There is a larger question that must be asked in discussing the recent introduction of new Claymore funds tracking global real estate and infrastructure: just how many asset classes does a portfolio need? How much more diversification benefits can real estate, infrastructure, commodities, or even more exotic asset classes provide? In Unconventional Success, David Swensen discusses asset allocation is great detail and provides some guidelines:
- A strong equity orientation.
- Sufficient diversification through many asset classes.
- At least 5 to 10 percent should be allocated to any individual asset class — enough to have an impact on the overall portfolio.
- An asset class should not dominate a portfolio — allocate no more than 25 to 30 percent to a single asset class.
- The asset allocation decision should be infrequently revisited.
Mr. Swensen’s reasonable guidelines suggest a portfolio composed of no less than 4 and no more than 20 asset classes. A strong equity bias means that the maximum number of asset classes should be far less than 20. The Sleepy Portfolio, for instance, has only eight asset classes — cash, short-term bonds, real-return bonds, REITs, Canadian equities, US equities, EAFE equities and emerging market equities — but the bulk of it (70%) is in equities. The other 30% is already divided between four asset classes; there is simply no room for new ones.
It seems to me (I don’t know of any studies to back up my claim) that once a certain level of diversification is achieved, adding more asset classes is likely to fall prey to the law of diminishing returns. So, as far as the Sleepy Portfolio is concerned, eight asset classes should be plenty enough.
