By Krishna Guha in Washington
Published: Financial Time May 16 2008
In the aftermath of the dotcom crash in 2002, Alan Greenspan famously argued that central banks had little power to stop bubbles inflating and then bursting. All policymakers could do, said the then Federal Reserve chairman, was to "focus on policies to mitigate the fallout when it occurs".
His most vocal supporter was Ben Bernanke, then a Fed governor. As an academic, Mr Bernanke had championed the view that central banks should ignore asset prices except insofar as they affect forecasts for inflation and growth. "Even putting aside the great difficulty of identifying bubbles in asset prices, monetary policy cannot be directed finely enough to guide asset prices without risking severe collateral damage," he said in 2002.
Mr Bernanke added a caveat: central banks could use regulation to reduce the incidence of bubbles and minimise the threat they pose. But this argument was never embraced by Mr Greenspan, who had little faith in the ability of regulators to prevent crises.
Six years and another bubble - this time in housing - later, Mr Bernanke is Fed chairman and the US central bank is taking another look at its hands-off approach to asset prices. Mr Bernanke is open to the possibility that it may have to make fundamental changes to its strategy.
Some economists - and some European central bankers - think the answer is to do what Mr Bernanke advised against in 2002: run monetary policy a little tighter when asset prices are rising without obvious justification. This tactic is known as "leaning against the wind".
The Fed intends to re-evaluate the evidence on this and Mr Bernanke is open to persuasion. But he remains sceptical about using interest rates in this way. He and his top colleagues at the Fed think there may be another option - one anticipated by the caveat to his 2002 speech - of "leaning against the wind" through regulation rather than interest rates.
What everyone agrees on is the desirability of finding a way to counteract bubbles more effectively. The US has seen two giant and disruptive bubbles in recent years. First was the dotcom bubble, which caused a recession and a deflation scare. Then the housing bubble took the economy to the brink of recession and threatened the stability of the global financial system.
"Given the events of the last eight months it would be foolish not to reconsider the Greenspan doctrine," says Kenneth Rogoff, a Harvard professor and former chief economist at the International Monetary Fund. "Maybe after careful consideration you might conclude that it is not possible to do anything better. But everybody is thinking about it."
Alan Blinder, a Princeton professor and former Fed vice-chairman under Mr Greenspan, says: "I think I am still with the orthodoxy but I have to admit that recent events are sowing some seeds of doubt."
The Fed's traditional laisser faire approach to bubble formation is under attack from two related but distinct critiques. The first says ignoring bubbles as they inflate but cleaning up the mess afterwards results in a bias in monetary policy that, over time, will cause rising inflation (see below).
The second holds that central banks should, in any event, tackle bubbles more aggressively than the Fed has done in the past. Besides creating misleading price signals that result in productive resources being diverted to unproductive ends, bubbles can also cause tremendous macro-economic volatility. When a bubble bursts, sharp falls in asset prices threaten financial stability and a central bank's growth and inflation objectives.
All central banks would like to find ways to avoid these threats. But to do so requires overcoming two basic objections set out by Mr Greenspan and Mr Bernanke in 2002: first, that bubbles are in practice impossible to identify until they pop; and second, that even if central banks could identify bubbles, they need to find a tool with which to address the problem.
"It is difficult to judge whether you are seeing a bubble or not. Usually one does not know until afterwards," says Lars Svensson, deputy governor of Sweden's Riksbank.
Some experts saw bubbles in technology as early as 1996 and in housing as early as 2003 - even though most analysts now believe that prices were not then unrealistic. Efforts to suppress bubbles that do not really exist could do enormous damage.
Although central banks do not have a clear informational advantage over the market in judging when an asset is overvalued, this argument looks less compelling given the experience of the past decade.
Moreover, some economists believe progress has been made in diagnosing bubbles. "When you have a rise in indebtedness or capital inflows accompanying a run-up in asset prices, it is more likely to be something to worry about," says Prof Rogoff. At a minimum, rising indebtedness suggests the economy would be more exposed to a sudden fall in asset prices if it turned out that there was a bubble.
In a similar vein, the credit crisis has strengthened the European Central Bank's conviction that it is right to put weight on the so-called "monetary pillar" - the money supply and credit data out of favour at other central banks - for signs that lending is expanding too rapidly. Even Mr Bernanke agrees, moreover, that a central bank does not need to identify a bubble with 100 per cent confidence in order to justify taking steps against it.
This leaves the second Greenspan/Bernanke objection of 2002 - that monetary policy is not the right tool for the job.
Not all experts agree with this. In a recent speech, Jürgen Stark, a member of the governing council of the ECB, said: "Leaning against asset price movements may be advisable in some circumstances."
He set out four conditions: policymakers would need to be confident that the bubble is a bubble, that it would not burst of its own accord soon, that it would respond to moderate rate increases, and that it would cause significant damage if left alone. Mr Stark admitted that these conditions would be met only "very rarely".
Prof Rogoff says that even if tighter policy does not ease the bubble, it would reduce leverage in the economy and make it less vulnerable if and when the bubble bursts. To the extent that the Fed's current approach has an inflation bias, the case for tightening rates when a bubble is inflating would be much stronger.
Fed officials say that for all the talk about leaning against the wind in Europe, there are few examples of it being actually practised. Sweden's Riksbank, however, cited rising house prices as a reason for bringing forward a rate increase in 2006 that would otherwise have taken place later.
Mr Svensson says "leaning against the wind" can be consistent with a flexible approach to inflation targeting that seeks to achieve an inflation objective over the medium term. "If a bubble increases the probability of a bad future outcome, it could be reasonable to raise interest rates and accept lower real activity and inflation in the short run in order to reduce the risk of a large bubble bursting, and causing too low inflation and real activity on a longer horizon," he says.
But he adds: "I think it would be a rare situation where we would have so much information that we would dare to do that."
Charles Bean, chief economist at the Bank of England, made a similar argument in a speech in Australia in 2003. He said that a central bank should "bear in mind those longer-run consequences of asset price bubbles and financial imbalances in the setting of current interest rates".
However, this approach only works if the bubble is susceptible to moderate increases in interest rates. Both Mr Greenspan and Mr Bernanke have argued that bubble dynamics are too powerful to be arrested by anything other than very large increases in interest rates that would devastate the broader economy. This is why Mr Bernanke concluded in 2002 that interest rates were not the right tool for the job. "One might as well try to perform brain surgery with a sledgehammer," he said.
Mr Bernanke and his colleagues today are attracted by the idea of dealing with bubbles in a more surgical way: using regulation selectively and aggressively to target specific excesses.
Regulation offers a separate tool to deal with a new objective: ensuring that asset prices are not widely out of line with fundamentals. Top Fed officials privately admit that if it and other regulators had been more aggressive during the upswing in house prices, the housing bubble would not have swelled to its current size and the financial system would have been better insulated from it.
Mr Bernanke believes that most bubbles originate in failures of microeconomic regulation - either bad regulation, or non-regulation of markets stricken by information and incentive problems. In this respect he fundamentally differs from Mr Greenspan, who sees bubbles as being deeply rooted in human psychology: greed and fear.
Many officials in Mr Bernanke's Fed also see potential in using regulatory policy dynamically to deal with bubbles as they emerge. Some policymakers favour putting the regulatory brakes on autopilot to some extent by adopting counter-cyclical capital requirements. This could help prevent bubble formation by forcing banks to hold more capital relative to assets in good times than in bad times.
Moreover, senior Fed officials are intrigued by the additional possibilities raised by Treasury proposals to give them new "macro-prudential" powers as part of a proposed overhaul of the US financial system. At present the Fed has largely "micro-prudential powers" - the ability to regulate large banks according to specific regulations.
The proposed macro-prudential authority would allow the Fed to order any financial institution to alter behaviour it believed jeopardised overall financial and economic stability. For example, it could require banks to hold fewer mortgage-backed securities or set aside more capital against them.
In principle at least, "macro-prudential" authority would give the Fed a powerful new tool with which to fight bubbles. "I would not rule that out," says Prof Blinder at Princeton. But he adds that the Fed would have to use any "macro-prudential" authority "very judiciously, because you could make a lot of mistakes".
There are other potential drawbacks. If the Fed were to tighten regulations to try to contain a bubble, financial institutions could simply move offshore. Moreover, while public opinion largely accepts the idea that interest rates should be set by experts and not politicians, central bank autonomy cannot be taken for granted.
Aggressive regulation attempting to curb a perceived bubble would risk a political backlash from those with interests at stake in continued asset price growth - such as tech entrepreneurs and investment banks in the late 1990s and home builders and home owners in the 2000s.
Using "macro-prudential" auth-ority to deal with bubbles remains, most economists believe, an idea well worth pursing. But it is not a silver bullet with which to dispose of bubble trouble.
Additional reporting by Ralph Atkins in Frankfurt
A sting in the tail on the way up
The Federal Reserve's current strategy of ignoring bubbles as they inflate but cleaning up the mess afterwards is under fire from economists who argue that it results in a bias in monetary policy that, over time, will result in rising inflation.
In technical terms, critics claim the Fed policy is "asymmetric" over the economic cycle. The Fed insists this is not true. Officials say the central bank responds in an even-handed way in both directions to take into account the effect that higher or lower asset prices have on the economy.
In practice, the US central bank tends to raise interest rates gradually when asset prices are rising but cut rates aggressively when asset prices are falling. This gives at least the appearance of an unbalanced or "asymmetric" approach.
However, Fed officials say it is the asset prices that behave asymmetrically, not monetary policy. The prices of all assets - whether technology stocks in the 1990s or house prices in the 2000s - tend to rise gradually for a long period and then fall sharply.
The Fed reaction to a given change in asset prices is the same whether the change is a price rise or price fall, they say. It moves rates more quickly on the way down than the way up only because asset prices fall more quickly than they rise.
This argument would be watertight if the Fed stuck to a rigid policy rule - moving interest rates mechanically in response to its base case forecast for inflation and growth.
But the Fed does not do this. Instead, it operates a risk-management approach to policy that puts additional weight on avoiding so-called "tail risks" - extreme events such as a deep recession or runaway inflation.
Critics argue that, in practice, it is this that makes policy asymmetric and creates an inflation bias.
John Taylor, a Stanford professor and former undersecretary at the US Treasury, goes as far as to argue that excessive risk management - deviation from the interest rate that would be expected based on growth and inflation - fostered both the tech bubble and the housing bubble.
The Fed, he says, took interest rates too low for too long following the Russian crisis in 1998, allowing the dotcom bubble to build, then repeated the mistake following the dotcom bust in 2000, which led to overheating in the housing market. "Rather than bursting bubbles they should make sure that monetary policy does not cause the bubble in the first place," he says.
Many economists dispute the view that Fed rate policy was responsible for causing either of the two bubbles. Fed officials, meanwhile, say they pay equal attention to the tail risks to growth and to inflation. The Fed took rates higher than the market expected in 2006, for example, because it was concerned that inflation could gain momentum.
But some fellow policymakers are unconvinced. A top European central banker says the emphasis on tail risks produces a bias towards easy monetary policy, since the tail risks are smaller on the way up - when asset prices change gradually - than they are on the way down.
Some argue that if the Fed wants to continue to have the freedom to respond aggressively to asset-price busts then it should run policy tighter than it currently does when asset prices are rising.
In effect, the Fed would deliberately undershoot its inflation objective a little each year during the upswing in the asset cycle, thereby banking enough inflation credit to justify cutting rates aggressively and risking an inflation overshoot when cleaning up after the bust. This would imply some willingness to "lean against the wind" with interest rates when asset prices are rising.
Alternatively, the Fed could apply tougher regulations when asset prices are rising, offsetting any asymmetry in monetary policy and eliminating any inflationary bias.
However, most officials at the Fed say the evidence does not bear out the claim that its policy has an inflationary bias. "I do not think that is empirically correct," says Alan Blinder, a Princeton professor. If it were, he adds, we would see core inflation rising over time.
Some economists, though, say it is too early to be sure, given the many other forces that influence inflation. "Ten or 15 years is not a long time in terms of understanding inflation," says Kenneth Rogoff, a Harvard professor. "Policy clearly is asymmetric. The inflationary bias may be small, but it is hard to argue that there is no bias."
Copyright The Financial Times Limited 2008
