The financial crisis has had its share of heroes, villains and scapegoats. One of the more prominent villains has been "fair value" accounting. Why? Well the argument runs like this: banks measured their investments at fair value, the crisis kicked in and market prices dropped, banks were forced to book losses on the investments, market prices dropped even further, banks marked their portfolio based on the new (lower) market prices and losses went up.
In short, an alarming, seemingly never-ending spiral. The effect: balance sheets were destroyed overnight and some of the biggest boys, including Citigroup, Bear Stearns, AIG and Lehman Brothers, either went belly up or came very close to failure. Let's go back to basics. First, why fair value? Well, companies have basically two options for reporting the numbers on their balance sheets - historical cost and fair value. Historical cost is just carrying the asset at the same price you bought it for. So if you bought IBM shares for US$200 million (Dh734.6m) in 1997, you would carry it at $200m even today.
The US-based Financial Accounting Standards Board (FASB) defines fair value "as the price received to sell an asset or the price paid to transfer a liability in any transaction taking place in an active market". The definition from the other major standard setter, the International Accounting Standards Board (IASB), is very close to this. Generally speaking, the asset value on your balance sheet should be the market value, or close to it. So for the same IBM shares, if the price today was $250m, you'd book a gain in your income statement of $50m and if the price dipped to $180m, you would be hit by a $20m loss. This is a very crude example as there are many ifs and buts.
Fair value is the major objective of standard setters worldwide. They see it as hugely important and the fair value concept has even spread from financial assets and liabilities to non financial assets; so much so that the IASB issued a standard a decade ago allowing companies holding property to measure these at fair value. Both methods have their supporters. The fair value camp has three strong points.
First, fair value shows the true value of the asset. This allows the balance sheet to show the true financial position. Otherwise you end up with artificially inflated, or depressed, asset values that are out of sync with reality. Second, they say that by passing any gains and losses through the income statement, the company's profit is less of a purely accounting creation with all the related issues and more of a realistic, economic profit.
Third, what is the alternative to fair value? They declare that historical cost is a far inferior measure and should be avoided. The historical cost camp have their points. The first is that fair value at worst caused the crisis and at best made it worse. They condemn it as pro-cyclical. This sounds plausible until you think of what really triggered the chaos, the subprime crisis, which was an enormous housing bubble fuelled by cheap money and speculation that burst due to declining affordability and demand. This has nothing to do with bad accounting and all to do with very bad banking.
The second complaint is that fair value makes the financial statements volatile by immediately booking all the changes in market value through the balance sheet and income statement. I find this argument a tad hypocritical, especially when you consider that no one raised a hue and cry when asset prices (and profits) were rapidly heading north between 1982 and 2006. Bit of a cherry-picking mentality, then.
Fair value gets a bad press not because of the concept, but the way it has been applied. There are three main problems. In any crisis, buyers suddenly develop cold feet and the market grinds to a halt, and hence there is no market value. In such cases, the standards allow the use of financial models to derive a value that is a proxy for market value. But this could get very dodgy as any model is based on a series of inputs, and by fiddling with these inputs you can get the value you want, still justify it and compare the book values with the model value. This is called "marking to model".
The second issue with fair value is that any change in asset fair value passes through the income statement. It's bad if the change is negative as profits take a hit. Actually, it's also bad if it's positive. How? Because all these are just accounting entries without any cash flowing in. Companies should be paying dividends only out of profits. But when an increase in fair value inflates profits but doesn't bring in any cash, the company risks paying dividends out of capital.
The third defect is related to the second. The company did not actually make $50m in cash on the IBM shares but investors look at the income statement think it did and start demanding higher dividends, and so on. Hence it is highly misleading. So what's the remedy? My vote is still for fair value, but with some tweaking. Whenever financial models are used the inputs should be as market-based as possible, not some theoretical number coughed up by another model.
Managers' estimates of fair values based upon their own judgments should not qualify as fair values. Independent experts (not the auditors, who typically lack the expertise needed) should review these models. All changes in fair value should be routed through the balance sheet and not the income statement so confusion is reduced. Investments that the bank intends and is able to hold to maturity should not be marked to market.
Now if only we could get all parties to agree on this. Binod Shankar is a CFA Charterholder. He is a writer and consultant and runs Genesis, a Dubai-based financial training company.