Sunday, October 19, 2014

Bubble mess and monetary policy

Am I the only one who is confused? We have debated for about a decade if monetary policy should take asset prices into account - going back to Greenspan's ""irrational exuberance"" speech at least. And very clever people, like Bernanke and Gertler, wrote clever papers arguing it's a bad idea. I looked at what happened when a central bank intervenes to bring down the stock market - as the German central bank did in 1927 (published here). My conclusion was that it was an unnecessary disaster - there was no bubble, and the intervention that sent the German stock market down by a quarter by year-end did real damage to the economy. Capital-raising by undercapitalized firms came to an end. Investment turned down the year after, and weakly-capitalized firms went under in the Great Depression much more quickly. My conclusion - targeting asset prices can be seriously bad for your economic health. Only ex post do we know if something is a bubble. Bernanke, when he became governor of the Boston Fed, mentioned my research in his first speech, concluding that interventions are a bad idea.

Well, a few years on, I am no longer quite sure. Let's take the different points of the ""no intervention"" idea in turn: a) you never know if it's froth b) even if it is, intervening may not work c) if it does, the collateral damage may be awful. True, if I say bubble, you can say ""bad model"" - every claim that prices are too high comes from some model, and the models can be pretty poor. But that doesn't mean that we have no idea about what periods of overpricing look like. If you see P/E multiples north of 30 (or infinity - no earnings, as on NASDAQ), or rental yields around 2% (Spain today), the prior should be that these are abberrations. I am also no longer so convinced that nothing can be done.

It's probably a bad idea to use interest rate policy, but there are other means. Instead of using interest rates, we can think about time-varying regulatory minima. Capital requirements could be made to fluctuate counter-cyclically, as per the Goodhart proposal. In housing markets, for example, we could use a simple rule that says ""raise the down payment requirement every time rental yields dip below 5%"", or something to that effect. While we have to live with a one-size-fits-all interest rate policy in Europe, mortgage markets are still national (try to get a German bank to finance your house in Spain - fat chance). We should encourage active guidance for national mortgage lenders, to take the different cycles into account. Finally, I think that stabilizing the prices of assets such as housing is beneficial overall -- there is much less risk of overshooting than in shares, and the welfare costs of pricing an entire generation out of adequate housing are phenomenally large. Next time you think ""disaster"" when the IMF tells you that Spanish house prices are set to fall by 20% or thereabouts, think of all the happiness it will buy for families that are living in cramped conditions. Combine that with very ugly redistributive implications of house price bubbles, and the pain of thousands of households defaulting on their mortgages and losing their homes, and I think we should consider a experimenting with better intervention just a little bit. Now, if only I could think of a good equivalent to raising the down payment in housing for equity markets...

No comments:

Post a Comment