Professor, Columbia University & Managing Partner, Sauma Capital LLC
Welcome to the second issue of the twelfth volume of The Journal of Investment Strategies, which contains three research papers.
The issue’s first paper, “Implementing mean–variance spanning tests with shortsales constraints”, underscores the shortcomings of Wald tests, which are frequently employed for short-sales considerations. Farid AitSahlia, Thomas Doellman and Sabuhi Sardarli illustrate the numerical instability of these tests. To address this concern, they harness the distinctive property of the stochastic discount factor when a risk-free rate exists, offering a more robust testing approach. A significant revelation from their study is the misalignment of several of the Wald tests used in the US retirement plan literature with the principles of mean–variance optimality. When correctly applied, these tests produce markedly different results. Emphasizing the prevalence of regression-based mean–variance spanning tests in empirical finance, AitSahlia et al delve into the intricacies of applying these tests amid short-sales constraints. This is especially pertinent for defined-contribution retirement plans, such as the 401(k) in the United States. Advocating for an enhanced testing approach, the authors exploit the unique mean discount factor, inherent when a risk-free asset is present, to alleviate potential numerical instabilities. They further spotlight the pronounced outcome disparities arising from erroneous implementations in earlier retirement plan research.
Our second paper, “What have we learned from 20 million historical US stock data?” by Mostafa K. Ardakani, delivers a comprehensive analysis of the US stock market from January 1995 to mid-June 2021. Unlike previous studies, which have typically focused on select indexes such as the Standard & Poor’s 500, this paper examines a vast data set comprising approximately 20 million end-of-day data points from 9071 stocks. The research unveils discernible calendar effects. For example, Mondays exhibit the lowest percentage returns, while Fridays register the highest. Interestingly, volatility is highest on Mondays and tapers by Friday, making Friday optimal for daily trading. In terms of months, January boasts the highest percentage returns, with October lagging furthest behind; however, both months show heightened volatility. Appendixes A and B of this paper (which are only available online) shed light on end-of-day and end-of-month percentage returns, respectively, offering valuable insights for potential investment strategies. Notably, day trading is discouraged, with weekly and monthly trades emerging as superior strategies. A pivotal observation is the rising correlation among stocks in recent years, possibly attributed to the surge in popularity of exchange-traded funds. This heightened correlation signifies that stock prices increasingly move in tandem, which can amplify fluctuations in the overall market, potentially exacerbating the snowball effect.
In “An empirical study of the contrarian strategy against US equities in the Japanese market”, the final paper in this issue, Yasuhiro Iwanaga delves into a rulebased contrarian strategy between the US and Japanese stock markets. Iwanaga’s core contrarian strategy posits that Japanese stock futures should be shorted at the opening price and closed at the closing price if the previous day’s daily return on US stocks is positive, and the inverse if US returns are negative. The empirical evidence confirms a reversal effect against US equities, showing that this contrarian strategy can yield statistically significant positive returns. To enhance the strategy, Iwanaga introduces the concept of the investment environment in both markets. For instance, if Japan-specific news is positive or its fundamentals are stronger than those of the US, Japanese stock index futures might experience elevated intraday returns compared with after-hours returns. Critically, a reversal effect against US equities is most pronounced when the US and Japanese markets operate under distinct investment environments. Incorporating abnormal after-hours return data further optimizes the strategy, outperforming the standalone contrarian approach against US equities. However, Iwanaga’s analysis focuses solely on Japanese stock index futures. Several avenues for future research are proposed, including investigating whether stock index futures in other countries show similar patterns to Japan regarding after-hours and daily returns; analyzing the reaction of Japanese stock index futures to stock indexes of countries other than the United States; and examining individual Japanese stock reactions to the previous day’s daily returns on the S&P 500 index. While Iwanaga’s analysis is foundational, it offers valuable insights for future investment decisions and advanced quantitative analysis.
On behalf of the editorial board, we extend our gratitude to you, our valued readers, for your unwavering support and interest in this journal. We are excited to present a growing collection of practical papers on diverse topics related to modern investment strategies, sourced from both academics and industry experts.
The authors demonstrate that Wald tests are prone to numerical instability when accounting for short sales.
The author offers a statistical characterization of the US stock market from January 3, 1995 to June 11, 2021.
This paper investigates the contrarian strategy against US equities, finding that for samples where the previous day's daily return on the S&P 500 is positive (negative), the next day's intraday returns on Japanese stock-index futures will be the inverse…