I’m going to talk about four reasons why the application of fair value accounting in a mark to market world, they have to buy all those things, could potentially amplify market price dynamics.  Then I’ll talk about two different rationales for accounting schemes.  And that probably will be enough for me today.

Now by way of review, I recognize that these issues have already been covered, but it’s useful if we want to speak the same language.  In particular, FAS Statement 157 provides guidance on fair value measurements, that is, fair value is the price that would be received to sell an asset in an orderly transaction between market participants at the measurement date.  What particularly helped me in the literature on this was thinking of fair value as an exit price, that is, what it will take to get out of a position. 

Now in also thinking about this, another aspect of 157 that’s important is the hierarchy of fair value measurements, and we don’t have to go much over that.  Level 1 quoted prices; Level 2 quoted prices for comparable instruments, effectively; and Level 3 based on unobservable inputs.  Now at this point, duty compels me to express some qualms in the comment period on 157, and these all were taken by speeches and testimonies by then Governor Sue Bies at the Federal Reserve Board about the overall applicability of FAS 157 across all financial firms, and they mostly, when you read those letters, talk about reliability and comparability issues, that is, fair value estimates for similar instruments can vary greatly in practice. 

Management can use significant judgment in selecting market inputs when prices are not available.  Minor changes and assumptions in a pricing model can have a substantial effect, and verification of prices based on unobservables will be difficult.  And, I think, Peter, this answers part of your question, why do you hear so many complaints about fair value accounting right now?  And that is for many market participants, they don’t trust their counterparts to do it the same way they are doing it.  The extent of discretion available in management judgments and sources of price quotes in the configuration of the model used to price at Level 3 all give scope for a range of potential answers to the same question, and that does lead into questions about counterparties.

Now what I’d like next to focus on is the economic consequences of full-blown mark to market of a trading entity complying with fair value accounting, particularly four mechanisms that, in fact, may amplify fluctuations in market prices.  Now, given the current circumstances, I’m going to be talking about amplifying a downdraft in market prices, and I’ll mostly be talking about securities - that’s the stuff on the asset side of these entity’s balance sheets.  But don’t forget these mechanisms were also in place supporting the updraft in financial prices.  And don’t forget that these mechanisms potentially apply to both sides of the corporate balance sheet.  Now, two issues are involved in fair value accounting at a mark to market firm.

First is how do you value your assets?  And you can think about them having three buckets corresponding to the level of pricing to market, to comparable or to model.  But the next important economic issue is how do changes in asset values pass through to the income statement and, therefore, have consequences for what you report as capital over time.  The first mechanism I’d like to point out is the potential direct adverse price dynamics of a balance sheet repricing. 

Suppose something happens in security markets so the value of some market asset goes down?  Let’s say it is an asset you hold directly on your balance sheet, and, therefore, it is easy to price.  The impairment of that asset value and reduction in the value of equity may force you to deleverage some.  But what happens there?  The sale of those assets in the market can further impair prices.  And, essentially, that’s what I have on the right.  The green line you should think of as the net demands for the real money sector. 

If prices are very high, they’re willing to sell some of their assets.  If prices are very low, they’ll come in and be a buyer.  When prices are near where they are currently, they’re pretty sensitive to order flow, that is, they can absorb some purchases or some sales, but there is a limit to how much they absorb.  If you show a lot to the market, prices will move a lot.  If you ask real money accounts to sell some of their securities and volume, prices will go up.  If you ask real money accounts to purchase some of the securities, some of those securities, prices will go down a lot. 

So the direct adverse price dynamics of marking to market is if a reduction in the value of an asset impairs your capital asset ratio and induces the firm to sell assets that may amplify the market downturn.  But fair value accounting implies indirect effects as well.  And that is if the price of that traded asset goes down, then the price of all assets comparable to that on your book will go down. 

Why?  Because that’s your Level 2 prices.  And if the value of those assets go down, and say market uncertainty increases, bid-ask spreads rise, measures of implied volatility rise, then anything you have priced at Level 3 will also go down.  But if that same economic mechanism is at work, that is, impairments in capital asset ratio lead the firm to sell assets into a declining market, then prices could go down by even more.  And the net effect of pricing everything on the balance sheet in this manner is providing scope for even more forced assets sales over time.  That is, there’s a potential for prices to go down by even more.

But the third consequence, and this is where actually putting in place 157 could actually have real consequences rather than merely reporting consequences, is herding can magnify these prices changes.  Why?  Everybody in the sector follows the same rules and uses the same pool of auditors, and, generally, the interpretation of rules becomes stricter at a time of stress.  Why?  Regulators worry, the general counsel worries that they’ll be subject to litigation risk if they’re not following similar prospects, and so there are more entities following the same strategies forced to them by the accounting. 

That is, 157, by clarifying the rules and making it easier to follow a certain set of rules, makes it more likely.  More financial firms are in a similar position in face of declines in market prices.  And that introduces a fallacy of composition, which is fundamental in macro, and that is deleveraging certainly would make an individual firm better off.  If you have an impairment of asset values that leads to some declines in your capital, and, therefore, your capital asset ratio looks worse, selling off some assets makes you more financially secure. 

But if everybody does it at the same time, no one is better off because prices go down by so much.  And that’s really what’s in that far right panel.  Indeed, we’re probably here today, that the herding induced by the price capital adjustment has forced so many sales or induced so many sales to the market that real money accounts are not there in sufficient volume and so the market disappears like a fist when you open up your hand.

Herding also applies to both sides of the corporate balance sheet because everybody in the market uses the Merton model.  What’s the Merton model?  Equity is a call option on the value of a firm’s assets given the current value of its debt.  Everybody buys the equivalent of KMV or its competitors to calculate the distance to default consistent with current equity values and the current composition in the balance sheet.  But that distance to default is generated by many firms and is used widely to set credit risk premiums and is the initial pricing for price default protection.  So when the values of assets decline, the Merton model will say the distance to default is closer, therefore, a decline in the fair market value of assets shortens the distance to default making new debt more expensive and increasing rollover risks, reinforcing the adverse credit dynamics. 

So in that sense, you can argue that fair value accounting amplifies price movements in financial markets, and that there may even be an asymmetric effect through its influence on credit risk spreads and rollover risk.  That is, in good times you’re so far away from default, small changes in prices have no real consequence.  But when you get down where we are now, default risks can escalate markedly introducing self-fulfilling prophecies of the denial of access to credit creating a failure to rollover.

I just want to introduce one other set of other thoughts, and that’s what’s the rationale for evaluation scheme?  We’ve heard today two different sets of view.  And the one is, I think, the securities market view, that investors need the information to evaluate each component of a firm’s balance sheet.  It’s basically atomistic.  You can price every bit on the asset and liability side of a firm’s balance sheet.  That will let you better predict the current market value of the firm.  It will also let you better predict earnings and dividends. 

It’s also then, in that sense, it’s useful to take discretion out of managers’ hand to element excessive smoothing of reported earnings when you mark everything to market.  And the banking view takes an approach looking more at strategy, and that is the decision to add an asset to a balance sheet is also a decision to fund that asset.  And the exit price of the asset should be what it returns when sold if the funding is not rolled over. 

So the horizon at which you price your asset should be linked to the horizon at which you funded that asset.  Assets funded short-term should be repriced frequently.  And one could also imagine the capital consequences in a regulated entity.  Assets funded long-term can repriced less frequently.  And I think that gets to the discussion of distressed sales.  Some people here viscerally object to pricing at a distressed sale because that’s an exit price headed to the exit too quickly.  They don’t view they’ll be doing that, therefore, they don’t view it as appropriate to use that price.  On the other hand, if you’re a trading firm and you face rollover risk, today is when you have to go to the door, therefore, there really is, in no notion, is there a distressed sale.  I think that finishes what I’ll say.

Peter Wallison:  Okay, thank you very much, Vince.  That was excellent.  Okay I think what we usually do now is give the - yes, yes, yes, yes - give the people from the panel an opportunity to talk to one another about things that were said that you disagreed with or want to clarify in some way, and I’m going to start with Leslie and we’ll move down the line.  Leslie, is there anything that you heard here that you’d like to clarify in some way?

Leslie Seidman:  I guess one area I’d like to probe is the idea that fair value accounting can actually drive a downward spiral.  I guess I’m curious how often you think companies report.  In other words, even in a public company situation where you’re reporting quarterly and then there’s a lag after that, it just raises questions to me about the staleness of the information and the direct cause and effect of fair value accounting and the consequences you described.

Vincent Reinhart:  So I think actually the story is marking to market can produce this adverse credit dynamic, and there it’s almost irrelevant at what cycle they report because their counterparties are doing that calculation in deciding of funding day after day after day.  And so the adverse credit dynamic associated with marking to market is a trading firm that feels if it’s seen as using too much leverage, it will be punished and pushed out of the business. 

The question about 157 and fair value accounting generally is, and I’m raising it as a question, did the scope of the entities who feel they must follow those rules widen because of the implementation of this rule?  And has it reduced the variability or rather increased the amount of herding among the firms that do do it just because in part because of the relative clarity and the fact it is implemented by a small set of auditing firms?

Peter Wallison:  So maybe another way to put that is, it’s pro-cyclical rather than counter-cyclical.

Vincent Reinhart:  There’s a lot of things in life that are pro-cyclical, including the behavior of regulators, the behavior of accountants, and also the behavior of, essentially, risk management, that is, risk management gets tougher exactly when prices are going down - exacerbating the downturn.

Peter Wallison:  And it’s also pro-cyclical on the upside.

Vincent Reinhart:  And so, therefore, marking to market just as well as value-at-risk probably added to the up leg because it provided better balance sheets that could support more expansion of that balance sheet because ratios just got better and better.  And, meanwhile, the value-at-risk said you can take larger positions because volatilities are so low you can support bigger tr