How Big is the Current Financial Crisis?

So how big is this financial crisis and how does it compare to other crises of the past? I have already alluded to the $1 trillion dollar estimate of Charles Morris. Elsewhere, Mr. Morris has said this estimate is a conservative one since it assumes an orderly unwinding in financial markets. This conservative estimate is consistent with Nouriel Roubini who makes a low-end estimate of financial losses at 1$ trillion. Martin Wolf discusses several estimates that go high as $3 trillion. He notes the following:

Losses of $2,000bn-$3,000bn would decapitalise the financial system. The government would have to mount a rescue. The most plausible means of doing so would be via nationalisation of all losses. While the US government could afford to raise its debt by up to 20 per cent of GDP, in order to do this, that decision would have huge ramifications. We would have more than the biggest US financial crisis since the 1930s. It would be an epochal political event.

Such high-end estimates are alarming. Now the IMF is weighing in on the matter. It has released its Global Financial Stability Report which shows estimated financial losses coming in just under $1 trillion. It also provides an interesting graph comparing these losses with previous financial losses:

Note that the IMF figure for the current financial crisis is based on the problems with subprime mortgages. As Charles Morris explains, though, the "subprime [crisis] is just the first big boulder in an avalanche of asset writedowns that will rattle on through much of 2008...Expect the landslide to cascade through high-yield bonds, commercial mortgages, leveraged loans, credit cards and -- the big unknown -- credit-default swaps." Figuring in these latter developments leads to the higher estimates of $2-$3 trillion dollars.

Someone asked me this past weekend how bad this recession will be going foward. Taking the consensus view, I said it would be mild and over by the end of 2008. Thinking about these estimates of financial losses, especially the higher-end ones, is making me less certain of that claim. For the same reason, Nouriel Roubini says the consensus view of when the recession will end is in fact wrong. I hope he is wrong.

Update
In a Bloomberg article discussing the expected $1 trillion loss put out by the IMF, this sobering fact is stated: "The [IMF] forecast signals the worst of the credit crunch may be yet to come, because banks and securities firms so far have posted $232 billion in asset writedowns and credit losses."