Recent Academic Research – substack.com

Welcome back to another issue of Recent Academic Research!
I just launched my paid offerings. You can view them and a sneak peak to paid content here. I spend a lot of time curating paid posts and these weekly free posts, and I would greatly appreciate any support. Other paid subscribers and I are confident of the immense value of the paid offerings to improve your trading systems and market knowledge.
Prices will be raised on January 18th. Consider subscribing now for a discounted price. With that said, let’s get into it.
Paper Title: Crowding Out and Market Declines: The True Cost of Congress
Authors & Date: Yosef Bonaparte | 01/03/2025
Link to Abstract & Paper
This paper examines how U.S. Congressional activity impacts stock market performance, finding that markets consistently underperform on days when Congress is in session. The study also highlights a strong link between Congressional approval ratings and market performance: a 10% increase in disapproval ratings correlates with stock market declines of up to 17.25%.
This effect is amplified by a crowding-out phenomenon, where media focus on legislative activity overshadows other market-relevant news, driving investor pessimism. Smaller-cap stocks and sectors like coal, healthcare, and utilities are particularly affected, emphasizing the significant role political sentiment and cycles play in market dynamics.
Clearly, politics are a significant driver of market performance. This relationship was evident recently when the Federal Reserve’s Minutes highlighted increased inflation risks tied to Donald Trump’s economic policies, such as tariffs.
The more media attention Congress receives, the more pessimistic investors become, leading to larger declines in stock performance. Also, equal-weighted indices are more affected than value-weighted indices, showing the greater impact on smaller-cap stocks. Clearly, the sentiment towards Congress by the American citizen has a large impact on overall market performance.
Paper Title: Perfect Recession Predictors
Authors & Date: Anthony M. Diercks, Daniel Soques, & Jing Cynthia Wu | 01/07/2025
Link to Abstract & Paper
This paper identifies perfect recession predictors—yield spreads that have signaled every U.S. recession since 1962 without false positives. By analyzing 645 million combinations of Treasury yield spreads, the authors find 83 perfect predictors, primarily based on forward spreads and moving averages.
One key example is the 4-year forward 1-month rate minus the current 1-month Treasury rate (with a 1 year moving average), which reliably predicts recessions when it inverts. These predictors outperform traditional measures like the 10-year minus 2-year spread, which often generate false signals. The study highlights the importance of focusing on long-term forward rates for short-term yields, combined with smoothing techniques, to achieve more accurate recession forecasts.
Many of you enjoyed last week’s paper on recession indicators, so when I came across this one, I knew I had to feature it. Honestly, it might be one of the most fascinating papers I’ve read so far.
Now, I wouldn’t treat these indicators as the ultimate crystal ball—there’s a chance of extreme overfitting here. That said, the ‘perfect predictors’ all share a strikingly similar structure: forward spreads starting just over 4 years out, smoothed with a 1-year moving average, and tied to short-term rates (1-5 months).
Oh, and in case you’re wondering, the authors predict a “strong signal” of a U.S. recession by August 2025, based on data through August 2023. Mark your calendars?
Paper Title: Causal Hangover Effects
Authors & Date: Andreas Santucci & Eric Lax | 12/30/2024
Link to Abstract & Paper
This paper examines how visiting "party cities" like New York or Los Angeles impacts athletic performance in the NBA and MLB. Teams underperform the day after visiting these cities, struggling to meet point spreads in the NBA and winning less often in the MLB.
The authors attribute this to nightlife-related fatigue and travel effects, with the impact dissipating after 24 hours of recovery. These findings highlight how off-court factors influence performance and even offer insights for profitable betting strategies.
I haven’t featured a paper on sports betting before, but when I came across this one, I knew it had to be included. There’s a statistically significant relationship showing that sports teams underperform after traveling to cities with a vibrant nightlife!
The authors suggest this could be due to fatigue and reduced cognitive performance from nightlife-related activities (e.g., partying, alcohol consumption, or sleep deprivation) and/or travel-related fatigue, including jet lag and short recovery times between games. While it may not be entirely due to excessive partying by players and coaches, it’s hard to believe that doesn’t play at least some role!
Paper Title: Time Series Feature Redundancy Paradox: An Empirical Study Based on Mortgage Default Prediction
Authors & Date: Chengyue Huang & Yahe Yang | 12/23/2025
Link to Abstract & Paper
This paper challenges the "more is better" approach in time series prediction, using Freddie Mac mortgage data to predict loan defaults. It finds that shorter training windows (e.g., 1 year) and key feature subsets significantly outperform models using longer historical data or excessive features.
Longer windows introduce outdated patterns, while non-critical features obscure essential predictors. The study highlights the importance of focusing on recent, relevant data and critical variables to improve accuracy and efficiency in financial risk modeling.
This paper highlights the importance of understanding the key drivers in a model. It also emphasizes how markets with frequent regime changes can make models that rely on older data lose predictive power.
The mortgage market is a clear example of this dynamic. Constant shifts in supply, demand, and interest rates likely disrupt prior relationships, making recent data far more relevant for accurate predictions.
Paper Title: Market-News Co-Moments and the Cross Section of Stock Returns
Authors & Date: Mohammadreza Tavakoli Baghdadabad, Girijasankar Mallik & Sriram Shankar | 01/05/2025
Link to Abstract & Paper
This paper examines how market-news co-moments—relationships between stock returns and market-wide news shocks—impact the cross-section of U.S. stock returns from 1928 to 2023. It highlights the role of factors like cash-flow beta (how a stock’s cash flows react to market news), co-skewness (asymmetry risk), and co-kurtosis (tail risk).
The study finds that portfolios with higher exposure to these risks earn significant premiums, such as abnormal annualized returns of around 7.5% for stocks with high cash-flow co-kurtosis. These findings suggest that incorporating co-moments into asset-pricing models and strategies can enhance risk-adjusted returns, particularly in volatile market conditions.
Some context for these terms:
Co-skewness: Stocks with negative co-skewness, which perform poorly during market downturns, demand a premium due to their undesirable risk characteristics.
Co-kurtosis: Stocks with high positive co-kurtosis, which are prone to extreme returns during market shocks, also earn a premium as compensation for their tail risk.
This is definitely a more complex paper, but its results are interesting nevertheless.
Subscribe if you want more summaries of academic papers!
Thank you for reading this week’s edition of Recent Academic Research. Remember to fill out the poll to let me know which paper was your favorite and like the post if you enjoyed it.
Feel free to follow up with any questions, comments, or ideas for the future!
Please consider sharing with someone who would enjoy these papers!
Share

The content provided in this newsletter, "Alpha in Academia," is for informational and educational purposes only. It should not be construed as financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities or financial instruments. Past performance is not indicative of future results. The financial markets involve risks, and readers should conduct their own research and consult with qualified financial advisors before making any investment decisions.
The interpretations, opinions, and analyses presented herein are those of the author and do not necessarily reflect the views of the original researchers, their institutions, or the full implications of the cited academic papers. While every effort is made to accurately represent the research discussed, readers should be aware that the summaries and interpretations may not capture the full scope or nuances of the original studies. The information contained in this newsletter is believed to be accurate and reliable at the time of publication, but accuracy and completeness cannot be guaranteed. The author and publisher accept no liability for any loss or damage resulting from reliance on the information provided.
This newsletter may contain links to external websites or resources. The author is not responsible for the content, accuracy, or reliability of these external sources.
By subscribing to or reading this newsletter, you acknowledge that you have read and understood this disclaimer and agree to hold the author and publisher harmless from any liability that may arise from your use of the information contained herein.
How do you feel about the merits of running multiple strategies at once? I spent the weekend creating a code to execute the trading strategy outlined in your last post (with paper money) and while a huge portion of that time was spent reminding myself about some forgotten python knowledge, I actually enjoyed it a lot more than I thought I would.
I thought I’d try to set a parallel environment up according to: Stocks Underperform with Congressional Disapproval (this seems even easier than the previous strategy) but my concern is that there’s a minimal amount of attention one has to give to a strategy for it to actually show representative results (monitoring for bugs, strange behavior, searching for explanations for abnormal losses/gains, etc.), and I don’t want to just have multiple set-it-and-forget-it strategies being tested at the same time with none likely to actually be worth much.
I guess my real question is, how much attention do you think a strategy should be given after it’s implemented to actually test whether there’s something to it?
No posts
Ready for more?

source

About The Author

admin

See author's posts