David R. Henderson  

One-Two Punch on Mark-to-Market

Marginal Tax Rates Will Rise... Misplaced Priorities...

Today, the Wall Street Journal published two excellent letters on the mark-to-market regs that banks are under. IMO, too little has been written about this by economists. The second letter tells a horror story; the first makes a constructive suggestion for getting mark-to-market out of the capital regs that banks are under.

That's the first punch. The second is a commenter who replied to a post by Alex Tabarrok in which Alex pointed that bank "nationalizations" are part of a bankruptcy process. The commenter, Andrew, wrote:

Great post, but don't normal business[es] only go bankrupt when they actually can't pay the bills, not just when some people call their balance sheet insolvent?

Andrew doesn't mention mark-to-market, but this is clearly what's going on: mark-to-market regs can lead regulators to say that balance sheets are insolvent even in the extreme case that all borrowers are paying their regular payments. An aide to my Congressman, whom I talk to occasionally, told me that this is what a banker in the Santa Cruz area told the Congressman was going on with some of his bank's loans on properties that mark-to-market had declared to be underwater but on which the borrowers had not missed a payment.

Comments and Sharing

CATEGORIES: Monetary Policy , Regulation

COMMENTS (23 to date)
Joe writes:

Mark to market measures the present value of ALL future payment streams. While historical payment streams are nice to have they do not necessarily mean future payment streams will actually be made.

The behavorial problem is of course how much you discount future payment streams. Mark-to-market attempts to place a value or chance on the fact that the person who has made all of their past payments may not be able to make their future payments.

Mark-to-market is not the problem. The problem is how much you leveraged up assets that have fallen drastically in value. If you did not leverage out assets and kept enough capital reserves you would not care about the mark-to-market valuations.

Kevin writes:

The discussion of MTM is a bit of a red herring. The policy question is whether and under what circumstances the government should be able to declare a bank insolvent and seize it. Suspending MTM is just a way of lowering the capital requirements for banks without admitting that that's what we've done, particularly if the banks are still required to report MTM. It would be better for everyone if the government just came out and lowered the capital requirements (or not), and moved on.

R. Pointer writes:

Dr. Henderson,

I don't know if you have had the chance to listen to the Econtalk between Dr. Roberts and Dr. Cochrane, but you should maybe look through the transcript. Dr. Cochrane talks a little about this fact of MTM and what the regulatory response should be. Namely, if loans are still performing, regulators should practice forbearance while still keeping a close eye on whether those loans continue to perform. We can have MTM but not force bankruptcy on performing institutions.


libfree writes:

Isn't the point that bond holders can force a company into bankruptcy on similar grounds to mark to market depending on the capital requirements behind those bonds. The whole point to this discussion must be that we think that the underlying assets have more value than MTM is showing them to have. That may be the case, but then again it might not be.

Les writes:

Carrying loans receivable at book value requires one to assume that all borrowers will repay all interest and principal obligations on time and in full. In a severe recession that seems to be a bold assumption.

MTM is a probabilistic estimate, based upon reasonable assumptions, about the extent to which borrowers will repay all interest and principal obligations.

Which is better? Book value, if you want to be precisely wrong. MTM if you prefer to be roughly right.

tom murphy writes:

Those who lived by MTM on the way up must now die by the regs. Nobody mentions the fact that these same guys were ringing the bonus bell on the MTM of a lot of these "toxic, impossible to mark" when they inflated their value to get paid in the halcyon days of 2004- 2007. I really think that if they want to forgo MTM on all assets that they put on the books going forward, that is fine but the ones they ve already bled, that;s the way the cookie crumbles.

Robert Simmons writes:

I don't think that this isn't a major point, but bankruptcy doesn't require missing payments. Covenant breaches are the same as default, and there are preemptive bankruptcies.

Dan Weber writes:

but don't normal business[es] only go bankrupt when they actually can't pay the bills,

That's right for "normal" businesses. Normal businesses sell goods or services and the cash flow is what allows that business to proceed.

Banks only trade in money. A bank without credit is a dead bank.

I'll buy bread from a bakery that's having cash flow issues. I'll even sell them ovens (cash on delivery, please). But when dealing with banks, it matters a lot both how solvent they are, as well as how solvent people think they are.

Gary Rogers writes:

I will qualify my response to this post, because I am neither an economist nor a banker. So, please feel free to quit reading if I am off base.

As I understand it, the securities involved are almost all unique products made up of income streams from bundled mortgages and other securitized debt. There is not just one layer, but multiple layers of bundled income streams with differing senority levels that are almost impossible to trace back to the source. These instruments were never intended to be resold and there is really no market for resale.

My understanding of mark to market rules is that the bank has to value their mortgage backed securities as though they could resell it on the open market. This leaves a lot for interpretation, but it can be done. The problem comes when everything becomes suspect, like it is today, and nobody knows what anything is really worth. Since nobody would buy the assets, the value drops to almost zero virtually creating a run on the bank without anybody withdrawing a penny.

If Joe were correct, the bank would continue to value its paper based on the present value of the income stream regardless of the uncertainty in the markets. This is a much better model. There can be disclosure involved that will tell investors that the bank holds paper that is at risk, but the value should not be reduced until the income stream starts shrinking. In other words base the value on what is known not what is feared.

As a comparrison, assume the bank is holding the loan on my house rather than a securitized income stream. If I lose my job, they do not mark down my loan as long as I stay current with my payments. The same should go for mortgage backed securities. The fact that a financial institution can go bankrupt without anyone withdrawing funds and without any reduction in the income stream seems crazy.

Mike Moore writes:

Gary, your understanding pretty closely mirrors my understanding, although I'm still an undergrad in econ.

What strikes me as odd is how these assets are being evaluated in a hyper modern way. We didn't formerly possess the technology to value assets day to day based upon market evaluations. The only problem I see with this, is that the assets were not originally issued with a continuous evaluation of their worth in mind. They were evaluated with the old system of accounting to think of. Joe is correct, but banks doing risk management to capital evaluations were working under old models pre-2007. When the rules changed, than they suddenly became insolvent.

I'm not a banking system expert, so I could be way off base.

dearieme writes:

One of the "excellent letters" includes "I cannot sell the house for what it is worth."

Jeff Harding writes:

I think that MTM is a good thing because it requires banks to fairly value their assets. Otherwise there would be no trust among banks and their depositors. MTM isn't something new. Basically it is a FASB rule that was adopted by the bank regulators.

According to the rules, companies only have to MTM those securities that are held for short-term investment or derivatives like subprime based securities. If securities are going to be held to maturity, they don’t have to MTM unless they believe that it is “permanently impaired.” Even then, they must only mark down values of the bad mortgages.

I therefore don’t see anything wrong with MTM.

Why should MTM be suspended in times of financial crisis? This is precisely the time when we should demand that banks should MTM.

Banks for years have been making bad financial decisions and now they want to paper it over as if they shouldn’t pay the penalty for their mistakes.

Bruce Wasserstein of Lazard said, “Accounting has now become an exercise in creative fiction. Saying assets are worth a lot doesn’t make them worth a lot.”

The problem is not MTM but bad decisions. And the banks know this. If you had bought a subprime security with money borrowed from them, I can assure you that they would require you to MTM and pay down the loan.

This is a red herring issue by banks who don't want to be told that their asset bases are unsound and that they must come up with additional capital.

I wrote on this issue in my blog: http://dailycapitalist.com/2009/01/10/whats-wrong-with-mark-to-market/

DC writes:

MTM is based on the belief that market prices are always correct. We just experienced two of the biggest market bubbles in the last 10 years (tech stocks and real estate). How can that possibly be true?

Instead, if you believe that market prices are always incorrect, and everything we see in prices is the price discovery process, I don't understand why you would put so much emphasis on what the market think. A couple of years ago, the market believed these securities were valued near par. Today, the market believes they are close to being worthless.

The reality is that no one knows what anything in this world is worth. The world has opinions which make up the buyers and the sellers. If these securities are truly worthless, then the holders of these securities will eventually suffer. Letting the "market" make this decision doesn't make any sense to me.

I would be in favor going back to cost basis accounting w/ asset value adjustments based on impairment. I agree that we can show mark to market values as footnotes for full disclosure. Why doesn't this seem like a reasonable solution?

Otherwise, when the markets do turn, we're again going to have a problem w/ mark to market b/c we're going to have overcapitalized banks when these assets are marked up.

Thomas Esmond Knox writes:

Shyam Sunder has written extensively on this subject. No-one seems to care.

"Shyam Sunder is the James L. Frank Professor of Accounting, Economics, and Finance at the Yale School of Management; Professor in the Department of Economics; and Professor (Adjunct) at the Yale Law School. He is a world-renowned accounting theorist and experimental economist. His research contributions include financial reporting, dissemination of information in security markets, statistical theory of valuation and design of electronic markets. He is a pioneer in the fields of experimental finance and experimental macroeconomics. Sunder has published six books and more than 150 articles in the leading journals of accounting, economics and finance, as well as in popular media. He is the immediate past president of the American Accounting Association and a winner of many research awards. Sunder’s current research includes the problem of structuring US and international accounting and auditing institutions to obtain a judicious and efficient balance between regulatory oversight and market competition."

Roger Cuddy writes:

The distortions caused by mark to market have been seen in both up and down trends now. At this point would it really be that difficult to get a consensus for a standard valuation methodology based on actual payment rate fluctuations? It seems wrong for a securitization or loan with little to no real impairment being valued at less than half par. Perhaps some middle ground between available for sale and hold to maturity classifications could also contribute to a more rational treatment.

nohype writes:

The important question is whether or not mark-to-market introduces destabilizing feedback into the system, so that small problems become larger problems. Most of the comments here seem oblivious to this possibility. The real-bills doctrine seemed like common sense in 1929-1933, but it introduced a destabilizing dynamic into the monetary system that caused the Great Depression. Are you sure that mark-to-market is not the real-bills doctrine of today?

Charcoal writes:

MTM has its pros and cons indeed. Without doing something related to MTM, our big banks will be dead or nationalized. I know their managements deserve it but these banks are envy of the world who entrust our big banks to manage their $. Losing our big banks will permanently damage our stature worldwide. Suspending MTM will be bad also because it loses transparency and thus credibility. Well, when we make $ in stock market, we pay capital gain tax, but when we lose $ in stock market, we are only allowed to take $3000 deduction per year. Why not asking the banks to do MTM but amortize their losses in accounting over the period until maturity? In this way, the pricing becomes less of an issue!

Pricing writes:

MTM isn't a problem, it is the inability for people to interpret MTM correctly. Consider the following example.

I have a risk asset that will pay $100 in 2 years plus market interest plus 30 bps. I can finance this asset by borrowing at 1%. So long as I believe the market interest rate plus 30 bps compensates for the riskiness of the asset plus my funding cost, I will purchase it.

I can now finance this same asset at 6%. What price would I pay, even assuming I will still receive timely interest and principal?

It is less than $100. The accounting standards require entities to discount cash flows based on another market participants view of risk. When the entire market financing cost increases, the corresponding asset prices decrease.

I buy the asset (for sake of argument, lets say at 85). Next year, I get paid libor + 30bps in interest. The asset price will pull to 100 so I recognize the interest income plus a market to market GAIN for bearing the risk of that asset.

Final year, I receive my principal back. Over time, I've recognized interest plus 15 in gains.

Marking to market a balance sheet for changes in discount rates puts everyone on the same footing.

This is easy if I assume that I'll receive the principal back. Assume the risky cash flows are much more risky than originally thought AND the market funding costs have increased. The asset may have to be purchased at a much lower rate to compensate over time for holding this risky asset and bearing the funding cost.

The entire process of marking the balance sheet can be seen as a rerisking of the entity. Either take the loss now and show investors the true return profile of your company, or suffer a negative net interest margin until the assets pay off.

The problem is that people are misinterpretting the balance sheet. A marked balance sheet is showing current values. An unmarked balance sheet is showing future values. Current values are far more valuable pieces of information to investors than future values.

cuOnTheOtherSide writes:

With all the talk about revoking Mark-to-Market not much in mentioned about the responsibility of the banks. They knew the rules going in. It is the same as "ignorance of the Law" as an excuse.

Some other aspects of the conversation not mentioned.

1. What were the assets on the books?

2. What was the quality of the assets on the books?

3. What was the leverage used for these assets?

4. With all the money these guys were paid, why did they seem to have little knowledge of the risk they were taking?

5. If they did not know or understand the risk they were taking why would you not want Mark-to-Market?

6. How much bigger the problem could have been if there was no Mark-to-market?

7. Not having Mark-to-Market is simply giving the banks permissions to overstate the value of assets on their books. Why would this be important to the banks?

Jeff Harding writes:

The purpose of MTM is to make sure that banks accurately state their balance sheets. We all know why that is good.

Most of us would probably say that one of the problems with our banking system is fractional banking, which, unless you know how to accurately predict business cycles, is inherently risky.

Thus, in general, we can't fault MTM for its pursuit of accurate reporting to create a more stable banking system.

All these issues raised about MTM deal with the fairness of its application to the valuation of assets. While it may be more difficult to value assets in chaotic times, it should be done.

Many of the complaints from banks that I have seen are about the difficulty of valuation, but they lump into that category normal difficult tasks with the most difficult tasks. They then say, "see it can't be done."

The house in Hawaii is a normal difficult task. Appraisers are asked to make judgments on these assets and do it all the time. Many clients don't like what they hear, but appraisal science hasn't changed.

Subprime MBS are on the more difficult scale of valuation. But many of them are valued by existing market means (Markit or OTC sales) so there is a methodology to do this. Again, the holders of these assets don't want to hear the results.

Look, this doesn't destabilize the banking system; it stabilizes it. Until these "toxic" assets on their books are properly valued and banks take the hit, we'll never recover. We'll have zombie banks and resulting stagnation like the Japanese.

Phillip Huggan writes:

I'm accounting ignorant, but accrual accounting seems to have no methodology for writing off projected cash-flow contraction.
I've seen many valid hypothesees about why MTM valuations are so painful:
1) Projected Great Depression.
2) Projected finance industry indefinite deflation; banks always wating til tomorrow to loan (my guess).
3) Projected future USA inflation (lowering value of long-term assets).
4) Projected future USD depreciation.
5) Projected poisoned well permanently of all CDO instruments.

I'd feel more comfortable modifying MTM if it is too draconian and the above 5 reasons aren't remotely possible. Say take some investor/business confidence index, and tie it to a MTM premium: when 90% of players are scared, induce a 9% premium on MTM values, when only 10% scared, only 1%. Accrual accounting looks backwards at a cash flow trend graph, and draws an optimistic line. With MTM you pop these bubbles.

Mark writes:

This discussion terrifies me. MTM has been greatly demonized in the press and getting rid of it is being viewed as some kind of a cure-all. But it is not the problem and the longer we focus on non-problems, the less we focus on real solutions. Wasting energy on demons increases the likelihood of a more serious recession or depression.

Taking Citigroup as an example, if Citigroup had fair valued all of its assets, we might have recognized their situation as being more critical six months to a year ago. The fact that Citi now trades at ~15% of tangible equity is not a result of mark-to-market accounting. It is because market participants don't trust Citi's valuation of loans (which are not marked-to-market).

Jon Hale writes:

The MTM that the banks want to suspend is SFAS 157, which was introduced in 2006 with an effective date for financial statements after 11-15-07. The FASB & SEC encountered objections and delayed the enforcement until 11-15-08. Before SFAS 157, assets that were held for trading were marked-to-market. SFAS 157 wants all assets and liabilities to be marked-to-market.

Part of what helped gain support for SFAS 157 is a clever use of the English language. If the FASB and the SEC had called it “Unfair Value Measures” and the subset “marked-to-trades-in-an-irrational-market”, no one would have taken it seriously. Instead SFAS 157 and 159 use the phrase “Fair Value” over 1,400 times to assure us that this method is fair and they tell us the new rules will help provide transparency.

Please remember that the FDR administration suspended a similar Mark-to-Market system in 1938. The SEC provided details on the 1938 suspension on page 44 of their 12-30-08 report to Congress in their “Study on Mark-to-Market Accounting”. The report is at http://www.sec.gov/news/studies/2008/marktomarket123008.pdf

In my opinion, some of the basic flaws in the MTM accounting measures are:

1. Bankers make loans based on projected cash flow. Once they’ve made the loan, the MTM measures ask the bankers to carry the loans at values equal to the trade prices of similar assets, not at cost or values based on cash flow. With disrupted markets, the trade prices can fluctuate over a wide range. Very few assets have escaped recent wide market price fluctuations. Assets that the banks might invest in, such as jumbo mortgages or credit card debt, have recently had wide trading price fluctuations. The change from cash-flow valuation to trade-price valuation has destroyed billions of dollars of equity. (See the 11-15/16-08 WSJ article on Fannie & Freddie. The chart shows the equity difference between the GAAP accounting in effect 9-30-08 and “Fair Value” accounting in effect for financial statements after November 15, 2008. The difference is roughly $70 billion less equity under the new accounting measures.) For assets that the banks intend to hold to maturity, MTM is useless and misleading.

2. MTM destroys predictability. The financial companies cannot mark-to-market until they have figures for the trade prices. This means huge surprises when the companies report their earnings. They can predict their cash flow, but not the end-of-quarter trade prices and resulting MTM valuation changes.

3. Under the new MTM rules, a financial institution can have a positive cash flow and a huge phantom loss from MTM adjustments. The lending patterns have been changed and many financial institutions have had to dilute their equity.

4. SFAS 157 wants all assets and liabilities to be marked-to-market. This means that a TARP contribution creates a phantom loss and reduces equity. (The TARP contribution improves the credit-worthiness, which increases the market value of the Company’s liabilities. Under SFAS 157 this translates to a loss, since the Company would have to pay more to buy-back their own liabilities. The loss reduces the equity.) This is counter-intuitive.

The SEC and the FASB have actively defended SFAS 157 and 159. The Statements themselves, the amendments and the SEC’s 12-30-08 letter to Congress total over 400 pages. It appears that they had good intentions, but the volume of amendments and explanations indicate that there were unintended consequences.

An interesting side question is who benefited from this recent accounting change?

1. Some of the regulators are hiring while the banks are laying workers off. Think about the FDIC as one example.

2. Articles have mentioned short sellers benefiting. (In addition to regular short-sellers and naked-short-sellers, also think about SKF and similar issues on the NYSE.)

3. Public firms are getting seriously hurt and some private and foreign firms are benefiting. (Think of the Lehman assets that were sold at distressed prices to Barclays and the Japanese securities company and the Merrill mortgage assets sold to a private TX hedge fund and the recent private offer for IndyMac assets, or the Nevada bank assets that went to PennyMac or the recent purchases by Wilber Ross etc.) Private and foreign firms are benefiting from the U.S. public companies’ MTM challenges. Many of the private firms do not have capital requirements or MTM requirements. In talking about profits, a PennyMac executive was recently quoted as commenting on his firm’s recent experience buying assets from the government. He said, “In fact, it’s off-the-charts-good.” (See page 5A of the 3-4-09 Sarasota Herald Tribune.)

Steve Forbes and others keep reminding us that some of the Bush-era financial policy changes contributed to the financial crisis. I appreciate Forbes and others reminding us that some of these policies, such as MTM, have been tried before.

“Those who do not learn from history are destined to repeat it.”

Comments for this entry have been closed
Return to top