- Our central philosophy is that the largest and most persistent active investment opportunity arises from long-horizon mean reversion.
- We have three core beliefs: investor preferences are broader than risk and return; prices vary around fair value; and a lack of conviction prevents investors from exploiting long-term value.
- The long-term capital market expectations presented on our Asset Allocation website represent our best forecasts and inform our behavior within the context of our investment beliefs.
- The beliefs underlying Research Affiliates’ asset allocation models and investment strategies are admittedly fallible but have proven useful in the past and should serve well in the future.
Mimicking the herd invites regression to the mean (merely average performance).
The public launch of Research Affiliates’ interactive Asset Allocation website1 this month gives us an opportunity to describe the investment beliefs underlying our models, expected returns, and investment strategies. To be clear, our beliefs are constructs that help us make sense of the capital markets. Though grounded in theory and supported by empirical examination, they are not facts scientifically proven beyond a shadow of a doubt. Authentically admitting the fallibility of one’s beliefs can be uncomfortable, especially in an industry all too often predicated on projecting unshakable confidence in predicting the future, but not to us. We prefer the honest acceptance that we (and anyone else we’ve come across so far) do not have perfect knowledge or information. For those willing to join us and embrace the discomfort, our beliefs draw a clear course of action that has served us well in the past and should continue to serve us well in the future.
Before describing each one of our core beliefs in their logical order, let us start at the end with our central investment philosophy:
The largest and most persistent active investment opportunity is long-horizon mean reversion.
Quite simply, systematically buying assets with relatively low prices (and correspondingly high yields) and having the conviction to hold these positions over potentially extended horizons allows investors to earn higher risk-adjusted returns, albeit at significant emotional cost. This central investment philosophy is the logical conclusion sitting atop three supporting beliefs.
Investor preferences are broader than risk and return.
The human psyche is rich and complex in ways well beyond those captured in standard finance textbooks. Yes, investors may weigh potential returns against risk. However, our experience teaches us that investors’ actions are also affected by the desired safety of following the herd and the corresponding fear of being an outsider; the pride and status derived from owning stocks of successful firms and avoiding association with assets that receive negative press; the allure of gambling and positively skewed distributions; the heightened aversion to losses relative to gains; and so forth. We readily acknowledge the validity of these preferences, rather than labeling them irrational, thus moving past the stale semantic debate about market efficiency.2 Our collective decades in investment management have provided us plenty of opportunities to observe these very human tendencies, and many of them are also now documented in the published works of behavioral economists.3
Prices vary around fair value.
In the short run, the market is a voting machine, but in the long run it is a weighing machine.
—Attributed to Benjamin Graham
Because of the complex and time-varying preferences of investors, at any given point in time, all asset prices are not unbiased estimates of fair value. Instead, prices can, and do, materially deviate from reasonable and transparent estimates of intrinsic value, often for prolonged periods. That is what we mean when we say there is noise in prices.4 But the noise we observe is not white noise, a neutral blanket containing no information. Quite the opposite: Indeed, we know far more than just current asset prices. We also know the path that prices have taken to get to their present levels. Based on this information, we can gauge which assets are likely overpriced and which are likely underpriced at the present time. Over longer periods, we expect prices to eventually revert toward estimates of intrinsic value, even as new errors in either direction are introduced. This mean-reverting process tilts the scale in favor of buying low and selling high, or at least buying low and selling at fair value!
What support can we provide for this belief? And what are the magnitudes at play, in terms of investment performance? In Figure 1, we reproduced De Bondt and Thaler’s (1985) analysis on the future performance of recent losers and winners in equities markets. This is one of the most direct tests of noise and mean-reversion. It also happens to show dramatic results. Sorting stocks on their most recent three-year performance (and labeling them losers and winners accordingly), we see that, on average, the loser portfolio outperformed the winner portfolio by nearly 25% over the following three years. Outperformance of this magnitude is not merely the cartoonist’s pennies in front of a steamroller, or even the economist’s $20 bills on the sidewalk. It is very large indeed.