Mark-to-Market Accounts Signal Caution for Investors

May 3 (Bloomberg) -- Mark-to-market accounting has long been viewed in academia as the gold standard for preparing financial statements. The rule makers, the Financial Accounting Standards Board and the International Accounting Standards Board, are coming to the same view. Yet shifting to those norms has some adverse consequences for investors.

For centuries, assets generally had been recorded on balance sheets at their actual “historical” costs. Critics argued that this method provided investors with stale information that was irrelevant to decision-making (“sunk” costs). Instead, they advocated marking assets to their estimated market prices, or “fair values.”

The greatest impact of the transition to mark-to-market accounting has been on financial assets and liabilities, which is where the best price information is found.

Banks and financial institutions are particularly affected. Their trading assets (debt, equity, Treasuries, derivatives and securitized loans) are on their balance sheets at market value. Gains and losses are booked in earnings each quarter. Because these entities are highly levered (typically more than 10-to-1), the effects on earnings are substantial. As an example, Citigroup Inc. had $292 billion in trading accounts as of Dec. 31, 2011, about 3.8 times its total market capitalization and 26.4 times its earnings for the year.

Current Prices

Clearly, mark-to-market accounting has a lot going for it, and current prices seem preferable to old. But several recent studies, including research that I conducted with Sudarshan Jayaraman, of the Olin Business School at Washington University in St. Louis, and Lakshmanan Shivakumar, of the London Business School, indicate it also has several limitations that investors need to be aware of:

-- Earnings expectations are more difficult to calculate: Under the new rules, published financial statements don’t reveal whether mark-to-market gains and losses are due to shocks to expected returns (yields) or shocks to expected cash flows, or both. This is important because yield shocks reverse in earnings over time, and cash-flow shocks don’t. Consequently, by omitting causes, mark-to-market accounting doesn’t provide enough information to form expectations of future earnings.

Take the example of a debt instrument bought for $100 and expected to pay a single $121 cash flow in two years, an annual yield of 10 percent. Imagine if a year after it is purchased, the bond is marked to its then market price, $105.22. This is bad news for the investor because the asset was expected to appreciate by 10 percent, to $110.

There is no way, however, for the investor to know why the price has dropped and this deprives him of the ability to accurately predict its future value.

If the bond fell because an increased risk of default reduced the expected cash flow to $115.74, and the instrument still is expected to yield 10 percent, then estimated next-period earnings would be $10.52 (calculated as $115.74 -$105.22, or as 10 percent of $105.22).

If, on the other hand, the bond fell because market forces had driven up the yield to 15 percent ($121/$105.22 - 1), then expected next-period earnings would be $15.78 ($121 - $105.22, or 15 percent of $105.22).

Interpretation is even messier if yields and default risk changed together. As the example illustrates, mark to market creates uncertainty about expected future earnings, and therefore also about the “surprise” content of these earnings when they are announced. The uncertainty only can be removed by knowing what caused the change in market price. Larger investors and insiders are in a better position to acquire that information. That means smaller “uninformed” investors are at a disadvantage in forming earnings expectations.

Bid-Ask Spreads

Our research shows that shares of banks that invest in trading securities are quoted at considerably wider bid-ask spreads than those that don’t. The spreads widen as banks’ trading portfolios increase. Widened spreads indicate that uninformed investors are attempting to compensate for their informational disadvantage.

Trading securities also are associated with diminished tracking by analysts, fewer earnings forecasts by management and less timely information environments generally.

Traditional accounting theory assumed that up-to-date prices provide all the information about company assets that investors need. That turns out to be incorrect. In the real world, earnings expectations play an important role, and forming them requires information about why mark-to-market prices changed.

-- Manipulated numbers: Under historical cost accounting, gains and losses would be booked only when assets were sold. This would allow companies to “manage” earnings by choosing what to sell, and when. Winners can be sold to book gains when earnings are short of consensus expectations, and losers can be unloaded in good years.

Mark to market books gains and losses as they occur, eliminating that timing discretion. But don’t be fooled: Companies can still manipulate the numbers. Because markets aren’t perfectly liquid, companies can make trades at the end of the quarter that influence mark-to-market prices. (Such window dressing is believed to be a reason for high quarter-end trading volumes.) Securities that aren’t actively traded attract “mark-to-model” accounting, which bases the reporting on subjective judgment and is subject to even greater abuse. In this case, too, uninformed investors who take reported numbers at face value are at a disadvantage.

-- Marking liabilities to market: FASB’s Statement 159 allows companies to record their liabilities at market price. Paradoxically, increased default risk reduces the value of debt and then triggers book gains. When Citigroup was bleeding during the first quarter of 2008, it booked a $12.65 billion pretax gain from deterioration in its credit quality. The gain was buried in the footnotes, and many shareholders wouldn’t have been aware of it. Even if they had, many would have found it difficult to interpret.

There are traps for investors in debt, too. Borrowers commonly agree to restrict risk by maintaining specified maximum leverage ratios, such as debt/assets or debt/Ebitda. When these covenants are violated, lenders obtain rights such as veto power over dividends, stock repurchases, new investment, mergers and acquisitions or more borrowing. The loan might become due immediately. Renegotiation frequently ensues. But if debt is written down on the balance sheet when asset quality deteriorates, the covenanted leverage ratio is rendered ineffective -- even if the assets are written down, too.

FASB introduced this rule in 2007, so researchers cannot untangle its effects from the financial crisis. Preliminary research shows a precipitous drop in the use of balance-sheet leverage covenants in Europe after the European Union adopted similar rules in 2005. No wonder: Balance sheets can no longer be relied on to record the actual amount that companies owe.

FASB is correct that mark-to-market accounting puts more up-to-date information in financial statements, but the theory on which that belief is founded places too little emphasis on what investors actually use the information for. Until standard setters change that mindset, investors need to exercise caution.

(Ray Ball, a professor of accounting at the University of Chicago Booth School of Business, is a contributor to Business Class. He is also a trustee of Harbor Funds and chairs its audit committee. The opinions expressed are his own.)

Read more opinion online from Bloomberg View.

Today’s highlights: the View editors on what’s missing from the U.S.-Afghanistan pact and better ways to fix the farm bill; Ezra Klein on how the U.S. isn’t like Greece; Amity Shlaes on why we should all back the gold standard; Caroline Baum on the lack of alternatives to austerity; Josh Barro on new arguments for a U.S. value-added tax.

To contact the writer of this article: Ray Ball at

To contact the editor responsible for this article: Max Berley at

    Before it's here, it's on the Bloomberg Terminal.