DFA – Has the Ivory Tower Collapsed?
The manager results table of this and recent issues of Insight show that a number of Dimensional Fund Advisor (DFA) funds have underperformed their relevant benchmarks, in some cases by a material amount, over the last 1–3 years. In response, some clients have justifiably asked whether the DFA funds or separately managed accounts are still an appropriate portfolio component. This presents a good opportunity to remind our clients of our beliefs regarding the role of public equities in clients’ investment portfolios generally and, specifically, DFA’s* role in implementing the public equity allocations.
You may also want to reference our recent article, “Unconventional Wisdom: Uniting Active and Passive Management”, which provides some important background on our investment approach, details the perils of traditional active management, and discusses how Aspiriant applies unconventional active management to clients’ portfolios.
Before addressing the DFA funds specifically, it’s important to be reminded of some of the key philosophies underlying the implementation of Aspiriant portfolios.
The real economy is the source of investment returns. The vast majority of investment returns are created not through financial engineering, derivatives, arbitrage, or leverage, but through the fundamentals of profitable economic activity. As investors, we can participate in that economic activity at various points throughout the value creation process – the production and sale of raw materials (commodities, such as oil and industrial metals), real estate (office buildings, factories, and shopping malls), debt financing (bonds), and equity financing, or ownership of the enterprise (stocks). Each plays a specific and important role somewhere in the value chain and earns a return over time for the use of investors’ capital.
Risk is priced. To entice investors to put capital at risk, investments must offer a return opportunity that is commensurate with the risk assumed. While you might loan money to Microsoft for the next 2 years at a rate of, say, 5%, you’d probably demand a much higher interest rate to extend a loan today to Chrysler or GM. Likewise with equities, owners of riskier enterprises must, in the aggregate and over time, be rewarded for the additional risk they’ve borne with higher investment returns. You might be comfortable buying Microsoft stock with the expectation of a 10% annualized return on your capital, but you’d have to expect a much higher return to entice you to purchase stock of a distressed company like Chrysler or Citibank or a small company that you’ve never heard of.
Markets work. In light of recent events, it might be difficult to convince you that markets actually do work! The current market environment is, in fact, an excellent example of how markets work to price risk. Big, formerly-stable companies like the domestic auto makers, many large banks, and even diversified industrial enterprises like General Electric, have seen their share prices plunge by over 90% as their perceived (and maybe actual) risk have increased. This is the capital market at work, identifying newly risky companies and adjusting their prices (downward) to provide a robust future return opportunity for those investors who accept the suddenly-increased (at least perceived) risk of investing in those companies. The market doesn’t get it right immediately or at all times, but over time, free capital markets do an excellent job pricing risk.
A (very) brief history of financial economics
In 1934, with the Great Depression as a backdrop, Columbia Business School professors Benjamin Graham and David Dodd wrote Security Analysis, which enumerated many examples of the stock market’s tendency to irrationally under-price out-of-favor companies which, in the real world, were generating strong earnings and had large (often dominant) market positions and other substantial competitive advantages. They established the idea that a company’s value and its price were not necessarily the same and, when a company’s stock price is much less than its intrinsic value, outsized return opportunities exist.
The Graham and Dodd approach remained the dominant approach to investing for decades until Harry Markowitz’s 1958 Modern Portfolio Theory uncovered the benefits of diversification across asset classes. Markowitz’s theory was further refined by Stanford professor William Sharpe, who in 1964 developed the Capital Asset Pricing Model (CAPM), which postulates that one factor – an asset’s sensitivity to market risk – determines its price. Sharpe further asserted that the CAPM explains over 80% of a diversified portfolio’s return.
Finally, in 1993, University of Chicago professor Eugene Fama and Dartmouth professor Ken French proposed the Three Factor Model, which agreed with the CAPM’s theory that greater exposure to the market increases return; however, Fama and French identified two additional factors that reliably and predictably explain stock performance – the “value effect” and the “size effect.” The Fama/French Three Factor Model explains, many believe, over 90% of a diversified portfolio’s return.
The size and value effects
Fama and French demonstrated that the size and value effects exist. Just as importantly, they hypothesized the reason these effects exist, why they are durable, and how they may predict excess equity returns over time. Simply put, stocks of small companies and “value” (i.e., distressed) companies are, in fact, riskier than stocks of large companies with strong earnings and balance sheets. As discussed above, investors quite logically demand higher returns from these riskier small and distressed companies and so, in the aggregate, those companies must provide the expectation of higher returns to attract capital.
How does the market ensure that these higher expected returns actually happen? Let’s say you would invest in Microsoft or ExxonMobil with the expectation of a 10% annual return on your capital. But would you invest in Citibank or GM today with the expectation of just a 10% annual return? Probably not. What about tiny companies (selected at random from a DFA fund) like Werner Enterprises, Skyworks Solutions or Zoran Corporation? You’d probably have to expect high returns to be enticed to invest in them, too. The global equity markets, incorporating the aggregate judgments of millions of investors worldwide, adjust the stock prices of the riskiest companies to a sufficiently low price so that today’s investors can reasonably expect high future returns to compensate them for taking the risk.
In fact, the equity markets are effectively giant risk pricing engines, constantly adjusting prices of all investments so that investors’ future returns are commensurate with the current risk taken. This price discovery process can be very ugly at times, and it is clearly imprecise and imperfect, often being heavily influenced by emotion, politics, and other outside factors. Observing this, Benjamin Graham famously said that, “In the short run the market is a voting machine. In the long run it's a weighing machine.”
The Size Effect chart, below, which separates US equities into ten size deciles (each with the same aggregate market capitalization), clearly demonstrates the size effect. Since 1926 (when reliable and consistent data begins) through 2008, the largest 10% of companies have generated an average annual return of 10.75%. The second decile has generated a 12.51% average annual return over the same period, and so on down the market capitalization spectrum, with each smaller decile generating a greater return than the larger decile before it. The smallest 10% of companies have generated an eye-popping 20.13% average annual return.
The Value Effect chart, illustrates the same idea, but for companies’ health, not size, as measured by “book-to-market ratio”, one measure of the market’s perception of their health.1 Decile 1, comprised of the 10% of companies with the lowest book-to-market ratios (i.e., “growth” companies), have delivered materially lower returns than “value” stocks, companies where the market price has been beaten too far down relative to the value of the company’s assets (i.e., a high book-to-market ratio). In other words, investors overpay for growth and punish distress too harshly. Of course, this isn’t a new idea -- it’s the very concept that Graham and Dodd identified 75 years ago in Security Analysis -- Fama and French just broadened it across equities generally.
While Graham and Dodd were right to notice the “value effect”, we think they were incorrect to characterize it as irrational. In fact, distressed companies’ low stock prices maybe an entirely rational response to those companies’ increased risk. The size and value effects are not random or ephemeral variations or statistical flukes, but rather a completely rational (and thus durable) reflection of the greater risk of small and distressed companies. Consequently, until investors stop demanding more return for taking more risk, these effects will persist.2
Putting theory to work
The Fama and French work forms the intellectual framework for Dimensional Fund Advisors.3 DFA structures investment portfolios to deliver very potent doses of the size and value effects by targeting smaller capitalization and higher book-to-market companies than standard indexes and most active managers. For example, the median market capitalization of the DFA US Micro Cap Fund is $95 million, versus $322 million for the Russell 2000 index. Thus, DFA delivers more exposure to the size effect. Likewise, stocks held by the DFA US Large Cap Value Fund have an average book-to-market ratio of 1.27, a much higher exposure to the value effect than the 0.66 average of the Russell 1000 index or even the 0.90 average of the Russell 1000 Value index.
The two preceeding charts divide US equities into the ten size and book-to-market deciles and illustrate the deciles in which three DFA funds invest. Note the far more focused range of holdings for the DFA funds compared to their benchmarks…evidence of their size and value tilts. Within the specified range for each fund, DFA does almost no “fundamental” analysis (e.g., visiting companies, speaking with management, forecasting profit growth); rather, after applying some basic screens for liquidity and financial viability, DFA will purchase just about any company within the desired range, (thus, a very large number of companies) reasoning that the market would have assigned a price to the company that is commensurate with its risk.4
One would expect that DFA funds, with their stronger exposures to the size and value effects, compare very favorably, over time, to funds investing in similar companies. And, in fact, that’s exactly what the data shows. The Small Core Universe chart, below, shows the annualized returns, over periods of 1-10 years, of three DFA funds we often use to implement the small cap blend exposure in clients’ portfolios. The returns of the three DFA funds (represented by the colored dots) and the Russell 2000 small cap benchmark (the black diamond) are arrayed against a vertical bar representing the performance range of the middle 90% of small cap managers, divided into quartiles. (The top 5% and bottom 5% of results exist, respectively, above and below the bars.)
Reading the chart from the right, the three DFA funds all clearly outperform both the benchmark and the majority of their peers over 8+ years, as the impact of their purer exposure to the size effect adds up over time. DFA’s exceptionally low management fees also help. Likewise, the Small Value Universe chart, illustrating the small cap value asset class, shows very strong performance of the DFA funds over time relative to peers, owing to DFA’s relatively stronger value orientation.5
In the recent past, however, the DFA funds compare less well against their benchmarks and peers. Importantly, given the recent severe difficulties for equities generally, this is not surprising. Because the DFA funds overweight “value” companies (including the financial sector) and smaller companies, DFA funds generally outperform the index and most peers when investors are willing to take risk with smaller companies and those perceived to be weak, and underperform when investors are not so willing to take on these risks. As the charts demonstrate, the cumulative effect of DFA's size and value exposures has resulted in material long-term outperformance; however, the price of that long-term outperformance is greater near-term volatility due to the portfolio tilts.
Perhaps ironically, we view DFA's recent underperformance as a sign that their strategy is working. Their funds have generally underperformed largely because they continue to target pure size and value exposures in a very consistent, disciplined fashion. This is exactly what we want them to do, with the confidence that we will be amply rewarded over time. Of course, we don’t simply take it on faith that the DFA funds… or, for that matter, any manager… are properly executing their strategy. We maintain a very close dialogue with senior managers, strategists, and executives at DFA, and receive from them very detailed quarterly attribution analyses so that we understand the exact causes of return variation in each fund.
The following charts demonstrate that the size and value effects exist in overseas equities, too. The Foreign Large Blend chart plots three funds we use to invest in large company stocks overseas. The Vanguard Total International Stock Index has generated the return of the benchmark over time (which is what we expect of an index fund); however, the DFA funds, which target value equities overseas, have trounced the index and their peers, landing in the top 5% of managers for all periods greater than 5 years. Again, recent fund performance lags, reflecting these funds’ greater volatility.
The final chart illustrates overseas small companies, where two DFA funds we use target the value and size effects. Not surprisingly, their exposure to the size and value effects has resulted in substantial outperformance versus their benchmark and peer funds over time. Interestingly, these funds have also performed very well recently relative to their peers, suggesting that actively managed overseas small company funds have performed particularly poorly in the recent past.
Returning to first principles
Returning to our first principles, we see that the real economy is the source of investment returns, so the foundation of one’s investment portfolio should reflect participation in the economy through a diversified portfolio of bonds, real estate, commodities, and public equities. Risk is fairly priced, so investors are rewarded for deliberately taking smart, diversified risk by investing in small and distressed companies. And we can rely on these effects to persist, because markets work to ensure that investments reflect their unique risks.
All of these factors suggest that the investment strategy you have adopted with our coaching and implemented in large part using proven DFA vehicles, is very likely to deliver attractive long-term investment returns.
Greg Schick and Bob Wagman
*DFA and the securities it offers are just a component of our currently recommended managers and securities used in portfolios. We will provide a complete list of recommended securities upon request. DFA’s funds or accounts are available only through institutional channels such as Aspiriant.
1Book-to-market ratio is calculated by dividing the book value of a company (i.e., the accounting valuation of a company’s assets, such as factories, intellectual property, brands, etc.) by the market price of the company. Companies that have experienced strong recent stock price appreciation will generally have low book-to-market ratios (the market price in the denominator is high) compared to companies that have suffered stock price stagnation or declines.
Book value strikes some people as being an archaic measure of value, particularly in a largely service-based economy, where companies’ largest assets are often intangible. DFA’s research shows that other measures of value (for example, price/earnings ratio) are also an effective measure of value, but they prefer book-to-market because it is more stable over time than many other measures.
2Not all risks are priced. Some risks, such as speculation on individual companies and market timing, have no expected return associated with them. Mutual fund and separate account managers who actively select stocks in an effort to outsmart the market will often attribute their outperformance to their superior stock selection. However, if an active manager is indexed to, say, the S&P 500 and outperforms the index simply by owning companies that are smaller and more distressed than the S&P 500 as a whole, then the active manager hasn’t really added value, as one could have achieved the same result, for much lower cost, with a combination of inexpensive index funds or DFA funds targeting small and/or value companies. True stock picking skill is evident only when the manager has achieved higher returns by means other than simply taking on additional priced risk… something that very few managers have done over time.
3Notably, Fama and French still actively consult with DFA and are co-owners and board members. DFA’s other owners and board members include many luminaries in the field of financial economics, including Roger Ibbotson, Rex Sinquefield, David Booth and Nobel Laureates Merton Miller (deceased) and Myron Scholes.
4While targeting the size and value effects are the most notable and distinguishing elements of DFA’s approach, their screening process also incorporates a host of other subsidiary factors including, for example, avoiding real estate investment trusts (a separate asset class) and utilities (which are so heavily regulated that non-market forces heavily influence their stock prices). Moreover, while DFA’s value-oriented funds invest in distressed equities, they avoid stocks that are so distressed that failure is a high probability (e.g., companies in bankruptcy proceedings or companies that don’t meet certain minimum exchange listing and liquidity requirements).
5The DFA funds’ performance would appear even better if the data set included separate account managers, were adjusted for survivorship bias (i.e., if closed and merged funds were included), and were adjusted for taxes.