Monday, 31 March 2008

The procyclicality of fair value

Broadly I am a supporter of the use of carefully estimated fair values for financial instruments: I think that accrual accounting hides a lot of information that the users of financial statements deserve to know. However, there is one aspect of fair value that is troubling: how it combines with leverage to provide another incentive to procyclicality. To see this, suppose we start with a bank that is Basel 1 adequate, i.e. its leverage is less than 12.5:1 (remember 1/8% = 12.5).

Start Year 1

This bank uses fair value for all of its assets, and the markets rise, so we find

Start Year 1End Year 1

Of the 100 of P/L 50 is dividended to shareholders and 50 is kept as retained earnings. Retained earnings are a component of capital so we have:

Start Year 1End Year 1

At this point management will be concerned that the ROE of the bank will suffer due to the falling leverage, so they acquire more assets to get the leverage back to 11:1, originating 450 of new assets.

Start Year 1End Year 1Start Year 2

Unfortunately now the market falls back and the 1650 of assets are now worth only 1600. This causes a loss of 50. Losses are a deduction from capital, so now we find

Start Year 1End Year 1Start Year 2End Year 2

At this point the bank is capitally inadequate and either has to sell 350 of assets or raise 28 of new capital to get its leverage back under the regulatory maximum of 12.5:1. In other words, when times are good, the bank extends more credit, and when they are bad, it either has to raise new capital (and until it has recapitalised it cannot lend further) - or perhaps even worse, it may well sell assets into a falling market, exacerbating the decline.

One could argue this is not the fault of fair value, but rather of the leveraging up that took place during the rising market. However in order not to encourage this, shareholders need to understand that, unlike an accrual accounted bank, a fair value bank will have low leverage in a rising asset price environment and higher leverage in a falling one. I am not convinced that is well understood at the moment so the temptation to overleverage may well be there.

Labels: ,


Blogger Andrew said...

The insurance industry can be equally dumb-measuring its ability ot underwrite business by volume of premium, so that the lower the premium rate the highe the volume of business that could be booked agaqinst a given capital base. Only recent;ly has a risk based measure been introduced, and that ineffectively. Sad when regulators do not understand the dynamics of the businesses they are regulating.An MBA is notr everything.

2:57 pm  

Post a Comment

Links to this post:

Create a Link

<< Home