Bailout

The very real moral-hazard trade

By Felix Salmon

Ezra Klein takes Jamie Dimon at face value when Dimon told the FCIC yesterday that the market didn’t consider bank debt to be risk-free, and that therefore there wasn’t a moral hazard too-big-to-fail trade.

But America’s biggest banks are too big to fail, and there is (and was) a moral-hazard trade. The key here is to look not at the risk-free yield on Treasury bills, but rather the cost of funds for small-enough-to-fail banks. If you do that, as James Kwak notes, you find that TBTF banks have a cost of funds 78bp lower than their smaller brethren.

Kwak concludes, reasonably enough, that the problem with the bank tax is that it’s far too small: TBTF banks are still much better off paying a 15bp tax and having a 78bp discount on their cost of funds than they would be if they had to lose both. In fact, the tax is lower than 15bp, since it’s not paid on deposits: Kwak calculates that for most banks it’ll be closer to 7bp.

Meanwhile, Jim Surowiecki has managed to convince himself that the moral-hazard trade doesn’t exist at all. But he’s looking for it in the wrong place. It’s not bank executives who put on this trade, it’s the institutional investors who lent to the banks. By reducing the banks’ cost of funds, they naturally enabled the banks to take on more risk.

Specifically, because the banks’ cost of funds was lower than the interest rate on the triple-A tranches of subprime CDOs, the banks had a huge incentive to borrow as much money as they could and invest the proceeds in those supposedly risk-free bonds. If the banks’ cost of funds reflected their real riskiness and wasn’t artificially kept down by the moral hazard trade, it’s unlikely that the CDOs would have looked nearly as attractive.

The moral-hazard trade is real, then, and it’s dangerous, and the bank tax doesn’t make it go away. For that, we’re going to have to look to regulatory reform, and specifically to significantly higher capital ratios for TBTF banks. There’s a good chance that will actually happen, but it’s by no means a done deal.

Full post as published by Macroprudential on January 14, 2010 (boomark / email).