I have often disagreed with the views of University of Chicago Professor John Cochrane over the paste five years. However, his latest oped in the Wall Street Journal is spot on.
In the oped Cochrane questions the rational for the increasingly common view that central banks should pursue “macroprudential” policies to reduce the risks in the financial sector.
This is Cochrane:
Interest rates make the headlines, but the Federal Reserve’s most important role is going to be the gargantuan systemic financial regulator. The really big question is whether and how the Fed will pursue a “macroprudential” policy. This is the emerging notion that central banks should intensively monitor the whole financial system and actively intervene in a broad range of markets toward a wide range of goals including financial and economic stability.
For example, the Fed is urged to spot developing “bubbles,” “speculative excesses” and “overheated” markets, and then stop them—as Fed Governor Sarah Bloom Raskin explained in a speech last month, by “restraining financial institutions from excessively extending credit.” How? “Some of the significant regulatory tools for addressing asset bubbles—both those in widespread use and those on the frontier of regulatory thought—are capital regulation, liquidity regulation, regulation of margins and haircuts in securities funding transactions, and restrictions on credit underwriting.”
This is not traditional regulation—stable, predictable rules that financial institutions live by to reduce the chance and severity of financial crises. It is active, discretionary micromanagement of the whole financial system. A firm’s managers may follow all the rules but still be told how to conduct their business, whenever the Fed thinks the firm’s customers are contributing to booms or busts the Fed disapproves of.
I completely agree with Cochrane.
Macroprudential policies might very well develop into ad hoc and completely discretionary policies rather than a rule based monetary and financial policy regime. That is certainly not the direction we would like to see monetary policy move.
Macroprudential policy explicitly mixes the Fed’s macroeconomic and financial stability roles. Interest-rate policy will be used to manipulate a broad array of asset prices, and financial regulation will be used to stimulate or cool the economy.
I think this is an extremely important point to make. The purpose of monetary policy is to provide nominal stability – either in the form of an inflation target, a price level target or an NGDP target and fundamentally the central bank basically only have one instrument to hit that – increasing or decreasing the money base. It is the Tinbergen rule - one policy instrument, one policy target. It is Econ101.
If the central bank starts to take into account macroprudential “targets” then it would have to put aside other targets. A good example of this kind of “mixed message” in monetary policy has been the conduct of monetary policy in Sweden in the last couple of years where the Riksbank increasingly has focused on macroprudential indicators – for example property prices and household debt – in the conduct of monetary policy.
The result has been that Swedish monetary policy has been tighter than it otherwise would have been if the Riksbank had only focused on it’s official inflation target. The result has likely been that Swedish unemployment is higher than it would have been if the Riksbank consistently pursued its official inflation target and there is very little – if any – evidence that the Riksbank’s policy has increased financial stability in Sweden.
In fact in a recent paper former Riksbank deputy governor Lars E. O. Svensson shows “that a higher policy rate leads to a higher debt ratio, not a lower one. This result may be surprising to some, at least at the Riksbank, which has apparently made a sign error in its assumptions. The result is actually quite easy to understand once one carefully considers how debt, GDP and inflation are affected by a higher policy rate.”
I think Svensson’s paper quite clearly shows the dangers of having macroprudential policy dominating monetary policy making.
Central banks have a lousy track-record on detecting “bubbles”
Back to Cochrane:
It’s easy enough to point out that central banks don’t have a great track record of diagnosing what they later considered “bubbles” and “systemic” risks. The Fed didn’t act on the tech bubble of the 1990s or the real-estate bubble of the last decade. European bank regulators didn’t notice that sovereign debts might pose a problem. Also, during the housing boom, regulators pressured banks to lend in depressed areas and to less creditworthy customers. That didn’t pan out so well.
Yes, markets are often wrong, but there is very little (no) evidence that central bankers are better at diagnosing “bubbles” and “systemic” risks. Hence, it is not a question whether markets are good or bad at forecasting bubbles. The question is whether regulators better than market participants in spotting bubbles?
It is also correct that moral hazard problems might mean that private investors will ignore or downplay risks and that is an argument for certain regulations of the financial sector. However, my own personal experience is that regulators often are extremely reluctant to utilize their regulatory powers when it is “obvious” that there is a bubble of some kind.
In 2006 I co-authored a paper forecasting a major macroeconomic and financial crisis in Iceland and in a number of papers in 2007 I made similar warnings about the risks to the Baltic economies. Unfortunately the warnings turned out to be correct. The “crash” predictions were essentially based on macroprudential analysis and in that sense one can of course say that this shows the effectiveness of macroprudential analysis. Macroprudential analysis have long also indicated that something could go badly wrong in India – however, the Reserve Bank of India has done little if anything to act to avoid these risks neither has the Indian government.
Even though I thought it was quite obvious that both the Icelandic and the Baltic economies were heading for a major crash the local regulators in both Iceland and the Baltic States were extremely hostile towards these warnings and they completely failed to act before it was too late. In fact the regulators in more than one case acted as “cheerleaders of the boom” rather than defenders of financial and macroeconomic stability. This of course added to the feeling among investors that nothing could go wrong.
Hence, even if macroprudential analysis can in fact diagnose “bubbles” and systemic risks it is in no way given that that will lead regulators to take the right actions (never forget the Iron Law of Public Choice). The Icelandic crash is sad testimony to that. So is the Baltic crisis.
Cochrane’s policy advice
Cochrane has good policy advice:
First lesson: Humility. Fine-tuning a poorly understood system goes quickly awry. The science of “bubble” management is, so far, imaginary.
Consider the idea that low interest rates spark asset-price “bubbles.” Standard economics denies this connection; the level of interest rates and risk premiums are separate phenomena. Historically, risk premiums have been high in recessions, when interest rates have been low.
…Second lesson: Follow rules. Monetary policy works a lot better when it is transparent, predictable and keeps to well-established traditions and limitations, than if the Fed shoots from the hip following the passions of the day. The economy does not react mechanically to policy but feeds on expectations and moral hazards. The Fed sneezed that bond buying might not last forever and markets swooned. As it comes to examine every market and targets every single asset price, the Fed can induce wild instability as markets guess the next anti-bubble decree.
Third lesson: Limited power is the price of political independence. Once the Fed manipulates prices and credit flows throughout the financial system, it will be whipsawed by interest groups and their representatives.
Hear, hear…Cochrane is right – central banks need to return to rule based monetary policies. Macroprudential policies on the other hand might risk moving us away from rule based policies and towards a regime where central bankers become firefighters.
PS I am not arguing that macroprudential analysis cannot be useful. It can be a good tool for market participants and (non-central bank) regulators, but monetary policy should focus on ensuring nominal stability. Nothing else.
PPS If central bankers really need a good macroprudential indicators then the best indicator might be to look at the growth and level of nominal GDP.