Our search for “a better way” following the Tech Bust at the turn of the millennium concluded with a decision to build our own risk-managed tactical investment program. To many advisors who grew up on a more passive, buy-and-hold approach informed by Modern Portfolio Theory, that conclusion amounts to sacrilege.
Or does it? A careful re-examination of Modern Portfolio Theory tell us that a more tactical approach to investment management is exactly what was originally envisioned. Moreover, our track-record over the past decade seems to substantiate its effectiveness. Our own Michael Kitces has done considerable work on this subject (see his blog titled The Rise of Tactical Asset Allocation which was originally published in the Journal of Financial Planning) and has been re-educating the wealth management business on the original message of Modern Portfolio Theory. We have borrowed some of his work today to support our conclusion. But nothing replaces Michael. We recommend that you attend his presentation if given the opportunity. He never disappoints.
What’s the Point of Modern Portfolio Theory?
Harry Markowitz developed Modern Portfolio Theory to explain how to identify the best allocation of a diversified multi-asset class portfolio. In short, he decided that there were multiple portfolios along an efficient frontier that represent the most efficient portfolio for a given level of risk. In other words, these portfolios should be allocated to achieve the maximum return for a given amount of risk. These efficient portfolios were to be constructed based on the optimum mix of return, volatility and correlation.
Implementation of Modern Portfolio Theory
In common practice, the wealth management industry has chosen to use historical means to serve as inputs: historical mean returns, historical mean volatility and historical mean correlation. Why not? We know that human nature changes very little over time and that the business cycle repeats over time as a consequence. Reversion to the mean is practical, feels scientific and avoids the issues associated with human emotion and heuristics.
Using historical mean figures for the inputs results in a relatively static efficient frontier as historical means change very slowly over time. Wealth managers can therefore select a handful of efficient portfolios that represent a cross section of client risk profiles. They can match an efficient portfolio to a client’s risk profile, set-it and essentially forget it – but for periodic rebalancing. During the golden years of wealth management (see The Golden Era of the Financial Advisor Has Come to an End), this investment strategy worked in spades for both clients and advisor practices. It marketed well due to its Nobel laureate pedigree, it worked well due to the secular bull market where asset prices only went up and it facilitated the practice of the solo practitioner because it required a minimal amount of investment work. Everything was good.
So What’s the Problem?
The problem is that without the tailwind of a secular bull market, a more passive, buy-and-hold approach to investment management has been considerably less effective. Indeed, its weak performance during the tech bust is what led to our own conversation about its effectiveness and our own search for a better alternative.
One of the first things we did was revisit our core assumptions. We took another look at Modern Portfolio Theory and, to our surprise, we learned that the common approach to strategic asset allocation was not as consistent with Harry Markowitz’s vision as we thought. Indeed, he did not recommend the use of mean return, mean variance and mean correlation as inputs. Quite the contrary. He recommended expected returns, expected volatilities and expected correlations. Exactly what common wisdom warned the industry against.
In his own words, Markowitz said “to use the E-V rule in the selection of securities we must have procedures for finding reasonable [estimates of expected return and volatility]. These procedures, I believe, should combine statistical techniques and the judgment of practical men. My feeling is that the statistical computations should be used to arrive at a tentative set of [mean and volatility]. Judgment should then be used in increasing or decreasing some of these [mean and volatility inputs] on the basis of factors or nuances not taken into account by the formal computations”.
Furthermore he said, “one suggestion as to tentative [mean and volatility] is to use the observed [mean and volatility] for some period of the past. I believe that better methods, which take into account more information, can be found.”
Do Expected Inputs Produce Better Results?
In our experience … Yes! We have undertaken the challenge to ascertain expected returns, expected volatilities and expected correlations for more than a decade. Do we make mistakes? Absolutely. That is par for the course. But results have been better for our clients and our business. We now offer a 10-year plus GIPS compliant and third party audited track record of beating the benchmark with less risk, net of all investment expenses and our management fee.
As we started out, we did not know that was possible. Research completed since then has confirmed that others have shared our success with a more active, risk-managed approach. This will be the topic of our next chapter.
Working through the Alternatives
Fortunately, we found the solution most appropriate for our practice. Our answer may or may not suit your practice. But the decision is very much the same. We have therefore decided to share our journey with you in this series of blog posts in the hopes that it helps clarify your own decision process. Then we plan to work through another series of “Due Diligence” posts that highlight some of the things you should be looking for (good and bad) when selecting a final solution.
The slower summer months are an ideal time to work through these types of decisions/projects. We hope you find this new series helpful. And, of course, you should always feel free to call. We are happy to help.