Weekly Research Recap
Latest research on investing and trading
This week’s recap highlights the most actionable investing insights from academic research, industry reports, blogs, and social media, with direct links to every source.
Crypto
Bitcoin Treasury Companies: Systematic Risk and Factor Validation (Krause)
Public companies holding Bitcoin behave as hybrid equity/crypto assets, with high market beta and strong Bitcoin exposure. Across 55 firms (2024–2025), Bitcoin betas are large and highly significant (β’s about 0.35–0.42), while alphas disappear once Bitcoin is added as a risk factor. Equal-weighted portfolios achieve Sharpe ratios of around 1.1, but Bitcoin miners lag badly (Sharpe about 0.34). Key takeaway: These stocks offer Bitcoin exposure, but no alpha.
The Impact of Bitcoin ETF Approval on Bitcoin’s Hedging Properties Against Traditional Assets (Hong, Feng, Wang, and Li)
After U.S. spot ETFs launched in 2024, Bitcoin began behaving much more like a risk asset. Time-varying correlations reveal a clear regime shift, with Bitcoin increasingly moving alongside the S&P 500. Its relationship with gold remains weak and inconsistent, while exposure to the U.S. dollar stays negative. Performance improved modestly after approval. Key takeaway: ETF access has pulled Bitcoin into the equity ecosystem, weakening its role as a hedge.
Equities
Dual Peer Effects and Cross-Stock Predictability (Avramov, Ge, Li, and Linton)
The paper introduces a Peer Index (PI) that decomposes firm characteristics into peer averages and within-peer deviations. PI predicts stock returns and future earnings surprises, consistent with slow diffusion of peer information. A value-weighted top–bottom portfolio earns 0.89% per month (pre-costs) with a 7-factor alpha of 0.68%, and predictability persists up to three years. Key takeaway: Markets underreact to peer information, driving return predictability.
Intraday Time Series Reversal (Iwanaga and Sakemoto)
U.S. equities display a clear intraday reversal: Overnight moves unwind in the first 30 minutes after the open. For SPY, the effect is strongest during high-volatility periods, supporting a first-half-hour reversal strategy earning about 8% annual excess return with a Sharpe exceeding one. The effect is strongest on the long side and has weakened since 2010. Key takeaway: Overnight buying/selling pressure reverses at the open, especially during periods of market stress.
Style Investing, Style Timing, and IPO Return Predictability (Ashour and Hao)
Using 8,999 U.S. IPOs (1975–2022), the paper shows that style momentum spills into IPO pricing. A one-SD rise in pre-IPO style returns lifts underpricing by 6.7%, but IPOs tied to hot styles then reverse, underperforming by 11%, 28%, and 63% over 3, 6, and 12 months. Firms also issue more often and faster when styles run hot. Key takeaway: Style momentum boosts IPO pops but predicts short-term reversals; fade hot-style IPOs.
How much is too much? Part 2: Why 60% in US equities might be just as crazy as it sounds (Battistella and McLoughlin)
A 60% U.S. equity weight is defensible only when returns are reverse-engineered to fit cap weights, implying U.S. returns near 7.1% (Sharpe about 0.24). But even small return errors (40–50 bps) make mean–variance portfolios unstable, pushing optimal U.S. weight toward 50% or less. Key takeaway: Under return uncertainty, risk-based, return-agnostic allocations appear more robust than cap-weighting.
Customer-Supplier Momentum Spillover and Its Cascading Effect (Wu, Shi, Luo, and Zhao)
Firms inherit momentum from their customers because markets underreact to supply-chain information. In China, a customer-supplier long–short momentum strategy earns 0.4–0.6% per month with strong significance. The effect is strongest when customer attention is low and propagates through indirect supply-chain links. Key takeaway: Overlooked economic networks generate persistent and tradable return spillovers.
Latency and the Look-Ahead Bias in Trade and Quote Data (Battalio, Holden, Pierson, Shim, and Wu)
The paper documents look-ahead bias in the NYSE TAQ database. Because trades and quotes reach the SIP with uneven latency, events are frequently reported out of order. As a result, around 20% of trades are mis-signed, and effective spreads and price impact are understated by more than 40%. Using exchange timestamps largely resolves these distortions. Key takeaway: TAQ-based microstructure analysis is unreliable unless SIP latency is explicitly corrected.
Fixed Income
New findings on the inversion of the 10Y-3M bond yield curve (Gomez Sanchez)
Studying 134 yield-curve inversions across 58 countries, the paper shows that equity drawdowns are linked to how quickly the 10Y–3M spread inverts. The key variable is T0, days from first inversion to the trough, which explains drawdowns better than total time inverted. The relationship flips after 2007, indicating a regime shift. Key takeaway: For equities, yield-curve inversion speed matters more than inversion length.
The paper times HYG versus IEF using two signals: (i) a price-ratio rule that compares the HYG/IEF ratio to its own moving average, and (ii) a return-spread rule that conditions on recent HYG–IEF return differentials. The preferred price-ratio specification (9-month lookback, +1% no-trade band) achieves 3.3% CAGR with Sharpe of about 0.5, versus 2.0% and 0.3 for a static 50/50 mix, net of costs. Key takeaway: Simple trading signals can materially improve bond allocation efficiency.
Hedge Funds
Not So Smart Alphas (Lhabitant)
The paper shows that hedge fund “alpha” is highly model-dependent. Using two hedge funds as case studies, monthly alpha ranges from near zero to about 1% depending on benchmarks, factor models, and timing controls, with wide confidence intervals and low R². It also highlights reverse alpha: Regressing market returns on fund returns can yield positive market alpha, a mechanical result of low explanatory power, not skill. Key takeaway: Alpha estimates are fragile and easily misleading.
Machine Learning and Large Language Models
Sector Level Equity Returns Predictability with Machine Learning and Market Contagion Measure (Peng and Yao)
Using 5-minute data, the paper constructs contagion measures by separating market–sector co-movements into jump correlations (shock-driven co-jumps) and continuous correlations (normal diffusion). Adding these signals improves monthly SPY and sector forecasts. Random forest cut forecast errors by 20–30% versus a random walk and support portfolios with a Sharpe of about 1.3. Key takeaway: Intraday contagion signals paired with ML deliver predictability of market returns.
Reinforcement Learning in Financial Decision Making: A Systematic Review of Performance, Challenges, and Implementation Strategies (Hoque, Ferdaus, and Hassan)
This paper synthesizes evidence from 167 studies (2017–2025) on reinforcement learning in trading, portfolio allocation, and market making. RL delivers its most consistent gains in market making, while performance elsewhere depends far more on data quality, implementation discipline, and domain expertise than on algorithm choice. Key takeaway: In finance, RL mainly succeeds through market understanding and execution quality, not model complexity.
Prediction Markets
Exploring Decentralized Prediction Markets: Accuracy, Skill, and Bias on Polymarket (Reichenbach and Walther)
Using 124.5 million trades on Polymarket, the paper finds that decentralized prediction markets are well-calibrated and slightly outperform bookmakers. Mispricing is temporary, mainly arising early in a market’s life and near resolution due to information lags. Only about 30% of traders profit, but profits persist, indicating genuine skill. Key takeaway: Prediction markets are broadly efficient, with brief, exploitable behavioral frictions.
Blogs
Alpha through equity-versus-FX trading (Macrosynergy)
Covariance Matrix Forecasting: Average Oracle Method (Portfolio Optimizer)
The FOMC Day Puzzle: Why Every G10 Currency Rallies Against the Dollar (Quantseeker)
Podcasts
BlackRock’s Rick Rieder: Warning Signs Are Flashing (Meb Faber)
Magnet Above. Trap Door Below | What the Options Market Tells Us About What Comes Next (Excess Returns)
Verified Trader: Here’s Exactly How He Made 40,000% Return On A Small Account (Words of Rizdom)
Tom Starke - Institutional Quants Think Differently (The Algorithmic Advantage)
Social Media & Industry Research
Institutions’ return expectations across assets and time (Alpha Architect)
Inside the ‘rolling thunder’ quant crises of 2025 (FT Alphaville)
Understanding Gold (Campbell Harvey)
Is Trend Following Better than “Buy the Dip”? (Alpha Architect)
Last Week’s Most Popular Link
Quantitative Evaluation of Volatility-Adaptive Trend-Following Models in Cryptocurrency Markets (Karassavidis, Kateris, and Ioannidis)
How Investors Pick Stocks: Global Evidence from 1,540 AI-Driven Field Interviews (Hwang, Noh, and Shin)
What Drives the Performance of Machine Learning Factor Strategies? (Esakia and Goltz)
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.