By: Frank Armstrong
By: Frank Armstrong, CFP, AIF
When my wife asks me “How do I look?”, believe me, the right answer isn’t: “Relative to what?”. It’s unlikely that she would appreciate a comparison to Britney Spears, or an observation that she might have put on a couple of pounds.
But, when you ask how your investment account has done, the only way to begin answer is to ask, “Relative to what?” An absolute number like +2.5% really doesn’t convey much useful information. After all, in one case a 2.5% return might be a disaster, while in another timeframe or for another investor it is an outstanding result.
We need a benchmark. But, its got to be the right benchmark.
The S&P 500 is the default de facto institutional account benchmark Twenty years ago, this might have been a defensible position. Then the S&P 500 pretty much was the market for most investors. Only a few hardy souls ventured into international waters, Modern Portfolio theory was virtually unknown, the benefits of small and value were not appreciated, and basic risk control techniques were little understood.
But, today, use of the S&P 500 as a standard benchmark is unfortunate, because as a comprehensive investment policy, the S&P 500 is a sub-optimal. The S&P 500 offers both lower expected return and higher risk than more sophisticated alternatives.
If you are focused on the S&P 500, you may very well overlook more sophisticated and appropriate investment policies. If you have adopted one of these more optimal investment strategies, then the S&P 500 is an inappropriate benchmark. Still, it’s the one with “top of mind” shelf space, and that causes mischief.
As investors add alternative asset classes to the portfolio to diversify and capture the long term additional return potential of smaller stocks and distressed companies, their portfolio performance will diverge from the S&P 500. The greater the exposure to foreign markets, and the more pronounced the style tilts, the greater the divergence.
Recent experience has demonstrated just how great this tracking error can be. During the period from 1995 to 1999 the global, small and value exposure caused significant underperformance relative to the S&P 500 (and various tech sectors). However, beginning in March 2000, the situation turned dramatically. And, throughout the entire period, the more diversified portfolio is the clear winner.
This divergence may create both a behavior problem for the investor and business risk for the investment advisor. Many investors will willingly tolerate long term systematic underperformance as long as their accounts look generally like everybody else’s. However, they become very uncomfortable if their performance doesn’t generally track their friends’.
In the long run, these diversified portfolios with appropriate style tilts will add value to the holder. But, only if held for the long run. In the short run, unless the reasons for any divergence from the S&P 500 are fully understood and anticipated, the investor may bail out during periods of under performance. Discipline and a long term perspective are essential keys to investing success, but difficult to maintain during periods when these styles are out of favor.
If investment advisors believe that they will be retained or fired based on a comparison to the S&P 500, their willingness to create alternative superior investment strategies vaporizes. What they will design and deliver are market weighted portfolios with token positions designed to give the impression of asset class diversification. The lesson to investment advisors is clear: look like everybody else, or run the risk of getting fired.
The only rational way to think about performance is relative to the investment policy. First and foremost, the investment policy must be the most appropriate for the client’s situation, goals and risk tolerance. If the investment policy isn’t right for the client’s long term needs, then minor variations in performance of asset classes are not relevant in any case.
Presuming that the investment policy statement is correct, then the next logical question is: “Is the account capturing the risk characteristics it was designed for? ” In our case, the investment policy calls for an equity global market basket tilted strongly towards small and value in every market balanced with a high quality short term bond portfolio. For good or evil, that’s the investment performance we should expect.
Beating the market is very difficult to do consistently (read that as just about impossible) and caries high cost and high risk. As passive investors, we have abandoned the Performance Fairy with his siren call of “beating the market” in favor of a more rational objective of capturing market returns and risk dimensions at the least cost, least risk position.
There remains the thorny problem of tracking error to the individual benchmarks within the asset classes. The DFA portfolios are constructed to passively capture dimensions of risk, which is a slightly different concept than acting as a true “index fund”. The problem with index funds is that an enormous amount of trading is required to fully replicate the index. So, DFA passively engineers portfolios in each market to captures these risk elements at lowest cost. However, anything less than full replication introduces some tracking error, especially where funds have unusually heavy cash flows relative to the index. This became an issue when massive cash flows poured into the new series of tax managed funds while the older series of companion funds had relatively little cash flows. In the short run these tracing errors will be totally random. In the longer run these tracking errors should prove minimal. We believe that the cost savings will more than compensate for tracking error.
In some cases DFA created asset classes to capture the benefits that academic research discovered rather than replicating an existing index. So, some of the DFA funds do not have widely published indexes to compare to. For instance, there is no generally recognized International Small Company or International Small Value index. However, the academic rationale for capturing the small and value premium on a global basis is sound, and the diversification benefit is significant. Again, the emphasis is on engineering portfolios to capture specified risk dimensions at the lowest cost rather than slavishly following an index.
Finally, even within passive strategies trading practices may impact performance. For example, the Micro-cap consistently outperforms it’s index due to DFA’s ability to opportunistically purchase large blocks of stocks below market cost. Likewise, the passive variable maturity strategy employed by the DFA short term bond funds has added value over a true index even though it makes no forecast of interest rate directions.
Investors and advisor have a shared responsibility. Investment advisors must clearly explain the benefits and limitations of their investment policy. Investors must understand and endorse those benefits and policy limitations and be prepared to endure the certain periods of relative underperformance in order to achieve the anticipated long term benefits.
If some investors still believe that they must track their neighbor’s portfolios even if they are inferior investment policies, then they should reduce or eliminate any portfolio tilt toward small, value and international.
In the simplest possible case, if your equity investments are solely inside the united states, and only large capitalization companies with no distinct style tilt, then the S&P 500 might be an appropriate benchmark. In this very simplistic case, if you are using an active manager, then measurement against the S&P 500 makes sense. After all, it is presumed that no one would hire an active manager unless he believed that the manager could outperform the appropriate index.
However, in a more sophisticated asset allocation plan with multiple asset classes and a distinct style tilt, comparison to the S&P 500 is useless. Where the portfolio is built using index or passive funds, each asset class is it’s own benchmark, and we would expect them to reasonably faithfully capture that benchmark. Industry weightings, style drift, average size, etc. are not likely to be an issue in a passive or index fund. Portfolio returns will be the blended index return minus costs, plus or minus a small tracking error. It’s highly unlikely that a simplistic benchmark like S&P 500, EAFE, Russell 5000 or any other benchmark that is published in the newspaper every day will describe your investment policy.
If you are obsessed with comparing a sophisticated multi asset class portfolio to some benchmark, your only solution is to perform a regression analysis against multiple global benchmarks weighted exactly like your Investment Policy Statement. In other words the investment policy statement becomes the benchmark, and the portfolio should very closely track it. For all practical purposes, the portfolio is its own benchmark.
So, when we ask how our investments are doing, the answer is more complex than just a number plus or minus the S&P 500. The only relevant comparison is to the investment policy statement.
In our case, the equity benchmark is the global portfolio overweighted in small and value. Because of the style tilt, our estimate for the average additional return over the global market index will be about 2.5% per year. We believe — based on academic research and past performance — that this is a superior equity portfolio for the long haul. If you understand and endorse the underlying assumptions that additional premium should be more than adequate incentive to live with tracking error relative to the S&P 500.