The narrow banking distraction

Two years later, we're still hearing about narrow banking as the solution to the problem of risky behaviour by financial institutions.

In fact, most of the problems that led to the 2008 bailouts were barely related to investment banking. The primary cause of the financial crisis was a collapse in the value of mortgages extended by deposit-taking institutions.

The sheer volume of mortgage lending was indeed partly enabled by investment banks, helping commercial banks to securitise the loans. But separating those from the deposit-taking banks would not have stopped this.

So is there a way to stop this from happening again? What was it that led banks to take these risks, and why did they pose a problem for the whole financial system? Why did we need to bail them out?

The size of banks is a potential risk factor. And the Volcker plan to impose a tax on wholesale borrowing and a cap on the size of individual banks would help with this. But if, instead of ten huge institutions lending too much on overpriced property, we'd had fifty smaller ones following the same policy, we'd still have ended up bailing them all out.

The real problem has been identified by Garett Jones in a prescient tweet earlier this year. Not "Too big to fail" but "Too correlated to fail":
In a financial crisis, TBTF seems to matter less than Too Correlated To Fail. Type-of-asset beats out size-of-liability.
The problem was that nearly every bank put its money into the same asset class. This both exposed the system to a much higher risk, and in itself contributed to the bubble in property prices.

But isn't capital supposed to diversify automatically? If too many people invest in one thing, its price should go up and other investors should find cheaper opportunities elsewhere. Unfortunately, these signals were not working properly. Partly this is a result of investors buying not to make a normal return on capital, but in the hope of selling to another investor later.

And partly it is caused by an network externality of information - that is, the fact that once a certain product is shown to make money for a bank, they are much more likely to invest in that product than others. In an efficient competitive commodity market, evidence that one product is profitable will marginally bias investors towards that product. But in a market which is information-driven, that little bit of extra evidence can completely transform behaviour, causing investors to pile into the profitable product and neglect others.

It's not hard to see that this happened in the financial markets throughout the period from about 2004 to 2007. Once mortgage-backed CDOs were shown to work, they became the asset of choice for investors controlling trillions of dollars; and the flow of money pouring into them made them even more profitable, reinforcing the belief that they were a smart thing to buy.

It's part of a regulator's job to combat externalities by putting the correct price on them. Sometimes an externality imposes a social cost - and highly correlated lending certainly does. If the market doesn't impose its own cost - and as we've seen above, it may not do - the state or regulator should impose a cost to correct the distortion.

So the key banking reform to make the financial system more stable is: putting a price on correlation.

How this can be achieved will be the subject of another article later this week.

Update: Niklas Blanchard's latest item reminds me that he's written along similar lines recently: "Too brittle to sustain".


PunditusMaximus said…
This is a weird discussion to me. Why aren't we just reverting back to 1978 rules, before our "crisis every 4 years" system, and going from there?

Popular posts from this blog

What is the difference between cognitive economics and behavioural finance?

Is bad news for the Treasury good for the private sector?

Dead rats and dopamine - a new publication