Archive for the ‘Uncertainty’ tag
Risk vs. Uncertainty
All “risks” have some “uncertainties”, but not the other way around. A good example from Wikipedia is:
We can be uncertain about the winner of a contest, but unless we have some personal stake in it, we have no risk. If we bet money on the outcome of the contest, then we have a risk. In both cases there are more than one outcome.
Taleb – The fourth quadrant
Class Connection – Decisions and Uncertainties
I recently read the book – The Black Swan – by Nassim Nicholas Taleb and was highly impressed by his thought process and his staunch empiricism. While going through his website I chanced upon another article that he wrote, as a sequel to the original book, called ‘The Fourth Quadrant’.
In this article,he talks about the limitation of statistics in providing a fair assessment of risk involved. He argues that current sub-prime crisis is a manifestation of inaccurately interpreting the information that statistics bore upon us. His question is – Are we using models of uncertainty to produce certainties?
For those who are not aware of this theory, here is a small definition to guide you – The Black Swan theory refers to a large-impact, hard-to-predict, and rare event beyond the realm of normal expectations. Unlike the philosophical "black swan problem", the "Black Swan" theory (capitalized) refers only to events of large consequence and their dominant role in history. Black Swan events may also be called outliers.
If you carefully study the fourth quadrant diagram as shown above, he talks about the fallacy of statistics only when the payoffs are large and the probability distributions unique (sometimes derived from events of the past and not necessarily futuristic looking). The problems and assignments that we have done in our class of assessing risks with small projects and payoffs do not necessarily falls into the fourth quadrant, but in the 2nd or 3rd. In those problems we were trying to measure and predict outcomes. But when trying to gauge effects, problems arise. Because here things work differently and the conditional expectation of an increase in a random variable does not drop as the variable gets larger (for examples read the article and the book).
Rare events are by nature, rare. And hence they are difficult to predict. Taleb, at the end of the essay provides some guidelines to follow when faced with a fourth quadrant scenario. Read it and avoid it.
