Sunday, May 31, 2009

Diversifying a China Oriented Portfolio

Glitzer asked me a little while ago to simulate a China oriented portfolio diversified with the Man-AHL managed futures fund. She specified two US listed stocks for the base portfolio - CHN - the China Fund and IFN - the India Fund with 80% allocated to CHN and 10% allocated to IFN and the remaining 10% in Man-AHL. If you invested a lump sum in these three securities at the beginning of October 1996 and rebalanced the portfolio monthly this is how they would have performed till now (ignoring transaction costs):

Everything has gone up very nicely. CHN gained 508%, IFN, 640%, and Man-AHL 724%. Glitzer's allocation would have gained 558%. But CHN and IFN have been very volatile with monthly standard deviations of 10.9% and 11.4%. Man-AHL's standard deviation is 5.2%, which is more than most developed country stock indices. Glitzer's portfolio would have had a monthly standard deviation of 9.4%. Can we improve on this?

Keeping the 8/1 ratio of CHN to IFN the allocation to Man-AHL that minimizes the portfolio standard deviation is 78% in Man-AHL (!), 19.6% in CHN, and 2.4% in IFN. This portfolio has a standard deviation of only 4.6% and ends up increasing by 805%. More than any of the funds due to the wonders of rebalancing!

Now Glitzer actually proposed dollar cost averaging - adding $1000 per month to the portfolio in the 80:10:10 proportions and not rebalancing. How would this have turned out since 1996?

We would have invested $153,000 and the current value of the portfolio would be $565,000. The current allocation would be: CHN, 80.1%, IFN, 11.4%. Man-AHL 8.5%. So there would have been a little drift from the original allocation. The total gain over the period would have been 490% and the monthly standard deviation is 9.1%. Dollar cost averaging a fixed allocation doesn't work here - it underperforms a lump sum investment with rebalancing. The optimal allocation to Man-AHL here is actually 83%. Investing a lump sum 80:10:10 in 1996 and not rebalancing would have resulted in a 463% gain with a 8.6% standard deviation. The current allocation would be 72:13:15.

The bottom line here is that based on historical performance very large allocations to managed futures are justifiable in order to improve volatility and return of equity oriented portfolios. Rebalancing helps a lot. Dollar cost averaging less so.

In the real world rebalancing costs money. But if you are investing regularly, adjusting your allocation to effect rebalancing might make sense. We don't know whether managed futures will perform as well in the future and I would not risk putting 80% of my portfolio in a single fund. I would use a smaller allocation to managed futures and distribute it across managers and use other asset types for additional diversification. And unless you live in Hong Kong like Glitzer or have your managed futures in a retirement account, you need to think about the tax implications of these funds.

No comments: