A version of this article was published in the August 2014 issue of Morningstar ETFInvestor . Download a complimentary copy of ETFInvestor here . The way I see it, exchange-traded funds have a lot in common with food. There are many different foods, but only a handful of essential nutrients. Likewise, there are lots of exchange-traded funds, but only a handful of distinct factors that drive their returns. When scientists look at why a food is good or bad for health, they look at its nutrients. Red meat isn’t bad in and of itself but because it contains trans fat, saturated fat, and cholesterol. Analogously, when those of a scientific mind-set look at investments, they look at factor loadings or exposures. From this perspective, a bond portfolio is a bundle of duration, term, and credit risk factors. Factors are to assets what nutrients are to foods. They’re what really matter. My goal when picking ETFs is to find the most efficient ways to obtain desired factor exposures, in the same way a dietitian picks meals to achieve a certain balance of nutrients. In equities, there are five major factors: market, value, momentum, quality (or profitability), and size. (There’s also a low volatility factor, but very few funds offer true exposure to it; minimum- and low-volatility funds obtain their superior risk-adjusted returns from their quality and value exposures.) These factors have historically earned excess returns. The market factor is simply defined as the return of the total stock market. Value, momentum, quality, and size are all defined as long-short portfolios that go long stocks with the characteristic and short stocks with the opposite characteristic. Market and size generate returns as compensation for their risks. Value, momentum, and quality, on the other hand, earn much of their returns by exploiting investor misbehavior. Stock-pickers have been exploiting these styles for decades. Think of the three as distinct stock-picking strategies that have been distilled into simple rules. When picking funds, I believe you should look for ones that offer efficient exposure to value, momentum, or quality–ideally, all three. At the very least, your fund should minimize exposure to expensive, poor-returning junk stocks, which historically have chewed through capital. Almost all equity ETFs offer some combination of these five factors. However, the price of exposure to these factors varies a lot. Some funds offer exposure to desirable factors like value and quality but fail to capture excess returns because of poor construction or high fees. iShares Select Dividend (NYSEARCA: DVY ) , for example, has historically shown a lot of exposure to value and quality but lost all of the theoretical advantage that it should have earned and more because of a disastrous 2008. The index’s yield-weighting methodology had it chasing falling knives while its quality screens weren’t stringent enough to discern firms temporarily down on their luck from firms in mortal danger. One way to rank funds is by running their historical returns through a factor model to identify the ones with the highest exposures to value, momentum, and quality. This is a tempting approach because it seems objective and scientific. However, factor regressions produce point estimates that mask the reality that factor loadings for most strategies change over time, sometimes dramatically. For example, high-yield stocks historically exhibited neutral to negative momentum loadings. The highest yielders, after all, are usually beaten-down, boring stocks. However, over the past three years or so, high-yield stocks have actually exhibited positive momentum loadings because investors have been chasing yield and defensive stocks. Most of the time, you can’t expect a strategy’s factor loadings to be constant. Assessing factor strategies requires some qualitative insights to take into account this kind of uncertainty. To extend the dietetic analogy, our factor models are like tools that offer unreliable readings on nutrient content. You can be more confident in your readings if you know certain things, like the provenance of the ingredients in your meal, the reputation of the supplier, and so forth. To ensure you’re getting the factor exposures you want, look for the following: 1) Simple selection and weighting rules. Some funds, like Vanguard Dividend Appreciation ETF (NYSEARCA: VIG ) , don’t disclose all their rules, or are vague about them, like iShares’ Enhanced series of ETFs. When a strategy is complicated or opaque, it’s harder to predict its behavior. 2) The use of multiple signals to reduce noise. While factors are traditionally defined by a single characteristic, such as price/book, practitioners often use multiple signals to capture a factor. The platonic ideal of “value,” for example, is not fully captured by price/book but also other measures like price/cash flow, price/earnings, dividend yield, and so forth. A phenomenal business like Philip Morris International (NYSE: PM ) has negative book value, so it’ll never come up as cheap using price/book, but it could come up as cheap on other signals like dividend yield. 3) Diversified portfolios. Factor investing is a statistical exercise where lots of low-edge bets are aggregated. There’s a trade-off. The deepest tilts are obtained through focused, high-turnover portfolios, but doing so introduces more noise and slippage and reduces capacity. 4) Economic intuition. If a strategy uses rules that don’t make sense, throw it out, even if its back-tested or historical returns look good. I’ve seen some crackpots use astrological signals like “Bradley turn dates” to time their investments. With a few glaring exceptions (that I won’t name here), strategy ETFs don’t use off-the-wall rules. Even if a sponsor has a fantastic strategy that fulfills all these criteria, you want to make sure the sponsor can execute it. To abuse the food analogy more, this is like making sure your meal is made by a capable chef with high-quality ingredients in sanitary conditions–it doesn’t matter if a meal’s mix of nutrients is perfectly engineered if it makes you puke out your guts the next day. Here are some operational characteristics I look for: 1) Low assets in relation to estimated capacity. If you squeeze through a big trade on the open market, you’ll push prices against you. This insidious cost is often bigger than the expense ratio–sometimes by orders of magnitude–but it’ll never show up in a prospectus or annual report. At a certain point, a strategy becomes too bloated to absorb more assets; trading costs eat up too much of the expected excess returns. Unfortunately, market impact costs are hard to predict because firms can mitigate them through block trades (where they hand over a big block of shares to another institution at a modest discount to current market price), internal crossing trades (where they simply swap the stock with another fund in the same firm), and other means. This capacity is a function of an index’s construction, the market’s liquidity, and the fund manager’s transactional capabilities. Vanguard and BlackRock/iShares are probably the best indexers, so all else held equal, you should expect a Vanguard or iShares ETF to have much more capacity than another firm’s ETF. 2) Low fees. Enough said. 3) Stable sponsors. You don’t want to invest with an ax-happy sponsor who might kill your ETF if it doesn’t maintain a certain size. Here are the funds I like, in order of preference: Schwab US Dividend Equity ETF (NYSEARCA: SCHD ) SCHD is my favorite U.S. equity fund at the moment, which might seem strange given that it’s lagged the broad market since inception. However, it has three things going for it. First, its loadings to value and quality factors are sizable, both since its 2011 inception and its 1999 back-tested index inception. That its live and back-tested returns show big and fairly stable value and quality loadings is reassuring. Over its full history, the index exhibited a value loading of 0.6, a quality loading of 0.3, and no momentum loading. However, the fund favors big, defensive stocks, which reduces its expected return. I expect SCHD will largely keep up with the market but with much lower drawdowns and volatility. Second, it’s dirt-cheap in all respects, with a 0.07% expense ratio. The strategy’s turnover averaged 15% since launch. Moreover, the fund is not bloated to the point where its trades push prices around. Finally, its index, the Dow Jones U.S. Dividend 100, is sensibly constructed, even if it’s not what I’d have come up with. The index screens for U.S. stocks that have paid a dividend in each year for the past 10 years. This screen weeds out smaller, less durable, and more-speculative stocks. It then ranks the stocks by annual indicated dividend yield and kicks out the bottom half. The remaining stocks are ranked by cash flow/total debt, return on equity, indicated dividend yield, and five-year dividend-growth rate. The four rankings are equal-weighted, and the 100 stocks with the highest composite ranks are selected for the index. Note that the ranking criteria penalize firms that 1) don’t earn adequate cash flow for each unit of debt they assume, 2) aren’t earning big profits for each dollar of invested capital, 3) haven’t grown their dividends, and 4) are expensive. By using value and quality criteria, the index focuses on cheap quality stocks. I have a few quibbles with the index’s construction. The 10-year dividend screen is arbitrary. Is a 10-year dividend history that much better than a nine-, eight-, or seven-year history? I’d be hesitant to rule out a stock based on a single metric. Apple (NASDAQ: AAPL ) in 2013 was astoundingly cheap and high-quality by numerous metrics but couldn’t have made it into this fund because it resumed its dividend in 2012 after a 17-year hiatus. Vanguard Dividend Appreciation ETF ( VIG ) This is the quintessential quality fund. Its value and quality loadings are 0.1 and 0.3, respectively. Like SCHD, VIG will lag in rallies and do better in downturns. The fund’s main screen or signal is 10 consecutive years of dividend growth. For the most part, only high-quality firms make the cut. However, there are plenty of high-quality firms that haven’t grown dividends for 10 years straight, including Wells Fargo & Co (NYSE: WFC ) . An array of weaker signals is preferable to one exacting signal. Another ding to VIG is its secret quality screens. Vanguard worked with Mergent to craft this index, so the rules are likely to be sensible, but I’d sleep better if I could see them for myself. VIG is cheap, charging only 0.10%. However, the fund’s track record has attracted a torrent of inflows in the past few years. With almost $20 billion in assets, the average stock position of VIG is about twice the stock’s three-month average daily trading volume. It’s only going to get bigger now that Wealthfront, a fast-growing robo-advisor, uses the fund in its default allocations. Despite my concerns, it’s hard to beat VIG’s low fees, capable managers, and sound quality strategy. If you own it in a taxable account and have lots of capital gains, do not sell it. The likely drag from turnover costs is probably less than 0.2% a year. But if you can swap VIG for SCHD at low cost, go ahead. iShares MSCI USA Momentum Factor (NYSEARCA: MTUM ) MTUM is the best momentum factor fund by far. It’s cheap, charging 0.15%, and well-constructed. MSCI provides back-tested history for the index starting in 1975. The index’s momentum loading averaged 0.3 with no value exposure. Its much bigger rival PowerShares DWA Momentum ETF (NYSEARCA: PDP ) charges 0.65% for lower momentum exposure (its higher recent returns are from its higher market beta), an opaque methodology, and a literally unbelievable back-tested history. MTUM buys and overweights stocks with the highest risk-adjusted returns over trailing seven- and 13-month periods, excluding the most recent month (so, months two to seven and two to 13). It rebalances semiannually but can rebalance monthly using the shorter signal if a volatility trigger goes off. This allows the fund to react more quickly to market rebounds (think 2009), when traditional momentum strategies often get slaughtered. iShares MSCI USA Quality Factor (NYSEARCA: QUAL ) QUAL’s back-tested index, which begins in 1976, has a 0.3 quality loading but a negative 0.1–0.2 value loading. In other words, if you pair QUAL with a value fund, you’ll offset some of your value exposure and get a weak quality exposure. QUAL is best for investors who don’t want much value exposure. QUAL buys and overweights U.S. stocks with high return on equity, low debt/equity, and low five-year earnings growth variability. This methodology is similar to GMO’s quality strategy but without a valuation screen. It rebalances semiannually. iShares MSCI USA Minimum Volatility (NYSEARCA: USMV ) USMV’s back-tested index begins in mid-1988. Contrary to what you’d expect, USMV is simply another way to get value and quality exposure. It shows value and quality loadings of 0.2, with below-average market exposure. However, because its methodology doesn’t explicitly target value and quality stocks, its loadings have swung around dramatically over time. The index uses a risk model to estimate variances and correlations for U.S. stocks and an optimizer to create the lowest-volatility portfolio possible given a set of diversification constraints. In layman’s terms, it takes advantage of the fact that some stocks neutralize each other to lower overall portfolio volatility. I have some general comments on iShares’ factor funds. First, the MSCI indexes they use are for the most part transparent, simple, well-diversified, robust, and high-capacity. Second, the funds won’t offer deep tilts when combined. For example, if you own MTUM and QUAL in equal portions, your aggregate quality and momentum loadings will be around 0.15 each, pretty weak. Third, even though these funds aren’t big, billions of dollars are already tracking the MSCI indexes in separate accounts and other vehicles. It’s hard to tell when the strategies will get crowded. iShares Enhanced US Large-Cap (NYSEARCA: IELG ) As far as I can tell, the Enhanced series of ETFs are the first to charge passive fees for true quantitative active equity management. The funds combine several value and quality signals, with some optimizations to target lower-than-average volatility, achieve sector- and stock-level diversification constraints, and keep a lid on turnover. IELG is the cheapest of the funds, charging a 0.18% expense ratio, a price tag that puts it well below other long-only multifactor funds like AQR Core Equity (MUTF: QCELX ) . To elaborate, the Enhanced funds look for stocks with low price/earnings, price/book, earnings variability, debt, and accruals (a measure of earnings quality). Each signal is sensible by itself. I like that no single signal will rule out a stock, an improvement over blunt (but effective) heuristics like 10 straight years of dividend growth. Unfortunately, BlackRock isn’t clear about how stocks are selected and weighted, how often the strategy is rebalanced, or the process by which the model driving the strategy is updated. Preliminary results suggest IELG has a sizable loading to quality but surprisingly little exposure to value. However, given the fund’s low cost, low asset base, logical signals, competent managers, and sound provenance, this young fund is worth betting on. For an investor with a strong belief in factor investing, this is one of my top picks. Disclosure: Morningstar, Inc. licenses its indexes to institutions for a variety of reasons, including the creation of investment products and the benchmarking of existing products. When licensing indexes for the creation or benchmarking of investment products, Morningstar receives fees that are mainly based on fund assets under management. As of Sept. 30, 2012, AlphaPro Management, BlackRock Asset Management, First Asset, First Trust, Invesco, Merrill Lynch, Northern Trust, Nuveen, and Van Eck license one or more Morningstar indexes for this purpose. These investment products are not sponsored, issued, marketed, or sold by Morningstar. Morningstar does not make any representation regarding the advisability of investing in any investment product based on or benchmarked against a Morningstar index.
