It may be tempting to skimp on due diligence for index funds and exchange-traded funds. After all, they typically employ transparent, rules-based strategies that are not susceptible to key-manager risk. But seemingly similar index funds can look and perform differently, so portfolio construction still matters. To evaluate these funds’ processes, we set out to answer two questions: 1) How is the index constructed? and 2) How is the portfolio managed to deliver high-fidelity tracking performance of its index? The former determines the composition of the portfolio and how the fund will behave. The latter focuses on both the fund’s replication strategy and structure. While it is important, it doesn’t move the needle as much as index construction, which is the focus of this article.
Index Construction Framework
The best index funds have portfolios that are representative of their actively managed peers—or in the case of strategic-beta funds, the style they are trying to capture. They are well-diversified, investable, transparent, sensible, and take steps to mitigate unnecessary turnover. Looking at these dimensions can help us differentiate between seemingly similar strategies and better evaluate their merit.
The more representative an index fund is of its actively managed peers, the more reliably its cost advantage should translate into better performance. If there are material differences—with respect to the sector, style, security, or country weightings—between the index fund and its active peers, its gross performance can differ meaningfully from the category average. And its cost advantage may not be sufficient to overcome underperformance when it does occur.
For example, Vanguard Total Bond Market ETF (BND, listed in the U.S.), which tracks the Bloomberg Barclays U.S. Aggregate Bond Index, has much greater exposure to U.S. Treasuries than its actively managed peers in the intermediate-term bond Morningstar Category, giving it far less exposure to credit risk. As a result, its holdings tend to have a lower average yield to maturity than its typical active peer. But the fund’s cost advantage exceeded the difference between its yield and that of the category average at the end of September 2016.
Diversification is perhaps an even more important consideration than representativeness because it can help mitigate exposure to risk that the market does not reward. Investors can gauge concentration risk with simple metrics like the percentage of a fund’s assets in its top 10 holdings. As a rule of thumb, if this figure is above 30%, firm-specific risk may be creeping into the portfolio. Diversification across sectors and countries (in the case of international strategies) is also important.
Even a well-diversified index may not be worth touching if it is difficult to track or has limited capacity, meaning that it is not investable. Fortunately, investability isn’t a problem for most equity index funds. However, it can be an issue for funds that invest in less liquid securities, like micro-cap stocks and thinly traded bonds. These securities are often difficult to access and expensive to trade. Consequently, index funds that cover them tend to have higher tracking error and transaction costs than those that invest in more-liquid securities.
Transparency makes it easier to anticipate changes to the portfolio, which is important because index strategies run on autopilot. It can also make it easier to understand how the index will likely perform. Indexes managed by committee, like the S&P 500 and Dow Jones Industrial Average, do not meet the transparency standard. Neither do index funds that employ complex optimization models to construct their portfolios. These strategies could still be worth investing in, but it is prudent to wait until they have established a record to get a better handle on how the process works in practice.
It goes without saying that a strategy should be sensible, but it’s not always obvious which strategies clear that hurdle. Price weighting, the approach that the Dow Jones Industrial Average employs, is a prime example of a process that falls short. There is no viable economic theory to support price weighting, which assigns larger weightings to higher priced stocks. Dow Jones simply adopted it because limited computing power rendered alternative approaches impractical when it first calculated the index in the late 1800s. While there are other such indexes that fail this basic sensibility test, there aren’t many of them and most do not attract much money.
Finally, good indexes take steps to mitigate unnecessary turnover. This is important because there are costs associated with turnover, particularly among less liquid securities. For example, the Dow Jones U.S. Small-Cap Total Stock Market Index does not remove constituents from the index unless they drift meaningfully outside of its targeted size range. This helps reduce turnover in instances where it wouldn’t materially affect the fund’s size orientation, thereby reducing transaction costs.
In contrast, the Russell 2000 Index does not apply buffer rules around its lower market-cap bound, where it can be expensive to trade. Additionally, there is a larger pool of money tracking this index, and it has greater exposure to less liquid micro-cap stocks than the Dow Jones index. The combination of these factors leads to higher transaction costs that can hurt performance. Over the trailing five years through September 2016, the Russell 2000 Index lagged the Dow Jones U.S. Small-Cap Total Stock Market Index by 84 basis points annually.
Index construction matters because it can have a significant impact on a fund’s performance relative to its peers over the long term. Aside from cost, it is the most important factor to evaluate before selecting an index fund or ETF.
