It's easier to invest $100,000 than it is $100 million. As hedge funds grow, the inherent hurdle they most often encounter is capacity. The larger the fund, the more difficult it becomes to capture the inefficiencies in the market for additional limited partners. It comes down to the semi-exaggerated phrase, "how can you beat the market, when you are the market." Nimbleness is a virtue of investment management. The easiest analogy is one of nautical term. A 22-foot ski boat can go from full speed in one direction and complete a 180 degree maneuver in a matter of seconds. A crash stop maneuver (from 'full ahead' to 'full reverse') on a fully-loaded supertanker can take almost two miles over a 14 minutes period. Its turning diameter alone is almost 1.2 miles. Ideally, one would rather be in the investment equivalent of a ski-boat.
The same can be said for mutual funds. Popular funds become victims of their own success. Large amounts of capital leads to over-diversification. If you have 20 great ideas, but $780 million to invest, you are either pushed into very large cap names, or large percentage holdings. What truly happens is that you invest a portion of your funds into your best ideas, and then surround those with second-, third-, or even fourth-tier ideas, but in lessor quantity. In the end, those funds end up owning the market and replicating their benchmarks. Each mutual fund prospectus highlights its top 10 holdings, but you must search for total number of holdings.
Look at Fidelity (I use them not because they are different than anyone else, but rather, I had their data handy). The median number of holdings in their actively managed funds is 195 positions. Position sizes range from 1,132 holdings in a $2.5 billion fund to 51 positions in their focused fund. Even the "Blue Chip" funds have over 300 names. You can find a list of holdings here. Although a fair portion of assets are invested in the top 10-20 holdings, the typical fund's average position size is often less than $1 million per name. There are as many as 100 names that each represent less than 1/2% of the fund holdings, and in aggregate represent 29% of assets.
Modern Portfolio Theory, and works including must-read A Random Walk Down Wall Street by Burton G. Malkiel, suggest that the benefits of diversification begin to marginalize at 10-20 positions.
The esteemed William Bernstein has made a counter-argument to that thesis. He notes that standard deviation of returns (the basis of Malkiel and others) is not the utmost benefit of diversification. Rather, he discusses a principal called Terminal Wealth Dispersion (TWD) coined by Edward O'Neal of Auburn University. Although standard deviation of returns in a 15- stock portfolio are equal to those with large positions, it's the relative returns that matter. TWD looks at the risk that continuous underperformance has on the ultimate value of a portfolio. His analysis (done in 2000) suggests that random, equally-weighted portfolios of 15 S&P 500 stocks in 1999 have dramatically different results with as many as three quarters of them failed to beat the S&P itself.
The reason is simple: a grossly disproportionate fraction of the total return came from a very few "superstocks" like Dell Computer (DELL), which increased in value over 550 times. If you didn't have one of the half-dozen or so of these in your portfolio, then you badly lagged the market. (The odds of owing one of the 10 superstocks are approximately one in six.) Of course, by owning only 15 stocks you also increase your chances of becoming fabulously rich. But unfortunately, in investing, it is all too often true that the same things that maximize your chances of getting rich also maximize your chances of getting poor.
Mr. Bernstein concludes that portfolios in excess of 100 names (potentially as many as 200 names) are more appropriate.
Although Mr. Bernstein's thesis supports the approach taken by most mutual funds, it differs from reality on two basic points - "equally-weighted" and "random." First, the portfolios of mutual funds are not equal-weighted. As discussed in the Stock Selector Fund above, vast holdings represent minimal percentages of assets. As such, a combination of ten under-performers can outweigh the gain of one great performer. Secondly, the key is to not randomly create portfolios, but seek to identify that "Dell superstock" in advance, or at least a portfolio of potential "Dell" stocks.
With a defined and tested methodology, an investor can reduce the randomness in their portfolio, and thereby increase the likelihood of superior return. By ranking the universe in order of investor-desired qualities, he or she can select "good" from "bad." Our results indicate the returns are diluted when "less-good" are added to the mix.
Using the Bloodhound System, we selected one of our model strategies - BH: Fundamental Growth Strategy. This particular strategy seeks fundamentally strong companies that have the potential for future growth and stock appreciation. The strategy identifies low debt companies with solid margins and has a fairly simple ranking methodology, favoring companies with high trailing twelve month returns on equity.
We created four equally crafted portfolios with the only difference being the number of stocks owned. The first portfolio owns 10 of the highest ranked stock, equally-weighted. The second portfolio owns the top 50. The third owns 128 (equal to the Fidelity Discovery Growth Fund), and the fourth holds 195 - equal to the median Fidelity fund. The Bloodhound simulator accurately recreates the portfolio (trade-by-trade) over the past 26 years using the user-defined ranking and buy rules on the stocks that existed at that time using the data available to the investor on those dates. The results are below.
As can be clearly seen in the data, a more concentrated portfolio of the "best" stocks outperformed the more "diverse" group over both the short-term and the long-term. In 2012, the portfolio of 10-stocks generated a 14% total return, slightly below its 26-year average of 16.9%. The strategy took a hit with Western Union (WU) - down 26.6% - but was buoyed with strong returns from AFC Enterprises (AFCE) - and Gartner Group (IT) - up 27%. The portfolio of 50-names generated a 10% return, which indicates the next forty names ran at an average of 9%. Names such as Apollo (APOL), Cypress (CY) and Herbalife (HLF) specifically hurt the return profile.
Throughout the grid, there remains a clear distinction between the 10-stock portfolios and the other portfolios. The difference between 50 names and 128 is not particularly meaningful. It might suggest that the marginal impact of diversification levels off after a certain number of positions. However, there remains a significant difference between each of the first three portfolios and that of the 195-stock portfolio. Over a 20-year period, the 10-stock portfolio generated an average return of 16%. The next 118 ranked names averaged a lesser 12.75% over the same period.
The real distinction comes from the remaining 67 stocks. As a reminder, these are not the same 67 stocks over 20 years, but rather the consequence of adding 67 of the lesser qualified names each year. There may be some overlap year to year amongst the list, and some names may be high on the list one year, and low another, but every year the portfolio rebalances. The average stock in that group used its proportional share of capital, but contributed miniscule returns taking down the portfolio's overall return. The cumulative pool of the 67 lowest ranked stocks yielded an average return of less than 1.5% per year compared to the 16% generated by the top ten. In 20 of the 26 years worth of data, the 10-stock portfolio beat the 195-stock portfolio, and often by a substantial margin.
We highlighted one of Bloodhound's model strategies; however, the consequence of adding lower ranked stocks is seen time and again in our analysis. Properly-ranked stocks benefit from concentration, and adding additional capital to the lesser ranks is not diversification, but rather di-wors-ification.