Alan Greenspan has learned nothing

By Felix Salmon
October 2, 2009
Atlantic event in Washington this morning, and it was quite shocking how little Greenspan seems to have learned from the crisis. Yes, he has decided that banks need more capital: "You cannot function with a financial system with capital as low as it was," he said, adding that "in retrospect, Basel II was clearly suboptimal. You can't have capital at 10% or less, because human nature changes too quickly."

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

David Leonhardt interviewed Alan Greenspan at the Atlantic event in Washington this morning, and it was quite shocking how little Greenspan seems to have learned from the crisis. Yes, he has decided that banks need more capital: “You cannot function with a financial system with capital as low as it was,” he said, adding that “in retrospect, Basel II was clearly suboptimal. You can’t have capital at 10% or less, because human nature changes too quickly.”

But on other matters — especially when it came to systemically-important things which didn’t fail in this particular crisis, Greenspan was alarmingly sanguine.

When Leonhardt brought up the subject of derivatives, Greenspan was at pains to differentiate interest-rate and foreign-exchange derviatives, on the one hand, from credit default swaps, on the other. He explained, quite rightly, that the percentage of the notional amount which players stand to lose is much higher in the CDS market that it is in say the rates market, but he never mentioned that notional amounts outstanding in the rates market are orders of magnitude greater than the CDS market ever was.

The degree to which Greenspan claimed not to be worried about the vast bulk of the derivatives business was highly alarming:

Nobody hears about problems in interest-rate or fx derivatives. We’ve had the most extraordinary stress test, and they came up clean. I’m saddened by the fact that the problems in CDS have been generalized to all financial derivatives.

For one thing, we do hear about problems in interest-rate derivatives: a man appearing at the conference later this afternoon, Larry Summers, contrived to lose $1 billion of Harvard’s money in just that market. Given the extreme measures which have been taken by the world’s central banks, and given the fact that global capital imbalances remain enormous, there’s clearly a lot of tail risk in the interest-rate and fx markets. If volatility there spikes in some unprecedented way — which is entirely possible — then no one knows who could end up becoming spectacularly unstuck in those markets.

Greenspan defended the CDS market, too, in familiar terms:

We have to make adjustments in the way that market is working, but fundamentally the concept of CDS is a very desirable one. You are distributing risk to those who are more interested in holding it or more capable of holding it, and that’s desirable.

Haven’t we learned that this isn’t true? Haven’t we learned that the people who end up holding the risk, once the banks have done their derivatives-based, structured-product thing, are precisely people who are not interested in holding it but who don’t realize what risk they’re holding? After all, those buyers will always be much more attractive, to risk sellers, than the kind of risk-hungry hedge funds who demand high returns for taking on that risk. The derivatives market turns out to have been one the best mechanisms ever designed for hiding risk, and I don’t see that as desirable in the least.

Greenspan, in other words, seems to have learned almost nothing from the crisis. When Northern Rock failed, he said, “the UK authorities were the best set of regulators in the world” — er no, they weren’t, as a glance at the leverage ratios at UK banks would tell anybody. So when Greenspan talks about inflation not being a problem until 2012, as he did at the end of his session, it’s not obvious why anybody should listen to him. He simply isn’t reliable or useful any more.

10 comments

Comments are closed.