Why valuation matters more than capital May 8, 2011 at 9:30 am

An article by Raihan Zamil on vox.eu makes – albeit a little unclearly – a point I have emphasised for a long time: valuation matters more than capital. (See also for instance here and here.)


Well, let’s take a typical bank. Say it has 100 of assets, supported by 90 of liabilities and 10 of equity.

Adding 2 or 3 to the equity is really controversial: asking for a Basel ratio of 12% is a hard sell. (15% is crazy, by the way. Just saying.) So going from 100/90/10 to 100/88/12 is difficult for supervisors.

But what if the assets aren’t really worth 100? If they are only ‘really’ worth 95, then what we really have is 95/90/5, and the ‘true’ Basel ratio is only 5.2%. Then increasing the equity by 2 points would still leave the bank a significant distance from being well capitalised.

Moreover, a 5% difference in asset valuation across something as big and complicated as a bank can easily happen. Ensuring that provisions in the banking book and marks in the trading book are accurate is really, really hard (even if you are not trying to pull the wool over the shareholder’s eyes).

What does this mean for bank supervision? It means that before worrying about capital, supervisors have to put a lot – and I mean a lot – of effort into checking valuation methodologies, both in theory and in practice. To be fair this happens to some extent in most juristictions already, but given how critical it is, and how hard it is to do correctly*, it would be far better to be over- than under-resourced here.

Now, once you are sure that the valuations are reasonable, you can then look at how leveraged the bank is and how quickly it might lose its capital. But you can’t look at that absent confidence in valuations as you basically know nothing about an institition if you don’t know that its valuations have been diligently determined.

Eat Me

* The above might be seen to imply that there is a ‘correct’ value which can be discovered with sufficient diligence for all assets and liabilities. I don’t believe that is true. In many ways the process – the process of testing valuations and reporting uncertainties, of checking methodologies – is more important than the precise answers.

5 Responses to “Why valuation matters more than capital”

  1. A good start would be insisting on consistent valuations ie If I record an asset at X the you must also mark it at the same ‘agreed’ value. It would be better if the marks were reasonable as well but on a macro basis consistency almost as important as at least the system wide capital calc would be close to the truth even if individual institutions were out

  2. Tim – interesting, but I am not sure that I agree. For unhedged assets that may be reasonable – although there is still a big risk if the consensus mark goes down and everyone tries to sell at once – but for hedged assets the mark is a function of the hedge strategy, so one would need to be mindful of that too. But certainly supervisors reviewing differences and asking banks to justify them makes a lot of sense.

  3. […] Why valuation matters more than […]

  4. But – err – the very fact that valuations are both so subjective and so volatile is one fundamental reason why banks should be made to hold a great deal MORE capital (the need to limit unfunded credit and so to ameliorate the boom-bust cycle is another).

    Hence, you have just undermined your own opinion of a 15% ratio as ‘crazy’ – or at least your implication that its madness lies in being too high and not, as is really the case, too low!

  5. […] A cookie jar is something that, whenever you want a cookie, you can get one out of. A reverse cookie jar, on the other hand, is something that, whenever you have a cookie, you put it in the jar and it disappears. (What a nasty idea.) Falkenblog suggests that banks right now are like reverse cookie jars. That might be right… […]