Higher Returns, Lower Risk, Lower Correlations
March 13, 2012
Concentrating a portfolio on a few choice assets dramatically increases an investor’s chance of superior performance. Nonetheless, most advisors and investors shun portfolio concentration as unacceptably risky. To a great extent, this is driven by the myth that adequate diversification is impossible unless one holds many stocks – a myth I will debunk below.
Most institutional and mutual fund equity managers hold more than 100 stocks, simultaneously reflecting and reinforcing the conventional wisdom that such fragmentation is necessary. But some highly prominent and successful managers have taken a different approach:
- In January 2006, Gabelli & Co. began publishing its quarterly Focus FiveTM to highlight the most compelling investment ideas from its institutional research team. The fund selects five companies based on current valuation and the potential for a near-term catalyst; no stock is held for more than two quarters. Over the five years and nine months that Gabelli has been publishing this report, the Focus Five has delivered a 22.1% annual compound return, vs. 0.8% for the S&P 500.
- Dennis Bryan is co-portfolio manager of the $1.2 billion FPA Capital Fund (FPPTX), the top-performing diversified U.S. stock fund over the past 25 years according to Morningstar. Despite the fund’s considerable assets, it holds only 21 stocks.
These are but two examples of concentrated equity portfolios that have generated superior returns. While it appears that the Gabelli researchers as well as Bryan are talented stock pickers, an equally important – if not more important – reason for their success is that each portfolio is concentrated in best-idea stocks.
A best-idea stock is one that a portfolio manager believes, based on research conducted by the manager and their buy-side analysts, has the best opportunity to generate a superior return. A growing body of research that shows active equity managers are superior stock pickers supports the existence of best-idea stocks.
Investors who avoid concentrated equity miss out on the triple benefits of excess returns, lower risk, and lower correlations. A portfolio concentrated in best-idea stocks has an excellent chance of generating excess returns. In turn, the cumulative excess return to investors lowers the risk of underperformance over time. Finally, a portfolio comprised of a small number of stocks is characterized by a low stock-market correlation. Thus the concentrated equity triple play: higher returns, lower risk, and lower correlations.
I will discuss each of these benefits separately. But first, let me address the widespread misconception that an investor can only properly diversify a portfolio by including a large number of stocks.
The diversification-volatility myth
Many believe that concentrated portfolios are much more volatile than are broadly diversified index portfolios. It turns out the diversification benefit of adding additional stocks to a one-stock portfolio is largely captured within the first few additions, as shown in Figure 1 below.
Based on methodology described in a 1968 paper by John Evans and Stephen Archer, Diversification and the Reduction of Dispersion: an Empirical Analysis; uses an average individual-stock standard deviation of 45%, an inter-stock correlation of 0.11 and equal weighting.
In this example, the typical stock has an annual standard deviation of 45%. Few would recommend a single-stock portfolio given the wide range of possible returns this implies – if the expected return for that stock were 10%, an investor might only expect with 95% confidence that his return would fall somewhere between 100% to -80%. In such a portfolio there is a substantial risk of losing everything!
Adding an equally weighted second stock provides 38% of the total potential diversification benefit (i.e., reduction in portfolio volatility as measured by standard deviation), as compared with a fully diversified market portfolio, which in this example would have a standard deviation of 15%. By the time a fifth stock is added, 70% of the potential benefit has been captured, and by the tenth stock 83% has been captured. The returns diminish quickly from there. Adding 10 more stocks captures another 8% of the benefit, and the maximum benefit is realized only when 270 stocks are included in the portfolio – the final 9% of the diversification benefit can only be had by adding 250 more stocks to the 20 you already hold.
This rapidly diminishing diversification value from the second stock to the 270th is a consequence of the mathematics of the portfolio standard deviation calculation itself, primarily driven by low correlations between individual stocks. How the portfolio is actually managed has little effect.
A 10-stock portfolio is, for all practical purposes, adequately diversified. An investor in such a portfolio will be subject to market volatility, but so are all equity investors.
Critics of this line of reasoning point out that individual stock volatility has increased over time, while inter-stock correlations have decreased, meaning that portfolio volatility declines less steeply today than Figure 1, which is based on a 1968 study, suggests. That may be the case, but the general conclusion remains the same: a relatively small number of stocks will provide most of the diversification an investor may seek.
Concentrated portfolios leave unaltered the recommendation that equities should not be used to fund short-term needs; whether you buy an index or just a few stocks, volatility makes an equity investment unattractive for meeting near-term needs.
Stocks are commonly used as a long-term growth engine for a portfolio, with their short-term market volatility viewed as an unfortunate side effect of their wealth-building potential. Since equity concentration adds to volatility only marginally, the decision of how much to invest in equities can be made largely independent of the degree of portfolio concentration. When it comes to returns, however, there are strong arguments for concentrated portfolios, a topic to which I now turn.
Would you like to send this article to a friend?Remember, if you have a question or comment, send it to .