Poor long-horizon investment outcomes invariably can be blamed on a handful of “usual suspects”: high fees and expenses; ignoring rebalancing opportunities; performance chasing; lack of diversification; and improper expectations. Rarely do these usual suspects operate as lone wolves. They typically travel together.
The optimist in us believes that—slowly but surely—times are getting tougher for these culprits. Expense ratios are coming down; no doubt helped substantially by more wide-scale adoption of passive equity investing. Similarly, smart beta is on the march, a more effective means of capturing long-term excess returns than traditional active equity management. Investors are also becoming more aware of the softer, hidden costs of trading and market impact expenses (Chow et al., 2017). Automated rebalancing is now commonplace. Kinnel (2017) and others have shown how fund investors can be on the wrong side of mean reversion by chasing performance. And accessing diversifying assets has never been easier than it is today with the growth of reasonably priced mutual funds and exchange-traded funds (ETFs) offering investment opportunities ranging from emerging-market currencies to commodities to bank loans.
But what about the last suspect: setting improper return expectations? This culprit hasn’t received as much attention as the others, reminding us of a favorite all-time film, 1995’s The Usual Suspects. In the movie, a customs agent, Dave Kujan, tries to figure out which of five criminals is responsible for a drug deal gone bad on a ship docked in San Pedro Bay. Agent Kujan interviews the sole survivor of the crime, Verbal Kint (played by Kevin Spacey), who is intent on pinning the blame on one of the other suspects. Nobody at the police station suspects Verbal Kint, who has a palsied arm and foot—in their view, an unlikely suspect—was responsible. But the ending (spoiler alert!) reveals that not only is Verbal responsible, he is actually Keyser Söze, the “world’s most notorious” criminal. Of course by then, Kint, er Söze, had walked out of the interrogation uncharged, never to be seen again. The movie ends with a quote from Söze: “The greatest trick the devil ever pulled was convincing the world he didn’t exist.”
As we survey the range of long-term return forecasts in the industry, we are shocked to see so many—including some (deservingly) well-respected asset managers—seemingly ignoring today’s low yields (and their downward pressure on future returns). Instead, many continue to use historical returns to forecast the future, one of the most common shortcuts in financial planning, and one we believe will not serve investors or advisors well.
How can advisors avoid this dangerous and common practice? Research Affiliates’ Asset Allocation Interactive tool illustrates the likely sources of future returns and provides an easily accessible way for proactive, client-oriented advisors to educate clients on probable future financial outcomes. In a subsequent article, we will show how the allocations in today’s normal portfolio can be shifted to offer more attractive long-term risk-adjusted returns.
Setting Investment Objectives: What We Can and Can’t Control
Bear with us as we do a 30-second review of Actuarial Science 101. Conceptually, retirement math has five variables: 1) annual savings rate, 2) return on savings, 3) retirement date, 4) retirement spending rate, and 5) life expectancy. The savings rate, retirement date, and retirement spending are individual choices, and as such, are entirely predictable. Clients make these choices themselves; some might prefer to work longer in order to spend more in their retirement years, or vice versa.
Sadly, we have little control over our lifespan, and practically speaking, little control over the ultimate return on our portfolios. Sure, we can try to influence these variables by making smarter choices—eating right and exercising will lead to a longer life, just as minimizing unnecessary investment expenses (including trading costs) and swearing off performance chasing should lead to higher net returns. At the end of the day, we just don’t have the control we’d like to have. Instead, we’re left with an uncomfortably wide range of expectations. Today, a 65-year-old male has a life expectancy of 19.3 years (i.e., will live to 84.3 years).1 Of course, that’s the average experience. Some men will live far longer, and some will sadly pass while still in their sixties.
Not only do we have to deal with a range of possible life expectancies, but a changing average life expectancy over time. A 65-year-old male in 1980 would have been expected to live until 79.1 years, a five-year shorter period of retirement bliss than he would be looking forward to today. The combined positive effects of advanced diagnosis, less smoking, more nutritious diets, better safety in occupational workplaces, vaccines and antibiotics, and seatbelt laws, among other factors, equate to a longer life expectancy. Fairly indisputable. Obviously, it doesn’t make sense to use a mortality table from 1980 to determine a client’s most likely retirement horizon in 2017 (37 years later).
Starting conditions matter in life expectancy, and as we’ll see shortly, also in expected returns. Advisors intent on producing better outcomes for their clients need to communicate realistic return expectations not only for the destination (investment-horizon retirement savings), but for the duration of the journey. Our Asset Allocation Interactive tool is designed for this purpose.
The Inevitable Delta: Past Realized Returns vs. Reasonable Future Expected Returns
Investor behavior indicates a profound belief that past is prologue. Any time an investor replaces a bottom-quartile performer with a top-quartile performer, by definition the investor is extrapolating the past into the future. The same can be said when an investor swaps out a recent laggard in asset class or style with a recent winner. A quick survey of the behavioral finance literature shows this tendency, a form of the “representativeness heuristic” (Kahneman and Tversky, 1972), is pervasive across a wide range of investment decision making. Debondt (1993) shows experimental and survey evidence that investors are subject to using representativeness to over-extrapolate recent price trends when forecasting future returns. Along with the representativeness heuristic, investors tend to use other related rule-of-thumb shortcuts, such as availability bias (Tversky and Kahneman, 1973), a heuristic (a mental shortcut that allows people to solve problems and make judgments quickly and efficiently) that leads us to favor what is most recent or relevant.
Unfortunately, these biases can be unhelpful or even dangerous in setting realistic return expectations. For example, past long-term returns of US equities are negatively correlated with their future long-term returns, whether we use a horizon of 10, 20, or 30 years. In other words, your clients’ experience in the capital markets (i.e., relatively high past returns), irrespective of the client’s age, is unlikely to be duplicated in the future—both the representativeness and availability heuristics lead us down the wrong path!