Normality comforts central bankers. Being forced to provide large quantities of liquidity to the commercial banks was abnormal and uncomfortable. But in abnormal situations, unconventional policies are the right ones. In ending the policy, the bank is taking a risk with something that matters far more than any discomfort, namely, the stability of the Japanese economy.

The BoJ introduced the policy of quantitative easing to augment the effectiveness of the zero interest rate policy, followed – with a short and disastrous break in 2000 – since 1999. The aim was to increase the liquidity of the commercial banks’ balance sheets and so increase the money supply and, indirectly, the commercial banks’ willingness to lend.

Today, the quantity of such reserves stands at Y35,000bn (£170bn). This is vastly above the Y6,000bn believed necessary to keep short-term interest rates at zero.

With real economic growth of 5.5 per cent, on an annualised basis, in the last quarter of 2005 and the year-on-year rise in the consumer price index (less food) at 0.5 per cent in January, the bank believed it was time to end those measures. It remains committed to the zero interest rate policy and to outright purchases of long-term government bonds. Only by Japan’s standards could yesterday’s change be regarded as a significant tightening.

Yet such exceptional measures are justified by exceptional conditions. Two points need to be stressed: the first is that deflation has barely ended; the second is that monetary growth remains feeble.

Note that the consumer price index (less food) in January was still 2.7 per cent below where it had been in December 1997 and was also no higher than in August 2002. Yet price levels as well as changes in price levels matter, since a long period of falling prices has increased the real burden of debt.

Moreover, over the last two years the average rate of growth of broad money has been under 2 per cent. This hardly suggests monetary conditions have been too easy. The withdrawal of excess liquidity may shrink the money supply or at least slow its growth. That hardly seems wise after so long a period of deflation.

Equally important, the BoJ has not accompanied the ending of quantitative easing with a decisive move towards inflation targeting. If Japan is to be a normal country with what is now considered to be a normal central bank, it should attempt to target a rate of inflation of 2 per cent, at least. Instead, we have merely an indication that it is looking at a range of 0-2 per cent.

One benefit from higher inflation is that it would help erode the overhang of debt still present in the economy. Another reason for such a target is that 2 per cent is probably close to an actual inflation rate of zero, given the biases in measurement.

But the most important reason for an inflation target at this level is that it would represent a clear commitment by the bank to end the spectre of deflation. That would be an immense encouragement to anybody thinking of borrowing.

The Japanese experience of the past 15 years has had one clear and overriding lesson: in an exceptional situation, orthodoxy is far more dangerous than its opposite. The Bank of Japan has now taken a step towards the orthodox and is taking a risk in doing so. Let us hope that it will not, as a result, damage the fragile recovery now under way.