Three of the most important pieces on risk:
1. the peso problem
2. fat-tailed distributions
3. Knightian uncertainty
The hat-tip belongs to falkenblog for bringing up three very important risk management issues in the context of finance and statistical theory. One of the current problems with risk management has been the focus on "black swans' without a good framework for discussion. It is also odd that the idea of "black swans" has been viewed as novel. Yes, the emphasis is important but we have seen much of this work before.
First, the peso problem has been well-documented across many asset classes. There is a risk premium associated with low probability large negative events. They exist and many market anomalies can be explained through these events. Using rolling histories of data may not include past negative events in our memory set. Bad things do happen even if they have not happened recently or have not happened in a given market.
Fat-tails are present. I have always viewed this as surprise risk. Surprises away from the normal can occur both on the up-side as well as the down-side. Investors have to be ready for both.
Knightian uncertainty exists. There are risk which are not easy measure. They are not countable, but they are present. Any investor has to be aware that risks which are not in the known distribution are possible.
These three themes are fundamental to a true understanding of risk.
No comments:
Post a Comment