No accounting for turbulent times?
Responding to Q4's conversation "Did innovation cause the credit crisis?" Rick Antle, William S. Beinecke Professor of Accounting at Yale SOM, puts accounting changes and their role in the current financial turmoil in context.
In a discussion with Q4, Yale SOM finance professors Frank Fabozzi, William Goetzmann, and Gary Gorton discussed the idea of securitization and its role in the current financial crisis.
Gorton made the comment:
[T]he press has only recently started to distinguish between losses and write-downs. So most of the stuff is not really losses, its write-downs. A write-down is an accounting event. The accounting profession has bought into the finance profession’s idea of efficient markets. So if the price of something goes down, the idea is that the value of it actually went down. And that’s a problem when you’re in a panic or a crisis like we’re in today where there are no prices. And to the extent that there are prices, they’re fire-sale prices.
So you’re writing things down under GAAP accounting using prices that we know are not exactly what they should be. And the problem is that GAAP accounting has real effects because the measured capital ratios for regulated financial institutions go down, and then those institutions have to raise capital at very punitive rates.
Fabozzi wondered what the accounting perspective might be. Here is a response from Rick Antle, William S. Beinecke Professor of Accounting at SOM.
Q: What’s your take on Gary Gorton’s comment?
To quote Stephen Ryan, formerly on the SOM faculty and now an accounting professor at the Stern School at NYU, “illiquidity is a tough nut to crack.” If there had been a market out there but now there isn’t, I don't know what you mean by “value.” Without a functioning market, we do not know what value actually is. So what do we want to show on a company’s books? It is going to have to be some sort of estimate, and surely it would include a big discount for illiquidity.
When Gary says that the press has only started distinguishing between losses and write-downs, a write-down is different from having actually taken a loss. But how different depends on what the item is, what the circumstances are, and how good that write-down number is. For example if the write-down is for marketable securities in the normal course of business, something like IBM stock, well, a write-down is as good as a loss.
One of the interesting things about this financial crisis is that companies have come out saying their write-down is $4 billion and two weeks later they say it's $8 billion. That's got to be telling you something about how fast estimates of valuation are changing.
When the numbers are changing that fast, other aspects of institutions have a sort of rigidity in them where they're not really processing the information in a fully intelligent way. For example, capital requirement regulations might have been more in tune with accounting practices that were not so quick to recognize gains and losses. It's kind of ironic to hear arguments from Wall Street that the accounting is too fast, because for years Wall Street made lots of money by engineering financial products to have certain accounting effects. People took advantage of the fact that accountants and the accounting rule makers didn't adjust fast enough to design financial products that had certain financial statement effects that may not have been real.
At any rate, you have to think about not only the problems associated with current practice, but what problems would be associated with other ways of accounting — out-of-date values have their own problems, which we had good lessons in with the savings and loan crisis. But now, the question is whether the pendulum has swung too far. Are we making accounts too sensitive to not just market values but all sorts of models of things that there aren't values for?
If you go all the way back to the justification for historical-cost accounting, it's rooted in the "going concern" notion. That is, in most circumstances we're doing accounting for firms as if they’re going to keep doing business. By doing too much marking to market and marking to model you're basically just viewing companies as portfolios of individual things that can be bought and sold separately. That is essentially liquidation accounting or bankruptcy accounting. I don't really agree with that because we recognize that organizations are there to create more value than they consume. The whole is worth much more than the sum of the parts, if you will.
Q: Could you explain the differences among amortized cost, mark to market, and mark to model accounting?
Thirty years ago if you bought an asset you would record its cost and you really didn't do much to that until you sold it. You might have amortized its value in a formulaic way based almost entirely on the conditions that were in effect at the time of the transaction. That’s the historical-cost model. It has the problem that a transaction has to take place before an accountant recognizes unexpected changes in value.
During the S&L crisis of the 1980s, people looked at the balance sheets of the financial institutions and found out that they had many items on their balance sheets at amortized historical cost. But since the value of those items had declined in an unexpected way, the institutions didn't really have the value in those assets that they appeared to have. What came out of that was a lot of pressure to show values using up-to-date information at any point in time.
The essence of mark to market is to recognize changes in values in the accounts before a transaction has occurred. We call it “marking to market” because the most reliable source of up-to-date value information is the market price. And so the first place they tried it was where information was readily available, that was marketable securities — something like the New York Stock Exchange. If you can't mark a marketable security to market, what can you mark to market?
While we started off with a situation where you've got market prices, we extended it to cases where there aren't such good markets and the assets are more specialized, like houses. You might be able to find out what your neighbor's house sold for, and one on the next block, but that's no guarantee, even when prices aren't moving much, that your house will sell for the same thing. Every house is a little bit different. We might start with comparables, but if we were going to mark your house to market, we still wouldn't know exactly what it would be worth.
The notion of fair value got introduced. The concept is to create some kind of ideal. It's the price at which a willing buyer and willing seller would exchange something. It's a much more difficult concept because what do you mean by a "willing buyer" and a "willing seller"? Is there only one price? Or a whole range of prices? And what if willing buyers and willing sellers can’t agree on a price — which is a pretty good way to think of illiquidity.
Pushing further from situations where there is a good market, in the extreme, what you get is mark to model, which means that there is no market but I have a model of the value and some idea of what the value might be, and I'm going to make an accounting entry to adjust the balance sheet books to bring the value in line with that model.
How do they figure out the numbers? If there's not a market for something directly, hopefully there are markets for parts of the model that let us construct pretty good estimates of the value. But, basically, the fewer market-related inputs you have, the more you're just making stuff up.
Q: Going back to Gary’s comment, can fear or panic artificially reduce value?
If everybody's forced to sell and there just isn’t the volume of buyers to buy right now, that's not a functioning market situation. That's a case where supply and demand breaks down, because there aren’t people equipped to take either side of the market, depending on where they think the best value is. They're being driven by something else. They don't have or they can't raise the money for some reason.
In well-functioning financial markets, if something is valuable, you can sell it for a fair price. And the expansion of the financial markets has greatly enhanced, I think, the welfare of society by allowing people to share risk and do a lot of other things. But in some sense we probably got spoiled
So this lack of liquidity is very perplexing. If the price of a security has been driven down too low, then private equity firms should buy it. Hedge funds should buy it. The argument that has to be made is that somehow, systemically, everybody is on one side of the boat — and I don't know why that would be, unless the volume is just too big. When you're talking about Fannie and Freddie, you're talking about trillions of dollars. There aren't enough private equity guys in the world to make that up. It may just be what's really going on is an order of magnitude bigger than what the system is capable of processing.
Q: Are there any areas where accounting regulations should or may change coming out of the financial crisis?
You could see some really big and interesting things. For example, I have seen a proposal to disclose losses separately from write-downs. So here are all the realized transactions and here’s fair value estimates. Or you could have an intermediate ground where things are portrayed as historical costs and everything else flagged in a different section, separated and isolated.
I think with decreasing costs of information acquisition and processing, the changes you might see will include increased disclosures like that, as opposed to thinking that what accounting is going to do is come up with one or three or four magic numbers that everybody can take as given.
Financial reports are already so complex, to think that the way out of this is more complexity and more disclosures is probably naive. In some ways that's just abandoning the fundamental problem of accounting, which is how you communicate in a more efficient way than just giving the raw data. And providing more disclosures won’t ease the regulator’s problem of defining something like capital ratios that financial institutions must maintain.
Q: Is there anything that accounting could have brought to the situation earlier, pre-crisis period, to say maybe we need to be looking at how we're approaching these mortgages or these securitizations?
Well, that's a very interesting question, because we've got to ask ourselves how much we can expect from the accounting profession. If finance experts can provide reliable ways to estimate value in illiquid markets, auditors could assess whether the methods and results make sense in a particular case. But the primary responsibility has to lie with the people who are designing, valuating, and executing transactions — if they can’t agree on values, it is too much to expect accountants to do it. If we want to make it the accountant’s job, then we have everything backwards.
Accounting has a lot of limits in its ability to head off big problems or anticipate too much. It's more a process that lets everybody know what the rules are — what's legit and what's not. I don't think accounting rule makers will ever be able to be out in front of something like this. You can imagine the outcry. Suppose some of these securities had really gone up, and the accountants had insisted that they use historical prices. You'd have people screaming from the woodwork "You idiots, just look at the market price. The value of this thing is way higher than the historical cost! Why are you doing this to us? We have much more capital than you're showing because you're not letting us mark up the assets, and you are just holding us back."
Interview conducted and edited by Ted O’Callahan.
For more on the concerns around the various models of accounting read a precient editorial in the Financial Times by SOM's Shyam Sunder and Stella Fearnley of Bournemouth University.