Weekly Research Recap
Latest research on investing and trading
Here’s this week’s briefing: The most useful investment research and market ideas I came across over the past seven days, sourced from academic papers, industry publications, blogs, and social media. Links to all sources are included.
Crypto
Are Day-of-the-Week Effects in Cryptocurrencies Real? Intraday Evidence from Active and Less Active Cryptocurrencies (Aalipour, Mehdian, and Rezvanian)
Crypto’s day-of-the-week effect isn’t really a full-day effect. Across 12 cryptocurrencies, daily anomalies are concentrated in just a few specific hours and differ by asset. Bitcoin’s Monday effect, for example, is driven by only two significant hourly intervals. Key takeaway: Predictability in daily returns can actually come from just a few specific trading hours.
The Bitcoin Polar Pricing Model: Cycles, Prices, and Predictability (Bonaparte)
Bitcoin’s four-year cycle isn’t just imposed by the halving calendar. An unrestricted search of the price data estimates a 3.86-year cycle, with two alternative methods finding 3.74 and 4.14 years. The pattern is much weaker at short horizons and becomes more informative over longer horizons, but overlapping returns can exaggerate this predictability. Key takeaway: Bitcoin’s long-run price cycle independently converges remarkably close to its protocol-driven halving schedule.
Equities
Text-Based Risk Similarity and Economic Links (Chen and Yang)
Firms don’t need to be in the same industry to transmit return-relevant information. Using 10-K risk disclosures to identify firms with similar risk exposures, this paper finds that stocks with the best-performing “risk peers” outperform those with the worst-performing peers by 1.78% per month. The effect is strongest across industries. Key takeaway: Shared risks create predictive links between seemingly unrelated stocks.
Buy the Rumor, Sell the News: When Is News Priced In? (Kargarzadeh, Khaledian, Parvini, Ghatak, and Khaledian)
Most stock news is already priced in by the time you read it. Studying 1.68 million U.S. stock-news events, the authors find that the price move is concentrated before and on publication day. But what happens next depends on the news: Quantified fundamentals keep drifting, while softer narrative news tends to reverse. Key takeaway: The post-news opportunity depends less on sentiment than on whether the news contains hard information or narrative.
Unexpected Gross Profit and Cross-Sectional Stock Returns (Yang, Cai, Rhee, and Wu)
A surprisingly simple profitability signal predicts stock returns: Current gross profit minus its 8-quarter moving average. This “unexpected gross profit” measure rivals standardized earnings surprises and is particularly strong among smaller stocks. Key takeaway: Profit surprises contain information about expected returns beyond traditional earnings surprises, with the effect attributed to systematic risk rather than mispricing.
Machine Learning & Large Language Models
MacroAllocAgent: From macro narratives to strategic asset allocation via a multi-agent LLM system (Wang, Deng, Li, and Yang)
LLMs work better for asset allocation when they form views, not portfolio weights. Across 13 Chinese macro assets, adding LLM-generated views to Black–Litterman modestly improved Sharpe: From 0.90 to 0.95, and from 0.67 to 0.76 with tighter allocation bounds. Direct LLM allocation achieved a Sharpe of just 0.32. Key takeaway: Let AI form the views; let conventional optimization build the portfolio.
Big Data Machine Learning Portfolios in Thirty Minute SPX Options (Maisch, Zhang, Ulrich, and Zimmer)
Machine learning finds predictability in 30-minute SPX option returns. Using 14M observations, the best models reach a 30-minute gross Sharpe of 0.72. Fewer inputs often work better: Theta, implied vol, bid/ask size, and vega dominate. Key takeaway: There are significant intraday mispricings in option markets, but they are hard to monetize after spreads.
AgonAlpha: Autonomous Alpha Discovery via Prompt Economy and Scalable Agentic Search (Ye, Sun, Ren, Yu, Yi, and Yang)
AI agents are getting surprisingly good at autonomous alpha discovery. AgonAlpha searched U.S. equities without humans writing factor formulas. Across 60 submissions, 17 earned WorldQuant BRAIN’s highest grade. The discovered signals used inputs such as option-implied volatility, trading volume, short interest, returns, and option positioning. Key takeaway: Alpha discovery can be automated from idea generation through verification.
Options & Volatility
Last Fifteen Minutes: Equity Options Volatility at the Close (Clark, Palepu, Poti, and Siddique)
The market close may no longer be the best time to measure SPX implied volatility. With the rise of 0DTE options, 3:45 PM implied vol predicts next-day realized variance far better than the raw 4:00 PM measure: Correlation 0.68 vs. 0.07. At expiration, collapsing time value makes implied-volatility estimates increasingly unstable. Key takeaway: In the 0DTE era, the last 15 minutes can distort rather than improve volatility signals.
One-Day Volatility Indices and the Variance Risk Premium in European Equity Markets (Wilkens)
Europe now has a VIX1D equivalent. This paper constructs one-day volatility indices for the EURO STOXX 50 and DAX from daily-expiry options. They add substantial information: Including them raises explanatory power for next-day realized variance by about 25 percentage points. But selling one-day variance is less attractive as quoted spreads absorb a large part of the premium. Takeaway: 0DTE options are potentially more useful for forecasting volatility than harvesting variance premia.
Blogs
Beyond Volatility Scaling: Does “Good” and “Bad” Volatility Matter? (Quantseeker)
Your Research Agent Is an Undisclosed Factor Exposure — And So Is Everyone Else’s (Jonathan Kinlay)
Sectoral Intramonth Momentum Cycle: Exploiting Turn-of-the-Month Patterns in Sector ETF Strategies (Quantpedia)
Podcasts
Toby Crabel - Short-Term Futures Trading with Size! (The Algorithmic Advantage)
Barry Ritholtz: How Not to Invest (Rational Reminder)
Social Media & Industry Research
VIX and Trend Following Revisited: Nearly a Decade of Out-of-Sample Evidence (Alpha Architect)
August Checklist (Citadel)
Setting the record straight: The truths about index fund investing (Vanguard)
Last Week’s Most Popular Links
Latent Characteristic Factors in Commodity Markets (Sakkas and Stavroglou)
Factor Neutralization and Risk-Budgeted Scaling: Evidence from Long-Short Anomaly Portfolios (Kosmakov)
International Yield Curves and Exchange Rates (Wang and Zhu)
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.