However, to ride this second wave, I invite readers to be grateful that Larry and Kevin are their guides.
As Larry and Kevin explain, the investable universe is much, much larger than what most investors realize. For example, there are countless sources of returns that have been available within financial institutions, whose lending, trading and underwriting activity has generated a steady stream of profits over centuries. A wide range of these exposures once considered “un-investable,” or, more fundamentally, not even thought of as sources of returns, are now increasingly available through new securities and funds.
Larry and Kevin guide the way with an engaging and actionable tour of alternative sources of risk premiums that are intuitive, pervasive and persistent—and their book is quite specific in exactly how, and how much, of this risk to take. Most importantly, they frame the landscape in the language of an investor: Will making changes to my portfolio increase my probability of being able to live the life I want, maybe even discover a life I didn’t know was available, with the peace of mind that comes from knowing my tail risk is reduced?
Larry and Kevin show us that, starkly, investors now have a choice to make. Do they comfortably fail or uncomfortably succeed? Failing comfortably means investing largely as before; succeeding uncomfortably means adopting a new mindset and fundamentally altering what portfolios look like. Making this change implies turning the traditional portfolio inside-out, where core holdings become the satellites and a new breed of alternative risk premiums become the core. This is going to be a huge change, and many investors won’t get there overnight. But the path forward has been illuminated, and the direction is clear.
Finally, a personal request: Before you read this book, please pause for a few minutes and take a quiet moment alone. Think about some of the things you have in your life that make you smile: family, friends, health, work. I find that a mental framework built around wanting what you already have is a better strategy for happiness than having what you want. Happiness is something we get to decide on in advance. After your pause, please then read the last appendix in this book, titled “Enough.” It captures what this book is really about, and shares Larry and Kevin’s deepest wisdom and most important messages. It’s one of the best pieces of financial writing I’ve read anywhere, ever. “Enough” also captures why folks like Larry, Kevin and I—who devote their lives to helping others achieve peace of mind through financial security—run smiling to work in the morning. We get to do what we love.
Enjoy!
Ross L. Stevens
Founder, Stone Ridge
January 2018
1 Duration is a measure of the sensitivity of a bond’s price to changes in interest rates, expressed as a number of years.
Introduction
This book was written for those looking to expand their knowledge of the evidence-based investing world. In this world, evidence and peer-reviewed academic research, not instinct, opinion, or ego, are used to design portfolios. Whether you are an advisor looking to better serve your clients, an investor looking to become more knowledgeable about the workings of your portfolio, or even a financial oracle, you will benefit from this book. While it is short in length, its content is heavy. It is data-rich and full of detailed examples. The empty rhetoric or the distracting noise often heard in the active investment world has no place here. Science and hard data make our case. There is no need for elaborate prose or the hyperbolic statements so frequently heard on the other side of the investing aisle.
You are about to embark on a journey that we hope will be both informative and of great value. It is a roadmap to the holy grail of investing—an investment strategy that can deliver higher returns without increased risk, or the same return with reduced risk. Consider your search ended.
Finally, if you tend to be daunted by data or unfamiliar terms, don’t worry. We have made great efforts to explain concepts in as simple terms as possible. Take your time, and know that by reading this book, you are taking steps toward becoming a better, more informed, hands-on investor. And that is something of which you can be proud.
Part I
* * *
Using the Science of Investing to Build More Efficient Portfolios
Chapter 1:
How to Think About Expected Stock Returns
An important part of the process of developing an investment plan is estimating future returns to stocks and bonds. Unfortunately, many investors make two big mistakes when doing so. The first is to simply extrapolate past returns into the future. This is a mistake because it ignores the fact that current stock valuations play a very important role in determining future returns. Consider the following.
From 1926 through 1979, the S&P 500 Index returned 9.0 percent. From 1980 through 1999, it returned 17.9 percent, raising the return over the full period by 2.3 percentage points to 11.3 percent. Investors in 1999 using the historical return of 11.3 percent as a predictor of future returns were highly likely to be disappointed because they failed to take into account the fact that the earnings yield (the inverse of the price/earnings ratio) had fallen all the way from 11.4 percent to 3.5 percent. And lower earnings yields predict lower, not higher, future returns.
While the historical real return to stocks from 1926 through 2016 has been 6.9 percent (9.8 percent nominal return minus 2.9 percent inflation), most financial economists are now forecasting real returns well below that level. No metric for estimating future returns is generally agreed upon as the best, but the Shiller CAPE 10 (cyclically adjusted price-to-earnings) ratio is considered by many to be at least as good as, if not better than, others because it explains a significant portion (about 40 percent) of the variation in future returns. The CAPE 10 ratio uses smoothed real earnings over the prior 10 years to eliminate fluctuations in net income caused by variations in profit margins over a typical business cycle.
The first to argue for smoothing a firm’s earnings over a longer term were value investors Benjamin Graham and David Dodd. In their classic text, Security Analysis, Graham and Dodd noted one-year earnings were too volatile to offer a good idea of a firm’s true earning power. Decades later, Yale economist and Nobel Prize winner Robert Shiller popularized the 10-year version of Graham and Dodd’s price-to-earnings (P/E) measure as a way to value the stock market.
In an attempt to minimize the impact of what might be temporarily very low earnings (due to a recession) or very high earnings (due to a boom), the Shiller CAPE 10 smoothes out earnings by taking the average of the last 10 years’ earnings and adjusts that figure for inflation. Let us assume that the Shiller CAPE 10 is at 30 (which it was as we wrote this), well above its historical average. To estimate future returns using this metric, you take the earnings yield—the inverse of the Shiller CAPE 10 ratio—and you get 3.3 percent. However, because the Shiller P/E is based on the lagged 10-year earnings, we need to make an adjustment for the historical growth in real earnings, which is about 1.5 percent per year. To make that adjustment, we then multiply the 3.3 percent earnings yield by 1.075 (.015 x 5), producing an estimated real return to stocks of about 3.5 percent, or 3.4 percentage points below the historical return. (We multiply by five because a 10-year average figure lags current earnings by five years.) Using other methodologies (such as what is called the Gordon Constant Growth Dividend Discount Model) deliver similar results, with most financial economists forecasting real future returns in the range of about 4 to 5 percent.
The second mistake investors make is to treat the expected return as “deterministic”—meaning they believe they will earn that specific return—rather than as just the mean of a potentially very wide dispersion of possible returns. The following illustration demonstrates why thinking of the expected return in a deterministic way is dangerous.
In a November 2012 paper, “An Old Friend: The Stock Market’s Shiller PE,” Cliff Asness of AQR Capital Management found that the Shiller CAPE 10 does provide valuable information. Specifically, he found that 10-year forward average real
returns fall nearly monotonically as starting Shiller P/Es increase. He also found that, as the starting Shiller CAPE 10 increased, worst cases became worse and best cases became weaker. And he found that while the metric provided valuable insights, there were still very wide dispersions of returns. For example:
When the CAPE 10 was below 9.6, 10-year forward real returns averaged 10.3 percent. In relative terms, that is more than 50 percent above the historical average of 6.8 percent (9.8 percent nominal return less 3.0 percent inflation). The best 10-year forward real return was 17.5 percent. The worst was still a pretty good 4.8 percent 10-year forward real return, just 2.0 percentage points below the average, and 29 percent below it in relative terms. The range between the best and worst outcomes was a 12.7 percentage point difference in real returns.
When the CAPE 10 was between 15.7 and 17.3 (about its long-term average of 16.5), the 10-year forward real return averaged 5.6 percent. The best and worst 10-year forward real returns were 15.1 percent and 2.3 percent, respectively. The range between the best and worst outcomes was a 12.8 percentage point difference in real returns.
When the CAPE 10 was between 21.1 and 25.1, the 10-year forward real return averaged just 0.9 percent. The best 10-year forward real return was still 8.3 percent, above the historical average of 6.8 percent. However, the worst 10-year forward real return was now -4.4 percent. The range between the best and worst outcomes was a difference of 12.7 percentage points in real terms.
When the CAPE 10 was above 25.1, the real return over the following 10 years averaged just 0.5 percent—virtually the same as the long-term real return on the risk-free benchmark, one-month Treasury bills. The best 10-year forward real return was 6.3 percent, just 0.5 percentage points below the historical average. But the worst 10-year forward real return was now -6.1 percent. The range between the best and worst outcomes was a difference of 12.4 percentage points in real terms.
What can we learn from the preceding data? First, starting valuations clearly matter, and they matter a lot. Higher starting values mean that future expected returns are lower, and vice versa. However, a wide dispersion of potential outcomes, for which we must prepare when developing an investment plan, still exists.
The following illustration shows the right way to think about the expected return of a portfolio or an asset class. Although stock returns do not fit exactly into a normal distribution (as the following bell curve depicts), a normal distribution is a close approximation. Thus, we think this graph will be helpful in explaining how to think about expected returns.
In the illustration, think of Portfolio A as a market-like portfolio (such as the Vanguard Total Stock Market Index Fund). Using the 3.5 percent expected real return to stocks (based on the Shiller CAPE 10) and an expected inflation rate of 2.0 percent, we arrive at an expected nominal return of 5.5 percent for the overall stock market. The right way to think about this 5.5 percent figure is as the mean (and median) of the wide dispersion depicted. In other words, there is a 50 percent chance the actual return will be greater than the expected 5.5 percent, perhaps a 30 percent chance it will be greater than 7 percent, a 10 percent chance it will be greater than 8 percent, and a 5 percent chance it will be greater than 10 percent. The possibilities are similar that it will fall on the left side of the distribution with returns below, and even well below, the expected rate of 5.5 percent.
Now consider Portfolio B, which has the same 5.5 percent expected return but a different potential dispersion of returns. As the following illustration shows, more of the weight of the distribution (its probability density) is closer to the mean expected return of 5.5 percent than is the case with Portfolio A. It is a taller and thinner bell curve, with less of its weight in the tails, both left (bad) and right (good).
Now consider both Portfolios A and B. They have the same expected return—in both cases the mean return is 5.5 percent. However, they have a different dispersion of returns. If you were faced with the choice of living with the risks of the potential return dispersions in either Portfolio A or B, which would you choose?
If you are like most people, you would choose to live with the risks of Portfolio B. Most investors are risk averse—given the same expected return, they choose the portfolio with the lower standard deviation of returns (Portfolio B). Said another way, if you are like most investors, you are willing to sacrifice the opportunity to earn the great returns in the right tail of the distribution of Portfolio A (that are not there with Portfolio B) if you also minimize, or eliminate, the risk of the very bad returns in the left tail of Portfolio A (that, again, are not there with Portfolio B).
Are you interested in learning how to create portfolios where the distribution of potential returns looks more like Portfolio B than Portfolio A? Learning how we have been doing this for about 20 years for our clients is the journey we will embark on next, beginning with a history of modern financial theory and asset pricing models.
Chapter 2:
A Brief History of Modern Financial Theory
The birth of modern finance can be traced back to 1952, when Harry Markowitz’s paper “Portfolio Selection” was published in The Journal of Finance. The most important aspect of this work was that Markowitz showed it is not a security’s own risk and expected return that is important to an investor, but rather the contribution the security makes to the risk and expected return of the investor’s entire portfolio. This contribution depends not only on the riskiness of the security itself, but also on how the security behaves relative to the behavior of the other assets in the portfolio (the correlation of their returns). Markowitz was able to show that you could add a risky asset, with higher expected returns, to a portfolio without increasing the portfolio’s overall risk if the asset’s returns were not perfectly correlated with the other assets in the portfolio.
William Sharpe and John Lintner are typically given most of the credit for introducing the Capital Asset Pricing Model (CAPM). The CAPM was the first formal asset pricing model, and it was built on ideas from Markowitz’s paper. The CAPM provided the first precise definition of risk and how it drives expected returns.
The CAPM looks at returns through a “one-factor” lens, meaning the risk and return of a portfolio is determined only by its exposure to market beta. It is important to understand that market beta is not simply the stock allocation of a portfolio. It is the measure of the equity-type risk of a stock, mutual fund or portfolio relative to the risk of the overall market. An asset (or portfolio) with a market beta greater than 1 has more equity-type risk than the overall market. If it has a market beta less than 1, it has less equity-type risk than the overall market. Thus, a portfolio with a 70 percent allocation to stocks and 30 percent allocation to Treasury bills could have a market beta of 1 if the stocks in the portfolio were highly risky stocks with a market beta of about 1.4. For example, these stocks might be high-flying tech stocks. Conversely, a portfolio with a 100 percent allocation to stocks could have a market beta of just 0.7 if the stocks it held were all less risky than the market as a whole. Perhaps they are “defensive” stocks, such as utilities, drug store chains and supermarket chains.
The CAPM was the finance world’s operating model for about 30 years. However, all models by definition are flawed, or wrong. If they were perfectly correct they would be laws, like we have in physics. Over time, anomalies that violated the CAPM began to surface. Among the more prominent ones were:
1981: Rolf Banz’s “The Relationship Between Return and Market Value of Common Stocks” was published in The Journal of Financial Economics. Banz found that market beta does not fully explain the higher average return of small (or lower market capitalization) stocks.
1981: Sanjoy Basu’s “The Relationship Between Earnings’ Yield, Market Value and Return for NYSE Common Stocks,” published in The Journal of Financial Economics, found that the positive relationship between the earnings yield (the earnings/price ratio) and average return is left unexplained by market beta.
1985: Barr Ros
enburg, Kenneth Reid and Ronald Lanstein found a positive relationship between average stock return and the book-to-market ratio in their paper “Persuasive Evidence of Market Inefficiency,” published in The Journal of Portfolio Management.
1988: Laxmi Chand Bhandari’s “Debt/Equity Ratio and Expected Common Stock Returns: Empirical Evidence,” published in The Journal of Finance, found that firms with high leverage have higher average returns than firms with low leverage.
Eugene Fama and Kenneth French’s 1992 paper “The Cross-Section of Expected Stock Returns” summarized all of these anomalies in one place. The essential conclusions from their paper were that the CAPM only explained about two-thirds of the differences in returns of diversified portfolios, and that a better model could be built using more than just the single factor of market beta.
The Fama-French Three-Factor Model
One year later, Fama and French published “Common Risk Factors in the Returns on Stocks and Bonds” in The Journal of Financial Economics. This paper proposed a new asset pricing model, called the Fama-French Three-Factor model. This model proposes that, in addition to the market beta factor, exposure to the factors of size and value further explain the cross-section of expected stock returns. The essential takeaway from this research is that small-cap and value stocks are riskier than large-cap and growth stocks, and that risk is compensated for with higher expected returns.
The authors demonstrated that we lived not in a one-factor world, but in a three-factor world. In so doing, they showed how the risk and expected return of a portfolio is explained by not only its exposure to market beta, but also by its exposure to the size (small stocks) and price (stocks with low prices relative to book value, or value stocks) factors. Fama and French hypothesized that while small-cap and value stocks have higher market betas—more equity-type risk—they also contain additional, unique risks unrelated to market beta. Thus, small-cap and value stocks are riskier than large-cap and growth stocks, explaining their higher historical returns and implying that such stocks should have higher expected returns in the future. Studies have confirmed that the three-factor model explains an overwhelming majority of the returns of diversified portfolios. In fact, the Fama-French three-factor model improved the explanatory power from about two-thirds of the differences in returns between diversified portfolios to more than 90 percent. An indirect, but important, implication of this finding was that if more than 90 percent of a diversified portfolio’s returns could be explained by the portfolio’s exposure to these factors, there wasn’t much room left for active security selection or market timing to add value. This suggests, in turn, that passive, evidence-based investing is the strategy most likely to allow you to achieve your financial goals. Subsequent research on active managers’ performance and persistence of performance supports that conclusion.
Reducing the Risk of Black Swans Page 2