Fiduciary Papers #11: Benchmarks for Comparing Investment Portfolio Strategies

Before choosing specific securities (stocks, bonds, mutual funds, etc.), a preliminary step in portfolio design is discerning the investment strategies that should be utilized. The choice of investment strategies, in turn, should be analyzed against a proper benchmark.

At its very core, investing is either about owning a portion of companies (equity investing, or stocks), or lending to a corporation or government (fixed income, or bonds). There are a few other type of assets that don’t fit nicely into either the “stocks” or “bonds,” such as real estate (either direct investment, or via REITs), commodities (via futures contracts), and various forms of options. (I would argue that “cash” belongs to fixed income investments.)

In assessing the choice of asset classes, or investment strategies, I suggest the following benchmarks be utilized:

For equities, a market-cap weighted “total stock market index.” Although “global” stock indexes exist, it might be more appropriate to utilize separate total stock market indexes for U.S., foreign developed, and emerging markets.

For bonds, a total bond market index, such as the Bloomberg U.S. Aggregate Float Adjusted Index (for U.S.-based investors).

For REITs, a broad-based index of publicly-traded REITs.

For each of the foregoing, investments can be made via very-low-cost mutual funds or ETFs that track the broad index. For example, VTSAX – the Vanguard Total Stock Market Index Fund Admiral Shares, currently possesses an annual expense ratio of only 4 basis points (0.04%), and it has a portfolio turnover rate (for the year ending 12/31/22) of only 3.4% (as computed using the SEC’s methodology; actual portfolio turnover rate may be higher). One advantage of this particular fund is that returns data exists since the fund’s inception – just over 30 years ago, in 1992.

The logic behind the use of such broad-based indexes, in evaluating investment strategies (a first step in the due diligence process), is simple. The returns of “the markets” are there for the taking, via very-low-cost mutual funds. Investment strategies should be measured against these returns (although even a lower return can be justified with a significant reduction in portfolio risk).

Investment strategies, when evaluated quantitatively, in essence are chosen because the investment adviser believes there mispricings relative to what market-weighted indexes may return, or a reduction of risk, or both, over the investment time horizons which are unique to the client.

In evaluating investment strategies against these benchmark indexes, the primary challenge is always possessing sufficient data, whether it be of returns over time, standard deviation (volatility) over time, correlations to other asset classes (or benchmark indexes), etc. Backtesting is often done with data extending back only a few decades (instead of, as might be desired, a few centuries).

Another challenge is that the future is not the same as the past. While the historical data for investment strategies informs us, it does not provide certainty of future results. For example, as knowledge spreads about a particular “new” investment strategy, the prices of securities in the market may adjust, thereby negating a portion of, or all, of the potential benefits of the strategy.

What we are left with is probabilities of chosen investment strategies outperforming the benchmark indexes (or index funds), over future periods of time. Investment advisers should search for investment strategies that possess a substantial probability of outperformance, over the time horizon established for that portion of the client’s portfolio. Keeping in mind the likely dispersion of likely returns and other measures (and forms) of risk.

Presenting the investment strategy(ies) chosen by the investment adviser to the client might best be undertaken in terms of probabilities over time of beating the benchmark indexes. For example, one might say: “This low-cost U.S. stock fund, which includes mid-cap and small-cap stocks and employs the price (value) and quality factors, possesses an 80% probability or greater of outperforming a U.S. total stock market index fund over any given 20-year period of time.” The historical probability might be greater, but since the future is not identical to the past, a more conservative projection might be utilized. Even then, it should be pointed out that the risk level of such a strategy might be different. For example, small-cap and value stocks performed worse during the early years of the Great Depression and during the Great Recession.

There are other steps in the process of designing investment strategies, and in choosing the securities and investment vehicles to implement such strategies. I do not suggest that the use of broad-based indexes is the appropriate method when choosing the index for benchmarking a particular fund or portfolio of securities. For example, if an advisor will use a “U.S. large cap value” asset class for a portion of the portfolio, then a further comparison of a fund (or portfolio) consisting of such equities – to an appropriate U.S. large cap value index – should be undertaken.

Investing other people’s money is not, and should not, be a game. Rather, in my view, it should involve informed judgments that possess a significant probability of investment success, whether such investment success is achieved through risk reduction, enhanced returns, or a combination of both, over the relevant time horizon of the client.