A Factor-Based Exploration of the ‘Sin’ Stock Anomaly
While socially responsible investing (SRI) and the broader category of environmental, socially responsible and governance (ESG) continue to gain in popularity, economic theory suggests the share prices of “sin” businesses will become depressed if a large enough proportion of investors choose to avoid them.
Such stocks would have a higher cost of capital because they would trade at a lower price-to-earnings (P/E) ratio, thus providing investors with higher expected returns. (Some investors may view those higher expected returns as compensation for the emotional “cost” of exposure to offensive companies. However, it’s important to point out that SRI and ESG—sometimes referred to as “double-bottom-line” investing—encompass many personal beliefs, not just one set of values.)
Thus, an investment strategy that focuses on the violation of social norms has developed in the form of “vice investing” or “sin investing.” This strategy creates a portfolio of firms from industries typically screened out by SRI and ESG funds, pension funds and investment managers.
Vice investors focus primarily on the “sin triumvirate”: tobacco, alcohol and gaming (gambling) stocks. Historical evidence on the performance of these stocks supports the theory—sin stocks have provided significantly higher returns than stocks in general.
Harrison Hong and Marcin Kacperczyk, authors of the study “The Price of Sin: The Effects of Social Norms on Markets,” published in the July 2009 issue of the Journal of Financial Economics, found that for the period 1965 through 2006, a U.S. portfolio long sin stocks and short their comparables posted a return of 0.29% per month after adjusting for the Carhart four-factor (beta, size, value and momentum) model. As out-of-sample support, sin stocks in seven large European markets and Canada outperformed similar stocks by about 2.5% a year.
The authors concluded that the abnormal risk-adjusted returns of vice stocks are due to neglect by institutional investors, who lean toward the side of SRI.
As further evidence that avoiding sin stocks comes at a price, Elroy Dimson, Paul Marsh and Mike Staunton found that, when using their own industry indexes that covered the 115-year period 1900 through 2014, tobacco firms beat the overall equity market by an annualized 4.5% in the U.S. and by 2.6% in the U.K. (but over the slightly shorter 85-year period 1920 through 2014). Their study was published in the 2015 Credit Suisse Global Investment Handbook.
They also examined the impact of screening out countries based on their degree of corruption. Countries were evaluated using the Worldwide Governance Indicators from a 2010 World Bank policy research working paper from Daniel Kaufmann, Aart Kraay and Massimo Mastruzzi, “The Worldwide Governance Indicators: Methodology and Analytical Issues.” The indicators comprise annual scores on six broad dimensions of governance.
Dimson, Marsh and Staunton found 14 countries that posted a poor score, 12 that were acceptable, 12 that were good and 11 with excellent scores. Post-2000 returns for the last three groups were between 5.3% and 7.7%. In contrast, the markets with poor control of corruption had an average return of 11.0%.
Interestingly, realized returns were higher for equity investments in jurisdictions that were more likely to be characterized by corrupt behaviors. As the authors note, the time period is short and the result might just be a lucky outcome.
On the other hand, it’s also logical to consider that investors will price for corruption risk and demand a premium for taking it. But it may also be a result of the same exclusionary factors found with sin stocks (investors boycott countries with high corruption scores, driving prices down, raising forward-looking return expectations).
A 5-Factor Analysis
With the introduction of the relatively new Fama-French five-factor asset price model (adding investment and profitability to beta, size and value) we gained incremental explanatory power in the cross section of returns.
In his January 2017 paper, “Fewer Reasons to Sin: A Five-Factor Investigation of Vice Stocks,” which covered the period October 1996 to October 2016, Greg Richey employed the single-factor (market beta) CAPM, the Fama-French three-factor (adding size and value) model, the Carhart four-factor (adding momentum) model and the new Fama-French five-factor model to investigate whether a vice stock portfolio outperforms the S&P 500 Index, a benchmark to approximate the market portfolio, on a risk-adjusted basis. Richey’s dataset included 61 corporations from vice-related industries.
Following is a summary of his finding
- For the period October 1996 through October 2016, the S&P 500 Index returned 7.8% per year. The “Vice Fund” returned 11.5%.
- The alpha, or abnormal risk-adjusted return, shows a positively significant coefficient in the CAPM, Fama-French three-factor and Carhart four-factor models.
- All models, including the Fama-French five-factor model, indicate the Vice Fund portfolio beta is between 0.59 and 0.74, in turn indicating that the vice portfolio exhibited less market risk or volatility than the S&P 500 Index, which has a beta of 1, over the sample period. This reinforces the defensive nature of sin portfolios; they have less correlation with economic activity. With the three-factor model and four-factor model, the vice portfolio has a statistically significant negative loading on the size factor (-0.17 and -0.18, respectively) and a statistically positive loading on the value factor (0.15 and 0.21, respectively), indicating that these exposures help explain returns. With the four-factor model, the Vice Fund loaded about 0.11 on momentum, and it was statistically significant. However, with the five-factor model, the vice portfolio’s negative size loading shrinks to just -0.05 and the value loading turns slightly negative, also at -0.05, and both are statistically significant. In the five-factor model, the vice portfolio loads strongly on both profitability (0.51) and investment (0.48). All of these figures are significant at the 1% confidence level.
- The annual alphas on the CAPM, three-factor and four-factor models were 2.9%, 2.8% and 2.5%, respectively. All were significant at the 1% confidence level. These findings suggest that vice stocks outperform on a risk-adjusted basis. However, in the five-factor model, the alpha virtually disappears, falling to just 0.1% per year. This result helps explain the performance of vice stocks relative to the market portfolio that previous models failed to capture.
Richey concluded that the higher returns to vice stocks occurred because they are more profitable and less wasteful with investment than the average corporation.
David Blitz and Frank Fabozzi add further insight into the performance of sin stocks with their study “Sin Stocks Revisited: Resolving the Sin Stock Anomaly,” which appears in the fall 2017 issue of The Journal of Portfolio Management.
In addition to examining the performance of sin stocks through the lens of the Fama-French five-factor model, they also used Andrea Frazzini and Lasse Pedersen’s low-beta versus high-beta betting against beta (BAB) factor, which goes long leveraged low-beta assets and short high-beta assets. Their study covered the U.S. market from July 1963 through December 2016 and the European and Japanese markets from July 1990 through December 2016.
Following is a summary of their findings for U.S. sin stocks.
- Consistent with prior research, the single-factor CAPM alphas are large and statistically highly significant. The monthly alpha is 0.47% with a t-stat of 3.6.
- The alpha remains highly significant after controlling for the size, value and momentum factors, with a significantly positive value loading largely offset by a significantly negative size loading. In the three-factor and four-factor models, the monthly alphas drop slightly to 0.45% and 0.40%, respectively, and the t-stats remain above 3.0 (meaning the results are statistically significant at the 1% confidence level).
- Exposure to the market beta factor is strongly and significantly negative—sin stocks tend to be low-beta stocks, and they have highly statistically significant loadings on the BAB factor. Accounting for beta, size, value, momentum and now BAB reduces the monthly alpha to 0.27% per month with a t-stat of 2.1 (the results are statistically significant at the 5% confidence level). In addition, after accounting for the BAB factor, exposure to the value factor becomes insignificant.
- When viewed through the five-factor lens, while the monthly alpha is still economically significant (0.13% per month), it’s no longer statistically significant (t-stat of 1.1) at the 5% confidence level. And, importantly, sin stocks load highly on the new “quality” factors of profitability (0.64, with a t-stat of 10.9) and investment (5.3, with a t-stat of 6.0).
- When all seven factors (beta, size, value, momentum, BAB, investment and profitability) are used, the monthly alpha drops further (to a still economically significant 0.10% per month) as does the t-stat (to 0.8). Sin stocks are found to have statistically significant negative loadings on the market beta and value factors, and statistically significant positive loadings on the investment, profitability and BAB factors.
Of particular interest is that, when Blitz and Fabozzi examined the more recent period from July 1990 through December 2016, the results were similar for regressions using beta, size, value, momentum and BAB.
However, after including the investment and profitability factors, the monthly alpha disappeared entirely and the explanatory power of the newer factors increased. In their out-of-sample tests (Europe and Japan), they found that sin stocks’ outperformance relative to the market was fully explained by exposure to the newer investment and profitability factors. In other words, the quality factors of investment and profitability explain the sin stock anomaly.
With advances in asset pricing theory to include the new investment and profitability factors, the outperformance of sin stocks appears to be fully explained. This is good news for investors who wish to avoid investing in sin stocks and until now have had to pay a price (excluding assets with above-market returns, and below-market risk in the form of low market beta). These findings point a way for SRI investors to both have their cake and eat it.
As Blitz and Fabozzi note: “Investors may restore their portfolios’ expected return by making sure that the portfolios’ factor exposures do not deteriorate when excluding sin stocks. For example, investors could increase the weights of stocks that are able to compensate for the loss in factor exposures that results from excluding sin stocks—that is, by investing more in non-sin stocks that have exposures to the same factors that drive sin stock returns.” Before you draw that conclusion, however, consider the following:
The premiums related to the newer profitability and BAB factors clearly are behavioral-based (as opposed to risk-based). Like the newer QMJ factor (quality minus junk), high-quality companies have the following traits: low earnings volatility, high margins, high asset turnover (indicating efficient use of assets), low financial leverage, low operating leverage (indicating a strong balance sheet and low macroeconomic risk) and low stock-specific risk (volatility that is unexplained by macroeconomic activity).
While it is true that premiums in behavioral-based factors can persist due to well-documented limits to arbitrage that prevent sophisticated investors from correcting mispricings, these limits to arbitrage exist mostly in stocks costly to trade and to short (they are hard to borrow and borrowing costs are high). High-quality stocks (meaning they load highly on the investment and profitability factors) tend to be large companies not as subject to limits to arbitrage.
Additionally, because the literature demonstrates the publication of academic research leads to the deterioration of premiums (about one-third on average, and more for easy-to-arbitrage stocks), it is certainly possible that the historical premium to low beta-stocks and profitable stocks could shrink.
On the other hand, as more investors express their personal beliefs through their investments, shunning sin stocks, it seems likely their prices would be further depressed, further raising their forward-looking return expectations.
Thus, it is possible the sin stock premium (relative to the market) could not only persist, it could increase, and the investment and profitability factors may no longer be able to fully explain sin stocks’ returns.
This commentary originally appeared December 22 on ETF.com
By clicking on any of the links above, you acknowledge that they are solely for your convenience, and do not necessarily imply any affiliations, sponsorships, endorsements or representations whatsoever by us regarding third-party Web sites. We are not responsible for the content, availability or privacy policies of these sites, and shall not be responsible or liable for any information, opinions, advice, products or services available on or through them.
The opinions expressed by featured authors are their own and may not accurately reflect those of the BAM ALLIANCE. This article is for general information only and is not intended to serve as specific financial, accounting or tax advice.
© 2018, The BAM ALLIANCE