Throwing the Baby Out with the Bathwater — a Rare Miss for Jack Bogle

Warren Buffet wrote in his annual letter this year: “If a statue is ever erected to honor the person who has done the most for American investors, the hands- down choice should be Jack Bogle.” Bogle, the founder of Vanguard Funds, helped launch an investing revolution when he created the first index mutual fund some forty years ago.

A core premise of index investing is that most active managers (investment professionals who would try to out-perform the index through stock selection) do not beat an unmanaged index. And extensive academic research has shown that after considering fees, expenses, and taxes, very few managers have been able to beat the index. Proponents of the efficient market theory, including me, would say that this conclusion is only logical: market prices reflect all available information about stocks, the market is generally efficient, so it’s tough to out-perform given the drag of management fees, transaction costs, and taxes. Some, perhaps even Bogle himself, might go as far as to suggest that any outperformance by an active manager is solely a result of luck, not skill. But not me. I believe that markets are generally efficient, but not entirely efficient. And while most managers do not exhibit enough skill to outperform the market on a consistent basis, some have and some do.

So I was not surprised, but I was somewhat disappointed, to read Bogle’s remarks made at the CFA Institute’s annual conference in Philadelphia last week. In commenting about “smart beta,” which describes a suite of strategies in which certain equity risk factors or “betas” are favored in an otherwise passive-style investment approach, Bogle said: “It suffers from the assumption that past data — heavily mined — will identify factors that provide sustainable performance and leadership in the future. Mark me as being from Missouri on that because it ignores the principle [sic] of reversion to the mean, or RTM, one of the most important things you need to understand about financial markets and stock returns and market returns and mutual fund returns. It’s a huge mistake to ignore RTM.” Bogle has a point. The investment profession tends to fall in love with investment fads and trends, creating new products that cater to those who wish to take advantage of the latest and greatest theme. He continued: “…popular fads drive product creation in the fund industry. That’s great for fund sponsors, and awful for fund investors.” The problem with this generalization is that it lumps together, then summarily dismisses, every strategy, structure, and idea that is not a plain vanilla equity index fund.

Certain risk factors that fall under the smart beta heading, such as “value” and “small cap,” have been well-documented to offer excess return compared to the broad market index. Surely Bogle knows this, since the same academic researchers who demonstrated the value of indexing subsequently highlighted the empirical out-performance of these risk factors. Eugene Fama was awarded the Nobel Prize for economics in 2013 for his work to develop the Efficient Market Hypothesis. During the past few decades, Fama has collaborated with Kenneth French to formulate the Fama-French three factor equity model. Based on extensive statistical research of U.S. (and subsequently global) stock prices, Fama and French demonstrated that a financial model that incorporates three factors, rather than the one beta factor in the Capital Asset Pricing Model (CAPM) — — the model that encapsulates the essence of the Efficient Market Hypothesis — better explains stock price movements. What are the three factors? Beta — as in CAPM — Value v. Growth, and Small v. Large risk factors. The three factor model does not discredit the intrinsic value of indexing, Bogle’s life’s work. It reinforces and extends it. I have had the privilege to hear Fama discuss the implications of the three factor model. It is not a free lunch — there is no free lunch in efficient markets — — but the additional two factors, value and small cap, do offer the possibility of higher systematic returns (along with higher systematic risk) than a broadly diversified portfolio without these risk tilts.

At North Capital, we have integrated the value and small cap risk factor tilts into our core equity portfolios, and our clients have benefitted (and, we believe, will continue to benefit) from the over-weighting of these factors. Is it absolutely necessary? No ~ the expected return available from plain vanilla indexing (core equity beta) accounts for the vast majority of the total expected return from a strategy that also incorporates a small cap tilt and a value over-weight. But if the empirical data shows the availability of excess returns, and there is a logical explanation for the empirical result (value: when you shop sales, you save money // small cap: human capital, which is not measured using conventional accounting metrics, is far more impactful in a small company than in a large company), then why not take what the market has to offer?

Obviously this type of analysis opens the door to…. systematic active management. If small cap and value factors offer additional returns for investors, what about other risk factors? In this context, Bogle’s point about mean reversion is a fair one. Smart beta comes in many different flavors, and there is no shortage of quantitative active managers peddling new factors in an attempt to harvest, package, and sell the hope of excess return to investors. Many such factors have a long-term expected return of zero, so a short-term period of out-performance offers little besides a marketing narrative — a precursor for the type of investment fad that Bogle warns about.

Think of new smart beta risk factors this way: there is a room of one hundred coin flippers. Each flipper claims a particular “skill” in flipping coins — call the skill “smart coin flipping.” Each flipper can make only one flip per year. On average, after five years, 1/32 or about 3% of the coin flippers will have called the correct side every single time. A large percentage of flippers will have have called 4/5 or 3/5 of the flips correctly. Are these flippers smarter than all of the others in the room? Not hardly. The key lesson here is that in analyzing statistical data (and the fundamentals underpinning the data) to identify sources of excess return, one must evaluate very large data sets and use intellectually rigorous analytical techniques to avoid bias and form logical conclusions. I believe Fama and French did this in their work, and subsequent academic research has supported their conclusions. The same cannot be said for the full myriad of new smart betas, some of which are cooked up with limited data, a shallow knowledge of statistics, and weak fundamental analysis — — the successful coin flipper, packaged as “smart beta” for the retail investor. When I evaluated hedge fund strategies for Bear Stearns, over a decade ago, we had a parade of quantitative managers pitch us backtested strategies that promised excess returns and high Sharp Ratios. We used to say that we had never seen a backtest that we would not want to invest in. If only it were so easy. I am confident that allocators today must have the same experience on a weekly basis. After all, market data is more abundant and accessible than ever, computers are fast and cheap, and any 23-year-old with rudimentary training in statistics knows how to program in R. How easy, then, to create a quantitative model that out-performs the index (on paper, and in the past)! The emergence of “smart beta” is, in the extreme, simply a repackaging of this same elixir for a new channel — — retail — — that is in love with indexing but cannot give up on the hope of higher returns.

Originally published at www.northcapital.com on May 28, 2017.