Hello! This week’s research recap brings you the key investing insights and resources from the past seven days, with direct links to every source.
Bonds
Common Risk Exposures of Bond Funds (Blitz)
Analyzing over 1,500 bond funds (2015–2024) shows that active returns, performance relative to benchmarks, are driven mainly by systematic risk exposures, not manager skill. For Aggregate and Corporate funds, about 55% of active returns come from underweighting duration and overweighting credit risk, while high-yield funds often reduce credit exposure. Key takeaway: Active return is not alpha; much reflects structural beta tilts, so investors must adjust for these risks to gauge true skill.
Crypto
Bitcoin and the Fama-French Model Rethinking Asset Pricing (Nahidi and Malekan)
Extending the Fama–French model to include Bitcoin shows that cryptocurrency risk is priced heterogeneously across firms. Bitcoin exposure is positive and significant only for crypto-related companies, while it remains irrelevant for the broader market. Interestingly, its inclusion weakens the traditional value (HML) factor. Key takeaway: Bitcoin acts more as a source of diversification than a hedge, partially subsuming traditional value effects.
Equities
Volatility forecasting for low-volatility investing (Conrad, Kleen, and Lönn)
Constructing low-volatility portfolios by weighting stocks inversely to forecasted volatility, rather than to trailing volatility, yields meaningful performance gains with Sharpe ratios rising from 0.76 to about 0.80–0.85. While still below the 1.00 Sharpe of an infeasible portfolio with perfect foresight, these forecast-based strategies lower volatility and bring allocations closer to the theoretical optimum. Key takeaway: Using volatility forecasts for inverse-volatility weighting improves performance compared to using historical volatility.
Equity Premium Prediction with Shrinkage Estimation Guided by Economic Theory (Liu, Liu, Wang, and Zeng)
Incorporating economic theory into predictive models substantially boosts equity premium forecasts. The paper introduces an empirical Bayes shrinkage method that pulls coefficients toward theory-implied values, such as requiring a non-negative equity premium and Sharpe ratios within 0 to 1.5, lifting out-of-sample R-squared from roughly 4–9% under OLS to about 5–12%. Key takeaway: Imposing economic constraints on predictive regressions materially enhances both statistical accuracy and economic value.
The Impact of U.S. Stock Buybacks: Theory vs Practice (Haghani and White)
U.S. stock buybacks, projected at $1.2 trillion in 2025, can meaningfully lift equity prices, defying classical corporate finance theory. A 3% buyback may raise prices by 3–5%, helping explain roughly 4 percentage points of annual P/E multiple expansion over the past decade. But higher prices compress future returns, and buybacks amplify market cycles. Key takeaway: Stock buybacks impact markets and returns in a significant manner, not just capital structure.
Investor Memory and Stock Returns: Empirical Evidence (Liu and Wang)
This paper introduces a new cross-sectional return predictor: Investor memory bias (MB), defined as the share of past negative daily returns investors are likely to forget, with older losses weighted more heavily. Stocks with the highest MB underperform those with the lowest by 0.90% per month (EW) and 0.81% (VW), before costs. Key takeaway: Selective forgetting by investors systematically distorts prices and leads to persistent underperformance.
Consumer Similarity and Predictable Returns (Shabanpour)
Using geospatial data on 150 million store visits, the paper constructs a novel consumer similarity network linking firms with overlapping customer bases. Stocks of firms serving similar consumers predict each other’s returns, revealing a persistent lead–lag effect driven by investor inattention. A long–short “consumer network momentum” strategy delivers 6–7.2% annual alpha (before costs). Key takeaway: Overlooked consumer linkages offer exploitable return predictability beyond standard risk factors.
Machine Learning and Large Language Models
The New Quant: A Survey of Large Language Models in Financial Prediction and Trading (Fu)
This is a comprehensive survey paper, covering over 50 studies on how large language models are transforming quantitative investing. It shows that LLM-derived signals from news, filings, earnings calls, and policy texts can predict returns, often outperforming traditional sentiment baselines, and can be integrated into trading systems.
Market Microstructure
Heterogeneous Return Predictability from Order Flow (Qian)
Retail order flow significantly predicts stock returns positively, while total order flow shows little signal. In contrast, retail flow has no predictive power for most ETFs due to higher noise trading. Yet, in banking stocks, predictability turns negative as extreme inventory risk causes market makers’ price impact to dominate. Key takeaway: Order-flow predictability is non-monotonic, positive when informed trading drives prices, negative when inventory constraints dominate.
Options
Peer Option Momentum (Jones, Khorram, Li, Mo, Yang, and Zhang)
This paper uncovers a new anomaly in options markets: Peer momentum, the predictability of a firm’s delta-hedged option returns from the past returns of peers linked by shared sell-side analysts. A 12-month peer momentum strategy delivers ~5.3% per month with a Sharpe ≈ 2.9 before costs and roughly half after costs, driven by option markets’ underreaction to peer volatility shocks. Key takeaway: Volatility information diffuses across linked firms, creating profitable cross-asset momentum unexplained by standard factors.
Volatility
Time-Varying Factor-Augmented Models for Volatility Forecasting (Zhang, Mo, Li, and Chen)
The authors propose a time-varying, factor-augmented volatility framework where dynamic factors are extracted from realized volatilities and plugged into HAR (and others). Out-of-sample R² rises by up to about 12% in U.S. tech equities and about 23 % in crypto. Moreover, in a crypto pairs test (ADA–ETH), returns flip from –5.5% to +7.3%, with Sharpe from –0.47 to 0.79. Key takeaway: Dynamic volatility factors materially improve forecasts and trading performance.
Blogs
Terms of trade as trading signals (Macrosynergy)
Why You Can’t Tell if Your Strategy “Stopped Working” (Statistically Speaking) (Robot Wealth)
Robust Skewness as a Source of Commodity Risk Premia (Quantseeker)
GitHub
Podcasts
Jay Rajamony – Beyond Factors: Reimagining Quant Equity for the Modern Era (Flirting with Models)
Mark Higgins: Learning From Market History (Rational Reminder)
Phil Goedeker Market Wizard The Next Generation (Confessions of a Market Maker)
Last Week’s Most Popular Links
All days are not created equal: Understanding momentum by learning to weight past returns (Beckmeyer and Wiedemann)
How to Use the Sharpe Ratio (Lopez de Prado, Lipton, and Zoonekynd)
Financial Machine Learning: An Engineering Problem (Lopez de Prado)
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.