The 1% Doctrine: Evaluating Risk
I've been struck by several recent and current policy decisions that involve high risk and (perhaps) low probability.
Cheney and the Threat of Nuclear Terrorism
The first such policy to catch my attention was the statement attributed to George Tenet and an un-named briefer in a briefing to then Vice President Cheney in Ron Suskind's book:
If there's a 1% chance that Pakistani scientists are helping al-Qaeda build or develop a nuclear weapon, we have to treat it as a certainty in terms of our response. It's not about our analysis ... It's about our response. (Wikipedia)
This type of threat is described by Cheney as a "low probability, high impact event"
Pascal's Wager: the Chance of Heaven
At the time I saw it as a parallel to Pascal's Wager, roughly stated:
Even if there is only a small chance of God's existence, one ought to act as a believer, because the cost is finite and the reward is potentially infinite (heaven).
To paraphrase Cheney, this is an "uncertain but high impact wager."
Investing: Value at Risk
The next high stakes decision to strike me was the risk anlaysis of potential losses caused by borrowing money (30-1) and betting that home values would not decline nation-wide.
The technique to estimate the risk involved in these complex transactions is called Value at risk. Value at Risk (VaR) estimates the highest loss that is likely to occur in a given time frame. Think of this as a:
- 100 year hurricane, or
- 500 year flood
The odds of these events happening in a given year are 1 in 100 and 1 in 500 respectively.
Value at risk is a way of calculating a similar loss to a financial asset. Two common VaR metrics are 1% Var and 5% Var, the maximum likelihood that an asset will loose value 1% and 5% of the time.
Thus, if I were selling you an asset with a one day 5% Var of $1 million, there is a 95 percent chance that any asset loss will be less than $1 million dollars on any given day. Restated, 1 day out of 20 will likely result in a loss of $1 million.
Value at risk was pioneered at JPMorgan, and then picked up by others. Eventually the federal government required that firms calculate a VaR and submit it to regulators. This resulted in the practice becoming widespread, and it started to be used for the opposite purpose that it was designed. Rather than serving as an aid to human intuition, as Goldman Sachs did, VaR became a crutch that obviated the need for human risk analysis and it produced a number that led firms to a false feeling of confidence.
Nassim Taleb, one of the biggest critics of Value at risk argues that "unlikely events" happen more often than our perception of probabilities, and when they do, their magnitude is greater than the dollar value stated by Value at risk.
Put another way, for a 1% VaR of $1 million dollars, one percent of the time the least you stand to loose is $1 million dollars, with the potential that it could be much worse--it could bankrupt your company and overflow into the wider economy. Value at Risk does not really help you prepare for the 1%.
In the case of climate change, I have heard of no such VaR. I have heard about some of the risks:
- drought (including an end to commercial agriculture in California)
- melting of the ice caps
- disruption of biological systems
- changing ocean currents
My assumption is that the risk is greater than 1%, 5%, or even 20%.
One "conservative" argument is that the cost of action is high and the likelihood of success is unknown. Until the likelihood is better known or there is a promising solution that is likely to yield the cooperation of all other interested parties, no American action should be taken.
What Guides Action?
In these four cases, we've seen widely divergent approaches to action, and it may have something to do with the amount of sacrifice required and the perception of the likely success.
- Cheney: Dire Warning: the risk is higher than the cost.
- Pascal: the reward is so high that is is worth it.
- Wall Street: a 1% risk is unlikely and success is measured quarterly
- Climate Change: "conservative" Americans don't act when the costs are high and success uncertain.
- Tom Friedman: Going Cheney on Climate: the "precautionary principle"
2009/02/01 # Permalink