Weekly Research Recap
Latest research on investing and trading
This week’s recap brings together the most valuable investing ideas from academic studies, industry commentary, blogs, and social media published over the last seven days, with full source links.
Crypto
The Impact of Spoofing on Bitcoin Market Microstructure (John, Li, Liu, and Yang)
Spoofing on Coinbase is widespread, with large orders quickly placed and canceled to move Bitcoin prices. It is highly profitable: Each one-unit rise in spoofing volume generates about +27 bps (bid) and +55 bps (ask) in short-term profit with limited risk. At the same time, spoofing widens bid–ask spreads and reduces liquidity. Key takeaway: Spoofing is both common and lucrative for manipulators, but it harms price fairness and increases costs for everyone else in the market.
Equities
Leverage Volatility Risk (Shentu, Wang, and Wu)
The authors measure volatility in financial intermediaries’ leverage as a proxy for financial uncertainty and estimate each stock’s beta to this volatility. A portfolio long stocks with high exposure and short those with low exposure earns about −5% per year, showing that stocks most sensitive to leverage volatility systematically underperform. Key takeaway: Investors should avoid high leverage-volatility exposure, as markets consistently penalize this risk.
The factor games: May the p values be ever in your favor (Stellet and Moraes)
The paper asks how to shrink the crowded “factor zoo” into a small, reliable set of pricing factors selected on a walk-forward basis, and whether this improves real trading performance. Using rolling LASSO shrinkage, factors are re-chosen each period using only past data. Portfolios built from this dynamic selection reach Sharpe ratios near 2.0 and outperform Fama–French benchmarks. Key takeaway: Systematic, walk-forward sparsity beats static multi-factor models.
International Investing: Diversification and Beyond (Kim, Korajczyk, and Neuhierl)
Using stocks from 27 countries (1966–2022), the study finds that isolating country-specific and mispriced global factors dramatically improves performance. Strategies targeting these sources deliver Sharpe ratios of 1.44–2.32 (before costs), compared to just 0.48–0.56 from naive international diversification. These gains persist even among developed G7 markets. Key takeaway: True international alpha comes from exploiting cross-country factor inefficiencies, not simply holding more countries.
The Intra-Day Stock Return Periodicity Puzzle (Haendler, Heston, Korajczyk, and Sadka)
Stocks show a repeating intraday return cycle: Price moves in a specific half-hour tend to repeat at the same time on the following days. The effect is strongest at the open and close and persists out-of-sample. It is mainly driven by predictable institutional execution, VWAP at the open, and index-related trading near the close. Key takeaway: There exist strongly significant intraday return patterns that largely reflect systematic institutional trading.
The POP Premium, Populism and the Cross-Section of Stock Returns (Colacito, Ding, Jiang, and Qian)
Using GPT-4o-mini on over 1 million WSJ articles, the authors construct a populism sentiment index and show that it is priced in the cross-section of stock returns. Stocks with low populist exposure outperform high-exposure stocks by 0.62% per month (~8% annually, before costs), with risk-adjusted alpha up to 0.79% per month (t ≈ 5.2), robust to standard factors and political cycles. Key takeaway: A long–short strategy based on firms’ populist betas delivers a meaningful cross-sectional return premium.
Fixed Income
Out-of-Sample Prediction of Treasury Yield Correlations (Rebonato and Feng)
Using PCA to forecast Treasury yield correlations, the authors show large economic gains over standard methods. In a Markowitz bond portfolio, Sharpe rises from 0.62 to 0.88, roughly a 40% improvement, with differences statistically significant and most pronounced during crisis periods. Expanding beyond the first three principal components adds noise and fails to improve results. Key takeaway: Better correlation forecasting strengthens fixed-income portfolio performance.
FX
Large Moves in the Foreign Exchange Market (Kostakis, McBride, Sarno, and Wang)
Large FX moves are not purely random. When short-term implied volatility (IV) exceeds longer-term volatility, the chance of a sharp currency swing rises materially. A 1% steepening of the 30–90 day IV slope adds roughly 13 bps to next-day absolute returns, and significantly increases the probability of a ≥3σ move. A strangle strategy triggered by extreme inversion earns strong, positive Sharpe even after transaction costs. Key takeaway: Inversion of the volatility term structure is a reliable signal for upcoming large FX moves.
Machine Learning and Large Language Models
The Training Set Delusion: Machine Learning Overfitting in Bitcoin Return and Volatility Prediction (Wang, Zeng, Li, and Wang)
Across extensive tests, most ML models for Bitcoin show severe overfitting, meaning strong training performance collapses out of sample. Return forecasts deliver negligible or unstable R², while simple ARIMA models remain more reliable. For volatility, GARCH-class models prove more robust than deep neural networks, which often overestimate risk. Key takeaway: In noisy markets such as Bitcoin, traditional econometric models are more reliable than complex machine learning.
A Practical Machine Learning Approach for Dynamic Stock Recommendation (Yang, Liu, and Wu)
The paper presents a rolling ML stock-selection system that trains five models, chooses the lowest-MSE model, ranks stocks by predicted quarterly returns, and buys the top 20% per sector. Using this process, the minimum-variance portfolio delivers a Sharpe of 0.462 vs 0.195 for the S&P 500 and a total return of 1181% vs 93%, with equally weighted reaching 2114%. Key takeaway: Dynamic, model-driven ranking plus proper risk control outperforms passive equity exposure.
Mutual Funds
Risk-Return Analysis of Mutual Funds (Vidal-Garcia and Garcia)
Do actively managed equity mutual funds truly outperform their benchmarks once risk is properly accounted for? Studying 21,671 funds across 35 countries (1990–2024), the paper finds returns move almost in lockstep with indices (average correlation ≈ 0.95). Roughly 68% of funds underperform and beta-adjusted alpha averages −0.7%. The authors argue this shortfall largely reflects management fees of around 2–3% per year. Key takeaway: Most active funds behave like costly index replicas, delivering negative alpha on average.
Blogs
QuantMinds London 2025 (Turnleaf Analytics)
Macroeconomics with Gaussian Mixture Models (Open Source Quant)
An Improved Approach to Momentum Investing (Quantseeker)
Podcasts
The U.S. Market Is Overpriced — What Could Happen Next Based On History (The Meb Faber Show)
CTAs After the Walls Come Down ft. Rob Carver (Top Traders Unplugged)
Building a Commodity Trader Multi-Strat with Tom Holliday of HIP Investments (RCM Alternatives)
Day Trader Turned $3000 Into $2.1Million Using Data & Statistics (Words of Rizdom)
Last Week’s Most Popular Links
The Repurchase Effect and Asset Prices (Chen, Gu, Xiong, and Yu)
Wordle (TM) and the one simple hack you need to pass funded trader challenges (Rob Carver)
Tactical Asset Allocations of Large Asset Managers (Ibert)
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.