April 14, 2008
Martin Hutchinson is the author of “Great Conservatives” (Academica Press, 2005) — details can be found on the Web site http://www.greatconservatives.com
Fair value accounting, by which debt and equity securities on a company’s balance sheet are “marked to market” — written up or down to their market price — has been hyped by accountants and regulators as the epitome of modern financial reporting, enabling investors to gain a completely true picture of their investment’s financial position. Indeed, Gerald White of the Chartered Financial Analyst Institute, speaking at an American Enterprise Institute conference Tuesday, believes it should be applied to all items on the balance sheet, not just financial instruments. There is just one problem: in the turbulence of the last nine months it has completely failed to work, and has indeed shown itself to be pro-cyclical, encouraging economically foolish behavior in both up and down cycles.
As someone who only thinks about accounting once a decade or so, I wasn’t really aware that much had changed from my business school days in the early 1970s. At that time, the values of assets on the balance sheet didn’t move much. Everything the company owned was dumped on the balance sheet at cost price and stayed there for decades while the world turned. The only exception was when the company held bonds or shares that had declined catastrophically in value (the occasional wobble was ignored) in which case they were declared “impaired” and their value written down. The fun for analysts was in finding companies whose downtown real estate was still held on the books at its value of 1926, when it had been bought, since there just could be a little teensy-weensy asset profit that might be unlocked from the company if one could figure out how.
This attitude to values was maintained through the inflation-accounting period of the late 1970s and early 1980s. Assets were assumed to be held for the long term, so buildings were written up by the movement in the consumer price index between the asset’s purchase date and the balance sheet date. US and British accounts differed in their approach to inflation accounting, which may have been one reason why it was abandoned fairly quickly once inflation returned to single digits, but neither system attempted to “mark to market.”
The “mark to market” approach had been used since 1940 by US investment banks, holders of large numbers of tradable securities, who needed to convince their regulators that their capital was adequate. It was not however used by British merchant banks, equally holders of substantial amounts of tradable securities. Only a small portion of merchant bank assets was held in a “trading account.” The remainder was held on a “back book” investment account and valued at cost. In this way, merchant banks were able to manage earnings very effectively; generally they built up large “hidden reserves” in good years which were amortized into earnings in years of unexpected dearth, so that the overall picture was smoothened. The result was to increase the confidence of the market in each merchant bank; people assumed that 200-year-old institutions had accumulated enough “hidden reserves” and undervalued real estate to smooth out any problems that might arise.
Mark to market accounting spread beyond the traditional investment banks around the late 1980s. Its great advantage, to executives who were for the first time paid a large portion of their remuneration based on profits, was that values could be marked up as well as down. No longer did you have to sell that illiquid investment in order to realize a profit on it and be paid a bonus; you could now recognize its increase in value on an annual basis. Conventional book value accounting was derided as being old-fashioned, and the original acquisition cost of a position scorned as being hopelessly irrelevant to an up-to-the-minute valuation.
The new accounting standard FAS157, propounded in September 2006 and coming into effect for fiscal years beginning in 2008, codifies this trend but does not materially alter it. Its most startling feature for a layman is that it allows companies to mark-to-market assets for which there is no market. Financial assets are divided into three “levels” according to their degree of marketability. Level 1 assets are those for which a ready market exists, Level 2 assets are those for which a market exists for comparable securities and Level 3 assets are those for which no market exists, which are to be valued by use of mathematical models.
In the use of this new standard by the investment banks, two problems have appeared. First the “Level 3” designation has given rise to all kinds of model-building creativity, by which “market” values can be artificially increased. Such is the ethical standard of Wall Street that this was done in a number of investment banks during the bull market years before 2007, with the resulting increases in value being taken into earnings and large bonuses paid to executives in actual cash based on the imaginary increases in value. This technique was particularly useful for private equity holdings, where the normal procedure of holding the position until it could be realized and booking the profit only then proved irredeemably boring to the impatient titans of Wall Street finance.
The second problem that has now emerged runs in the opposite direction. “Level 2” valuation techniques allow the institution to ignore prices received in a “distress sale.” However, in a bear market almost all sales are distress sales; the asset holder is distressed that his asset has declined in value and is only selling it because he needs the cash. Since the values of ABX indices on subprime mortgages have declined to a modest fraction of par, holders of this rubbish have decided that they do not represent the true market. Consequently, in their view, there is no true market, consequently the assets are “Level 3.” It is notable for example that Goldman Sachs’ Level 3 assets increased in the last quarter from $54.7 billion to $82.3 billion. Since it seems most unlikely that Goldman, a smart operator if ever there was one, has been deliberately loading up on $26.6 billion worth of illiquid rubbish, the change must result largely from strategic reclassification from Level 2 to Level 3. Indeed, Goldman’s Level 3 asset backed securities doubled during the quarter to $25 billion, presumably for precisely the reason that Goldman found unattractive the market prices prevailing for those securities.
At $82.3 billion, Goldman Sachs “Level 3” assets are more than twice its capital. This is not therefore a peripheral problem, which can be allowed to remain hidden within the arcana of accounting conferences. The reality is that, as was demonstrated in the true recessions of 1973-74 and 1980-82 but not in the mere dips of 1990-92 and 2001-02, the value of highly illiquid “Level 3” assets taken on at the peak of a bull market is pretty well a big fat zero. Since the Michigan Consumer Sentiment Index is now at its lowest level in 26 years, it is beginning to become clear even to investment bankers that the US and probably the world are in the early stages of a “proper” recession of the 1973-74 and 1980-82 pattern, albeit a recession with a considerable inflation problem attached. In that event, the “Level 3” assets on the balance sheets of Goldman Sachs and other financial institutions are worth only a small fraction of their nominal book value, and the institutions themselves will eventually be demonstrated to be insolvent. So much for the supposed greater transparency of “fair value accounting.”
The shaky state of the world’s major financial institutions is a matter of history; their shareholders and creditors have been deluded by the fictions of fair value accounting and the excitement and profitability of a prolonged asset bubble. Repeated bankruptcies are probably the only fair way out; in practice however most such institutions will be deemed “too big to fail” and the cost of their rescue will fall on the world’s long suffering taxpayers.
At that point, the taxpayers should have something to say about the accounting practices that made financial institution balance sheets and income statements so illusory. Fair value accounting must go, and be replaced by the much sounder principle that pertained previously, that assets can and should be written down, but must never be written up until they are sold. If banks and investment banks wish to make a memorandum account of the fair value of their investment positions, they should certainly be free to do so but increases in that fair value should not be brought through the income statement until they are sold, and bonuses should not be paid on the basis of such increases.
Accounting was invented in fifteenth century Florence (well, actually in Dubrovnik, a dependency of Florence at that time.) Another Florentine of that period, Niccolo Macchiavelli, would instantly have recognized the attractions of fair value accounting, enabling investment bankers to maximize their short term returns, while leaving the long term problems of an illiquid and clogged balance sheet to be sorted out by regulators at the expense of the public. In the long run, institutions that are “too big to fail” are also too big to take large risks at public expense, and both their operations and their accounting should be scaled back to the most conservative basis. Their place will be taken by smaller institutions, not subject to bailouts by the general public, who will remain relatively unrestricted and will be able to assume risks in order to innovate, albeit only on a modest scale as befitting their smaller balance sheet.
In other words, the structure of finance, like the principles of accounting, needs to revert at least 40 years; probably, in the United States, 80 years.
Reprinted from The Bear’s Lair at PrudentBear.com