Good and bad financial innovation

By Felix Salmon
August 27, 2009
Simon Johnson and James Kwak have an important article on financial innovation in the latest issue of Democracy. The broad thrust of the article I agree with -- as you might expect, given that after listening to my last debate on the subject, James Kwak came to the conclusion that "obviously I agree most with Salmon". (Thanks, James!)

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

Simon Johnson and James Kwak have an important article on financial innovation in the latest issue of Democracy. The broad thrust of the article I agree with — as you might expect, given that after listening to my last debate on the subject, James Kwak came to the conclusion that “obviously I agree most with Salmon”. (Thanks, James!)

That said, there are significant chunks of the Johnson and Kwak article that I disagree with, and I feel that it’s probably long past time that the “financial innovation: good or bad?” debate allow itself to make some nicer distinctions than have generally been made until now. So with the clear proviso that I’m on the “financial innovation: bad” side of the broader debate, here’s where I take issue with Johnson and Kwak.

First, they address securitization:

Securitization—the transformation of large, chunky loans into small pieces that can be easily distributed among many investors—was a beneficial innovation, because it expanded the pool of money available for lending. And securitization on its own, before the new products of the late 1990s and 2000s, did not produce the colossal boom and bust we have just lived through.

Sure, nothing, on its own, produced the colossal boom and bust we’ve just lived through. But securitization is as much to blame as anything else, if not more so. Securitization absolved lenders from sensible underwriting, since they knew they were just going on onsell that debt anyway. And it made bond investors comfortable with the idea of buying structured products tested only by models, as opposed to actual analyzable liabilities of real-world entities. You can’t phone up the CFO of a special-purpose entity and ask him how things are going. And lending in general works when there’s a relationship between the borrower and the lender. Securitization severs that relationship, which is harmful.

I’d also take issue with the idea that anything which expands the pool of money available for lending is, ipso facto, a good thing. To the contrary, things which expand the pool of money available for lending can serve only to inflate credit bubbles. In general, lending shouldn’t be easy to come by; and it should in principle always be just as easy to issue equity as it is to issue debt. We’re nowhere near that point right now, and securitization only serves to drag us further away from it.

Johnson and Kwak then attack the credit default swap:

Another paradigmatic product was the credit default swap, which insured a security (like a CDO) against the risk of default. But by underpricing that risk, it essentially tricked investors into buying securities that they would not otherwise have bought. The losses were borne by the companies that underpriced the credit default swaps, such as A.I.G., and by the government, which had to bail out A.I.G.—leading to the misallocation of capital to value-destroying investments.

The CDS, pace financial innovation, did not in and of itself underprice credit risk: it was simply a measure of credit risk. And indeed for most of the history of the CDS market, the basis on CDS was positive: as you’d intuitively expect, the cost of insuring a certain credit against default was higher than the spread on that credit’s bonds. You couldn’t lock in a risk-free return by simply buying a security and insuring it against default. So if the CDS market was underpricing risk, the bond market was underpricing risk even more. It was only after the credit market imploded that the negative-basis trade started becoming possible. So you can’t really blame a negative CDS basis for any part of the crisis.

Yes, it’s clear, in hindsight, that AIG, in particular, was underpricing credit risk. But that has nothing to do with the structure of the CDS market more generally. Instead, the problems with AIG surrounded the fact that it only ever sold credit protection, and never bought it; and that once it had sold protection, it used its triple-A credit rating to avoid having to put up any collateral against those positions or otherwise be forced to protect itself against loss. AIG in general, and AIG Financial Products in particular, did a lot of things wrong. But that’s not the fault of the CDS market.

Johnson and Kwak are right that regulators should be inherently suspicious of financial innovation; they’re possibly too polite to mention that this is largely because most financial innovation comprises, at its heart, some kind of regulatory arbitrage. (Securitization being no exception.) I agree also with the idea of standardizing CDS documentation, although it should be said that that is already happening to a large extent. I’m not at all sure, however, that standardized CDS will be much easier to regulate than the customized CDS of old. It’s not the customization which is the problem, it’s trying to get a grip on net positions, in a market which is constantly in flux. The way to solve that problem is to simply let the market continue down the road of the past six months, where CDS are increasingly being shunned as an asset class in favor of good old-fashioned bonds.

Johnson and Kwak then finish with a list of good financial innovations we might encourage: banking the underserved (a no-brainer), reforming health insurance (yes, but let’s not debate that here), and finally this:

We need innovation in financial education. A large part of our regulatory system relies on consumers being able to make intelligent choices when faced by an ever increasing and ever more complex set of financial choices. The recent crisis has shown that even large and supposedly sophisticated investors, such as municipalities and pension funds, did not fully understand the products they were buying. Economist Robert Shiller has proposed government-subsidized financial advice; this may not be a sufficient solution, but it is a start. Obama’s proposed Consumer Financial Protection Agency (first proposed by Elizabeth Warren in Democracy, Issue #5, “Unsafe At Any Rate”) could also go far in improving consumers’ understanding of their financial options.

This I’m much less sure about. You can be sure that no matter how good the financial education provided, the consumers of that education won’t end up being better educated about financial affairs than large and supposedly sophisticated investors, such as municipalities and pension funds. The problem with investors who made bad choices during the boom wasn’t that they were insufficiently educated: it was rather that they were educated too much. Financial education breeds overconfidence, and overconfidence was a much more important cause of the crisis than insufficient education was. If a strong CFPA prevents truly harmful products being sold to consumers, that’s the best we can hope for: a mass education program isn’t practicable and wouldn’t work even if it were implemented. The last thing I want is Robert Shiller being unleashed on the public, telling them that they can hedge the value of the equity in their houses by buying derivatives on house prices.

So thank you, Simon and James, for fighting the good fight. But this clearly isn’t the last word on the subject.

9 comments

Comments are closed.