Exchange-traded funds held $810 billion in U.S. stocks as of January 31, 2013, up 21.6% in the trailing 12 months. Our new ETF holdings dataset allows investors to track the underlying U.S. equity holdings of ETFs on a daily basis.
One promising area for research is a comparison of the weight of each stock in the ETF universe to its weight in major indices. Since most ETFs are cap-weighted, we would expect most stocks to have weightings in the ETF universe that are similar to their weightings in the major indices.
But this is not always the case. We compared the U.S. equity holdings of ETFs to the Russell 1000, which is a good proxy for the U.S. stock market. ETFs overweight some stocks and underweight others. For example, Microsoft (NASDAQ:MSFT) is the third-largest holding of ETFs (1.53%), but it has only the seventh-largest weight in the Russell 1000 (1.38%). In contrast, General Electric (NYSE:GE) has the third-largest weight in the Russell 1000 (1.52%) but is only the tenth-largest holding of ETFs.
Top 10 Weight in Russell 1000
Top 10 Weight in U.S. Equity ETFs
International Business Machines
International Business Machines
Procter & Gamble Co
Johnson & Johnson
Johnson & Johnson
Procter & Gamble
ETF managers with cap-weighted benchmarks are obligated to re-balance their portfolios so that they closely match the benchmarks they track. Therefore, we would expect that the weight of underweighted companies will eventually rise, while the weight of overweighted companies will eventually fall. One strategy to exploit these weighting discrepancies is to be long the stocks that are most underweighted in ETFs and be short the stocks that are most overweighted in ETFs.
Using data for the past three months, we calculated each stock's "ETF weight" and compared it to its weight in the Russell 1000 on December 3, 2012, the first trading day of December 2012. For example, Apple (NASDAQ:AAPL) was 0.056% overweighted by ETFs relative to the Russell 1000. On the one hand, the most overweighted stock, Simon Property Group (NYSE:SPG), was overweighted by 0.4165%. On the other hand, Exxon Mobil (NYSE:XOM) was underweighted by 0.5686%, the most underweighted stock relative to the index, followed by General Electric (0.4923%) and Berkshire Hathaway (BRK.B, 0.3479%).
The average weight deviation for all Russell 1000 stocks is 0.0001%, which shows that the stock holdings of ETFs are generally the same as those of the overall market. The standard deviation of the difference is 0.0509%. Below we consider only the stocks whose weight in ETFs is one standard deviation from the mean. This process narrows the universe to 103 stocks: 56 underweighted and 47 overweighted. We call the underweighted stocks Group 1 and the overweighted stocks Group 2.
We constructed a paper portfolio of stocks for each group-one equal-weighted and one cap-weighted. In the past three months, the equal-weighted portfolio for Group 1 outperformed the Russell 1000 by 1.47%, while the equal-weighted portfolio for Group 2 underperformed by 2.87%. The cap-weighted portfolio for Group 1 underperformed by 0.12%, while the cap-weighted portfolio for Group 2 underperformed by 8.08%. If an investor were long Group 1 and short Group 2, he would have outperformed the Russell 1000 by 4.34% for the equal-weight portfolios or 7.97% for the cap-weighted portfolios.
Relative Performance of Group 1 and Group 2 to Russell 1000 in Past Three Months
Group 1 (Stocks Underweighted in ETFs)
Group 2 (Stocks Overweighed in ETFs)
It is disappointing that the cap-weighted portfolio for Group 1 underperformed in the past three months, even if only by 0.12%. One explanation is that ETF managers overweight or underweight particular stocks for fundamental reasons. For example, if there are two companies in an industry with similar market caps and similar growth prospects, managers may overweight the one that pays higher dividends.
As we continue to explore the idea of exploiting weight differences between ETF holdings and stock indices, there is one piece of good news for beleaguered Apple shareholders. Three months ago, Apple was overweighted by 0.0560% (it was the 39th most overweighted stock in the Russell 1000, and the overweighting exceeded one standard deviation). This week, Apple is 0.1048% underweighted (it is the 24th most underweighted stock, and the underweighting exceeds two standard deviations). This data suggests ETF managers sold Apple disproportionately to other stocks in their portfolios amid the 39% drop in Apple's share price since September 2012. Hence from this relative weight perspective, the stock is ready to outperform in the next few months.
Disclosure: I am long AAPL. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.