Over the past few years, investors and advisors have been presented with a question: Can smart beta be made smarter?
Smart beta, as you no doubt know by now, is an investment approach — usually wrapped in an exchange-traded fund — that’s a sort of middle ground between active and passive management. A smart beta fund employs a mechanistic index methodology designed to play up an investment factor, or attribute, such as low volatility or high dividends. Some funds’ index strategies synthesize a couple of these attributes. Low volatility stocks together with high dividend payers, for example, is an equity strategy exploited by both the PowerShares S&P 500 High Dividend Low Volatility Portfolio (NYSE Arca: SPHD) and the Legg Mason Low Volatility High Dividend ETF (Nasdaq: LVHD).
The first true multifactor ETF was launched in 2011. “Factor,” in this context means a tilt in the direction of high book-to-market value and small capitalization consistent with the Fama-French model.
A multifactor ETF differs from style-tilted portfolios, such as the iShares S&P Small-Cap 600 Value ETF (NYSE Arca: IJS), which were floated a decade before by dint of scope and dynamism (see “Growth Or Value? Which Way To Lean?”).
IJS tracks a committee-determined index of small-cap stocks exhibiting low price-to-book, price-to-sales and price-to earnings ratios; IJS’ investable universe is limited to small-cap stocks. A multifactor fund, such as the FlexShares Morningstar U.S. Factor Tilt Index ETF (NYSE Arca: TILT), quantitatively culls the entire market — across all capitalization tiers — to find those stocks that, when combined, best express the small-cap and value factors. You won’t find large-cap stocks in IJS, but you will find them in the TILT portfolio.
Being the first to market, TILT is the granddaddy of domestic multifactor ETFs. TILT manages 32 percent of the $3.3 billion now invested in domestic multifactor ETFs.
Eight more multifactor ETFs were launched between February and November 2015, each one sifting stocks through its unique screen. Some use three factors, others four or five. The greater the number of factors, the more diversification is obtained provided the factors are not highly correlated to one another. As you might expect, these disparate portfolios produce differing investment results.
Different Strokes
The Hartford Multifactor U.S. Equity ETF (NYSE Arca: ROUS) scores large-cap stocks for value, quality and momentum in an integrated process which tends to produce deep factor exposure. ROUS carries significantly more active risk than many other multifactor ETFs. Over the past year, ROUS matches up best as a value play.
The best alpha producer of the lot is the SPDR MSCI USA StrategicFactors ETF (NYSE Arca: QUS). Unlike ROUS, the QUS portfolio is built in “sleeves” with its three factors — quality, value and low volatility — represented by sub-portfolios, each sleeve accorded equal weight. Low volatility has been its standout factor since 2016.
There’s great similarity in the behavior of the iShares Edge MSCI Multifactor USA ETF (NYSE Arca: LRGF) and that of the Hartford ROUS portfolio. Both funds bring value and, secondarily, momentum upfront to the same degree. LRGF optimizes its four factors — quality, momentum, value and size — to keep the portfolio’s expected risk close to its benchmark.
Next to neutrality, low volatility is the prominent exposure in the four-factor Global X Scientific Beta U.S. ETF (NYSE Arca: SCIU). While not a factor explicitly sought by this or any other multifactor ETF, neutrality is an attribute (and a substantial one at that) of the broad market. About 30 percent of the factor exposure embedded in the SPDR S&P 500 ETF (NYSE Arca: SPY) is neutral. SCIU nearly mirrors that.
The only domestic multifactor ETF that’s cranked out negative alpha over the past year is the Goldman Sachs ActiveBeta U.S. Large Cap Equity ETF (NYSE Arca: GSLC). GSLC is built on four equal-weighted sleeves: value, quality, low volatility and momentum. The fund’s underperformance may be due to the current expression of a low-value — or growth — tilt. GSLC is big; it owns a 45 percent share of the domestic multifactor ETF segment.
Nearly half of the factor bias articulated by the John Hancock Multifactor Large Cap ETF (NYSE Arca: JHML) is value. Given the fund’s three-factor stool — size, value and profitability— that’s no surprise. What is surprising is the unintended influence of momentum over the past year, given that the fund doesn’t screen for that factor. Momentum accounts for nearly a third of the fund’s current factor attributions.
Large-cap stocks are screened for three factors — value, quality and momentum — in the JPMorgan Diversified Return U.S. Equity ETF (NYSE Arca: JPUS) then weighted by inverse volatility on a sector-by-sector basis. Giving the least volatile stocks larger weightings lends a defensive posture to the fund.
The Deutsche X-trackers Russell 100 Comprehensive Factor ETF (NYSE Arca: DEUS) scores large-cap stocks for high value, momentum and quality as it screens for low volatility and smaller size. The net result is a heavy overweight: more than half of the fund’s factor exposure is value. Of all the multifactor ETFs described here, in fact, DEUS is the big value play.
… and different outcomes
Set against SPY, multifactor funds have collectively produced better-than-market returns, albeit with a tad more volatility. Nearly all produced positive alpha. All this would seem to make a compelling case for an allocation, but care must be taken to fully appreciate the portfolio impact. Each fund contributes tracking error, or active risk, which can pull a portfolio in unwanted directions. Portfolio owners, too, need to watch out for inadvertent factor overloads when they allocate to these ETFs. It’s not enough to assume that the factor tilts sought by an ETF will be actually expressed. Sometimes, the synthesis of the factor search is an unexpected tilt. Often, that’s a result of the fund’s methodology.
There are two basic scaffolds for a multifactor fund: the sleeve approach and the integrated approach. Quants argue endlessly about the merits of each, but there seems a consensus around the notion that the integrated approach — a bottom-up methodology that simultaneously screens the investment universe for factors rather than building sub portfolios — yields superior risk-adjusted returns when high exposures are sought or higher tracking error is countenanced. Stock-specific risk tends to be more pronounced in the sleeve approach.
A fund’s tracking error should inform an investor’s deployment decision. ETFs with low tracking error consume smaller chunks of a risk budget and thus can be more readily used to complement other portfolio exposures; multifactor products with more active risk are better suited as core holdings.
Two-thirds of our multifactor ETFs are best described as value tilts (see Table 2). These, too, are the funds that have cranked out the highest alpha coefficients recently. So it seems that there’s something to be said for ramping up a portfolio’s value exposure. Well, yes, for now. Factors are cyclical creatures. When value’s a winner, momentum often lags. It works the other way ‘round, too. Combining value and momentum attributes provides a more pronounced portfolio diversification benefit.
Multifactor ETFs can provide superior risk adjusted returns, but they also add a level of complexity to portfolio management. It’s vital that investors know what they're buying, especially the impacts these products have on their active risk levels and aggregate factor exposures. Investors should also appreciate the persistency of factors. While factor effects are cyclical, rotations can be lengthy, producing extended periods of underperformance or outperformance.
Brad Zigler is WealthManagement's Alternative Investments Editor. Previously, he was the head of Marketing, Research and Education for the Pacific Exchange's (now NYSE Arca) option market and the iShares complex of exchange traded funds.
