Policy shift might yet save Japan

Although the Bank of Japan has rejected `monetisation', it now seems to have moved in this direction
MUCH OF the Group of Seven economic gathering over the weekend was spent discussing the Japanese economy, and in particular the issue of "monetisation" by the Bank of Japan (BoJ).

A policy shift in this direction has been urged on the Japanese authorities by Paul Krugman, professor of economics at MIT, for over a year, and in recent weeks the US Treasury seems to have warmed to the idea - admittedly as a last resort option - reflecting the fact that Japan seems to be lurching towards outright depression.

Despite the fact that the BoJ has repeatedly rejected the option of monetisation as too dangerous in the past few months, their policy board meeting a week ago seems to have moved substantially in this direction, and the Ministry of Finance - although interestingly not BoJ governor Hayami - has been openly talking as if this button has now been pressed. It is still dubious whether in fact it really has been pressed, but if it has, what does it mean?

Monetisation means exactly what it sounds like - the central bank would directly expand the rate of growth of the money supply, in effect "printing money". In a modern economy, this would probably not involve physically turning over the printing presses more rapidly, although it might. Instead, the central bank would simply buy bonds from the commercial banks, crediting the latter with an increase in their bank balances at the BoJ. The rate of growth in the "monetary base" - the total of cash plus bankers' deposits at the BoJ - would thereby be directly boosted.

Monetarist economists such as Milton Friedman have always contended that such a policy is appropriate when an economy, like Japan today, is headed into deflation. In such circumstances, cuts in nominal interest rates cannot keep pace with the decline in prices, so real interest rates rise unintentionally. Interestingly, non-monetarists such as Professor Krugman have recently taken up a similar battle cry, believing that this is the only way of increasing inflation expectations enough to get real interest rates down. Professor Krugman believes that real rates in Japan today are higher than the real rate of return on capital, so investment in plant and equipment will continue to shrink until real rates are cut.

As we have seen, the most likely way of implementing such a policy would be for the BoJ to step up its direct purchases of government bonds. But central banks often buy government bonds in the marketplace, so how would this be different from usual? The difference lies in the crucial fact that the BoJ, under a policy of monetisation, would not sterilise the effects of the bond purchases by subsequently selling short-dated bills back into the market.

In the past 12 months, the BoJ has been a heavy buyer of bonds, but has always sterilised these purchases by subsequently selling enough bills into the market to keep overnight interest rates in the region of 0.25- 0.5 per cent. The net effect of these operations has been that private sector banks have held fewer long term bonds than otherwise would have been the case, and have instead held more short term bills. This has reduced long term interest rates, and has increased the liquidity of the banking system. But the monetary base has not been affected.

Under "monetisation", the BoJ would still buy bonds, but would not sell bills back into the marketplace to drain the money markets. As a result, money would be left sloshing around the system, the overnight rate would fall to zero, and the growth of the monetary base would increase sharply. For example, if the BoJ "monetised" by buying bonds worth 1.5 per cent of GDP from the banks, this would boost the growth of the monetary base by 15 per cent. Something of this order might be under consideration by the BoJ for the coming year. But the key question is whether even this would have a dramatic effect on the economy.

Certainly, the simple act of cutting short rates from 0.25 per cent to zero would not make much difference to the growth of demand. So what other effects might there be?

First, the banking sector would be left holding money instead of very short dated bills as a result of the shift towards monetisation. This would further increase the liquidity of the banking system, but only by a slight amount. It is possible that banks might be induced to on-lend this money to the corporate sector, or perhaps to buy foreign assets which would devalue the yen, but it is unclear whether these effects would be very large. After all, does it really matter to a private Japanese bank whether it holds a seven-day bill drawn on the central bank (earning interest at 0.2 per cent), or an overnight cash deposit at the central bank (earning zero interest)? Probably not - in which case monetisation will have little effect via this route.

Second, the rate of growth of the monetary base will increase under a policy of monetisation. As we have seen, this may make very little difference to the lending behaviour of the banks, in which case the growth of the broader monetary aggregates - "M2 plus CDs" in the case of Japan - will not accelerate. The so-called money multiplier - the relationship between broad and narrow money - will therefore shrink, and there will be little change in bank lending which might boost the spending of households or corporations. However, the very fact that the monetary base is expanding more rapidly might trigger an increase in inflationary expectations, which would in itself tend to increase the growth of demand - especially capital expenditure. This "Krugman effect" would clearly be a useful step in the right direction, though in current deflationary circumstances it would surely not be very large.

Third, and probably most important, a shift toward monetisation would free the hands of the budgetary authorities to engage in a much bigger programme of fiscal expansion by cutting taxes and/or increasing government expenditure. The most recent dose of fiscal expansion, announced late last year, was heavily undermined by a sharp increase in bond yields, reflecting, in part, concerns about a surge in debt issued by the public sector. The Japanese government's internal national debt is now close to 100 per cent of GDP, and this total is increasing by 10 percentage points per annum - an explosive rate of growth which exceeds even that in Brazil or Russia.

There are strict limits to the scope for further fiscal action if this is to be financed in the normal way by the issuance of government bonds to the public. But if instead it is financed directly by central bank monetisation, there need be no increase in the bond issue as the government increases its budgetary injection, and therefore no problem with either the debt burden or rising long term bond yields. This means that a money- financed increase in the budget deficit becomes feasible, and is clearly the preferred way to restore growth in the economy.

Paradoxically, it therefore turns out that the main effects of a shift towards a monetarist solution are not monetary at all, but could still turn out to be crucial via their links to the Keynesian wing of policy - budgetary expansion. Provided the Japanese authorities recognise this, a move towards monetisation might be the start of better times.