Indexing pioneer Vanguard Group may be stock pickers' last hope.
Investors are increasingly turning their stock picking over to computers. So-called smart beta exchange-traded funds track indexes that replicate traditional styles of active management such as value, quality, and momentum. Investors handed $184 billion to smart beta ETFs from 2015 to 2017 while pulling $308 billion from equity mutual funds, according to data compiled by Bloomberg Intelligence.
It's not surprising. Smart beta ETFs are cheaper, and investors are skeptical that human stock pickers can beat the bots by more than the difference in fees. According to Morningstar data, the average expense ratio for smart beta ETFs is 0.47 percent a year, and the asset-weighted average expense ratio — which accounts for the size of the ETFs — is just 0.26 percent. That compares with 1.13 percent and 0.7 percent, respectively, for actively managed mutual funds.
Vanguard, however, is changing the calculation. It recently rolled out actively managed ETFs with fees that are comparable to those of smart beta funds.
Vanguard's U.S. Value Factor ETF, for example, charges 0.13 percent a year, compared with 0.06 percent for Vanguard's smart beta Value ETF. That's a big deal because the stock pickers behind the U.S. Value Factor ETF don't have to overcome a daunting fee hurdle to compete.
But that doesn't mean the funds are interchangeable. Even though they are both "value" funds, the U.S. Value Factor ETF gives investors more exposure to the value style than the smart beta Value ETF. That's an important difference.
One way to gauge a fund's exposure to a given style, or factor, is to look at the stocks it holds. Bloomberg ranks stocks based on various factors. Value, for example, uses metrics such as price-to-book ratio, and growth uses measures such as earnings growth. Stocks are ranked on a scale of negative 3 to positive 3 for each factor. A value score of 1 means that a stock is cheaper than roughly 84 percent of stocks — or, in technical terms, 1 standard deviation cheaper than the average stock.
Here's how it works in real life. JPMorgan Chase & Co.'s price-to-book ratio is 49 percent lower than that of the S&P 500, so it's likely to rank high for value. At the same time, the growth of JPMorgan's earnings per share has lagged that of the S&P 500 by 3.5 percentage points over the last five years, so it's likely to rank low for growth. By contrast, Facebook Inc.'s rankings on value and growth are likely to be just the opposite.
With those factor scores, a weighted-average score can be calculated for any fund. The numbers are instructive. Vanguard's U.S. Value Factor ETF has a value score of 0.99, according to Bloomberg data. By contrast, its smart beta Value ETF has a score of 0.31, or far less value exposure.
Value investing has historically delivered higher returns and volatility than the broad market. Because of its higher exposure to value, the U.S. Value Factor ETF is likely to deliver a bumpier ride and possibly higher returns over time. All else equal, it's also likely to beat the smart beta fund when value outperforms and lag when value underperforms.
Those are crucial insights. Consider the Zevenbergen Genea Fund. Its institutional share class was the best performing U.S. large-cap growth mutual fund over the last year through Tuesday, according to Morningstar, with a total return of 66.3 percent. That compares with a return of 25.9 percent for the Russell 1000 Growth index.
It's not a coincidence. Growth outpaced the market over the last year, so more exposure to growth helped boost returns. Zevenbergen's institutional share class has a whopping growth score of 2.25, compared with just 0.42 for the growth index.
Inversely, funds with less exposure to growth paid a price. The Fairholme Fund was the worst performing U.S. large-cap value mutual fund over the last year. It returned a negative 9.4 percent through Tuesday, compared with 7.8 percent for the Russell 1000 Value Index. Part of the reason is that Fairholme is lighter on growth than the value index. It has a growth score of negative 1.46, compared with a negative 0.52 for the index.
Investors have become smarter about fees. It's only a matter of time before they're smarter about the factor exposures in funds. And I suspect that smart beta ETFs will soon allow them to choose not only among styles, but also the amount of exposure to those styles.
The question then will be whether stock pickers can beat a bot armed with comparable factors. With fees out of the way, Vanguard's stock pickers are set to offer the first clues.