September 25, 2008
Mark-to-Market Isn’t to Blame
Blaming fair-value accounting for banking misadventures is like criticizing the newspaper for reporting a murder.
The credit meltdown has spawned a few false villains, and one is “mark-to-market” accounting. Too many observers—including Congress’s Republican Study Committee
—think that if we just suspend the accounting rule that requires financial firms to report certain investments at fair market value, then the crisis will go away. But accounting isn’t the culprit here. You may ask: What the heck is
“mark-to-market” accounting? The people who write and enforce our national accounting standards mandate that publicly traded financial companies must report some of their assets (things like mortgage-backed securities) and liabilities (like money they’ve borrowed from other institutions) at “marked-to-market” values. That is, if you bought a certain security at $100 last year, but you can’t find anyone to buy it this year for more than $40, you’ve got to report its worth as $40, since, unless you plan to build a time machine, that’s the “fair-value market price” it would command. Contrary to popular belief, mark-to-market accounting isn’t new. Financial institutions have reported many financial instruments this way for decades, usually because they wanted to. Their desire makes intuitive sense. If much of your business is borrowing huge amounts of money and then buying and selling securities to exploit and magnify small price changes in those securities, then you’ve got to have a way of telling the world how much money you’ve made this quarter doing just that. Informing your investors that Joe’s mortgage is still worth the same $250,000 it was 10 years ago (when somebody else lent him the money), minus Joe’s previous payments, doesn’t advance that goal. But telling them that you made $50 million on Joe’s mortgage and thousands of others like it this month by exploiting small changes in interest-rate expectations does. Financial institutions even fought for, and eventually won, the right to report some of their really
long-term assets at day-to-day “fair values.” Firms like Macquarie report the value of toll roads on such a basis so that they can take regular profits from them. Enron won the right to report long-term energy contracts on a “fair-value” basis, allowing it to do much the same thing, with a big dose of fraud mixed in. This stuff sounds crazy (and it might be), but it’s what the industry wanted. Especially over the past decade or so, banks thought that anything
—whether it’s a toll road in the first world or a power plant in India—could be monetized and treated as a perfectly liquid, instantly tradable financial instrument, like 100 shares of Exxon stock. But contrary to another popular belief, no financial firm has to report all
of its assets at “fair value.” What must
be reported that way are things like derivative securities, securities purchased temporarily—or what an institution hoped was temporarily—through constant trading techniques, and securities that companies have designated as “available for sale.” This, too, makes sense: if a bank always planned to sell a particular security pretty fast, and was, in fact, depending on the money from that sale to continue funding its operations and churning out profits, it should value that security at what it’s worth right now. When it comes to most long-term investments, financial firms have more discretion. If a bank had always planned to hold a security “to maturity”—that is, hang onto a mortgage-backed security for the entire life of all of the mortgages bundled into it, for example—it doesn’t have to pay much attention to those fair values. There is a big exception, though: if the bank believes that that long-term investment is permanently impaired, it must subtract that impairment from the long-term value. A security could go bad like this for many reasons: if the market thinks that 40 percent of the borrowers in a mortgage-backed security are going to default, that’s a permanent impairment, because the institution holding the security will suffer those losses even if it holds the security until maturity. But even then, an institution doesn’t have to write these long-term securities down to immediate market values; it only has to write them down to, say, the expected value of the remaining good
mortgages in the security. So what’s all this talk about fair-value rules being new and somehow precipitating or at least exacerbating the current crisis? In truth, the only new wrinkle came last November, when the accounting-standards people issued new guidelines for how to measure fair value. First, they reminded people that though companies usually have a choice about whether to report an asset in a “fair-value” category, once they decide to do so for a particular investment, they can’t change their minds. Second, the standards folks told companies how they should mark affected assets to “fair values” when there wasn’t much of a market for them. The board assigned three simple categories. “Level one” denoted assets that financial institutions could accurately compare with identical assets trading in active liquid markets, meaning that investors could feel confident that the “fair value” prices were current and reasonable. “Level two” contained assets whose fair values were harder to determine, but for which institutions could still find somewhat comparable assets trading in somewhat active markets. This designation gave companies a way to warn investors that they should treat those prices with some skepticism. “Level three” designated assets for which there was no market activity—like lots of mortgage-backed and derivative securities, starting last year. If a company labeled an asset “level three,” it was a signal to investors that it was next to impossible to assign a fair value to that asset. The companies didn’t have to report such values as zero, though. They could still provide their best guess as to the securities’ value, as well as the reasoning behind that guess. But the accounting board clearly meant for investors to treat any such “level three” asset values with a large helping of salt. Around the same time, of course, the markets for the most opaque, complex securities were seizing up, and hundreds of billions of dollars’ worth of residential mortgage securities, commercial mortgage securities, and the like have moved straight from the sterling “level one” corner office to the dreaded “level three” basement. Whereas once you could instantly find thousands of buyers for mortgage-backed securities close to the prices at which the securities were issued, nobody wants them anymore—at least not without weeks of scrutiny, and certainly not at the price the banks want to sell them. In demonizing fair-value rules, critics say that the standards have spawned write-down after write-down, causing yet more losses at financial institutions that use their peers’ values as guidelines, and causing more investors to flee, escalating the losses and causing big firms to fail. No one doubts that we’re experiencing a crisis in confidence in asset values, but fair-value accounting didn’t cause it. It could have been stopped only by the banks themselves. They could have chosen, starting more than two decades ago, to be in the long-term investment business rather than in the short-term, exotic-security creation and trading business. People who say that itâu™s not proper to value a long-term asset at today’s value miss the point. Most such assets were never meant to be long-term investments for the banks that had just issued them or still held them when the credit crisis struck. Moreover, fair-value accounting isn’t exacerbating the crisis, and suspending the rules won’t slow it. First, the problem for investors isn’t that banks are blindly, slavishly adhering to some arbitrary rules. It’s that with or without the rules, nobody knows what certain securities are worth. Investors didn’t short Lehman Brothers’ stock because it had written its “level three” securities down to zero (it hadn’t). They shorted Lehman partly because they didn’t think that it had written such securities down far enough—some were still valued at 70 percent of original value, while Merrill Lynch had sold what seemed to be similar securities at 22 percent of their original value. As for the charge that it’s ridiculous to value some mortgages at 22 percent on the dollar, and that fair-value accounting helps to create the absurdity: maybe, maybe not. The stark truth is that when you consider that banks wrote mortgages against houses that may have been more than 100 percent overvalued, and when you consider how much it costs such institutions to foreclose on a house and maintain it for a few months or longer before sale in a tough market, it’s easy to see how values get down to less than half. Subtract some more money for uncertainty—which markets do all the time—and you’re down to 22 percent, more or less. Finally, even if standard bearers and regulators suspended fair-market rules today, banks would still be wedded to fair-market principles, at least until all of today’s complex securities are unwound. Consider credit-derivative securities, a form of insurance against debt default. AIG, which holds half a trillion dollars in such obligations, would have gone bankrupt last week without government help. But AIG’s problem wasn’t some accounting rules. Even without them, AIG’s trading partners would have demanded higher cash collateral from the firm as ratings agencies downgraded the firm, due in part to their own private assessment of the chance that AIG would actually have to pay out on those claims. The same was true at firm after firm: risks increased and counterparties demanded more cash, as called for in private contracts. Changing the accounting rules midstream can’t change that. In the end, the only thing that was wrong with “fair value” accounting was that it was a mirror of the modern financial industry. Financial institutions thought that they could trade anything, anywhere, at any time, safely and virtually risk-free and for an instant profit. It turns out that they couldn’t. Fair value’s sin was in exposing that failure spectacularly. But the anti-“mark-to-market” crowd may well get its wish, not because the accounting and regulatory world will throw away fair-value rules, but because investors will regard the business behind such rules much more carefully. After all, financial institutions needed money from the outside world to create all of those fair-value investments in everything from mortgages to toll roads; it’s unlikely they’ll replenish their now-depleted coffers in the future, because investors now understand what complex securities and assets structured to trade instantaneously do not only on their way up—but on their way down.
Nicole Gelinas is the Searle Freedom Trust Fellow at the Manhattan Institute and a contributing editor of City Journal. Gelinas writes on urban economics and finance, municipal and corporate finance, business issues, and crime. She is a Chartered Financial Analyst (CFA) and a member of the New York Society of Securities Analysts. Gelinas has published analysis and opinion pieces on the op-ed pages of The New York Times, The Wall Street Journal, the Los Angeles Times, the San Diego Union Tribune, the New York Sun, the New York Daily News, the New York Post, the Dallas Morning News, the New Orleans Times-Picayune, and the Boston Herald. She has also written for Crain's New York Business and National Review Online. Before coming to City Journal, Gelinas was a business journalist for Thomson Financial in New York, where she covered the international syndicated-loan and private-debt markets. She also wrote a regular op-ed column for the New York Post. Gelinas graduated from the Newcomb College of Tulane University with a Bachelor of Arts in English literature. She and her husband live in Manhattan.