Four hundred billion dollars; $900 billion; $1.4 trillion; $2.2 trillion. The International Monetary Fund has estimated the likely deterioration of assets due the sub-prime crisis in the US on four occasions, starting September 2007 and the latest one last week; its estimates have multiplied more than five-fold over this period and, its Global Financial Stability Report suggests, it's still counting.

Given this, it should be no surprise that, despite the $350 billion that the US and UK authorities have already put into their banks, and despite the promise to put in much more, the bad news from US and other industrialised countries continues to come in. Citi has just declared quarterly losses of $8.3billion, Bank of America $1.8 billion and Deutsche Bank $6.4 billion. RBS is looking at between pound 7 and 8 billion, and Wells Fargo $2.6 billion.

Judging from how the estimates of write-downs are ballooning, governments are just playing catch-up, and not very well at that. Indeed, according to the IMF, these banks will require at least half a trillion dollars in capital (beyond what they can hope to earn) over this and the next year.

One problem is that, not only is the capital being injected proving woefully inadequate, a lot of it is simply going back to the US Fed, the European Central Bank or the Bank of England by way of deposits. That is, banks continue to fear the worst and prefer to earn low interest on their capital instead of lending it to businesses (seeking an assured return of capital rather than a return on capital).

Much the same thing happened when the US gave a tax payback to citizens, in the hope that they would go out and spend the money. Instead, most of them used the money to repay debt or build their bank balances. Clearly, the lack of confidence in the near-term future is affecting the behaviour of both consumers and bankers.

How long will things continue in this manner, and what will bring about a turnaround? The IMF's update of its World Economic Outlook suggests that an uptick in the global economy may take place next year (its 2010 global output growth forecast is 3 per cent, compared to 0.5 per cent for 2009), but this has to be treated with some scepticism.

The IMF itself talks of downside risks; in any case, till as recently as November, the IMF's 2009 growth forecast was 2.2 per cent growth! In any downturn, the turnaround comes only when investors feel asset prices have bottomed out; when consumer demand has fallen so much that it has nowhere to go but up; and when bankers feel that businesses (or those that remain) are on even keel.

In other words, there has to be at least a quarter or more without further bad news before positive sentiments set in and cause a recovery.

Governments in the G-20 are trying to help by adding to demand, and their fiscal deficits are projected to rise almost 3.25 percentage points (to 7 per cent of GDP) in 2009. Whether this will be enough depends on various things. If the recession is in line with classical Keynesianism, the impact will be straightforward, and easy; if the recession is Hayekian and requires structural changes (undervalued Chinese exchange rates and excessive US twin deficits will take a longer time to fix), the pump-priming won't help.