A look at finance professor Wesley Gray's new exchange-traded fund.
This article was published in the November 2014 issue of Morningstar ETFInvestor. Download a complimentary copy of ETFInvestor here.
ValueShares US Quantitative Value QVAL has an unusual history. Its creator, Drexel University finance professor Wesley Gray, began his career as a money manager while studying for his Ph.D. at the University of Chicago under Eugene Fama. Friends and family gave Gray around $3 million to run in a systematic value strategy. Gray also wrote the Empirical Finance Research blog. Eventually, he began providing free stock screeners based on academic research. He graduated to publishing his own research on stock-picking rules. It culminated in a brick of a book fittingly called Quantitative Value, coauthored with value investor Tobias Carlisle and published in late 2012. In it, the two survey the literature on stock-picking signals and synthesize their own strategy. Around the time the book was published, Gray's firm began offering tax-managed separate accounts implementing the strategy. Gray soon realized there was an even more tax-efficient vehicle: the exchange-traded fund.
Many ETFs make heavy use of a tax arbitrage enabled by the in-kind creation/redemption mechanism. When an authorized participant redeems a basket of ETF shares for the underlying securities, the ETF sponsor can select shares with the highest tax liabilities (that is, the lowest cost bases). The transaction is done "in-kind" and so is not considered a taxable event by the IRS. A clever ETF sponsor can reduce or even eliminate year-end capital gains distributions to investors. Many high-turnover ETFs have never distributed capital gains.
At least one quantitative hedge fund noticed and briefly contemplated launching high-turnover ETFs itself. However, doing so would have required daily portfolio transparency. Gray's firm made the plunge when it launched QVAL on Oct. 22, 2014.
Few funds come with expository tomes. Fewer still are comprehensive and reveal a penetrating mind. Quantitative Value demonstrates a common-sense understanding of what academics call anomalies. Academics still puzzle over how the historical outperformance of value stocks can be reconciled with asset-pricing models grounded in efficient markets. (Contrary to what some would have you believe, the notion that value stocks outperform solely because they are riskier is not the academic consensus.) Gray believes value strategies exploit behavioral biases. His true mentor is not Fama, but Benjamin Graham.
Gray and Carlisle understand the pitfalls of using historical data. People who don't deal with data every day don't realize that it does not appear in a state of perfection but is often incomplete, inconsistent, and error-ridden. Almost every quantitative investor has to make judgment calls about how to deal with these problems. In general, the two show a good understanding of the problems affecting historical stock market data, an understanding that does not seem to be shared by most analysts running around with back-tests.
Around the time I read Quantitative Value, Gray's firm filed to launch QVAL. I was impressed enough to write at the time that "if the ETF comes out with an expense ratio of less than 0.50% and gains at least $100 million in assets, I will likely reallocate most of the [ETFInvestor's model portfolio's] U.S. equity position to it." ("Swaps," ETFInvestor, December 2013.)
The bad news is the fund is more expensive than I'd have liked (I'm cheap), charging 0.79% on assets per year. The price is, as Warren Buffett puts it, within a big zone of reasonableness, when you consider what you get in return for what you pay. One way to see QVAL is as a hedge-fund-style quant value strategy that just happens to be an ETF. Most quant ETFs use dead-simple rules and do not have managers actively tending their models. Before I get into whether you should buy it, let's discuss its methodology.
In a nutshell, the fund does the following each quarter:
1) Start with large- and mid-cap stocks listed on the NYSE, Nasdaq, and AMEX. Exclude stocks with market caps below the 40th percentile of the NYSE. Exclude mortgage REITs, royalty trusts, ETFs, ETNs, CEFs, ADRs, and SPACs. Exclude financials.
2) Attempt to clean the universe of frauds, manipulators, and bankrupt companies. This includes stocks with operating cash flows that persistently lag net income, rapidly changing financial statement ratios, lots of leverage, and rapid sales growth.
3) Sort the remaining stocks by earnings before interest and taxes (EBIT) over total enterprise value (TEV), or EBIT/TEV, where TEV = market cap + debt - excess cash + preferreds + minority interests. Pick the top decile.
4) Assign economic moat and financial strength scores. The economic moat score is determined by long-term measures of free cash flow, return on capital, and margin. The financial strength score is a modified Piotroski F score, composed of 10 items: current profitability (3), stability (3), and recent operational improvements (4). Average the two. Pick the stocks scoring in the top half.
In back-tests, much of the strategy's value comes down to its use of EBIT/TEV, a measure popularized by investor Joel Greenblatt as a component of his "magic formula" in his 2005 The Little Book That Beats the Market. (Look beyond Greenblatt's knack for picking cringeworthy names; his book You Can Be a Stock Market Genius is lauded by many hedge fund managers as the bible on special-situations investing.) EBIT/TEV has in back-tests more discriminatory power than traditional valuation ratios like price/earnings and price/book. EBIT/TEV and its cousin EBITDA/TEV (earnings before interest, taxes, depreciation, and amortization) indicate the yield you'd get from acquiring all the equity and liabilities of a firm, or the yield you'd get if you took the firm private. EBIT/TEV is a clear improvement on P/E in that it adjusts for distortions caused by different capital structures and tax regimes.
EBIT/TEV is very hard to improve upon on paper. Standard practice for testing a valuation signal is to sort all the stocks in a universe, bucket them into deciles (lowest 10%, the next 10%, and so on), weight the stocks by market value (or equal-weight them), and repeat the procedure each year. A signal will ideally show two characteristics: 1) a wide spread in returns between the highest and lowest deciles, and 2) monotonicity, or steadily increasing returns as you move from the low to high buckets. EBIT/TEV beats signals like price/book and price/earnings in both respects, and handily.
I am less concerned about EBIT/TEV's historical results being a statistical fluke because it has logical appeal (but it's not clear to me why it should be better than EBITDA/TEV), has been identified by multiple independent researchers, and has been found in various markets and over long periods.
As far as I know, no ETF uses EBIT/TEV as a trading signal. Greenblatt launched mutual funds in 2009 that followed his "magic formula," but he merged those funds into the the new long-short funds under the Gotham name that charge 2% expense ratios. Live performance of both the Formula Investing and Gotham funds has been promising. The Gotham funds have posted stunning results over their short histories. However, if I had to choose between Greenblatt's Gotham funds and QVAL, I'd pick QVAL hands down for its 1) greater transparency, 2) lower fees, and 3) crushingly superior tax efficiency.
Of course, the ETF investor's next-best alternative is not the Gotham funds, but low-cost quant-lite ETFs like Schwab US Dividend Equity SCHD. Here we enter the quagmire of more qualitative judgments. My procedure so far for assessing quant equity ETFs is to look for ones that offer high loadings to the trifecta of what I consider solid factors: value, momentum, and quality. The funds that offer the deepest, most-efficient loadings win. By this standard, SCHD is near the top, aided by its 0.07% expense ratio, hefty value and quality loadings, and neutral momentum loading.
In back-tests, the quantitative value strategy would not pass this hurdle on a pure efficiency basis. While factor regressions show decent value and quality loadings, they also show a big and persistent negative momentum loading. If we ignore the large, statistically significant alpha, the strategy is not an outstanding value like SCHD.
But the evidence behind QVAL's strategy is strong enough to take the positive alpha estimates seriously. Once you accept the existence of some anomalies, you slide down the slippery slope to at the very least contemplate the existence of more anomalies and improvements to the way you capture known ones. After all, the original evidence for the size, value, and momentum effects was no better than--in fact, considerably worse than--the evidence for newer anomalies like profitability/quality. Historical data was less clean, shorter, and originally limited to the United States.
So, should you invest in QVAL? For investors who believe markets aren't efficient, find Gray's view of the world persuasive, and can tolerate tracking error, by all means jump in--but please read his book first. However, most investors should at most begin with a modest starter allocation until they truly believe in the process and have high confidence in Gray and his team.
Disclosure: Morningstar, Inc.’s Investment Management division licenses indexes to financial institutions as the tracking indexes for investable products, such as exchange-traded funds, sponsored by the financial institution. The license fee for such use is paid by the sponsoring financial institution based mainly on the total assets of the investable product. Please click here for a list of investable products that track or have tracked a Morningstar index. Neither Morningstar, Inc. nor its investment management division markets, sells, or makes any representations regarding the advisability of investing in any investable product that tracks a Morningstar index.