As disruptors of traditional businesses, companies such as Facebook, Amazon, Netflix and Alphabet (collectively, FANG) have made big headlines as their stocks soar into mega-cap territory. Recently, Goldman revised the acronym to FAAMG to include Apple and Microsoft (although excluding Netflix since it has only a small weight in the S&P 500) to better represent the mega-cap tech stocks driving the rally in the broad indices. There has been significant market chatter about how narrow the rally has been, with FAAMG creating $660 billion of the total $1.8 trillion in incremental market capitalization (as of June 7, 2017).
Back in early 2016, Cliff Asness at AQR looked at the impact of individual stocks on the S&P 500 from 1994 to 2014 and compared the results with the FANG-driven rally in 2015. Though not an exhaustive study, the exercise involved removing top-performing N stocks that had the biggest impact on the calendar year’s index return. By quantifying the impact on index returns by removing those N stocks (starting from 1 stock up to 20) and re-weighting the index over the remaining 500 minus N stocks, Asness pointed out that 2015 was 'staggeringly normal', as can be seen from the chart below.
A fascinating paper published by Dr. Hendrik Bessembinder of Arizona State University, entitled Do Stocks Outperform Treasury Bills?, emphasizes diversification as a key tool in portfolio management in the absence of the skills needed to consistently identify mispriced stocks. Using the CRSP database, he found that of the $32 trillion (as of December 2015) in lifetime shareholder wealth creation from 26,000 stocks on U.S. exchanges that have appeared in the CRSP database since 1926, 86 top performing stocks (just 0.33% of the total) collectively accounted for roughly half of the overall wealth creation. Moreover, merely 1,000 top-performing stocks (4% of the total) accounted for all of the wealth creation. That is, the other 96% of stocks have collectively generated lifetime dollar returns that match the one-month treasury bill; though this does not necessarily mean they did not generate any value.
Since actively managed portfolios tend to be poorly diversified, these studies have important implications for investors in strategizing the number of stocks or industry sub-groups the portfolio should constitute. Very high positive skewness of some of the stocks and tendency of majority of the stocks to return less than zero excess returns, results in median excess return in negative territory. This holds a lot of significance since this conflicts with assumptions made in the capital asset pricing model (CAPM assumes positive mean excess return). Simply put, very large positive returns of a few stocks offset the negative returns to more typical stocks due to the aforementioned positive skewness in individual stock returns. This lottery-like payoff is associated more with private equity or venture capital investments. Studies such as these show that it exists in public equity markets as well; masked only by the large number of listed securities which allow investors to obtain diversification more easily than PE or VC investments. The ability to diversify across listed securities helps the investors earn excess returns, which historically has been ~6%.
These results highlight the importance of portfolio diversification, especially for investors who view performance in terms of the mean and variance of portfolio returns. Not only will a poorly diversified portfolio experience higher volatility (variance of portfolio returns), it will be subjected to the risk of excluding those relatively few stocks that, in retrospect, generate large excess returns. Underperformance of an actively managed portfolio can thus be attributed to poor diversification as well, and not necessarily only on fees, transaction costs and negative skill.
Despite the possibility that poorly diversified portfolios will more likely underperform the overall market, active managers may choose to maintain concentrated portfolios (concentrated portfolios occasionally deliver excellent returns). This may be done to take advantage of perverse incentive structures (where fees are based only on quantum of returns, excluding any risk based metric) or merely a preference for skewed payoff distribution. Since diversification reduces skewness, these investors may choose to hold portfolios that are not highly diversified.
To summarize: Despite a preference for concentrated portfolios by active investors, studies have shown that diversification helps in not only reducing volatility of returns, but also enhancing returns primarily due to positive skewness in stock returns. Concentrated portfolios will more likely underperform the overall market in the absence of the ability to systematically identify mispriced stocks.