Wolfgang Münchau: What central banks should do to deal with bubbles

The world has conducted two controlled experiments on how to fight financial bubbles in the past decade. Both failed.

The first was to ignore the bubble and to mop up later. The idea seemed plausible to a lot of people. But it was based on the false premise that the costs of mopping up would be bearable.

The second experiment has just concluded in Sweden, also with calamitous results. There, the central bank did the exact opposite. It had previously raised interest rates to rein in a domestic housing bubble. In doing so, it generated deflation and raised unemployment. It recently corrected that policy error by cutting the interest rate back to 0.25 per cent.

These two experiments present the opposite ends of our thinking: either ignore bubbles or ignore everything else. What should central banks do?

The consensus view is that they should rely on macroprudential regulation. The Hong Kong Monetary Authority, for example, imposed restrictions on loan-to-value ratios for mortgages. The Bank of England recently placed caps on mortgages with very high income multiples.

Central bankers love macroprudential tools because they are in thrall to an old idea that is simultaneously true and useless. The Tinbergen rule, named after a Dutch economist, states that you need one policy instrument for each policy target. If you have two targets – price stability and financial stability – you need two instruments. Monetary policy deals with prices, macroprudential regulation takes care of bubbles. Problem solved.

Or is it? For a start, the instruments are not entirely separate. Monetary policy affects not only retail prices but also the prices of financial assets. If a central bank commits to keeping interest rates at zero for the foreseeable future, it sets a benchmark for the price of risk-free securities directly and other securities indirectly.

There is also a more fundamental problem. Consider Spain’s housing bubble. The country has an above- average share of brilliant economists and bankers yet hardly any of them expressed concern about pre-2007 house prices. So what would macroprudential supervision have accomplished in those years? Even if our hypothetical macroprudential regulators had correctly identified the risks, they would still have focused on the banking sector. Yet the real tragedy of post-bubble Spain occurred in the household sector. The country’s conservative bankruptcy rules meant that many mortgage holders have been saddled with huge debts for the rest of their lives. Macroprudential regulation might have saved the banks but it would not have saved Spain.

Central bankers are fooling themselves if they think macroprudential regulation is a potent independent monetary policy tool. It is a useful supplementary tool, nothing more, nothing less.

Our best hope lies in a unified framework. Unfortunately, the workhorse models used in mainstream economics do not have a concept of finance. Default cannot happen in these models because their fundamental building block is the “representative agent”, jargon for “your average Joe”. But the average Joe cannot simultaneously default and be defaulted on. You need two Joes for that – none of them is average. Specifically, you need a financial sector in those models, one that includes what we have seen in the past decade – default, credit crunches, rent-seeking, extortion of governments, antisocial behaviour, unethical behaviour, criminal behaviour – to mention just a few.

The great James Tobin, another Nobel-prize winning economist, produced a model with an explicit role for asset markets as long ago as 1969. But rather than building on his work, the economic mainstream rode off in a different direction. The financial crisis gave rise to new approaches but they are still not mainstream. Central banks do not actually use them.

I have been intrigued by some pioneering work by Markus Brunnermeier and Yuliy Sannikov at Princeton, who constructed a model in which the world has two states: one in which banks lend freely and one in which they do not. The policy prescriptions of this model are not fundamentally different from what central banks have done recently. But with the possibility of a future credit crunch integrated into the model, interest rate policy cannot be blind to asset price developments. Such a model would suggest an earlier rise in interest rates in the UK, for example, compared with standard models. For the eurozone, the model would justify aggressive policy easing because a fall in the rate of inflation or outright deflation would harm financial balance sheets and add to instability.

There are several competing approaches. Modern monetarists focus on money, as opposed to credit, as the driving force. But despite their huge differences, both ideological and practical, none of them supports the experiment that just failed in Sweden or the one that failed 10 years ago. And none is particularly keen on macroprudential regulation.




Wolfgang Münchau


Fonte: FT