Arnold Kling  

Bad Bets or Co-ordination Failure?

From Poverty to Prosperity<... Financial Reform, 2...

James Kwak discusses an issue that has been dogging the analysis of the financial crisis since the beginning, and which has resurfaced now that people are looking at what steps to take to reform the financial system.

I put the question this way: Was the financial crisis just a bunch of bad bets, or was it a co-ordination failure?

The bad bets would be speculative housing purchases and mortgage loans made to finance those purchases. The co-ordination failure is like a bank run. As Kwak puts it,

A repo, or repurchase agreement, is a transaction where one party (the "shadow bank") sells some securities to another party (the "depositor") in exchange for cash and simultaneously agrees to buy those securities back at a predetermined (higher) price at some date in the (near) future (like tomorrow). In effect, the depositor is lending cash to the shadow bank, and holding the securities as collateral; the difference in the two prices is the interest. It wants the collateral because nothing else is guaranteeing its loan to the shadow bank (as opposed to ordinary FDIC-insured deposits). The collateral is generally worth at least as much as the amount of the loan, to minimize the risk to the depositor; but the remaining risk is that the shadow bank won't make good on the repo and the collateral will fall in value.

...So when the "depositors" got nervous about investment banks like Bear Stearns, they refused to roll over their repo agreements (that is, when the shadow bank closed a repo by buying back the securities, the depositor refused to lend new cash via a new repo), or they imposed a larger "haircut" - they lent less cash for the same amount of collateral. The result is a bank run - only this time the run is on the shadow bank.

I call a bank run a co-ordination failure because if the depositors could all get together and co-ordinate, they could agree not to withdraw funds unnecessarily, and the bank would be fine. In the shadow banking system, if everyone would just agree not to demand extra collateral, then everything would be fine--supposedly.

Since the crisis broke last year, the policy insiders have been acting as if what we were facing was a horrible co-ordination failure. On the other side of the fence, a number of us in the peanut gallery have been saying that the problem is one of bad bets.

One argument in favor of co-ordination failure is the point that if you look at the stock market, the total loss of wealth is much more than the loss of housing market wealth. However, I could make a couple of counter-arguments. First, the stock market was discounting future profits from mortgage finance. So the loss in the market reflected the discovery of losses on bets already made plus the realization that the betting game was not going to be profitable in the future. In addition, you had Paulson and Bernanke scaring the bejeezus out of everybody, and the emergence of a political environment very hostile to market capitalism.

To me, the co-ordination failure story implies that mortgage assets were massively undervalued because of liquidity problems. Any hedge fund could have made a ton of money buying mortgage securities. What Paulson and Bernanke were doing was saying that government could act as that hedge fund.

For example, take Freddie Mac and Fannie Mae. They were not viable last year, in part because investors were demanding 100 basis points more than they used to in order to lend to them. If that was a co-ordination failure, that means that if all the investors had agreed to lend at rates closer to Treasury rates, Freddie and Fannie would have survived. If you believe that, then the taxpayers should make a profit with "conservatorship." If it was not a co-ordination failure, and investors were appropriately reacting to the risk that Freddie and Fannie were insolvent, then the government will not make a profit out of conservatorship. My impression is that the latter is the case.

Everyone knows there were bad bets in the housing and mortgage markets. But this question of whether on top of those bad bets there was a co-ordination failure is a really, really fundamental issue. If there was no co-ordination failure, then all the talk about "shadow banking" and the need for a "systemic risk regulator" is beside the point.

Comments and Sharing

COMMENTS (5 to date)
Mattyoung writes:

[Comment removed pending confirmation of email address. Email the to request restoring your comment privileges. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Vasile writes:

Well, another option, of course, is that the coordination failure happened at an earlier moment.

A hypothesis quite compatible with the one of bad debts.

Arun Eamani writes:

This could be semantics, but what about other Systematic issues like every body using the same Model(with different calibrations may be) or market irrationally competing for traders ratcheting their salaries leading them to take asymmetric risk. These couldn't be solved by market because of limited information and time of the shareholder/bondholder if not the herd mentatility

Bill Woolsey writes:

What Kling calls coordination failure suggests that what Citibank and the investment banks wasn't really foolish. It was, rather, that those funding them began to behave foolishly.

The "bad bet" approach suggests that the real estate "bubble" was being supported because Cittbank and the other politically connected invetsment banks were willing to put their capital (such as it was) on the line, funding portfolios of mortgage backed securities with short term lending.

The first problem suggests some kind of govenrment intervention to get them back into business. How can the govenrment convince the panicky investors to lend short term to investment banks so that they can again hold portfolios of mortgage backed securities?

The second scenario suggests that those who made these bad decisions need to take losses--probably bankruptcy.

Alan writes:

It is important to distinguish between co-ordination failure and bad-bets, but the exercise is riddled with identification problems.

One could argue that bad-bets were the problem with the "leveraged loan" markets (subprime, Alt-A, private-equity loans etc), but the spread of these problems to other markets like prime mortgages, and highly-rated corporate and municipal debt were co-ordination failures, particularly in securitisation markets where the 'originate and distribute' model exacerbates this co-ordination failure.

Also, co-ordination failures can quickly turn into "bad-bet" problems. e.g. the liquidity squeezes in the municipal bond market and money market funds in middle of 2007 quickly became a "bad-bet" for investors that faced losses after these funds were forced into fire-sales to repay investors.

The interaction between these factors is what I meant by 'identification problem', and it means that we're unlikely to have much success in trying to disentangle these two effects.

Comments for this entry have been closed
Return to top