Expected return is shown on the vertical axis, and expected volatility (standard deviation of returns) is displayed on the horizontal axis. Assets lying to the left and higher on the graph offer a superior tradeoff between expected return and expected volatility. The efficient portfolios (those combinations of assets that provide the highest return for exposure to a given level of risk) are shown as grey dots, with the dashed line sketching out the efficient frontier, illustrating the typical return premium; that is, buying portfolios with higher volatility provides higher expected returns. Clicking on the portfolios on the efficient frontier brings up a summary card with the portfolio’s return and risk characteristics. For example, based on data as of August 31, 2017,3 the 10.0% volatility portfolio shows a 5.5% expected return with a 10.2% standard deviation.4
Of course, very few clients understand the concept of standard deviation, so it’s useful to transform volatility, or standard deviation of returns, into a discussion tool to convey how widely shorter-term returns can vary around long-term expectations. Under normal distributions, investors can expect to receive a return within one standard deviation of the mean approximately 2 years out of 3, and to receive a return more than two standard deviations beyond the mean about once every 20 years.5 Because market returns have “fat tails,” crashes occur more frequently than expected and in more dramatic fashion than would be predicted by a normal distribution.6
For a portfolio with a 5% expected return and 10% volatility, we should explain to our client that they will likely receive a return worse than −5% every 6 years or so, and at some point during their investing lifespan, an annual return worse than −15%—and even potentially much worse! Suppose your 35-year-old client Nancy has $100,000 to invest. Although she believes she has a high risk tolerance, she draws the line at losing more than $10,000 in any given year. Should Nancy invest in the 10% volatility portfolio? This would not be a good idea because at some point over the next 40 years (a realistic life span for her), Nancy should expect to lose more than that amount in a single year.
Clients should not just be comfortable with the possibility—but should expect—that losses of this magnitude will occur, and should be (pleasantly) surprised if they do not. We are continuously amazed at how frequently investors are “shocked” by negative returns that fall well within the normal distribution of outcomes.
Tracking Error: Being Wrong and Alone
Most risk discussions between financial advisors and their clients stop after the analysis of volatility. But this misses a crucial element in how human beings think about risk. Much of our happiness is centered not on absolute success, but rather our position relative to our neighbor’s success.7 The risk of being wrong, and alone, can be quantified by tracking error, which is a benchmark-sensitive measure of risk: How widely does a portfolio’s return vary around the return of its specified benchmark? A portfolio that consistently delivers a return close to its benchmark’s return generates low tracking error, and a portfolio producing outcomes falling far from those of the benchmark have high tracking error.
This concept, a form of maverick risk, is driven by the psychological tendency toward “herd behavior.” Humans have an atavistic tendency to stick with their herd, seeking safety in numbers. Early humans who became separated from their tribes, or animals separated from their pack, often met with early death. In today’s world, this evolutionary tendency manifests as a natural discomfort when we receive different outcomes from our peers, which modern society has described through the recently coined term “Fear of Missing Out” (FOMO).
This fear can be particularly powerful in the financial markets. Just ask anyone who has hesitantly watched as their friends and neighbors got rich investing in Chinese internet stocks, or crypto-currencies, or flipping real estate in 2007, or riding the dot-com bubble in 1999. We can also look to no less a luminary than Sir Isaac Newton, who fell prey to FOMO long before it earned that acronym. The South Sea Company garnered a monopoly from the English government on trading with the emerging markets of the day in the South Seas. In early 1720, a bubble began to inflate in the company’s stock. Sir Isaac, an early investor, quickly doubled his initial investment over the course of several months, and exited happily. His resounding success quickly turned dark, however, as he watched the stock price of the South Sea Company double and triple from his exit point. His friends who remained invested reaped even greater rewards, no doubt regaling their good fortune to the crowded cocktail parties of the day. What had been a great success for Sir Isaac turned into a painful failure. Sir Isaac, unable to bear the psychological damage inflicted by his friends’ good fortune, bought back into the South Sea Company near the height of the market and subsequently lost nearly his entire life savings. This brought him to utter the alleged lament, “I can calculate the movement of heavenly bodies, but not the madness of men.”
Maverick risk, then, presents opportunities to investors who are willing to bear the discomfort of acting differently from their peers. This can take the form of shunning popular, but expensive, investments as Sir Isaac tried, but failed, to do, or by investing in unloved, discounted assets that may be unpopular, but often because of that unpopularity are priced to deliver attractive expected returns.
Take the example of Warren Buffett. In addition to having a keen sense for determining the fair value of corporate and financial assets, Buffett also benefits from a willingness to take maverick risk, along with secure control of capital. His track record as an investor is very well known; the return on Berkshire Hathaway Class A shares since 1980 has been an annualized 20.4%, far in excess of the S&P 500 Index return of 8.8% for the same period. This return has, however, been generated with a tremendous tracking error of 22%, meaning that roughly one-third of the time Buffett’s returns have differed from the S&P 500 by more than 22%! Clearly, many of these instances have been on the upside, but Buffett has also held on steadfastly through many down periods.
From mid-1998 until March 2000, Berkshire Hathaway underperformed the market by 54% (close to three standard deviations). How many clients would be able to stomach those extended losses without throwing in the towel? Barron’s questioned “What’s Wrong, Warren?” and warned “Warren Buffet may be losing his magic touch” (Bary, 1999). But Buffett has never been one to let short-term market movements change his convictions, proclaiming that “If you aren’t willing to own a stock for 10 years, don’t even think about owning it for 10 minutes.” Buffett stuck to his maverick philosophy of investing, and by July 2002 had recovered all the shortfall and more.
Commensurate with Buffett’s willingness to be different is the security of his capital, based on collected insurance premiums that must remain invested. Advisors who recommend portfolios and asset allocations with higher levels of tracking error must ensure that their clients fully understand the nature and extent of this risk and not fall prey to the temptation to pull out of their investments during the inevitable drawdowns when compared to the “herd portfolio.” As Arnott (2003) notes:
No decisions are infallible. Decisions that leave an investor alone carry the inherent risk of being both wrong and alone. If an investor is wrong and alone, a strong likelihood is that the assets’ owner will not have the patience to see the investment decision through. The decision, even if correct in the long run, will be reversed before it can succeed.
A quick look at the Asset Allocation Interactive tool helps frame the decisions an investor needs to make.