Monday, November 24, 2008

Double jeopardy for financial policymakers

(Wolfgang Münchau @ FT.com) In the current turmoil, accidents can pull us in vastly different directions. The unforeseeable bankruptcy of Lehman Brothers, the US investment bank, transformed a lingering financial crisis into a near-systemic meltdown. Depending on how other unpredictable events turn out in the next few weeks and months, we could end up with a deflationary depression, an inflationary boom or even one followed by the other.

In such an environment, economic forecasts are useless – worse than useless, in fact, because they give us a sense of certainty where there is none.

One path-changing accident would be the bankruptcy of one of the large US carmakers. The probability of such an event has clearly risen in the past week. Naturally, this would be bad for the US car industry itself. But it might be even worse for the banks, especially those that got involved with credit default swaps – probably the most dangerous financial products ever invented. CDSs are unregulated shadow insurance products that investors buy to protect themselves against default of corporate and sovereign bonds. Protection against a default by General Motors was among the most sought-after contracts.

The housing market is another potential time-bomb. Until recently, most of the housing experts were content to predict a 25-30 per cent fall in US prices – peak to trough. Such a fall would bring prices back in line with the long-term trend. But this was before an expected mild downturn turned into a big recession, and credit froze up.

Under such conditions, one would expect house prices to overshoot, say at least 10 or 20 percentage points beyond the trend line. So we may be talking about a peak-to-trough fall of 40-50 per cent on average in nominal terms – and more in real terms. There is no reason to see why the downturn should be any different in the UK, Ireland and Spain.

The path might take us elsewhere, however. Suppose the Detroit Three get their bridging loans. Suppose further that the CDS market does not collapse, and governments find effective measures to prevent an extreme overshooting of house prices. Then we might find ourselves in a completely different world. The recession in the western economies might end in the middle of next year. With interest rates close to zero, borrowing would pick up fast. Oil and commodity prices would rise as fast as they came down. I would expect the central banks to be too slow in raising their interest rates for fear of killing off the incipient recovery.

The result would be a sudden rise in inflation, perhaps the mother of all bond market crashes and, quite possibly, a dollar crisis.

So there are risks both ways – asymmetric perhaps, but surely significant, both in terms of their impact and their probability.

Statisticians distinguish two types of errors: type one and type two. Suppose we believe that another Great Depression is about to happen. A type-one error would be to reject our depression scenario when it is true, while a type-two error would be to accept it when it is false.

The US Federal Reserve’s policy is about avoiding a type-one error – underestimating the threat of a depression – at all costs. I was quite surprised last week – though perhaps should not have been – when I learnt that the Fed had quietly adopted a policy of “quantitative easing”.

The Fed conducts open-market operations normally with the goal of keeping the actual Fed funds rate close to the target rate set by the Fed’s open-market committee. The Fed funds rate is the rate at which banks lend their balances to each other overnight. But, more recently, the actual Fed funds rate has fallen much below the target rate, which is 1 per cent. Under a strategy of quantitative easing, the Fed does not care about the rate. The goal is to increase the money supply, by swamping the Fed funds market with liquidity. The calculation is that this would give banks an incentive to buy higher yielding securities, which would reduce long-term interest rates, over which the central bank has no direct influence.

For a central bank, this is comparable to the deployment of the nuclear option – your last or last-but-one policy option. Ben Bernanke, the chairman of the Fed, once co-wrote a paper on the subject of what a central bank can do when interest rates hit the “zero bound”* – a zero rate. The answer is that there are a few options, quantitative easing among them. It is interesting, though, that he has already deployed his weapon of mass desperation while still some distance away from the zero bound.

The US policy establishment regards this crisis principally as carrying a “one-tailed”, or one-sided, risk of a deflationary depression, to be avoided at all costs. But there are also grave risks associated with making a type-two error. A subsequent rise in US inflation could trigger a mass flight out of dollar assets and a large rise in US market interest rates, followed by a huge recession. The main difference is that the policy options would be a lot more constrained under such a scenario. In fact, a type-two error could also give rise to a depression – only later. I still think it is best to treat the crisis as an event with a “two-tailed” risk.

*Monetary Policy Alternatives at the Zero Bound: An Empirical Assessment, B. Bernanke, V. Reinhart, B. Sack, Federal Reserve Board, Finance and Economics Discussion Series, 2004-48

No comments: