Updated Januray 2021
ARE DFA STRATEGIES SUPERIOR?
Abstract: Imitation is the sincerest form of flattery. DFA did most of the quantitative research examining possible sources of additional market return and has created multi-factor based offerings in the passive and index space since 1981 based on this research. Since 2013 numerous mutual funds have been introduced with fund names similar or identical to DFA's asset class funds and with various alleged asset class tilts and factor tilts to capture asset class premiums. They are often termed "smart beta", "intelligent", "fundamental" or "factor" funds. Conventional index funds are also sometimes promoted as alternatives to DFA's models and they underperform DFA's fund offerings. This article compares the strategies, composition, expense ratios, and returns from DFA's funds with factor tilted funds and index funds offered by major brokerage and fund firms. DFA strategies are relatively unique and have produced higher returns long-term but can underperform in selected asset classes for shorter time-frames as asset classes move in and out of favor.
HOW DFA DESIGNS AND MANAGES PORTFOLIOS
DFA's (Dimensional Fund Advisors) funds and investment models evolved from academic research on investment returns and efficient portfolio management which began in the 1970. DFA offered its first fund in 1981. DFA's approach to portfolio construction and management is fundamentally different from asset class tilted and index funds which are currently promoted by major brokerage firms and fund companies.
DFA developed its fund models and manages its portfolios from very detailed statistical analysis of every possible factor or strategy which might influence expected portfolio returns. First, it examined historical investment returns going back to 1927 or as far back as possible for asset classes with briefer histories. From this research DFA used factor analysis and other complex statistical procedures to define and construct mutual funds which best capture asset class returns. In non-statistical terms, the factor analytic statistical methodology DFA employed to develop its strategies and portfolios took massive amounts of historical returns data and boiled that data down to the minimum number of factors which explain and capture as much as possible of all the returns. It then includes those factors in its fund portfolios.
DFA pioneered several portfolio strategies including the now well-known asset class tilts towards value, company size, and profitability. These strategies are discussed elsewhere on this website in "Portfolio Design" and "Risk and Return" and included in DFA's models and funds. DFA's research is well supported by Dimson et al.'s exhaustively analyzed data base of 25 countries since 1900. See Credit Suisse's Annual Yearbook online for annual updates of the data. DFA considers itself a research house focused on investment and portfolio returns as well as an active manager of passively invested funds. DFA's passive portfolios are not simple in structure or execution, continually evolve. DFA employs many research Ph.D.'s and has Nobel laureates on its board.
In addition to historical analysis of returns data DFA also micromanages portfolio strategies in all its funds to capture even small increases in expected returns. These microstrategies include patient buying strategies which bid below market and will wait to do a trade until a price target has been hit, avoiding trades around index reconstitution, a committee dictated process which reduces returns by about 0.45% per year, and also securities lending. DFA also avoids purchasing or owning positions with excessive volatility and continually conducts research into every detail of investment markets and portfolio management. When the benefits of including a strategy in fund management outweigh its costs DFA includes it in its fund portfolios. A comprehensive study by Sunil Wahil examined all of DFA's equity trades from 2007 through 2008, over 1.3 million trades representing about $89 billion. He estimated DFA enhanced returns by 0.6% to 0.8% annually due to patient buying strategies, the use of alternative trading markets in dark pools, and securities lending.
The main difference between DFA's approach compared to other recent offerings with factor tilts is that DFA's offerings are entirely research driven and based only on the longest term empirical data on market returns. DFA's portfolios are wholly unlike sales driven brokerage products which rely on committee opinions or simplistic factor models. We examine DFA's models in more detail in "Portfolio Design" and "Risk and Return" on this website. DFA sometimes applies its factor models to analyze securities distributions in indexes and style tilted funds and map the sources of returns in different funds offered by different fund companies. Examples are presented below in this article. DFA's models should be considered a "meta-model", a model that can model other investment portfolios and explain where those portfolio returns have come from. This article will examine index and ETF structures, then compare DFA's funds with index and ETF offerings from Vanguard and Schwab.
INDEXES
Indexes are committee designed portfolios which aim to capture certain slices of the equity or fixed income markets. The most well-known of these are the Dow Jones Industrial Average, the S&P 500 and the Nasdaq 100 but there are numerous indexes covering just about all categories of equities and fixed income. Indexes often overlap substantially with DFA's asset class portfolios and sometimes carry the same names as DFA's funds. Indexes are not the same as DFA's asset class funds in design or operation and DFA's asset class factor analyses can be used to understand and define how indexes differ from DFA's asset class based portfolios.
Indexes should not be termed "benchmarks" since marks on benches don't move around while the composition of index funds varies continuously over time due to removals and additions to indexes, share buybacks, private equity buyouts, index "reconstitution", company mergers, and differential share price growth in various holdings. Indexes are not constants. Index committees usually meet biannually or annually to examine possible changes they might want to make in their index portfolio holdings and it usually adds and eliminates several portfolio holdings every year.
From the Nasdaq 100 internet bubble top in March 2000 through end 2002 a surprising 80% of the stocks in the Nasdaq were removed. It was not at all the same index in 2002 it was in 2000. In a typical year 22 of 500 companies in the S&P 500 are changed. During the 2008 liquidity crisis and market decline 6% of the S&P 500 stocks were sold and replaced. The Dow Jones Industrials (DJIA), probably the best known of the indexes, has changed 51 times since its inception and today holds only 1 of the 30 stocks that were in the DJIA in 1929. An index that continually changes should not be considered much of a guide to anything that is a constant. It typically shifts 3 or more of the 30 DJIA stocks in the DJIA index each year. Somewhat surprisingly, research has shown removed stocks doing better than newly included stocks across some time frames.
Indexes carry very low expense ratios, down to 4 bps. in many Vanguard funds, particularly if they reside in ETF fund structures. In ETF's the expense ratio can be kept extremely low because little need be done to construct the portfolio other than follow the list of intermittent changes made in index committee recommendations. In ETF's, investors simply own a number with assets pledged by contracts from large authorized participants, typically large brokerage firms. Open ended funds offered by DFA and Vanguard purchase and hold securities within a portfolio and buying and selling occurs. This increases the expense ratio a little versus an index. In ETF's not much needs be done other than obtaining pledges of securities from authorized participants and tracking retail investor shares. One characteristic of ETF's is that they can trade at a discount or premium to the market price of the pledged securities and create losses that don't reflect market prices for the pledged securities. This is not the case with open ended mutual funds which do not suffer from discounts or premiums. Although we assemble all ETF portfolios for clients if they wish we prefer avoiding ETF's due to the discount/premium problem.
DFA has calculated that avoiding index reconstitution costs for 3 months after index reconstitution increases investor returns by between 0.45% and 2.21% annually, significantly more than lower cost index ETF expense ratios save for investors since the difference in expense ratios between DFA and index ETF's runs about 0.1% to 0.25% annually depending upon the asset class. DFA's funds benefit since most index funds must reconstitute on the same day that the changes in index composition are announced and volume explodes when reconstitution occurs. This drives down the price of securities sold in the index and drives up the cost of securities purchased and impairs index returns.
DFA illustrates the difference between indexes and its portfolio strategies by comparing the return of the Russell 2000 small cap index with a small cap buy-and-hold portfolio for 1990 through 2015. Index rebalancing cost an average of 2.2% per year. For the 10 years ending in December 2015 DFA's small-cap fund returned 7.9% annually to the Russell 2000 index's 6.8%. DFA has examined index reconstitution in detail and found that in day 49 after index reconstitution reconstituted indexes trailed DFA's equivalent portfolios by 0.6% in US large value and 1.35% in small value.
As opposed to committee designed indexes DFA's fund portfolios are rooted in statistical analysis of asset classes and the returns and risks associated with them and with asset allocation mixes. DFA's asset class portfolios include every traded stock in any given asset class (or mix of asset classes) which passes various criteria for belonging in a portfolio as defined by DFA's research. After securities are purchased and included in portfolios DFA performs numerous daily screens on all portfolios to identify companies or bond issuers which are shifting asset classes, experiencing serious legal or regulatory problems, have abnormal price volatility or are changing in other ways which will affect expected returns. Companies with difficulties are sometimes sold.
Examples of the additional return DFA's strategies provide are further illustrated by comparing the trailing 10 year returns through end-2007 in DFA's funds and their equivalent indexes. DFA's US large value fund returned 6.11% annually compared to the Russell large value index return of 4.42%. DFA's US small value fund returned 9.62% compared to 7.47% for the Russell small value index. DFA's international value fund returned 9.54% compared to 8.04% for the Russell index, and DFA's emerging markets value fund returned 22.2% compared to the Russell emerging market index's 15.5%. These are all annual return advantages for DFA and compound over time. DFA's strategies clearly captured the value premium better than equivalent committee designed indexes.
EXCHANGE TRADED FUNDS (ETF's)
ETF's are legal and operational structures which differ substantially from open-ended mutual funds. DFA offers only open-ended funds while Vanguard offers both open-ended funds and ETF's. ETF's are created from pledged (promised) shares by authorized participants to ETF fund managers. ETF shareholders own a pledged asset, not a fractional share of an asset as is the case with open-ended mutual funds. It is a contractual pledge to provide the shares but does not provide true share ownership and during a market crisis contracted pledges could be abrogated. On countless occasions ETF's have traded at share prices far below the market value of the assets in the ETF, sometimes more than 60%. ETF investors do not own the pledged shares, the value of what they own is whatever the exchange traded shares trade for and that may be well above or below the actual market value of the underlying pledged securities in the ETF portfolio. ETF share owners own shares in a derivative structure though the ETF industry has argued ETF shares are not derivatives in the sense that futures and options are derivatives. This is technically true because futures and options derive their prices mathematically from the value of the underlying assets, not imprecisely from whatever the market will pay for an ETF pledge. ETF investors do not own an asset and the ETF does not own assets. That is the reason ETF's can have extremely low expense ratios, sometimes as low as 0.04%. Little more than a computer is required once they are legally set up and funded. Investors merely own a number based on a promise and contract. The ETF industry controlled about $5 trillion of assets as of Spring 2018 and continues to add assets as of January 2021.
ETF promoters have advanced the narrative that arbitrageurs will quickly close the spreads between traded share prices and the underlying value of the assets, the net asset value. And, ETF promoters emphasize that, unlike open ended mutual funds, ETF's allow intraday trading and instant liquidity with stops. However, a casual perusal of the historical share prices of ETF's versus their net asset value often reveals that ETF's have traded at very substantial discounts to the underlying pledged securities for periods of months or or even years. And, ETF's are heavily traded by high frequency trading computers in large amounts, trades which can change from buying into selling in nanoseconds. The examples below illustrate that ETF share prices can differ dramatically from the value of the pledged securities. In December 2014 the New York Federal Reserve Bank, Wall Street's top regulator, announced that extreme price movements in ETF's holding stocks, bonds and other assets had prompted it to take a closer look at the inner workings of ETF's. Little had been done as of 2019. The Fed was also concerned that ETF share prices do not always reflect underlying securities values and that authorized participants who pledged the shares might not honor their pledges during periods of heavy redemption requests.
Risk in ETF's was dramatically illustrated in the so-called "flash crash" on May 6, 2010. The Dow fell almost 800 points in 25 minutes. The Wall Street Journal reported that while the Dow lost 9.2% of its value at its worst, many ETF shares lost almost all their value, some dropping to pennies per share. Stop losses triggered massive selling and due to the rules imposed by regulators, investors who lost between 10% and 60% were not compensated and had large losses. The Nasdaq canceled over 10,000 trades that took place more than 60% below the pre-crash price. When the damage was tallied about 70% of the canceled trades during the flash crash involved ETF's (CBS Marketwatch). The Wall Street Journal (10-10-10) also reported that hundreds of ETF's performed differently than the indexes they were supposed to track and had an average annual tracking error of 1.25% in 2009 and 2010.
On June 20th, 2013, a day where equity indexes declined substantially, the Financial Times reported, "The losses for ETF's today were far beyond what the most sophisticated financial risk models could have predicted for worst case scenarios". In August 2015 ETF declines were again serious and more than a fifth of all US ETF's were forced to stop trading. ETF liquidity disappeared instantly thanks to HFT (high frequency trading) computers. HFT ETF traders, however immediately blamed NYSE Rule 48 and pushed for its removal. Pressure to remove regulations on ETF's remains to this day. The Financial Times commented that HFT traders did not have enough information to run their algorithms, backed out of the market instantly and withdrew liquidity in August 2015. Trading was littered with more than 1000 trading halts that inhibited the alleged essential arbitrage mechanisms of ETF's and caused their prices to fall well below the indices they were designed to track. Futures, cash and ETF's are priced and arbitraged off each other and according to the Financial Times HFT showed "evidence of cracks in the plumbing that underlies the world's largest equity market." The Bank of England warned that "the risks associated with ETF's are not being made clear." Vanguard founder John Bogle has also warned about ETF's even though some Vanguard funds are ETF's. In 2016 (Financial Times, 12-13-16) Bogle called for politicians to re-examine ETF's and noted that ETF investors trailed returns in conventional open-ended index mutual funds by a large 1.6% annually. He also complained that annual turnover in the largest ETF tracking the S&P 500 sometimes reached a stunning 3000% of assets, very significantly driving up trading costs. This was occurring even though the S&P committee certainly didn't change 3000% of the assets when they updated it periodically.
The Financial Times (7-25-13) reported that high levels of settlement failures in ETF's were reviving debates about their structure and liquidity. Settlement failures occur when banks pledge assets to ETF's and then fail to deliver them. The powerful Bank of International Settlements, the global central banker's bank, warned that settlement failures in ETF's pose potential systemic problems to the global financial system. Hint: Think 2008. Said one critic, "This is an unrealized risk which could morph into an operational risk in nanoseconds. What is the actual risk of a security if it takes five days to find the security to make settlement?" In May 2015 Reuters reported that the country's largest ETF providers were arranging billion dollar credit lines just in case an ETF sell-off turned into a credit market melt-down and they need to come up with money for shareholders since they won't be able to sell their pledged shares.
A Financial Times article on ETF's (4-24-18) states that ETF's are a "A $5T market that balances precariously on outdated rules" and "is a regulatory backwater" even though recent data shows that 7 of the 10 most actively traded US securities were ETF's. The article further states that ETF's create special risks and an ETF's price is only as good as the contracts that link everything together, collectively known as the "arbitrage mechanism". In times of stress this mechanism is fragile and has sometimes failed dramatically. According to the FT, the SEC largely improvises regulations for each ETF and existing disclosure rules fail to address potential problems with the arbitrage mechanism. Functionally identical funds are often subject to disparate rules or opaque rules. For example, The US's second largest ETF that tracks the S&P 500 lost 20% on 8-24-15 even though the S&P 500 that the index is suposed to track lost only 5%.
It should also be noted that one common narrative for promoting ETF's is that they don't pass through yearly capital gains to investors. This is so since investors don't actually own shares, they own a number representing shares. Although DFA does pass-through capital gains in its funds, DFA's capital gains pass-throughs typically are very low, running in the 0% to 2% range annually in most funds though volatile markets and smaller illiquid foreign markets can increase capital gains and increase pass throughs. DFA's tax-advantaged core-vector funds are so tax efficient that they sometimes pass through no gains in a given year. Indexes tend to be somewhat less tax efficient than DFA style passive asset class funds since they are reconstituted every 6 or 12 months but still are far more tax efficient than actively managed funds that trade frequently.
To summarize, ETF's create many risks which do not exist in open-ended mutual funds, particularly a liquidity risk, and most ETF's are based on indexes which have their own inherent costs and flaws described above. Evanson Asset Management® does not recommend ETF fund structures or indexes though if our clients wish advice on selecting them and purchasing them we will do so.
VANGUARD
Vanguard offers open ended and ETF structured funds along with index funds and active funds. If prospects or clients are considering self-managing their investments we recommend Vanguard. However, Vanguard is not a research house like DFA and bases its open-ended index and ETF offerings on committee designed indexes. It offers several funds which carry asset class names identical to DFA's funds but these funds often differ from DFA in asset class composition and asset class capture under the surface.
Let's compare some of DFA's offerings with Vanguard's offerings on three dimensions, expense ratios, asset class composition, and returns. Expense ratios are somewhat lower in Vanguard's index tracking funds and index ETF's than DFA's funds. Open-ended mutual funds and ETF's based on indexes do not have the costs involved when securities must be bought or sold since ETF portfolios are derivative and investors own a number and a pledged asset, not a share of the securities that were purchased by the ETF. Vanguard's open-ended equity and fixed income mutual funds, when based on indexes, carry very low expense ratios, typically 0.04% to 0.10%. Little research is required from index fund managers other than tracking securities that are selected by an outside index committee as something to be bought or sold. Index tracking funds suffer, however, from index reconstitutions and other structural issues. DFA's research found ETF's reduce annual returns versus conventional open-ended mutual funds between 0.4% and over 2% annually, far more than the lower costs of operation in ETF's.
Vanguard funds with names identical to DFA funds typically tilt towards somewhat different equity asset classes than DFA's funds and are not in most cases directly comparable. For example, during December 2008 Vanguard's US large value index had a higher price/book ratio, 1.79, and higher average market cap, $43.5 billion, than DFA's US large value portfolio with a 1.30 price/book ratio and $23.1 billion average market cap. A higher price/book ratio indicates a portfolio with less capture of the value premium but it would produce better returns when growth is outperforming value in the markets as had been the case between 2012 and 2018. When comparing returns numbers between different funds from different fund companies it is always of utmost statistical importance that investors accurately compare identical asset class mixes for identical time frames. Otherwise, it's an apples to oranges comparison. Asset classes go in and out of favor in completely unpredictable time frames so if growth is doing better then value Vanguard's US large cap value fund might do better than DFA's since the Vanguard value fund tilts more towards large growth than DFA's.
Seeking Alpha (October 2016) matched DFA and Vanguard funds by asset class for 1998 through May 2016 and fournd that DFA outperformed Vanguard indexes by 0.1% in US large cap growth, 1.4% in US large cap value, 1.2% in US small cap growth, and 0.9% in US small cap value or about 1% overall annually. DFA specializes in capturing value and small cap premiums in addition to other factors boosting expected returns. DFA's non-US equity funds outperformed Vanguard's equivalent offerings by 1.5% or more annually.
Expense ratios in Vanguard's non-index actively managed fixed income funds run in the 0.1% to 0.2% range and 0.2% to 0.5% range in equities. Expense ratios in DFA's non-index passive fixed income funds run in the 0.12% to 0.23% range and expense ratios in DFA's non-index passive equity funds run from 0.08% in US large growth, a highly liquid asset class with small bid/ask spreads, up to 0.72% in emerging market small caps, an asset class with limited liquidity and very high bid/ask spreads. Vanguard does not offer a fund in emerging market small caps. Expense ratios for most DFA equity funds run between 0.2% and 0.5%, about the same as Vanguard's active equity funds. After the drag in returns created within index based portfolios from index reconstitution and other factors are included DFA's funds come out well ahead of Vanguard's but not for every year and every time frame sampled.
SCHWAB'S "INTELLIGENT" FUNDS
Schwab's "intelligent" ETF funds, introduced in 2015, mimic DFA's funds to some degree by tilting towards factors DFA has identified from research as do several recently created funds from other fund companies that promote offerings in "smart beta", "fundamental" and "factor" tilts in funds. Although these funds may sound similar to DFA's funds in marketing materials or even carry nearly identical names to DFA's funds they are not at all the same as DFA funds. Let's compare some of Schwab's intelligent factor funds with DFA's on three key dimensions: asset class composition, expense ratios, and returns.
Schwab's intelligent funds use RAFI methodology which ignores prices when constructing portfolios and bases them on a few fundamental measures of a company's financials, five year average sales, dividends, and cash flows. Accounting variables that are used by Schwab's funds do not contain any information that is critical in predicting expected forward returns. DFA constructs portfolios which include several variables which statistically affect expected returns including current equity, current liabilities, and other accounting variables and it reviews portfolio variables daily since market prices continuously reflect these. RAFI's fundamentally weighting approach increases exposure to value stocks and DFA has found that exposure to the value premium explains much of the performance of Schwab's fundamental portfolio weightings over standard indexes. However, DFA's portfolios have a stronger value tilt than Schwab ETF's in all markets and that leads to a higher expected return. Of course, if growth stocks outperform value stocks for some period of time, as they sometimes do and have in a few years prior to 2019, Schwab's fundamentally weighted approach may produce higher returns than DFA for some period of time, it just won't be based on capturing the value premium.
A few comparisons are illustrative. Schwab's fundamental US small cap index correlates 0.55 with the small cap premium while DFA's correlates 0.79. Schwab's target index correlates 0.09 with price while DFA's factor fund correlates 0.55 with price. Schwab's portfolios are not constructed with the same asset class tilts as DFA's and DFA tilts significantly more towards small and value where higher expected returns are located. Schwab's fundamental international small cap index carries an average market capitalization of $1.78 billion to DFA's $1.475 billion and a price/book ratio of 1.56 to DFA's 0.95. Higher average market capitalization means that Schwab fundamental funds hold stocks in somewhat larger small cap companies and a higher price/book ratio means Schwab's fundamental small cap growth fund is tilted more towards growth than DFA's. Schwab's fundamental emerging market ETF fund holdings carry an average market cap of $27 billion while DFA's is only about $2.7 billion, a very different asset class size sample. Fund factor tilts are significantly different between Schwab and DFA funds and DFA can provide these if requested.
Expense ratios for Schwab's US large cap, US small cap, international market, emerging market and other ETF's based on equity indexes run between 0.04% and 0.17% annually depending upon asset class, very low and in the same range as Vanguard's index based ETF's. As noted in the section on ETF's above, ETF indexes are extremely low cost since most are based on indexes dictated by committees and ETF's are built from pledged assets and are derivative in structure, making portfolio management relatively easy and cheap. ETF investors own a number, not a share which can be surrendered for the value of the underlying securities at any point in time.
In a full comparison of expense ratios on DFA's offerings with Schwab's "intelligent" indexes, DFA's expense ratios were about 0.1% to 0.25% higher depending upon asset class. DFA, however, more than makes up for that by avoiding the costs of index reconstitution. Long-term return comparisons of Schwab's intelligent funds with DFA's equivalent funds aren't available since Schwab's funds relatively recently initiated. And, close examination of Schwab's asset class fund structures indicates that they are not identical to DFA's funds and are biased toward growth stocks and larger market capitalizations and Schwab's micromanagement strategies are minimal compared to DFA. Data from DFA and from Dimson et al. for a century or more clearly shows that value outperforms growth long term by about 2% to 5% annually in almost all countries. See "Portfolio Design" elsewhere on this website.
For the prior 3 years through Fall 2017 DFA's funds have returned about 0.2% to 1.5% more per year than Schwab's intelligent funds with similar but only slightly better returns in a few asset classes. This was during a period when Schwab's growth and large cap portfolio bias should have given them somewhat better returns than DFA. For example for the prior 3 years from Fall 2017 DFA's large company fund returned 0.21% annually more than Schwab's equivalent ETF, 0.5% more annually in small caps, 1.19% more annually in international large cap core-vector funds, 1.45% more annually in international small value, and 1.37% more annually in DFA's emerging markets core-vector fund. DFA's core-vector funds are multi-asset class funds which are comparable to Vanguard's multi-asset class offerings.
The Financial Times (11-26-17) published a whole section on smart beta which included an article entitled "Lack of Benchmarks Fuels Concerns". In this article, the FT notes that smart beta strategies had been the fastest growing section of asset management over the past five years since 2012 and lists ten major companies that offer them including Blackrock, State Street, Vanguard, Schwab, First Trust and Wisdom Tree. Curiously, DFA is not mentioned in the article even though it was managing factor tilted portfolios long before the other firms and did the research that led to factor portfolios. The FT article included several industry insiders who had similar comments about smart beta funds to those expressed in our website article here. Several warned that smart beta funds take a modified passive strategy but returns may not be as strong as forecast due to data mining problems in back tested data that may not stand up to rigorous analysis and inaccurate comparisons of asset classes and sampling times. Rob Arnott, CEO of Research Affiliates, it developed some of the earlier smart beta funds, warned that some ETF products could go "horribly wrong" and that investors should not be duped by high recent returns.
A close relative of smart beta funds are the robo-funds. All employ some degree of active management and deviate from market cap weighted portfolios through fund selection or sector selection. Backend Benchmarking, BEB, has tracked the performance of 60 robo portfolios for varying short term time frames and compares their performance to a benchmarked 60% equity/40% fixed income portfolio. It then deducts 30 bps. annually to normalize for lower expenses charged by robos. A key takeaway is that robos in the last four years through Q12020 underperformed their respective benchmarks by between 0.88% to 1.33% annually or about 25% to 38% of the total return annually versus a passive 60/40 benchmark portfolio which can be expected to return about 3.5% annually. Asset classes go in and out of favor with no predictability. From 2017 through March 2020 the Russell 3000 Growth Index outperformed the Russell 3000 Value Index by 51.7%, gaining 46.7% versus a decline of 5.0% in the value index. This was entirely due to changes in the p/e ratio of the two indexes with the growth index rising from 23.0 to 23.4%, very high, and the value index p/e falling from 18.5:1 to 13.0, slightly below average. Long-term, value stocks outperform growth stocks by about 2-4% annually.
CONCLUSIONS
Claims regarding investment returns need to be carefully evaluated and DFA's statistical models are ideally suited to doing so as well as essential in designing DFA's fund offerings. We consider it highly likely that various funds not examined in this article which purport to compete with DFA in addition to Schwab's ETF fundamental indexes and Vanguard's committee based indexes suffer from the same reductions in return which funds from these two broker/custodians do. Unfortunately, narratives and data that sell financial products and services are usually constructed to capture revenues in a highly competitive investment marketplace and sell "product" rather than accurately describe and capture available returns. Narratives change often and are often sales focused and forward returns are always unpredictable.
In Q22020 a few articles have appeared in the mainstream financial press which have questioned the need for any asset class diversification at all. That's mostly because since 2012 investors obtained the highest returns by owning only US large and small growth stocks and avoiding other all equity asset classes domestic and foreign. For example, on 8-5-20 the WSJ published a piece by Meir Statman entitled "5 Myths About Stock Diversification". In it the autheor argues that diversification of an equity portfolio beyond 12 to 18 stocks in total offers little to no additional benefit, owning a portfolio with a handful of stocks is safer than a portfolio of thousands of stocks investors are unfamiliar with, owning three index funds (Vanguard S&P 500, Vanguard small-cap index, and Vanguard Total Stock Market) will capture just about all global market returns, and that US and international stocks are closely correlated so their is no reason to own overseas stocks. Another study entitled "Smart Beta is Actually Still Dumb Money" and published in Summer 2020 by U. Washington academics found that the average return of smart beta indexes dropped 2.77% per year on paper before ETF listing to -0.44% per year after ETF listing. Extensive research by DFA and others, some mentioned in this article, soundly contradicts these claims and shows long-term investing exclusively in US large and small growth stocks underperforms globally diversified equity portfolios and asset class diversification by at least 2% annually. However, very low fee ETF's may pass on a little more of the return to investors with some ETF's carrying a very low annual expense ratio of only 0.04%. ETF's, however, can trade at a discount or premium to the underlying share price and that presents other problems for portfolio returns.
DFA has become a major player in the mutual fund business since 2000 with over $400 billion in assets under management as of 2018 and DFA's founders have conducted detailed research on market returns and portfolio construction since the 1970's. DFA's core framework remains academic, research driven and empirical and is not marketing, product and sales driven like its competitors. We recommend that investors look very carefully at funds which carry the same or similar names to DFA's asset class funds or claim to capture factor returns similar to DFA or indexes which claim to capture returns similar to DFA. There is little evidence that alternatives to DFA's offerings marketed by other firms, when accurately matched for asset class and the time period sampled, outperform DFA's more analytic and empirical strategies.