Markets and Martingales
I don’t know a whole lot about the world of high-finance, but I do understand a bit about probability and in probabilistic terms, the government’s proposed $700 billion bailout of U.S. financial markets bears some uncanny similarities to the use of a “martingale” strategy to make money. That is not a good thing.
Gamblers have unsuccessfully tried to win using martingale strategies since at least the late 18th century. Here’s how the basic version is supposed to work, applied to roulette as an example…
- Start by betting $1. If you win, pocket the $1 winning and bet $1 again.
- If you lose, play again, but double your bet to $2. If you win, your total winning is $1, $2 won on the second bet, minus $1 lost on the first. Pocket the $1 winning and begin again, with a $1 bet.
- If you lose the $2 bet, play again, doubling your bet to $4. If you win, you will again have won a total of $1, $4 won on the third bet, minus $2 lost on the second bet, minus $1 lost on the first. Pocket the $1 and begin again.
- If you lose, at this stage or any stage, double your previous bet and begin again…
The problem is that the odds favor a single catastrophic losing streak – “catastrophic” defined as bad enough to swallow your initial stake – eventually wiping out any gains made from building your winnings $1 at a time. Unless you have an infinite amount of money (in which case, why are you bothering to bet?), you will reach a point where you can’t make the bet that’s necessary to cover your losses.
The signs of martingale-based thinking in the recent history of U.S. financial markets are disquieting. We’ve seen the government, twice in the past 20 years, appear to make two late-stage bets to cover a run of large losses, first the S&L bailout in the 1980s, now the proposed $700 billion bet to cover losses from the current housing and mortgage crisis. In between, there was the bailout of Long-Term Capital Management in the late 1990s, where some big Wall Street firms — including names like AIG and Lehman Brothers — put up big money to rescue a multi-billion dollar hedge fund that suddenly went on a big losing streak. They won that bet, kept playing, and later went on a losing streak that they were unable to cover.
The government’s resources for covering market losses are not infinite. Congress can’t decide to temporarily suspend the law of large numbers or any other of law of chance and probability. If the government keeps pumping funds into a financial system that requires continual doubling-down in order to stay ahead, isn’t it only a matter of time before the government experiences the catastrophic loss that wipes out all of its (er, that would be our) stake?