Still, empirical studies generally turn up positive results from central-bank asset purchases. They appear to move interest rates, for example. The BoJ’s foray into QE in 2001 quickly cut short-term rates to zero and is generally thought to have had a small but meaningful downward impact on medium- and long-run interest rates. Early reviews of crisis-era asset purchases are likewise modestly positive. A recent survey of QE research findings by the Federal Reserve Bank of San Francisco indicated that $600 billion of asset purchases could be expected to reduce long-term rates by 15-20 basis points, equivalent to a 75-basis-point cut in the federal-funds rate.
Causation is harder to discern for equity prices. Some of the expected impact may be priced in before QE is announced, as happened when Ben Bernanke, the Fed chairman, hinted at QE2 in the summer of 2010. Yet QE programmes in Japan, Britain, and America appear to have been associated with rising equity prices.
The trillion-dollar question is whether QE boosts the broader economy (see chart on previous page). Early assessments answer with a cautious “yes”. Research published by the Federal Reserve Bank of San Francisco indicates that the Fed’s asset purchases have probably taken 1.5 percentage points off America’s unemployment rate. Real output by late 2012 may be 3% higher than it would have been in the absence of QE1 and QE2, and payroll employment could be as much as 3m workers higher than it would otherwise have been. Research by some BoE economists on the impact of its first £200 billion in QE purchases suggests that it may have raised Britain’s real GDP by as much as 2% and inflation by 1.5%, an impact equivalent to a three-percentage-point cut in the main interest rate. Although a different BoE study found a more modest impact, the data so far suggest that QE helps the real economy.
A sea of troubles
Yet in the minds of many critics, even such gains do not justify the risks, great and small, of large-scale asset purchases. Three dangers stand out. The first threat is to the function of some financial markets. The Bank for International Settlements (BIS) argued in its recent annual report that huge growth in bank reserves was driving overnight-lending rates to zero, causing the market for unsecured overnight lending to atrophy. Since the unsecured overnight rate has been the principal policy lever for central banks, this development could, the BIS warns, make it hard for them to rein in inflation in the future.
The risk can be mitigated, however. In late 2008 the Fed began paying interest on the reserves held by banks over and above the minimum required. Raising the rate of interest paid on excess reserves can make new bank loans less attractive, thus tempering overall credit creation. This tool represents a means to check inflation despite the breakdown in unsecured markets.
A second risk from QE is of distortions in the market for government debt. The borrowing costs of some governments are extraordinarily low—an auction of ten-year Treasuries on July 11th produced record-low yields. A flight to safety is a contributing factor, but it seems that markets either anticipate decades of abysmal economic growth, or the risk premium for holding long-dated bonds is unsustainably low, thanks in part to central-bank purchases. Any adjustment may be sudden and have unpredictable consequences.
A related concern is that QE is reducing market pressure on sovereigns that would otherwise face higher interest rates and a corresponding need to deal responsibly with their public finances. This is not a concern to take lightly. A central bank can lose control over inflation if the market has lost confidence in the sovereign and the bank is forced into buying government debt. On the other hand, a central bank that neglected its duties to play fiscal watchdog could risk its independence.
Clearly, there are risks to unconventional policy (abroad as well as at home, where emerging markets have long complained that expansionary rich-world policy has caused waves of capital inflows). But the critical judgment is whether uncertain risks of uncertain magnitude outweigh the benefits of doing more. Cumulative output gaps across the rich world now run into the trillions of dollars. Tens of millions of people are unemployed, and many have been so for several years. The bar to not doing QE should be high.
The gains from asset purchases would seem to be clearest and largest in the euro area. Although the ECB is prohibited by law from providing direct fiscal aid to member governments, it may buy debt in the market to ensure the proper conduct of monetary policy. It has on occasion bought the debt of troubled peripheral sovereigns.
An ECB programme of QE would probably aim to spread purchases across all member governments, but could nonetheless benefit troubled governments. If QE successfully reduced government-borrowing costs, the pressure for ever-stricter austerity measures would ease marginally. Portfolio rebalancing could bring down private borrowing costs. Purchases of foreign assets might weaken the euro, helping exports. And the possibility of a swifter end to the recession could encourage euro-zone countries to push ahead with structural and institutional reforms.
For economies that have already used QE, the problem is one of diminishing returns. The Fed’s first round of asset purchases between late 2008 and 2010 reduced corporate-borrowing rates by nearly a percentage point; its QE2 programme of $600 billion in Treasury purchases, rolled out in late 2010, succeeded in bringing down corporate rates by 13 basis points. In Britain, banks still face high borrowing costs and remain nervous about lending more to small and medium-sized enterprises: a new credit-easing programme about to be introduced by the BoE and the Treasury is long overdue.
For additional QE to prove effective in both Britain and America, central banks must change their approach to inflation. Temporary, higher-than-normal inflation can facilitate wage and price adjustments and help erode the real value of household debts. Most importantly, when nominal interest rates can go no lower, a higher inflation rate corresponds directly to a lower, and more stimulating, real interest rate.
The pale cast of thought
Both the BoE and the Fed target an inflation rate of 2%. Even a modestly effective new round of QE should quickly lift inflation expectations back to target. But if markets think above-target inflation will prompt a reversal of the policy, then new QE will have very little impact.
The Japanese experience of QE illustrates the problem. In 2006, with inflation barely above zero, the BoJ brought QE to an end, rapidly reducing the monetary base and then raising its benchmark interest rate. Real interest rates never wandered far into negative territory. Although the BoE in particular has been prepared to tolerate inflation above its formal target, neither it nor the Fed has moved away from a goal of 2% inflation.
One answer may be a temporarily higher inflation target. Charles Evans, the president of the Federal Reserve Bank of Chicago, has proposed a plan in which the Fed would announce its intention to accommodate inflation of up to 3% a year as long as the unemployment rate remained above 7%. The plan would allow the Fed to gin up a bit more inflation now without committing it to a permanent rise in the target rate, something the inflation-averse central bank is loth to consider.
Relaxing inflation targets is hard for central bankers with intellectual roots in the stagflationary 1970s. In Mr Bernanke’s view, central bankers’ victory over the runaway inflation of that decade is a momentous achievement. But that stability is now being purchased at a very dear price.