Commodity markets saw increased volatility with both the Japanese earthquake and the Libyan crisis creating tail events which drove market flows to safety. Both events were unexpected in size and scope and caused an increase in the market risk premiums and herding behavior as a reaction to the unknown. Nevertheless, strategy and market diversification provides an easy way to protect against tail events especially those that are concentrated to one market or asset class.
Tail events, however, can have a mixed effect for active trading if there is limited follow-through or sudden reversals. Active trading may cut exposures at the wrong time through automatic stop-loss methods; consequently, it is critical that traders carefully assess unique market situations. As portfolio managers, it is all the more important to understand the mix of trading strategies within an active portfolio.
What makes a tail event?
The current buzz in risk management has been on hedging tail events, yet there seems to be little discussion on what constitutes a tail event and tail risk. Should a tail event be defined as a unique or surprise market situation which does not occur frequently, or should it only be a price move that falls in the extreme left hand side of the sample distribution. The distinction is important. Rare events do not always generate large price moves and not all price moves are tied to rare events.
Broad generalization concerning the chance occurrence of “black swan” events does not have meaning without acknowledging that surprise events have to be tied to market reaction and to the underlying knowledge about the probability of rare events. One man’s tail event may have been fully anticipated by another. Tail events and risk are related to the knowledge or scenarios generated by a portfolio manager. Additionally, tail events have to be placed in the context of probability and size. It is the exposure not the chance that determines risk.
In the case of the Japanese earthquake and the nuclear disaster afterwards, the sequence of events was not anticipated by the market. By that definition, those were tail events, but a close examination of price action suggests that they were not broad tail events as measured through longer-term time horizons or across all asset classes. The shocks from the financial crisis of September 2008 were deeper and more broad-based. Still, it may be early to see the longer-term implication of these events.
Of course, the immediate reaction to the earthquake was a strong negative move. For some companies and localized markets, this natural disaster tail event is unquestionable and will be long-lasting. However, for a broader global portfolio, the tail event impact of a localized shock is less clear. Many markets by the end of March were higher and look as though they have not been affected by the earthquake. In that sense, was the earthquake a true tail event?[1] Similarly, for longer-term investors, the definition of a tail event is unclear. Monthly returns were in many cases not out of the ordinary. Hence, tail events have to be looked through the context of time horizon, trading activity, and portfolio diversification. Tail event risk is most easily mitigated through the broadest form of diversification.
[1] An oil price shock from the Libyan uprising could be considered a different type of tail event. While the place and timing of the oil price shock was a surprise, the fact that there will be a price spike on Middle East unrest is well-known and the expected price response is more predictable. The trading behavior should be different in this situation.
No comments:
Post a Comment