Japan's failure to recover from the 1990 collapse of its financial bubble is one of the most puzzling and threatening aspects of the world economy. Fiscal stimulus, which pushed Japan's central government budget deficit to a whopping 11.4 per cent of gross domestic product, did little to get the economy off the ground. Repeated operations to remove bad debts from the banking system have proved sisyphean: no matter how much debt was cleared, the amounts of remaining bad debt have seemed only to rise. The looser monetary policy adopted by the Bank of Japan last month at least points in the right direction, though it will require much more boldness and careful manoeuvring to make a difference.

Japan is invariably blamed for its current plight but this is only part of the story. Global politics has also played an important role. As Japan's efficient manufacturing sector was making deep inroads into the world's automobile, electronics and semiconductor markets in the late 1980s, Europe and the US closed the doors to further export growth with explicit trade barriers and implicit political threats. One result was a significant deterioration of Japan's terms of trade in the 1990s.

Japan was told to grow through domestic demand, not through exports to world markets. This was deeply flawed advice, because Japan is a structural net savings surplus economy, in which its ageing population wants to save more than its companies can profitably invest within Japan. The result is a structural current account (and trade) surplus and a net supply of savings to the rest of the world. In the normal course of events, the bursting of the financial bubble should have weakened the yen and raised the net trade surplus but Japan was warned sternly by the US and Europe against exporting its way out of recession.

Amazingly, the yen strengthened sharply between 1990 and 1995 while the Japanese economy was grinding to a halt. The US government began to lecture Japan about the virtues of large budget deficits on the erroneous grounds that fiscal stimulus would produce a domestic-demand-led recovery. Instead, nervous households simply increased their own savings alongside the widening budget deficit.

The limited yen depreciation between 1995 and 1998 might have helped Japanese recovery but for the east Asian financial crisis (and some domestic policy mistakes). The weakening yen and the 1997 crisis were related. The exchange rate should have increased Japan's exports, mainly of capital goods, to the developing countries of Asia and the Asian countries should have financed that increased flow of Japanese exports through increased capital inflows from Japan. Increased capital flows would have been the financial counterpart of Japan's larger trade surplus.

The problem came when the capital flows from Japan to the rest of east Asia suddenly ceased in mid-1997, so that the other Asian countries could no longer finance their purchases of Japanese exports. Japanese capital flows had been highly concentrated in short-term bank loans, partly because such loans were accorded more favourable regulatory treatment than long-term bank loans under the Basle standards for bank capital adequacy. When bank regulators in Japan started to crack down on the banks' bad balance sheets, the Japanese lenders abruptly called in their loans from the rest of east Asia, helping to trigger the broader regional crisis in 1997. The yen strengthened again in the midst of the crisis.

Four years and one regional crisis later, Japan is seemingly returning to a weaker yen policy. Even if Tokyo is not targeting a weaker currency, a properly exapansionist monetary policy combined with econonic weakness will produce that effect. The currency has already moved from Y105 to the dollar at the start of January 2000 to Y124 to the dollar today. On macroeconomic grounds, it is likely to weaken much more, perhaps to Y140-Y150 in the next year or so, if Japanese monetary policy is as expansionary as it needs to be.

Interestingly, when Japan's recent shift towards easier money and a weaker yen was being debated, US lobbies such as the car industry publicly opined that this time a weaker yen would be acceptable to the US, as if confirming that the overly strong yen of the past decade was heavily influenced by US politics.

Asignificant monetary-expansion-cum-currency-depreciation in Japan would have two beneficial effects. First, to the extent that the Japanese domestic price level rises in proportion with yen depreciation, the main effect would be to lower the real value of bad debts in Japanese balance sheets but without changing Japan's international price competitiveness. Economic activity would pick up as the debt overhang was eased, as would Japan's demand for imports from the rest of the world.

Second, to the extent that yen depreciation exceeds the increase in domestic prices, the dollar price of Japanese goods in world markets would decline, shifting global demand towards Japan. The combined effect would raise Japanese aggregate demand, with a small net effect on aggregate demand in the rest of the world.

The rest of the world, especially in Asia, could absorb higher net exports from Japan if the increased excess of Japanese savings over investment were diverted into capital outflows in a smooth manner, so as not to replay the Asian disaster of 1997. Intuitively, Japan needs more exports to keep its economy fully employed, while the rest of Asia needs Japanese capital goods and technology to keep its own growth on track. But increased Japanese exports to Asia can work only if Japan can reliably provide the financing and only if the rest of Asia can reliably manage the increased capital inflows from Japan.

Once again, we are led back to the Asian financial system as the nub of the issue. Japanese monetary expansion makes good macroeconomic sense but it must be combined with greatly improved financial intermediation both in Japan and in the borrowing Asian countries for the monetary expansion to produce the desired effects. Part of the answer lies in channelling much more of Japan's capital flows in the form of stable foreign direct investments rather than volatile bank loans. The other part, however, lies in re-creating a banking system and other financial intermediaries that can once again lend for the long term, without frequent upheavals in the direction and magnitude of lending.

The writer is director of the centre for international development at Harvard university