Weekly Research Recap
Latest research on investing and trading
Welcome to this week’s Tuesday roundup, featuring a curated mix of the most relevant investing ideas from the past week, sourced from academic papers, industry research, blog content, and sharp insights shared across social media, with links throughout.
Equities
Retail limit orders (Anand, Samadi, Sokobin, and Venkataraman)
Limit orders outperform marketable orders on an all-in basis: The implementation shortfall is 8 to 9 bps lower, even after accounting for a 16 bps opportunity cost embedded in the measure. High fill rates (65%) help contain execution costs, and this advantage is even more pronounced in smaller stocks. Key takeaway: Patient, liquidity-supplying limit orders consistently reduce trading costs for retail investors.
Expected Returns with Trends and Cycles (Hillenbrand and McCarthy)
Valuation ratios (like price-dividend) seem weak as return predictors because they blend two effects: Discount-rate variation and shifting cash-flow dynamics. The latter introduces noise and dilutes the signal in standard regressions. Once you separate these components, a clear return signal emerges: Out-of-sample R² is around 9% at one year and over 20% at five years, while raw ratios underperform a simple historical average. Key takeaway: Valuation ratios contain return signals, but they only become visible after stripping out cash-flow noise.
One Hundred Years in the U.S. Stock Markets (Bessembinder)
From 1926 to 2025, U.S. equities delivered 10% annual returns and created $91T in wealth, but outcomes are extremely skewed. The typical stock lost money, and only 41% beat T-bills. Net wealth creation is entirely driven by a small minority; about 3.7% of firms account for all gains, with concentration rising sharply over time. Key takeaway: Equity investing is a power-law game; missing a few big winners can derail long-term performance.
Dinner Table Alphas (Cao, Chen, Cohen, and Zhao)
Fund managers linked to senior executive spouses consistently outperform. The edge is highly concentrated: In spouse-linked industries, stocks they buy outperform those bought by managers with non-executive spouses by 4% per quarter (with sold stocks underperforming similarly). These trades also anticipate earnings surprises and corporate events. Key takeaway: Alpha isn’t just skill, it’s who you’re connected to.
FX
Geopolitical Risk in Currency Markets (Melone and Stathopoulos)
Sorting currencies by their sensitivity to a text-based geopolitical threats index, constructed from the share of news articles covering war, terrorism, and global tensions, yields a clear return spread. A long–short strategy (high minus low exposure) earns 3.3% annually (Sharpe about 0.39) with a 2.8% alpha beyond standard risk factors. Key takeaway: Geopolitical risk is a distinct priced factor in FX; high-exposure currencies earn a premium.
Machine Learning and Large Language Models
Empirical Asset Pricing via Learning-to-Rank (Lin, Su, and Zhu)
Learning-to-rank produces superior portfolios by directly optimizing stock rankings rather than estimating expected returns. It delivers roughly 1.5 to 2.3% monthly excess returns with Sharpe ratios up to 1.2, versus 0.35 to 0.7 for standard return models. The gains come from more accurate identification of top and bottom stocks and better hedging. Key takeaway: Optimizing relative rankings is a more powerful way to build long–short strategies.
Understanding the Convolutional Neural Network’s Stock Return Predictability (Guan, Li, and Si)
A convolutional neural network trained on 5-day price charts delivers strong cross-sectional predictability: A long–short decile strategy earns 54.5% annually with a Sharpe of 3.74 (2020–2024), rising above 5 over longer samples (before costs). Performance is driven primarily by the structure of OHLC bars; removing them roughly halves Sharpe, while adding features such as 52-week levels degrades results. Key takeaway: Simple price patterns, not added complexity, contain powerful, exploitable short-term signals.
Agentic Artificial Intelligence in Finance: A Comprehensive Survey (Aldridge et al.)
This literature review synthesizes research on agentic AI in finance, showing a shift from rule-based systems to autonomous, goal-driven agents that plan, learn, and coordinate. Multi-agent models can deliver strong risk-adjusted performance, with Sharpe ratios above 2 and relatively small drawdowns. Key takeaway: Autonomous systems can improve performance, but they also amplify risk.
Large Language Models and Stock Investing: Is the Human Factor Required? (Crisostomo and Mykhalyuk)
LLMs can generate stock recommendations, but only under strict guidance. Simple prompts produce weak results, while imposing a multi-factor framework (valuation, growth, quality, momentum, macro, sentiment with weights) improves performance. Adding human review lifts returns to 3.0% monthly alpha, and incorporating regulatory filings pushes performance toward 4%. Key takeaway: LLMs can assist in stock selection, but only when constrained by a disciplined factor model and human oversight.
Portfolio Construction
Complex Modern Portfolio Theory (Hellum, Jensen, Kelly, and Malamud)
Modern Portfolio Theory (MPT) breaks down when the number of assets (N) approaches the number of observations (T), but the paper shows that performance recovers once N ≫ T. In this high-dimensional regime, Sharpe ratios increase and can surpass those in standard settings. The gains come from implicit regularization in the covariance matrix, not better return estimates. Key takeaway: MPT can benefit from very large asset universes, provided the number of assets greatly exceeds the available number of observations.
Blogs
From Hype to Reality: Building a Hybrid Transformer-MVO Pipeline (Jonathan Kinlay)
AI Will Create Millions of Quants (RobotWealth)
More of the Disease, Faster (What happens when you ask an LLM to find you an edge) (RobotWealth)
Unlocking relative value across asset classes (Macrosynergy)
Timing Value vs. Growth: Evidence from 100 Years of Small Value–Large Growth Spread (Quantpedia)
Podcasts
How the World’s Largest Oil Derivatives Trading Firm Is Navigating the Iran War (Odds on Open)
The Hidden Cracks in Systematic Strategies No One Talks About ft. Nick Baltas (Top Traders Unplugged)
The Christina Qi Episode (The Alternative Data Podcast)
Social Media & Industry Research
The Return of the King: Trend Following Is Back – But Will It Last? (Alpha Architect)
A Trend Following Deep Dive: The Optimal Market Mix for a Trend Follower (Man Group)
Last Week’s Most Popular Links
Beyond Prompting: An Autonomous Framework for Systematic Factor Investing via Agentic AI (Huang and Fan)
An Index of Commodity Futures Returns Since 1871 (Janardanan, Qiao, and Rouwenhorst)
Save The Date: Analyst/Investor Days as a Trading Signal (Cabrera, Kolokolova, and Zhang)
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.
Paid subscriptions may not be available in all jurisdictions and may change without notice.