As noted in prior chapters, many researchers have found that some anomalies or factors have occurred largely during the month of January, or during the turn of the month. Marc Reinganum (considered the founding father of stock market anomalies) found that January accounted for nearly half of the total returns for the smallest cap stocks, but less than that for other small-cap stocks.1 In the 1980s Bruce Jacobs and Ken Levy found that a large percentage of the returns between large and small companies occurred in the month of January. Additionally, half of the January returns occur in the first few days of January. Some researchers refer to this as the "small-firm-in-January" effect. Jacobs and Levy also found that during some periods, the entire dividend effect occurred in January and that the Low P/E, small size, and neglect anomalies all existed independently.2
A relatively recent study on interrelationships of factors was by Malcolm Baker, Ryan Taliaferro, and Terence Burnham. They computed optimal allocations to four tilts using 1968–2014 data, with value, small size, high profits, and low beta receiving shares of 20%, 26%, 23%, and 24% respectively.3
Using data from 1980 to 2015, Timotheos Angelidis and Nikolaos Tessaromatis argued that you can use factors to design portfolios by country rather than by individual stocks and achieve better than benchmark performance. They noted that research on individual stocks supports the view that small-cap, value, high-momentum, and low-risk factors have outperformed.4
Campbell Harvey, Yan Liu, and Heqing Zhu published a paper in 2016 evaluating hundred of published papers, plus working papers resulting in over 300 different factors that explain returns (some are highly correlated). They proposed a multiple test framework that creates a much higher hurdle for a factor explaining stock returns. That led to their conclusion that most of the research claiming to explain returns was likely false. Value and Momentum were among the few that exceeded the hurdle.5 But some other researchers have suggested momentum is not a distinct risk factor since it aggregates autocorrelations found in other factors.6
In February of 2016 Davis McLean and Jeffrey Pontiff published a paper analyzing stock factors of 97 variables out of sample from their publications studies, and after publication. Returns were 26% weaker out of sample and 58% lower after publication.7 In other words, more than half of expected excess returns from the factors were not found after the publication of the factor. Regarding trading costs and fund size, a 2019 paper estimating the impact of transactions costs on factor investing. They estimated funds will incur 0.3% of market impact for every 10% of the respective stocks average daily volume (momentum strategies generally have the highest costs among the factors).8
The Value Line Anomaly & Implementation Shortfall
Value Line is a service that historically ranks stocks from 1 to 5 for timeliness. As a group, each rating has historically outperformed the next lowest rated group (the ones have outperformed the twos, which outperformed the threes, etc.). The impressive performance of the rating system led many to refer to it as the "Value Line Anomaly" or the "Value Line Enigma."
Some researchers have argued that Value Line's outperformance resulted from their use of earnings surprises and price momentum while others have suggested that the model portfolio's outperformance resulted from assuming higher risk.
The initial Value Line results were so impressive that Value Line was the subject of a complimentary 1973 article in the Financial Analysts Journal by Fischer Black titled "Yes, Virginia, There is Hope: Tests of the Value Line Ranking System." In the article, Black confessed that previously he had been a strong believer in the efficient market hypothesis and passive management. Yet his research of Value Line's rating system, confirmed that the system did produce significant excess returns over a five year period. Excess returns would have resulted even after taking two percentage points out for round trip transactions costs (turnover in the rankings is high).
In an attempt to match the returns of the top rated stocks, Value line established a mutual fund. The results of the Value Line Centurion fund (which invested in 100 Group 1 stocks and the top 100 of 300 Group 2 stocks) may serve as an important lesson for investors. Not only did the real-money fund not keep pace with the paper returns from the top rated stocks (which continued to outperform on paper), it didn't even outperformed the market.
According to David Leinweber, the ValueLine paper portfolio had an annualized return of 26%, but the real ValueLine fund had an annualized return of only 16% from 1979 to 1991.9 In other words, while Value Line seemed to have an ability to pick stocks well, the paper returns weren't realizable by the mutual fund (and possibly Value Line subscribers). Robert Salomon Jr. concluded in “Value Line's self-defeating success" in Forbes on June 15, 1998, "Value Line's rankings are a prisoner of their own success: They work so well that too many people try to act on them."
Some have theorized that the failure of the Value Line fund to keep up with the model portfolio demonstrated implementation shortfall. Transactions costs can significantly reduce returns particularly in portfolios with high turnover. Investors also have to account for the bid-ask spread and mutual funds historically had an added burden of not being 100% invested because of the need to maintain cash reserves (more recently derivatives and other tools have been used to reduce or eliminate underexposure from cash reserves). It’s also important to note that a court order in the 1960's mandated a delay between publication of Value Line rating changes and trading in the portfolio. Investors could interpret this as an example of how difficult it may be to profit from an anomaly.
The Value Line experience is one of the many examples of "implementation shortfall" (previously discussed in chapter 7) demonstrating that finding a market inefficiency doesn’t necessarily translate into outperformance. Andre Perold and others have written extensively on the reality that strategies that seem to offer an investor an advantage may not work in real world conditions because of transactions costs and other costs. Perold has also argued that the larger a portfolio is, the harder it is to exploit any informational advantage.
The difference between paper portfolios and real-money portfolios is a separate issue from the question of whether past outperformance will continue in the future. Investors that attempt to utilize strategies going forward based on back-tested models must deal with both "implementation shortfall" and the performance persistence issue. Strategies that continue to outperform with real money after being initially discovered via back-testing are perhaps the exception rather than the rule.
Notes
1. Marc R. Reinganum, "The Size Effect: Evidence and Potential Explanations," Investing in Small-Cap and Microcap Securities, Association for Investment Management and Research, 1997.
2. Bruce I. Jacobs and Kenneth N. Levy, CFA, "Disentangling Equity Market Returns," Equity Markets and Valuation Methods, The Institute of Chartered Financial Analysts, 1987.
3. Malcolm Baker, Ryan Taliaferro, and Terence Burnham Optimal Tilts: Combining Persistent Characteristic Portfolios Financial Analysts Journal Fourth Quarter 2017 https://www.cfapubs.org/doi/pdf/10.2469/faj.v73.n4.1
4. Timotheos Angelidis and Nikolaos Tessaromatis, Global Equity Country Allocation: An Application of Factor Investing, Financial Analysts Journal, Fourth Quarter 2017 https://www.cfapubs.org/doi/pdf/10.2469/faj.v73.n4.7
5. Campbell Harvey, Yan Liu, and Heqing Zhu "... and the Cross-Section of Expected Returns", Review of Financial Studies, January 2016 https://faculty.fuqua.duke.edu/~charvey/Research/Published_Papers/P118_and_the_cross.PDF
6. Sina Ehsani and Juhani Linnainmaa, Factor Momentum and the Momentum Factor, 2019 https://www.nber.org/papers/w25551 or https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3014521
7. R. David Mclean, Jeffrey Pointiff, Does Academic Research Destroy Stock Return Predictability? February 2016 Journal of Finance http://onlinelibrary.wiley.com/doi/10.1111/jofi.12365/full
8. Feifei Li, Tzee-Man Chow, Alex Pickard, and Yadwinder Garg, Transaction Costs of Factor-Investing Strategies, Financial Analysts Journal, Second Quarter 2019 https://www.cfainstitute.org/en/research/financial-analysts-journal/2019/0015198X-2019-1567190
9. David Leinweber, “Using Information From Trading in Trading Portfolio Management," The Journal of Investing, Summer 1995
I am offering the online version of The Peaceful Investor using "The Honor System." If you don't plan to purchase a version of the book, yet you think it was worth your time and you learned a significant amount, you can tip or compensate me in a number of ways. This will probably not be tax deductible for you, but I will report and pay taxes on any payments.
- Paypal me at gkarz@aol.com
- Pay me by credit card. Email me (proficient at aol.com) with an amount you'd like to donate and I will email you an invoice via Square.com. I will not have access to your information - you would enter it on the secure square website.
- Before you make purchases at Amazon, link through one of my links. As an Amazon Associate I earn from qualifying purchases.
- Contact me to inquire about a review or second opinion of your finances.
If you don't think the material was worthwhile, I would appreciate emails letting me know what you read and whether you disagree with anything in particular. I would also appreciate anyone letting me know if they find any typos, mistakes, or suggestions how to improve the material.
