Weekly Research Recap
Latest research on investing and trading
Hi there, lots of interesting papers this week. Below is your weekly research recap, with direct links to the full articles.
Commodities
Commodities and Conundrums: Decoding Behavioural Finance in Market Dynamics (Till)
Investors often underestimate the influence of psychological biases in trading, particularly in commodity markets. This paper examines real-world cases, such as the collapse of MF Global, where overconfidence, loss aversion, and confirmation bias led to significant failures. It also explores common commodity trading strategies like calendar spreads, seasonal trades, and momentum, and highlights how they depend not only on market structure but also on anticipating investor behavior.
Crypto
Liquidity Provision and Its Information Content in Decentralized Markets (Chen, Deng, Nie, Zou, and Fu)
This paper develops a theoretical model of liquidity provision on decentralized exchanges like Uniswap, showing that supplying liquidity is only profitable when volatility stays within a specific range; too little or too much volatility leads to losses. The authors test the model using Uniswap tick-level data and find support for its predictions. They also show that net liquidity provision, which captures the flow of capital into or out of pools, helps predict future volatility.
Building Crypto Portfolios with Agentic Ai (Castelli, Stefano, and Piergallini)
The authors show how a multi-agent system built with Crew AI can improve crypto portfolio management. Each AI agent handles a specific task, such as data loading, cleaning, optimization, or performance evaluation. Using only past returns and volatility, the system applies a rolling monthly mean-variance optimization that adapts to changing market conditions. This dynamic approach significantly outperforms a static strategy, delivering higher Sharpe ratios.
Equities
Price Pressures from Daily Mutual Fund Flows: Does it pay to go with the flow? (Suominen and Tuovinen)
Daily fund flows affect stock prices and the broader market. This paper shows that heavy mutual fund selling often leads to sharp price drops that later reverse, while strong buying tends to persist due to continued inflows. The authors develop two trading strategies, one exploiting mean reversion after outflows, the other riding inflow-driven trends, and find both deliver strong performance. They also show that these flow signals help in timing the overall market.
The Market Reaction to Free Cash Flow News (Lem, Koski, and McVay)
Most investors focus on earnings, but free cash flow may offer a more informative signal of firm performance. This paper shows that stock prices, particularly for growth firms, respond more strongly to free cash flow surprises than to earnings surprises. Free cash flow also offers a clearer signal of value for firms with volatile earnings, one-off charges, or deferred revenues. The market reaction is even stronger when it is explicitly mentioned in earnings reports. Hence, investors may benefit from tracking both signals.
The Truth of Silence: Bad Signal of No Analyst Report (Shi and Yang)
Investors often rely on analyst reports for guidance, but what happens when analysts stay silent? This paper finds that extended gaps between reports, referred to as analyst silence, often signal trouble ahead. Stocks with high silence tend to underperform, as analysts may withhold bad news to protect relationships with firms. While the long-short strategy is tested on Chinese data, the idea could possibly apply elsewhere. Hence, simply tracking time since the last analyst report can potentially serve as a useful predictor of future stock returns.
The case for low-risk equity investing: evidence from 2011-2025 (Leote de Carvalho, Andreis, and Laplenie)
The low-volatility anomaly is well established: Low-risk assets tend to deliver higher risk-adjusted returns. By combining traditional volatility with additional measures like financial stability and debt levels, this paper builds a low-risk equity strategy that consistently outperforms the market from 2011 to 2025. Hence, low-risk strategies can offer strong performance, especially when designed with thoughtful sector and risk controls.
A tug of war across the market: overnight-vs-daytime lead-lag networks and clustering-based portfolio strategies (Lu, Zhang, Reinert, Cucuringu)
Most studies focus on how individual stocks reverse overnight gains during the day. This paper shows that these effects spill over across stocks. By analyzing how one stock’s overnight return predicts another’s daytime move, the authors build a lead-lag network and cluster stocks into leaders and laggers. Trading laggers based on leader signals yields strong returns and can potentially offer a new source of short-term alpha for investors.
Levered Single-Stock ETFs (Bessembinder)
Many investors are drawn to levered single-stock ETFs for amplified exposure. But this paper shows they tend to underperform expectations by around 0.70% per month due to daily rebalancing and trading costs. Positively levered ETFs can outperform in strong trending markets, but performance deteriorates in volatile or mean-reverting conditions. Inverse ETFs perform even worse.
Anomaly Disagreement (Luan)
Most equity strategies rely on anomaly signals, but these signals often conflict. This paper shows that when anomalies disagree more for a stock, return forecasts become less reliable, and both institutional and short-selling investors react more cautiously. Still, stocks with high disagreement earn higher future returns, consistent with a priced risk premium. A long-short strategy that buys high-disagreement stocks and shorts low-disagreement stocks delivers statistically significant returns.
Machine Learning and Large Language Models
Foundations of Large Language Models (Xiao and Zhu)
This book offers a clear and structured overview of how Large Language Models are built, trained, and applied. It explains how they learn from vast amounts of text through self-supervised learning, how they are fine-tuned and prompted for specific tasks, and provides a comprehensive guide to techniques for using them effectively.
Machine Learning for Timing: Evidence from U.S. and Chinese Stock Markets (Zhu, Jin, Zhu, and You)
Investors often try to time the market to boost returns, but it’s unclear whether this adds real value. This paper shows that cross-sectional stock selection using machine learning on fundamental data consistently outperforms timing strategies based on price trends and moving averages. While adding timing signals can offer a small boost, mainly from the short side, these gains disappear once trading costs are considered.
Options and Volatility
Long-Term Benefits of Call Overwriting (Medvedev)
Selling covered calls to generate income is a popular strategy, but does it boost long-term returns? This paper finds that call overwriting only makes sense when options are sufficiently overpriced to offset the loss of upside and the greater downside risk in equity returns. Using real market data, the study shows that these conditions are rarely met, meaning the extra income often doesn’t justify the trade-off.
Time-Varying Factor-Augmented Models for Volatility Forecasting (Zhang, Li, Mo, and Chen)
This paper introduces a flexible method that extracts time-varying risk factors from realized volatilities across assets using a rolling principal components approach. These dynamic factors are integrated into standard models like AR, HAR, MIDAS, and LSTM, leading to more accurate volatility forecasts for both stocks and cryptocurrencies. The improved forecasts enhance trading performance, notably increasing returns and Sharpe ratios in a volatility-scaled pairs trading strategy.
Trend Following
Enhancing Traditional Betas with Managed Futures Strategies: A Risk-Based Hedging Framework (Sadik)
Traditional portfolios like 60/40 or equity-only often underperform during market stress, offering limited protection when both stocks and bonds decline. This paper rigorously shows that an allocation to managed futures can improve Sharpe ratios, lower drawdowns, and reduce tail dependence. Even modest allocations (e.g., 5–25% of portfolio risk) deliver meaningful gains, especially for equity-heavy portfolios. The benefit is strongest during crisis regimes, where managed futures provide diversification exactly when traditional assets fail.
Blogs
Overnight Crypto Returns (Eric Falkenstein)
A Quant's Guide to Covariance Matrix Estimation (Open Source Quant)
Do Gold Miners Beat Bullion? What the Data Really Says (QuantSeeker)
GitHub
Book on Natural Language Processing
Podcasts
The TRUTH About TRADING No One Tells You | Jack Schwager (Words of Rizdom)
$4.5M | Process, Patience, and Prop Firms · Kyle Ng aka JadeCap (Chat with Traders)
Edward Yu – Bringing OTC On-Chain and the VariationalOMNI Perp Dex (Flirting with Models)
Trend Following in a World That Loves Bubbles ft. Mark Rzepczynski (Top Traders Unplugged)
Social Media
Introduction to Prompt and Context Engineering (Carol Alexander)
Understanding Return Expectations, Part 6 (AQR)
A Wild Year for Markets Hits Trend-Following Hedge Funds - WSJ (via Jerry Parker)
Diversification for Ultra-Long-Term Investors (AQR)
Last Week’s Most Popular Links
How To Bet On Winners (and Losers) (Brownlees and Souza)
Non-Linear Factor Investing in the Era of Machine Learning (Chin)
The Power of Position Sizing in Portfolio Management (Blotnick)
Disclaimer: This newsletter is for informational and educational purposes only and should not be construed as investment advice. The author does not endorse or recommend any specific securities or investments. While information is gathered from sources believed to be reliable, there is no guarantee of its accuracy, completeness, or correctness.
This content does not constitute personalized financial, legal, or investment advice and may not be suitable for your individual circumstances. Investing carries risks, and past performance does not guarantee future results. The author and affiliates may hold positions in securities discussed, and these holdings may change at any time without prior notification.
The author is not affiliated with, sponsored by, or endorsed by any of the companies, organizations, or entities mentioned in this newsletter. Any references to specific companies or entities are for informational purposes only.
The brief summaries and descriptions of research papers and articles provided in this newsletter should not be considered definitive or comprehensive representations of the original works. Readers are encouraged to refer to the original sources for complete and authoritative information.
This newsletter may contain links to external websites and resources. The inclusion of these links does not imply endorsement of the content, products, services, or views expressed on these third-party sites. The author is not responsible for the accuracy, legality, or content of external sites or for that of any subsequent links. Users access these links at their own risk.
The author assumes no liability for losses or damages arising from the use of this content. By accessing, reading, or using this newsletter, you acknowledge and agree to the terms outlined in this disclaimer.