Undercover Economist

Boomtime for trying to fathom the bust

Economists from MIT and Yale attempt to untangle a crisis that has already lasted longer than the first world war

What is the best description of Gavrilo Princip’s notorious act, almost 98 years ago? “Moved his finger.” “Murdered the Archduke.” “Started the Great War.” “Ruined the 20th century.” None of them quite captures what happened, and the story of the war that followed can be – and has been – told in many different ways.

It is with something of this spirit that the economists Andrew Lo of MIT and Gary Gorton and Andrew Metrick of Yale attempt to untangle a financial crisis that has already lasted longer than that war. They do so in a pair of articles in this month’s Journal of Economic Literature.

Lo reviews 21 books about the crisis, 11 by academics, one by the former US Treasury Secretary Hank Paulson, and nine by journalists, including my colleague Gillian Tett. Gorton and Metrick take on 16 more technical pieces of research. Do they get any closer to the bottom of this complicated affair? A little, but the sheer variety of sources demonstrates the difficulty of the task – as does Lo’s decision to evoke Akira Kurosawa’s 1950 film Rashomon, an exploration of the shifting nature of truth. Unlike Kurosawa, though, all three academics believe that the truth can be uncovered with enough care.

One of the central questions concerns the nature of the initial credit crunch, in which financial institutions suddenly struggled to roll over short-term loans, or “repo” agreements that functioned very much like loans. Why?

One explanation is that repo markets suffered the wholesale version of an old-fashioned bank run; while banks had suffered real losses, the panic was out of all proportion to those losses. Instead, it was a self-fulfilling situation: quite rationally, nobody wants to be at the back of the queue when a bank run breaks out.

The alternative account is that the “originate-to-distribute” business model under which companies made loans, repackaged the loans and sold them on to others, was fundamentally flawed. It led to a catastrophic decline in lending standards, with both borrowers and lenders agreeing loans that were careless at best and fraudulent at worst.

Gorton – who developed risk models for American International Group, which was later bailed out by the US Federal Reserve – has been a proponent of the “bank run” view. With Metrick, he surveys research papers studying the panics. In the asset-backed commercial paper (ABCP) markets, it is possible to track many different “bank runs” – events when lenders refused to refinance maturing loans for particular bundles of assets. While these runs were often connected to fundamentals, in the initial credit crunch of August 2007, many runs broke out apparently at random. This, say Gorton and Metrick, was a panic phase of the crisis. It explains why trouble kept breaking out in parts of the financial system that were unrelated to the subprime mortgage sector that was the initial source of trouble. Lo is only partly persuaded by this – self-fulfilling panic had a role to play, he says, but that hardly puts the originate-to-distribute model in the clear.

Lo takes the time to debunk several pieces of received wisdom: that bankers were being paid to take excessive risks; that the efficient markets hypothesis led investors astray; and that in 2004 the SEC changed its rules and excessive leverage at banks was the result. All sensible claims, says Lo – but contradicted by the facts.

Not all is lost in the fog of war, though. As Gorton and Metrick point out, the institutional details of the crisis may be new and intricate, but its macroeconomic pattern – in particular, the housing bubble that preceded the crisis – is very familiar. The really striking thing about the crisis is not its familiar shape, but its gigantic scale.

Also published at ft.com.