High-cost, low-return bank capital

By Felix Salmon
February 17, 2010
William Wild has a great little paper up, disputing the widely-accepted idea that the best form of regulatory capital, for banks, is common equity. It isn't, he explains: shareholders have a strong incentive to maximize leverage and risk. They don't make banks safer, they make them riskier.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

William Wild has a great little paper up, disputing the widely-accepted idea that the best form of regulatory capital, for banks, is common equity. It isn’t, he explains: shareholders have a strong incentive to maximize leverage and risk. They don’t make banks safer, they make them riskier.

In its place, Wild proposes a form of regulatory capital which, far from having the large potential upside of shares, has no potential upside at all. Essentially, it’s just cash, invested in risk-free assets, throwing off a very modest income stream for its owner, and which can’t be repaid or redeemed. Yet if the bank gets into trouble, all that cash would go towards absorbing losses before there’s any default.

Such capital would not be attractive to investors. But, says Wild, that’s a feature, not a bug:

Irrespective of form, there is no doubt that the required upĀ­front premium or discount would be substantial. The perpetual instrument would provide, at best, an ongoing risk free rate of return but with clearly material levels of risk. To bemoan this high cost is, however, to miss the point. The highest quality form of regulatory capital will be expensive, given what it is required to do. As has been amply demonstrated, cheaper forms of regulatory capital are simply poorer at performing the function.

Yes, says Wild, the high cost of this capital will end up in higher borrowing costs for the people that banks lend to. But that’s also an increase in transparency: the hardest part of any bank’s loan-pricing operation is trying to come up with a cost of existing equity, while the cost of this new instrument will be much more obvious and easy to calculate.

Finally, says Wild, this new instrument is much less pro-cyclical than bank equity. When banks get into trouble, they tend to stop lending, rather than increase their capital by issuing dilutive new stock. And since banks generally get into trouble when the economy is doing badly, the decrease in lending only makes macroeconomic matters worse.

By contrast, Wild’s new form of regulatory capital could be issued much more easily, with a much smaller hit to shareholders — doing so might not even require board approval, if the cost of buying new capital could be incorporated into the terms of some new loan to a customer.

In general, I like the idea of a high-cost, low-return form of bank capital. When it comes to macroprudential regulation, you can’t let market rules like the link between risk and return drive everything.

Update: If the SSRN link doesn’t work, the paper can also be found here.

7 comments

Comments are closed.