How to Use ETFs for Factor Investing
By Lewis Braham
May 3, 2019
Financial planner Robinson Crawford has invested his entire individual retirement account in just four exchange-traded funds tracking cheap stocks as well as hot ones in the U.S. and overseas. His logic: Academic research indicates that stocks possessing value and momentum qualities, or “factors,” do well, so why not invest solely in them? He wouldn’t dare adopt that approach for his clients, however. “It’s a high volatility strategy,” says the 28- year-old founder of Montebello Avenue, a Scottsdale, Ariz.–based advisory firm. “You’ve got to really understand what you’re buying.”
Unfortunately, most investors can’t understand because too many factor based products have flooded the market. According to Morningstar, there are 771 factor ETFs and mutual funds, with more than $1 trillion in assets. Meanwhile, academics have identified some 400 factors, many via data mining or back-testing. “Many of these factors have been so-called discovered, but it’s just basically lucky—if you try enough things, something is going to look like it works within [a study’s] sample,” says Campbell Harvey, a partner at Research Affiliates who has published papers critical of the factor data-mining explosion.
Given that, how can you build an effective factor-based portfolio? First, identify factors that have real-world track records and significant academic support. Harvey, in a paper titled “Alice’s Adventures in Factorland,” tracks the historical provenance of the key factors and their efficacy. Even the estimable value factor identified by Nobel Prize winners Eugene Fama and Kenneth French dates back only to their 1992 “three factor model” paper, a long period by market standards, but only a twinkling from a scientific one. (Climate science, for instance, dates back to the early 1900s.) That said, a nonacademic version of value investing has existed since the publication of Security Analysis by Benjamin Graham and David Dodd in 1934.
In addition to value, size, momentum, and low volatility, factors have long track records of success. Interestingly, the small-cap, or “size” factor, identified in 1975, works, but academics now believe that’s because it’s a composite of other factors—value and illiquidity. Also, low-volatility (1966) or, as Harvey terms it, “low beta” stocks historically don’t beat the market but instead equal it, making for better risk-adjusted returns, i.e., advantageous for conservative investors. High operating profitability has been one of the few factors to work well since the 2008 crash, yet the results prior to 2013 are back-tested. High quality stocks have also won, but the factor dates back to only 2003.
Once you’ve chosen factors, you must decide how to invest in them. For investors who don’t like straying too far from the market and thereby risk under-performing it—what is known as tracking error—a market-capweighted factor index such as Vanguard Value (ticker: VTV) might be best. But by weighting stocks that fit the factor criteria by their market value, you do not get the purest expression of the factor. The largest value stock isn’t necessarily the cheapest one. A better alternative if you can handle the additional tracking error is to buy a fund that weights the stocks based on their factor exposure, such as iShares Edge MSCI USA Value Factor (VLUE). Crawford owns Alpha Architect U.S. Quantitative Value (QVAL) because he sees it as the purest expression of the factor. That ETF holds just 40 stocks screened specifically for their value characteristics. By contrast, Vanguard Value holds 338, and iShares Edge MSCI USA Value Factor holds 147.
But concentration raises stock-specific risk and, often, volatility. The QVAL ETF’s standard deviation—a volatility measure—is 53% higher than the S&P 500 index’s. “Our strategies have about 300 holdings in them, and we have some stock-specific risk,” says founder Gregg Fisher of Gerstein Fisher, an advisor that has run factor portfolios for more than 20 years. “If you’ve got a portfolio of 30 to 50 stocks, you’re really an active manager.
You can’t have reliable exposure to these factors in a broad diversified way with 30 to 50 names.” Unfortunately, the most difficult question to answer with factors is how much to allocate to them. While some, such as Harvey, advocate tactical, contrarian shifts to factors based on macroeconomic conditions and which factor has underperformed recently or is “cheapest,” others, like Fisher, argue that timing one’s factor allocation is as difficult as timing the market overall.
If doing the allocation yourself seems too complicated, you could opt for a multifactor ETF, which does the factor selection for you. Two top performers are SPDR MSCI USA Strategic Factors (QUS), which equal weights value, quality, and low-volatility stocks, and Hartford Multifactor US Equity (ROUS), which allocates 50% to value, 30% to momentum, and 20% to quality. But doing so may be too cookie-cutter for your needs.
Fisher advocates a customized “life cycle” allocation based on age and risk tolerance. In a 2018 paper, he examined the returns and volatility for various hypothetical factor portfolios and found that shifting as you age from the most volatile factor strategies—small-cap value, which combines size and value factors, to the least volatile—quality—produced the best results. Fisher’s approach accounts for the fact that a retiree’s time horizon is quite different from a recent grad’s.
To succeed in factor investing requires an honest appraisal of your ability to stomach long periods of under-performance. The impatient should have the lowest factor exposure. The worst thing would be to have already missed the market’s return and then dump a factor fund from frustration before it has had a chance to win.