Market Overview
How indexes are made – and how to make your own
Some time ago, we discussed in this space how the Dow Jones Industrial Average matured over the course of 130 years. American Cotton Oil, American Tobacco and U.S. Leather gave way to Curtis-Wright, Mack Trucks and Union Carbide. These in turn yielded to Allied Chemical, Anaconda Copper Mining and Bethlehem Steel. Today, Apple, Cisco Systems and Salesforce are among the list of Blue Chips.
Which raises the question: How meaningful are the Dow, the S&P 500, the Russell 2000 and all those other indexes? And do we really need – by our count – 26 of them tracking U.S. stocks?
Index curation
Making an index involves four distinct steps according to S&P Global:
- Construction: selecting securities to achieve the intended exposure
- Weighting: determining the factors that determine how much of the index should be apportioned to which security
- Calculation: Dividing the cumulative market cap for all components by a divisor that is a product of the weighting
- Review: Changing components and weighting over time to stay current.
Each step is fraught with controversy. How do we know the Dow Jones Industrial Average truly reflects the 30 most prominent companies in the U.S.? Should an index be weighted by the value of the individual shares of each company or by some other metric? While the Calculation step’s divisor is easy enough to backsolve, how was it arrived at and how do we know that’s more valid than some other number? And when one component is retired and another takes its place, how do we know that this won’t artificially raise or lower the index?
These are all good questions and, since the development of these indexes is proprietary, we don’t have clear answers.
We do know that, as soon as a stock is added to an index, its price tends to jump. This suggests an inorganic rise in demand because a fund which excluded that stock is now bound by its charter to buy it. Indexes, then, are evidence of that scientific truism that the mere act of observation and measurement skews results. In other words, we have no idea how a component stock would be valued if it had never been included in the index.
Why we need them
Still, indexes are useful, if for no other reason than they give us benchmarks. If a portfolio manager can return value to investors in excess of the broader market, then that portfolio manager is truly worth keeping. If those investors could do just as well passively, then why bother paying a management fee?
They’re useful for another reason: as templates, so you can build your own. Maybe we need more indexes, not less.
Start with, say, the Russell 2000. Filter it for some other criteria – R&D spending, inside ownership, exposure to a set of macroeconomic factors, whatever – then re-weight it as you see fit.
This process is called custom indexing, and it is becoming increasingly popular. This may sound like overkill, but custom indexing has become one of the most successful investment tools in recent years. It gives you the diversified portfolio that any index fund would, but with the flexibility to omit underperformers. Because this strategy is by definition “custom,” it’s up to the investor to define the objective. It might be to generate current income, or it might be to preserve or grow wealth. It might be to support companies that share your values and to avoid those that don’t.
And, because you can chop out the laggards at any time, you should be able to reduce market risk and outperform established index funds. More importantly, you can take the proceeds from selling the laggards and reinvest them in better-earning companies with similar profiles. This affords you an excellent opportunity for tax harvesting, which involves selling securities at a loss, then replacing them in a portfolio with other securities that are similar but not identical. While investors and their advisors have been doing this for years, it has historically been an ad hoc, end-of-year activity. Custom indexing drives tax harvesting algorithmically and on an ongoing basis, so it should tend to be more efficient for this purpose.
At the close
Tax harvesting isn’t for everybody or for every portfolio. This strategy isn’t relevant for qualified retirement plans – IRAs, 401(k)s and the like – or other tax-advantaged accounts. Further, as Institutional Investor recently reported, mileage varies based on which index you start with to capture the same slice of the market.
Custom indexing and tax harvesting, as you can imagine, gets complicated quickly. It’s best not to make these tactical moves without the benefit of a trusted financial advisor.
