Tuesday, April 21, 2015

Over-diversification and hedge funds



The gains from diversification have been called by some as the only "free lunch" in finance. It is almost like apple pie - no one should be against diversification, but it is possible that you can have too much diversification. (No different than having too much apple pie.)

Adding more stocks will diversify away risk until you get the market portfolio which is non-diversifable. The same also applies to hedge fund strategies. As you add more hedge funds within a strategy, there will be diminishing diversification. In fact, there will be diminishing gains based  on just abut any feature. There will be a decline in drawdowns but it will not be eliminated. There will be a smoothing of returns to an strategy norm. At that point, paying fees for a fully diversified portfolio makes no sense. Adding another manager, paying their fees but not getting any portfolio value is a loser's game. There will be a performance drag. The question is determining what is the optimal number of hedge funds after which there is minimal value-added. There is a law of demising marginal return for hedge funds.

This is not a simple question. There has been significant controversy over the optimal number of stocks. Some have argued that that it can be as low as ten while others will say the number is closer to  40. Most will say that a set of 30 is a good number.

With respect to hedge funds, there has been less research overall but enough for us to make some generalizations. A number greater than 10 is pushing the limit on where diversification is maximized. The outer limit to diversification comes at a number around 20. This will vary by strategy and through time but the limit is lower than what you will find with an individual stock portfolio. An example would put this in perspective. If you move from 1 to ten managers in managed futures, you will see a decline of about 35 percent in volatility. If you add another 10 managers, you will see an additional reduction of between 5-10 percent in volatility. If you have to track twice the managers and only receive a marginal reduction in risk, you are not receiving a good return to risk trade-off.

If you add managers across strategies, the diversification may be greater, but the principle is the same. Adding more managers will have a marginal benefit. If you are conservative and want the maximum diversification, the simple rule of an even dozen is a workable first pass on the problem.  

No comments: