Is Diversification A Bad Thing?

My friend Eddy Elfenbein raised an interesting point on his Crossing Wall Street blog about the distribution of stock returns (need to tip my Red Sox cap to Mebane Faber and Nick Gogerty). The takeaway (and there is data to back this up) is that most individual stocks don’t do much; actually the study linked above says that only one in five stocks is a “significant winner.”

The argument being made is that too much diversification can be a bad thing (diminishing returns). I come at this a different way. Most of the accounts I manage have 40-45 holdings when fully invested. Most of those, about ¾, are stocks with the rest being ETFs. Obviously I’m talking about the equity portion of the portfolio.

I build portfolios from the top down going sector by sector. Before stocks are selected, decisions are made about overweighting or underweighting the sectors and then what countries to buy (there is more to this but for space sake…). Top down theory says that stock selection is the least important part of the process, the idea being that if you figure out, for example, that you want to own a Norwegian oil stock and get Norway right and get the sector right that the few stocks you would choose from will all correlate closely and to the actual stock chosen becomes a shade of gray.

There are no guarantees but it works often enough.

I have found that in going through this process and then selecting 40-45 holdings that one or two of the stocks selected will skyrocket over the course of a year; more than 50% up. Often the one or two that go up that much are not the ones I would expect. This happened to me several years ago with Advanced Auto Parts, I mention that one because it was the biggest surprise ever, this has also happened with banks and other often “docile” stocks. This is not meant to brag because I am convinced that any reasonably researched portfolio would include a couple of stocks that skyrocket and hopefully it is clear I am talking in the context of a bull market.

With 40-45 names you may have figured out that most of the holdings get a 2-3% weight in the portfolio. If you start with the idea that the stock market has an average annual return of 9-10% then one stock (out of 40) with a 2% portfolio weighting that doubles will add 2% to the portfolio’s total return. Then if one other stock (out of 40) goes up 50% that would add another 1% to the portfolio’s total return. The two would deliver 1/3 of the annual average which is a lot of heavy lifting for the portfolio. It is easier, IMO to own 40-45 stocks with the expectation that most will be a little ahead or a little behind the S&P and get the sort of boost described above from a small number of holdings.

If you only have 15-20 holdings then you have more riding on being exactly right about those 15-20 and you reduce the chance of getting an Advanced Auto surprise. 15-20 is simply a more difficult path.

Where possible, and it is not always possible, I prefer to make things as simple as possible.