End of the world scenario

Of what does a 20th century financial meltdown will look like? As far as Willem Buiter is concerned, this is the end of American capitalism as we know it. But not being a financial market participant I cannot quite imagine it. Let's put it this way: we've withdrawn all of our money market accounts. Here's the White House scenario:

More banks could fail, including some in your community. The stock market would drop even more, which would reduce the value of your retirement account. The value of your home could plummet. Foreclosures would rise dramatically. And if you own a business or a farm, you would find it harder and more expensive to get credit. More businesses would close their doors, and millions of Americans could lose their jobs. Even if you have good credit history, it would be more difficult for you to get the loans you need to buy a car or send your children to college. And ultimately, our country could experience a long and painful recession.

However, if I were a market participant I may be "too close to the problem" in the sense that I may think the entire universe revolves around financial markets. In any case, I want to play devil's advocate because it is more fun:
The devil in me claims that there is no credit crunch or if there is, it is minimally affecting Main Street. The credit crunch is only in the financial sector (interbank lending, overnight loans) and my assumption is that these are only to meet margin calls between financial institutions due to deleveraging and marking-to-market. Assuming that the government intervenes, will we actually observe a world where all the financial institutions fall like dominoes? Aren't financial institutions rational enough to realize that after one or two failures (okay, maybe five or six), they need to save themselves by renegotiating all of their positions?

Yes, there is a coordination problem because no one wants to be the first to say they are going to write down their losses and actually put their own valuation of their toxic securities because if they do realize the losses, the price of the security may increase due to perhaps less downward pressure as some of the possible supply is taken off the market. This is assuming that there is a downward spiral not only due to actual sales but in anticipation of future sales of these securities as institutions deleverage. Consequently, the volume of trade in OTC derivatives such as CDS and CDOs will shrink.

In my warped view then, the only participants will be hurt are those in the financial sector with very little spillover to the real sector except in the following cases:
1. Those employed in the financial sector will be the hardest hit as their employers go bankrupt. Does the payroll of the financial sector amount to $700 bn?
2. Retirement accounts and investments in the financial sector will see their returns fall.
3. A mild recession will follow as banks try to recapitalize.

What are the wild cards?
The intangibles like trust and confidence, i.e. Keynesian animal spirits. All the talk about calamity and end of the world will actually make people believe the end of the world is coming. Like our withdrawal from the money market funds. :) The foreign sector, i.e. all the SWFs that are holding U.S. assets. What would they do? With all the talk about systemic risk and the financial system "seizing up" there really has not been enough clarity about what it all really implies. The White House scenario doesn't lend itself to clarity, it's a scenario. What I mean by clarity is the following:
If we don't do the bailout then 80 percent of the banks will go bankrupt. 4 million employees in the financial sector will become jobless. Banks will stop lending to companies and entrepreneurs because .... (why? I actually can't think of any reason).

But continuing on,
Banks will stop lending and businesses will be unable to operate for lack of cash flow. An additional 50 percent of the economy in various sectors will close. Another 80 percent of the population will become unemployed. We will all fall to disease, pestilence and death, etc.

In the end, this bailout and a lot of others are justified on the assumption of an end-of-the-world scenario: a small probability of large negative losses. (Some have referred to this as the peso problem.) What we still lack is a good way of analyzing such situations in particular, estimating probabilities of catastrophe.

Another way to look at it is the following: Is it cost effective to use $1 trilliion to prevent a crisis that never occurs? (Assume that all of $1 trillion used to buy toxic securities/bailout is not recovered.) I think we can get a partial answer on the costs using analyses of lost output from previous crises by making the assumption that a crisis will occur. Assuming US GDP is about 14 trillion, 1 trillion of 14 trillion = (1 x10^12)/(14 x 10^12) = 1/14 => 7 percent of GDP. Assume that this is only a banking crisis that does not result in a currency crisis, and that the crisis lasts for 5 years and for each year of the crisis, output falls by 5 percent with no bailout. Now assume that the bailout cost is 10 percent per year (no one really believes that $1 trillion will be the final word on this) and manages to "cushion" the crisis so that output falls by only 4 percent per year and the crisis last for 4 years - is this still money worth spending? From this point of view then my opinion about the bailout can change. For instance, what is the breakeven probability in a series of changing assumptions to the above? Am I comfortable with the probabilities?

An old post on Catastrophe: Risk and Response uses cost-benefit analysis.