Indeed, many of the instruments at the heart of the current credit storm barely existed before this decade - which means that computers can only model these markets based on the benign conditions of the past few years. Another big problem is that computer models do not always take account of the way that their own behaviour is affecting markets.
Another big problem is that computer models do not always take account of the way that their own behaviour is affecting markets.
Actually, this is true only for the credit derivatives. The equity (stock) derivatives are "almost" safe I can say (well, they will shake if the whole economy goes down, but they will not trigger another crisis of their own). And t hat's the vision of a pessimistic outsider who decided a few years ago to move inside the matrix and check the small cogs.
They did trigger the 1987 crisis, which was the first crisis after the mass-marketing of stock options. It was clearly caused by the primitiveness of the Black-Scholes model, the lack of a real shock in the historical data to calibrate the model, and indeed the very first program trade automata, which clearly did not mind for their own impact on the market.
State-of-the art equity trade automata now make 1-st, 2-nd and n-th order estimates of the impact of their moves on the market, and they won't make another 1987 crash. (besides, they have better safety limiters now)
I think the reason 1987 went on so well, in spite of being the most spectacular one-day drop in decades, is that it really was decoupled from the real economy. It really was the result of programs valuing deals with the wrong model and trading in unchecked volume, in a death spiral. When humans pulled the plug, the market came back to normal.
B&S has been at least amended (and completely superseded for many options). And history now has a nasty stress-test through which all new equity models must pass.
Credit derivatives, on the other hand, have never been tested for a systemic crash. 1929 was believed to be too old and irrelevant. The "dissemination of risk" thing can actually work for on punctual default.
Remember Enron ? losses where close to 20 B$ with a duration in months - I'm talking AAA bonds and credit lines only, not the equity. This is something unheard before: -in France, we already had Credit Lyonnais which came up on the same order of magnitude but the state bailed it out and the duration of the liability was more like 10 years). - even LTCM only lost a few B$ over a couple of years, after the initial 120B$ short-term refinancing panic, which is why banks agreed to bail it out themselves once they figured out there was not so much risk in refinancing .
Well, Enron did not trigger any domino effect, although it would have if it did happen in the time of LTCM. Why ? precisely because it's lenders had offloaded the risk to larger markets.
Today it won't work, because the risk is systemic, not a one-off: banks have been trading their respective risks with each other, thinking "I'm smartly dumping my shit over my neighbor", when they should have thought forward "Hey, but we're all throwing the same kind of shit around". So basically they hold portfolios full of crap that is no better than what they sold away.
Subprime, alt-A will suffer two-digit losses on 600B$ of AAA bonds. Then consumers will choke, corp profit too, and a greater amount of AAA bonds, CDOs, CDS backed by commercial real estate and corporate junk bonds will be wiped out. Some emerging countries will go into civil war and default on a smaller amount of sovereign debt (but still pretty big).
We are talking of a shock of the magnitude of a sudden default by a major sovereign borrower (the size of France or Germany), in the tune 1000+ B€ of AAA grade. Calyon (french peasants turned investment bankers) just reckoned 150B$ of global losses, goldman sachs now 300B$ (they minimize, they may be sued), it will turn out to be bigger after ARM-october, 50B$ in loans to reset. And that's just subprime.
Recall that total US$ M3 is probably about 13 000 B$. Of which 30% is "classic" money. So we really are about to see some +15% of "wall street money" (chimp money) burn in flames. Guess there will be fewer banks on the Street in two years. Pierre
The problem is that a number of them have been built on the assumption that counterparty risk and/or liquidity risk could be, to a large extent, ignored - i.e. they are valuable so long as people thing they are but not if everybody actually needs to cash them in - because the whole house of cards folds, then.
That's what we're seeing now. With nobody willing to take one new risks on these, the existing ones suddenly cannot find buyers, and thus have no apparent value - the value will appear only if and when they are actually called on, and the counterparty pays, or not, the requisite amount. but in the meantime, they are a huge black hole. But a lot of finance was built using as foundations these products, so there are whole edifices that suddenly don't know what they are standing on.
Fun stuff. Good time to be a litigation lawyer, I expect. In the long run, we're all dead. John Maynard Keynes
