Swedroe: Investor Mistakes Writ Large

Pension funds often make the same common mistakes as individual investors.

Pension funds often make the same common mistakes as individual investors.

Extrapolating the recent past into an indicator of an investment’s future performance is one of the strongest investor biases documented in behavioral finance research. This research has found that the related mistakes of “recency” and “herding” often lead individual investors to buy yesterday’s winners and sell yesterday’s losers. By doing so, these investors act as if they can purchase yesterday’s returns, when, in fact, it is possible only to purchase tomorrow’s.

Evidence of this bias has shown up in studies demonstrating that the returns earned by individual investors tend to be lower than the returns earned by the very investments they make. My colleague, Carl Richards, called this “The Behavior Gap.”

The gap between investor returns and investment returns appears whether we are talking about asset classes, strategy styles, single stocks or funds. It’s generally thought that pension plans, as opposed to individual investors, are highly disciplined in their investment process. An example of disciplined investing would be to establish an asset allocation plan and then adhere to that plan through rebalancing.

Pension Funds Make Mistakes, Too

However, Andrew Ang, Amit Goyal and Antti Ilmanen—the authors of the 2014 study “Asset Allocation and Bad Habits”—found that pension plans’ reputation for strict discipline is “a gross simplification of industry practices.” Their study, which covers the 22-year period from 1990 through 2011, uses data from CEM Benchmarking. CEM Benchmarking collects pension fund information through yearly questionnaires, and its data represents the broadest coverage in North America.

The study includes 573 U.S. pension funds, with median assets near $3 billion and average assets near $10 billion. These funds hold 30 to 40 percent of total assets under management by U.S. pension funds.

The authors found that, on average, pension plans do let their allocations drift with relative asset class performance. For example, they found that when averaged across all funds, policy weights—or strategic target asset class allocations—were 57 percent for equities (equally weighted) over the entire 22-year period from 1990 through 2011.

However, following the bull market of the late 1990s, policy weights for equities rose from 54 percent to a peak of 61 percent from 1999 through 2001. And following the financial crisis of 2008, the allocation to equities had fallen back down to just 46 percent by 2011.

There is strong evidence of short-term momentum (measured in months, not years) in equities. But over the longer term, reversion to the mean is much more likely. Unfortunately, it appears that pension plan sponsors are subject to the same mistakes as individual investors. They tend to buy the last three to five years’ winners, and sell multiyear laggards.

Given that pension plan sponsors are likely more aware of the evidence, you would hope they could avoid the mistakes common to many individual investors. However, plan sponsors are still human beings, and thus subject to the same behavioral errors as individuals. And it appears that behavioral biases are hard to overcome on any level. Being aware of such biases is the first step on the road to overcoming them.