QuantSeeker

Weekly Research Recap

QuantSeeker — Tue, 08 Sep 2026 21:56:59 GMT

Welcome to this week’s Tuesday roundup, your curated selection of the most actionable investing ideas from the past week, spanning academic research, industry reports, blog posts, and insightful discussions across social media, with links throughout.

Asset Allocation

From Strategic Policy to Tactical Trades: Dynamic Asset Allocation with Predictable Returns and Trading Costs (Kolm and Ritter)

This paper develops a dynamic framework for tactical asset allocation that jointly accounts for expected returns, portfolio risk, trading costs, and how today’s positions affect future decisions. It derives when optimal trades should be smaller or larger than standard one-period TAA and how quickly portfolios should adjust. Key takeaway: Optimal tactical trades depend on both today’s opportunity and the cost of repositioning tomorrow.

Commodities

Correlation and Commodity Risk Premia (Fan and Zhang)

Commodity momentum struggles when commodities start moving together. Across 29 futures markets, scaling down momentum exposure during high-correlation regimes reduces drawdowns and improves risk-adjusted returns. Combining correlation, volatility, and stop-loss rules raises net Sharpe from 0.42 to 0.67. Key takeaway: Cross-commodity correlation can be a useful signal for timing momentum exposure.

Crypto

Bitcoin option expiration, gamma exposure, and intraday price reversals (Weiss, Gaudiosi, Zhou, and Webb)

Bitcoin options appear to move the underlying around expiration. The paper finds that when ATM open interest is high and dealer gamma exposure is negative, Bitcoin tends to fall before the 8:00 UTC expiry and then fully reverse within two hours. Key takeaway: Bitcoin option positioning can create systematic, short-lived price pressure around expiration.

Equities

Financial-Statement Signals and the Cross-Section of Stock Returns (Dong)

The accounting “factor zoo” may be much smaller than it looks. This survey catalogs 155 financial-statement signals, but argues that many reflect overlapping economic ideas such as valuation, profitability, accruals, investment, financing, and fundamental momentum. Key takeaway: Don’t treat every accounting signal as an independent factor. Combine signals by economic concept to avoid repeatedly loading on the same information.

The Implied Equity Term Structure (Baele, Driessen, and Jankauskas)

This paper asks how the price of equity risk changes with the horizon of corporate cash flows. Using U.S. stocks since 1980, the authors find that longer-dated cash flows generally command higher expected returns. But the pattern reverses for value firms and companies with high credit risk. Key takeaway: Equity risk is priced differently across horizons, and the shape depends strongly on firm characteristics.

Rethinking Predictability: Distributional Forecasts of the Market Risk Premium (Sinha)

Expected-return forecasts may miss much of what matters for investors. Using S&P 500 data, this paper finds limited ability to predict next-month excess returns, yet forecasts of the full return distribution contain useful risk information. Distribution-based allocation cuts max drawdown from 56.9% to as low as 16.0%. Key takeaway: Forecasting uncertainty can be more useful for controlling downside risk than forecasting returns.

Factor Time (Bowles, Reed, Ringgenberg, and Thornock)

Fama–French factors may be more time-sensitive than they look. Standard portfolios are refreshed annually, leaving accounting information increasingly stale. Monthly updates materially change value and investment factors, but the greater improvement comes from modeling the returns generated by these updates separately. Key takeaway: How often factor portfolios are updated significantly affects measured alpha and performance.

Timing the Market: A Skewed Perspective (Chan)

“Time in the market beats timing the market” is not universal. Across six developed markets over 25 years, waiting for a 20% dip to buy beat monthly investing in Japan, the UK, and Germany, but produced 12.6% less final wealth in the U.S. Key takeaway: The cost of waiting to buy a dip depends on the market drift. The stronger the long-run return, the more expensive it is to sit in cash waiting for a better entry.

Blogs

Gold and macro factors (Macrosynergy)

Are Covered Calls Just Expensive De-Risking? (Quantseeker)

Not another one! My fifth book... (Rob Carver)

A Sharpe of 2.1 From Nothing: The Second Number Your Agent Doesn’t Log (Jonathan Kinlay)

Podcasts

How Top Prop Traders Find and Scale Their Edge · Jeff Holden (Chat With Traders)

Professional Trading Roundtable: “A 50% Market Crash Is Mathematically Inevitable!” (Words of Rizdom)

Social Media & Industry Research

What Daily Stock Returns Tell Us About the Economy (Alpha Architect)

What I Learned About Asset Pricing (Pedro Santa Clara)

Investing in private equity: A decision framework for drawdown versus semiliquid funds (Vanguard)

Last Week’s Most Popular Links

Agentic Quantitative Trading: A Survey of Workflows, Systems, and Evaluation (Hua et al.)

Expected market risk premiums in the international cross-section (Berkman and Malloch)

Two Heads Are Better Than One: t-Statistics and Monotonicity in the Factor Zoo (Fan)

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.

Are Covered Calls Just Expensive De-Risking?

QuantSeeker — Mon, 07 Sep 2026 23:00:26 GMT

Covered-call strategies promise an appealing combination: Equity exposure, lower volatility, and regular income from selling options.

But covered-call strategies also have substantially lower market beta than the equities they own. That raises an interesting question: Does option writing actually improve the risk-return tradeoff, or could investors do better by simply holding less equity and making withdrawals themselves?

I test this using long histories of S&P 500 and Nasdaq-100 covered-call strategies, covering both benchmark indexes and investable ETFs.

The results challenge some of the conventional arguments for covered calls, not only on returns, but also on downside protection and the income they generate.

Weekly Research Recap

QuantSeeker — Tue, 01 Sep 2026 16:18:41 GMT

This week’s briefing covers the investment research and market ideas I found most valuable over the past seven days, across academic research, industry publications, independent blogs, and social media. Links to all original sources are included.

Equities

Expected market risk premiums in the international cross-section (Berkman and Malloch)

Options markets contain a surprisingly strong signal for international equity allocation. Across international markets, option-implied expected market risk premia predict next-month returns: A 1%-point higher risk premium corresponds to roughly 1.1% higher subsequent excess return in the cross-section. Key takeaway: Local index options help predict next-month index returns.

Calendar Effects of Asset Pricing Factors (Fan, Li, and Miao)

Factor returns have calendar effects, but almost exactly opposite to the market. Across 153 U.S. equity factors, premia are weaker overnight, on Fridays, in January, around month-end, and near major macro announcements. Key takeaway: Market seasonality and factor seasonality tend to run in opposite directions.

Has the Correlation Risk Premium Been Competed Away? Evidence from Implied and Realized Correlation, 2006-2026 (Bawa)

The economics of dispersion trading have changed dramatically. In 2010–14, option-implied correlation exceeded subsequent realized correlation by 10.7 points on average. By 2022–26, that spread had flipped to -1.9 points. Key takeaway: The historical premium for selling correlation across S&P 500 stocks has largely vanished.

Two Heads Are Better Than One: t-Statistics and Monotonicity in the Factor Zoo (Fan)

A high t-stat isn’t enough to identify robust equity factors. Across 150 U.S. characteristics, t-stats and monotonicity, the share of months returns rise across factor quintiles, are nearly unrelated (correlation −0.08). Combining both better predicts out-of-sample returns and alphas. Key takeaway: Factor robustness is better judged by both statistical strength and the shape of returns across portfolios.

ETFs

Volatility and Returns to Leveraged ETFs (Bessembinder)

Leveraged ETFs don’t necessarily suffer from high volatility. What matters is serial correlation in the underlying returns. When returns tend to persist, more volatility can boost expected leveraged returns; when they reverse, it hurts. Without serial correlation, volatility itself has no effect. Key takeaway: Serial correlation, not volatility alone, determines the expected impact of daily rebalancing.

Twenty Years of Leveraged and Inverse Exchange-Traded Products: What Have We Learned? (Chang and Madhavan)

Twenty years after leveraged ETFs were launched, the authors review what investors have learned. Daily leverage works as advertised, but long-run returns are highly path-dependent. Hidden financing costs can create additional performance drag. Key takeaway: With leveraged ETFs, the return path and hidden costs matter greatly for realized returns.

Machine Learning & Large Language Models

Agentic Quantitative Trading: A Survey of Workflows, Systems, and Evaluation (Hua et al.)

This survey paper covers how agentic AI is being applied to quantitative trading. Across 20 systems, signal discovery is nearly universal, but only 2 consider all five core areas: Factor research, signals, portfolios, execution, and risk management. Key takeaway: Strong AI forecasts and backtests do not reliably translate into live trading performance as risk control, liquidity, and execution costs matter greatly.

Option-Implied Signals and Crash Risk: Predictability and Machine-Learning Evidence from U.S. Equity Options (Li and Wang)

Machine learning is uncovering option signals that traditional measures miss. Using U.S. equity options from 2015–2026, XGBoost beats linear models for next-month return prediction in the recent 2023–26 regime: 1.29% vs. 0.07% OOS R². The most important predictors also change across market regimes. Key takeaway: Option-market predictability is regime-dependent, with nonlinear ML adding value in the most recent period.

Options

Volatility Salience: Evidence from the Options Market (Ding, Li, Shi, and Zhai)

Investors appear to overpay for options when a stock’s recent volatility stands out relative to market volatility. Across U.S. equity options from 1996–2025, greater “volatility salience” predicts lower subsequent delta-hedged returns. The effect is strongest in OTM options: High-minus-low spreads reach −1.11% for calls and −1.83% for puts. Key takeaway: Unusually attention-grabbing volatility appears to predict option overpricing.

OPIN and the Cross Section of Equity Option Returns (Chen, Hu, and Yang)

Informed options trading appears to be priced. A new measure of informed trading intensity (OPIN) strongly predicts delta-hedged option returns: Options with more informed activity subsequently earn lower returns. For ATM calls, the high-minus-low OPIN spread is about −2.3% over 52 days. Key takeaway: Option order flow contains information about subsequent option returns.

Portfolio Construction

The Marginal Cost of Diversification (Sanford)

Diversification can eventually increase portfolio risk. In U.S. equities, nearly half of stock additions that reduce idiosyncratic risk also increase systematic risk. As portfolios grow, the idiosyncratic benefit shrinks faster than the systematic-risk effect. Key takeaway: More holdings do not necessarily mean less total risk; what matters is how each addition changes factor exposure.

Prediction Markets

The Cross-Section of Price Changes in Prediction Markets (Hou, Li, and Zheng)

Prediction markets exhibit short-term reversal. Across nearly 97,000 Kalshi contracts, recent 1–6 day price moves tend to reverse, with the strongest effect at the 1-day horizon: Contracts with the largest prior increases underperform those with the largest declines by 0.77 cents the next day. Key takeaway: Prediction-market prices show short-run mean reversion.

Blogs

Does Short-Term Mean Reversion Work Across Asset Classes? (Quantseeker)

Can ChatGPT Forecast Stock Price Movements? (Alpha Architect)

Podcasts

David Booth: What True Wealth Actually Means (Rational Reminder)

Jack Schwager: 40 Years of Market Wizards (Capital Horizons)

The Art of Simple Investing | Tips from Ben Carlson (Bogleheads)

Social Media & Industry Research

September Setup: The Asymmetry Has Changed (Citadel)

Academic Alpha (AQR)

Last Week’s Most Popular Links

Rethinking Time-Series Momentum in Equity Portfolios: Common Signals, Long-Only Positioning, and No Need for Volatility Timing (Zakamulin)

On the Anatomy of Trend (Kjaer)

Prior Sentiment and Returns Around Earnings Announcements (Kazemi and Makridis)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Does Short-Term Mean Reversion Work Across Asset Classes?

QuantSeeker — Thu, 27 Aug 2026 20:23:29 GMT

In a previous post, I tested several simple short-term mean-reversion signals on liquid equity ETFs. The results showed attractive risk-adjusted returns, particularly during periods of high volatility.

This week, I extend that analysis and ask: Does the same type of short-term mean reversion work across asset classes?

Using intraday data for liquid ETFs covering equities, bonds, currencies, commodities, real estate, and Bitcoin, I test the same signal across markets.

The results reveal large differences across asset classes, and an interesting pattern in when those returns are earned.

Weekly Research Recap

QuantSeeker — Tue, 25 Aug 2026 20:46:35 GMT

This week’s briefing brings together the most useful investment research and market insights I found over the past seven days, spanning academic studies, industry research, blogs, and social media. Links to every source are included.

Crypto

Skewness Illusion in Cryptocurrency Portfolios (Horvath and Rigamonti)

Crypto’s lottery-like payoff largely disappears with diversification. Across 2,359 coins, 96% have positively skewed weekly returns. But that skewness largely disappears once portfolios reach around 25 coins. As in equities, diversification washes out the extreme positive skew of individual assets. Key takeaway: Positive skewness is largely an individual-coin phenomenon, and not a feature of crypto portfolios.

Equities

Prior Sentiment and Returns Around Earnings Announcements (Kazemi and Makridis)

Investor sentiment before earnings predicts returns after the announcement, even after accounting for the earnings surprise. Across 83,293 U.S. earnings events, low-sentiment stocks had 1.38 percentage points higher abnormal returns than high-sentiment stocks over the next 10 days. The pattern holds after both earnings beats and misses. Key takeaway: Prior sentiment contains return information beyond the earnings surprise.

Rethinking Time-Series Momentum in Equity Portfolios: Common Signals, Long-Only Positioning, and No Need for Volatility Timing (Zakamulin)

Across U.S. size, value, investment, and profitability portfolios, a single 12-month market momentum signal works better than calculating momentum separately for each portfolio. The best Sharpe comes from staying long when market momentum is positive and moving to cash otherwise, without volatility scaling. Key takeaway: Within equities, trade the market trend, not each portfolio’s own trend.

Sentiment and Return Predictability in Global Value Chains (Heckmann)

Across global equities, FinBERT sentiment from a firm’s customers and suppliers predicts its next-month returns beyond the firm’s own tone. Optimistic customers predict higher returns; optimistic suppliers often predict lower returns, consistent with greater input-cost pressure and bargaining power. Key takeaway: Customer optimism can be good news for a stock; supplier optimism can be bad news.

FX

Option-implied Hurst exponent for FX timing (Petric and Schadner)

FX options can tell investors when to take currency risk. Across 9 major currencies since 2001, using an option-implied Hurst measure derived from the volatility term structure to time FX exposure cut average volatility from 10.6% to 5.3% and max drawdown from 41.4% to 20.3% versus passive FX exposure. Key takeaway: FX option term structures seem useful for timing downside risk.

Options

The Price Impact of Hedging Expiring Index Options (Amaya, Garcia-Ares, Pearson, and Vasquez)

0DTE options can move the S&P 500 into the close through dealer hedge unwinding. On the highest- vs. lowest-dealer-delta days, S&P 500 futures returns differed by 17 bps over the final 30 minutes and 12.5 bps over the final 15 minutes. Key takeaway: Dealer delta in expiring options can create systematic price pressure into the close.

When Does Intraday Options Trading Predict Stock Returns? (Lin, Luo, and Shao)

Options traders seem to know more near the closing bell. Across 3,317 stocks, bullish options order imbalance in the final 30 minutes predicts higher next-day returns more than twice as strongly as the average 30-minute options order-imbalance measure. Key takeaway: For predicting next-day stock returns, late-day options positioning contains the strongest signal.

Trend Following

On the Anatomy of Trend (Kjaer)

Trend-following’s crisis performance is distinct from its average timing ability. Across 81 markets over 25 years, trend provides increasingly strong protection during large market declines. Yet, its signals show much less evidence of consistently timing subsequent returns. Key takeaway: Trend’s most robust feature is its downside protection, not its average timing ability.

Blogs

Wealth management with macro factors (Macrosynergy)

The Sharpe Stability Ratio: Evaluating the Sharpe Ratio Temporal Consistency (Portfolio Optimizer)

Interest Rate Surge? (John H. Cochrane)

Podcasts

Mebane Faber Interview with Michael Covel on Trend Following Radio (Michael Covel)

Why Trend Following Feels Wrong (But Could Boost Your Portfolio) (Meb Faber)

Learning from Buffett and Munger (Masters in Business)

Social Media & Industry Research

Elastic Expectations (Citadel)

Crazy days in the stock market (Acadian Asset Management, Owen Lamont)

Passive-Investing Boom is Making Life Harder for Stock Pickers (Justina Lee, Bloomberg)

What’s Pushing Long-Term Bond Yields Higher? (PIMCO)

Last Week’s Most Popular Links

Unexpected Gross Profit and Cross-Sectional Stock Returns (Yang, Cai, Rhee, and Wu)

VIX and Trend Following Revisited: Nearly a Decade of Out-of-Sample Evidence (Alpha Architect)

AgonAlpha: Autonomous Alpha Discovery via Prompt Economy and Scalable Agentic Search (Ye, Sun, Ren, Yu, Yi, and Yang)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Weekly Research Recap

QuantSeeker — Tue, 18 Aug 2026 20:10:59 GMT

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)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Beyond Volatility Scaling: Does “Good” and “Bad” Volatility Matter?

QuantSeeker — Sun, 16 Aug 2026 18:40:29 GMT

Traditional volatility scaling, reducing exposure when realized volatility is high and increasing it when volatility is low, has been shown to improve risk-adjusted returns for some assets and strategies. The intuition is straightforward: Risk is generally more predictable than returns, so adjusting exposure as risk changes can improve portfolio efficiency without requiring a directional market forecast.

The evidence appears particularly strong for market and momentum portfolios. For example, Schwarz (2025) studies nine factors across 45 international equity markets and finds that, after realistic transaction costs, the clearest benefits remain for market and momentum portfolios.

Harvey et al. (2018), using more than 60 assets with data extending back to 1926, find that Sharpe-ratio improvements are concentrated in equities and credit. For bonds, currencies, and commodities, volatility scaling does little for Sharpe ratios, although it can still reduce tail risk.

In short, volatility scaling appears most useful for equity-like risk assets and much less reliable elsewhere. That raises a natural question: What if we distinguish between upside and downside volatility rather than treating all volatility equally?

In this post, I discuss recent research on this idea and test it across asset classes.

Weekly Research Recap

QuantSeeker — Tue, 11 Aug 2026 19:02:07 GMT

Welcome to this week’s briefing, your curated roundup of the most actionable investing insights from the past seven days, drawing from academic research, industry reports, blogs, and social media, with links to everything.

Commodities

Latent Characteristic Factors in Commodity Markets (Sakkas and Stavroglou)

Many commodity signals tend to reflect the same underlying forces. Using 11 characteristics across 25 commodity futures, this paper finds that much of their variation can be reduced to just two factors: One related to how risk is shared among market participants, and another related to broader macroeconomic conditions. Key takeaway: The commodity factor zoo seems to be much smaller than it appears.

Crypto

Agentic AI Nowcasting and Cryptocurrency Return Predictability (Sun, Wang, and Zhang)

Agentic AI can turn real-time public information into investable signals. In a real-time test across coins in the CoinMarketCap 100 Index, AI-generated scores predicted crypto returns from 1-hour to 1-week horizons. A Top 10–Bottom 10 portfolio earned 1.78% at the 1-day horizon. Key takeaway: AI can uncover return-relevant information not yet fully reflected in crypto prices.

Equities

Factor Neutralization and Risk-Budgeted Scaling: Evidence from Long-Short Anomaly Portfolios (Kosmakov)

Factor neutralization can improve anomaly portfolios, but only when the risk reduction outweighs the return sacrificed. Across 133 long–short strategies, removing exposure to the Fama-French five factors plus momentum raised average Sharpe from 2.25 to 3.57. Key takeaway: Neutralization is most effective when factor exposure contributes more to portfolio risk than to expected returns.

Fixed Income

Overnight Corporate Bond Trading (Abdi, Rossi, and Wu)

Corporate bond trading contains an unusual overnight signal. When public trade reporting is temporarily paused, institutional trades become much larger, and their order flow carries substantially more information about future prices than daytime trading. The signal even spills across bonds from the same issuer. Key takeaway: Overnight order flow contains information about future bond returns that is only gradually incorporated into prices.

Machine Learning and Large Language Models

International Yield Curves and Exchange Rates (Wang and Zhu)

Exchange rates are more predictable than commonly thought. Using machine learning and yield curves from 17 countries, this paper beats the random-walk benchmark OOS for 8 of 9 G10 currencies at 1 month, and all 9 at longer horizons. Key takeaway: Currency predictability depends on information across global yield curves, not just bilateral interest-rate differences.

Machine learning for realised volatility forecasting (Rahimikia and Poon)

Machine learning can improve volatility forecasts, but its advantage disappears when volatility becomes extreme. Across 23 NASDAQ stocks, ML models using order-book, news, and historical volatility data beat traditional HAR models during 90% of the out-of-sample period, but underperformed on extreme-volatility days. Key takeaway: Combining ML with traditional models is more robust than replacing them.

On the predictability of ETF returns with technical predictors (Gong and Muller)

This paper trains a random forest model on decades of individual-stock data, then transfers the learned relationships to international equity ETFs. The resulting top-minus-bottom decile portfolio earned 0.76% per month out of sample, with momentum and volatility signals particularly effective. Key takeaway: Predictive patterns learned from individual stocks can transfer to ETFs.

Yield Curve Prediction with Machine Learning: Forecasting Approaches and the Role of Macroeconomic Predictors (Kang)

Machine learning can improve Treasury yield-curve forecasts, but the gains are highly concentrated. Using 2015–2025 out-of-sample data, the greatest improvements appear at short maturities and in curve-slope forecasts, especially when macro data are included. Key takeaway: Macro information is most valuable for predicting changes in the shape of the yield curve, particularly its slope.

Do Market Regimes Improve Machine-Learning Stock Ranking? (Pascual Miralles and Alfeus)

Machine learning can rank stocks effectively without elaborate regime timing. An XGBoost model ranking 200 U.S. stocks produced a 34.2% CAGR and 1.28 Sharpe out of sample from 2021–2025. Adding market regimes or portfolio optimization failed to improve results. Key takeaway: Adding sophistication to a strong investment signal doesn’t necessarily improve performance.

Prediction Markets

Prediction Markets as Event Hedges: Can They Actually Be Used in Practice? (Kirillov)

Prediction markets can potentially hedge event risk, not just forecast it. A Trump-win contract reduced the one-day loss on a $100k solar ETF position from −9.32% to +0.53%. But limited liquidity made the approach impractical at institutional scale. Key takeaway: Prediction markets can provide targeted event hedges for smaller portfolios.

Blogs

Revisiting Intraday Momentum (Quantseeker)

Bond indices and systematic duration management (Macrosynergy)

Most Asset Managers Already ‘Use AI.’ Few Turn It Into Alpha (CFA Institute)

Podcasts

How a 50-Year Veteran Thinks About Risk Management · Peter Brandt (Chat with Traders)

Trading 80 Synthetic Markets with Trend, Carry & Skew: Jiro Fujisawa, Asset Management One USA (RCM Alternatives)

Portable Alpha: Asking the Questions That Matter ft. Harry Moore (Top Traders Unplugged)

Nicolas Mirjolet, CEO of Quantica on Why the Most Profitable Trends Face the Most Skepticism (StrandGlobalMacro)

Social Media & Industry Research

Traders on Defense (Citadel)

Skewness as a Hidden Driver of Anomaly Returns (Alpha Architect)

Portable Alpha: Ask the Hard Questions (Man Group)

Last Week’s Most Popular Links

Disagreement on Volatility (Yang and Zhu)

A Seat at the Table: Turning AI Macroeconomic Views into Portfolios with a Factor Framework (Elkamhi and Lee)

The Costs and Benefits of Leveraged ETFs (Murray and Sammon)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Revisiting Intraday Momentum

QuantSeeker — Thu, 06 Aug 2026 23:37:58 GMT

Intraday momentum isn’t discussed nearly as often as long-term momentum, yet markets sometimes continue moving in the same direction over short horizons within a single trading session.

A few years ago, researchers proposed that this pattern isn’t simply random noise. They argued it reflects systematic trading flows from market participants who are forced to rebalance as markets move, creating predictable short-term price pressure.

The results attracted considerable attention, but one obvious question remains:

Does the effect still hold today, and is it actually tradable after accounting for implementation costs?

In this post, I test the original idea, focusing on SPY and QQQ, extend the sample well beyond the original study, and test whether the edge has survived out of sample. I also examine whether simple filters improve the strategy and how much transaction costs it can absorb.

Weekly Research Recap

QuantSeeker — Tue, 04 Aug 2026 18:29:56 GMT

Every week I sift through academic papers, industry research, market commentary, and investing discussions to find the ideas that matter. Here are this week's highlights, with links throughout.

Crypto

Crypto Futures Risk Mitigation: Dynamic Rebalancing, Carry and ARP Strategies (Georgiou, Eigenheer, and Sapuric)

Crypto futures often offer unusually high carry, but capturing it is harder than it looks. This paper moves beyond documenting the anomaly by testing implementable long-spot/short-futures strategies with dynamic rebalancing and option overlays. Key takeaway: The persistence of a risk premium doesn't guarantee it's easy to harvest.

Equities

Disagreement on Volatility (Yang and Zhu)

Investor disagreement isn't just about expected returns, but also about future risk. This paper measures disagreement about future stock volatility and finds that a long–short portfolio buying the lowest-disagreement stocks and shorting the highest-disagreement stocks earns about a 1% monthly return spread. Key takeaway: Risk disagreement is priced.

The Costs and Benefits of Leveraged ETFs (Murray and Sammon)

Leveraged ETFs aren’t inherently good or bad. This paper shows their success depends on what they leverage. Broad index LETFs can create investor value, while single-stock LETFs face much higher volatility drag and financing costs. Key takeaway: A 2× ETF on an index is economically very different from a 2× ETF on a single stock.

Short Interest Data is Short (Welborn)

Official short interest is widely used to gauge bearish positioning. This paper suggests it misses part of the picture. Comparing FINRA data with securities-lending records, lending exceeded reported short interest on 47 to 61% of reporting dates. Key takeaway: Public short-interest data probably understate true short exposure.

FX

Covered Interest Parity: The Long Run Evidence (Accominotti, Cen, Chambers, and Degorce)

Covered interest parity (CIP) didn't just break after 2008. Using data from 19 currencies over 1963–2025, this paper shows violations were common for decades. The real exception was 2000–06. Key takeaway: Even textbook arbitrage can persist for years when financial intermediaries face balance-sheet constraints.

Machine Learning and Large Language Models

A Seat at the Table: Turning AI Macroeconomic Views into Portfolios with a Factor Framework (Elkamhi and Lee)

Everyone is asking how AI will change investing. This paper argues AI's biggest role may not be stock picking, but helping investors form disciplined, probabilistic macro views. Those views are translated into candidate portfolios through a transparent, rule-based framework while humans retain final control. Key takeaway: The greatest value of AI in asset allocation is augmenting human judgment, not replacing it.

Options

The Factor Structure of 0DTE Option Returns (Dorion, Orlowski, and Song)

0DTE options now account for a huge share of SPX trading, but are they mispriced? This paper shows that six systematic risk factors explain most intraday 0DTE option returns. Compensation comes from bearing gamma and jump-skewness risk, while any remaining alpha disappears after realistic transaction costs. Key takeaway: Apparent 0DTE mispricing largely reflects systematic risk exposures.

Portfolio Construction

Is Correlation Neglect Bad for Portfolio Diversification? (Chen)

Investors are often told to pay more attention to correlations. This paper argues the real challenge is noisy correlation estimates. Preserving only the direction of co-movement produces more robust out-of-sample portfolios. Key takeaway: Robust diversification comes from preserving the direction of correlations, not their noisy magnitudes.

Private Equity

Buyout Performance with Assets Valued at Market (Ennis and Rasmussen)

This paper challenges one of private equity's biggest narratives. Using market prices instead of sponsor-reported NAVs, the authors estimate a beta of roughly 1.5–1.6 and find no statistically significant alpha. Key takeaway: Smoothed valuations can make private equity appear less risky, and more skillful, than market prices suggest.

Volatility

Beyond Current News: A Momentum-Augmented Model of Future Volatility (Zakamulin)

This paper uses 12-month momentum to forecast next-month volatility and finds a robust U-shaped relationship. Future volatility is highest after prolonged declines, lowest after moderate positive trends, and rises again after exceptionally strong rallies. Key takeaway: Momentum contains information about future risk, not just future returns.

Blogs

Stablecoins and the Future of Treasury Markets (CFA Institute)

Raise rates? Warsh’s conundrums (John H. Cochrane)

Conformal Prediction in Quantitative Finance (Vertox Quant)

A Window into Bond Investors’ Uncertainty About R‑Star (New York Fed, Guillaume Roussellet)

Podcasts

Stacie Mintz – Turning Qualitative Fundamentals into Quantitative Factors (Flirting with Models)

The 6-Year Test 10,000 Tickers Failed: with Eric Crittenden of Standpoint Funds (RCM Alternatives)

Private Equity Unwind with Dan Rasmussen, Founder & Managing Partner of Verdad Advisers (Planet MicroCap)

Social Media & Industry Research

Quick Take: Follow the Earnings (Acadian Asset Management)

Flying below the radar: Insider trading by executives below the top (Alpha Architect)

The frenzied serenity of the stock market (Acadian Asset Management, Owen Lamont)

Last Week’s Most Popular Links

The Science and Practice of Trend-Following Systems (Sepp and Lucic)

Learning to Trade I - Greek, Parameter and Statistical Hedging (Buehler)

Visual Recovery (Hu and Wu)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Weekly Research Recap

QuantSeeker — Tue, 28 Jul 2026 21:54:45 GMT

Here’s this week’s Tuesday roundup, featuring a selection of the most practical investing ideas from the past seven days of academic research, industry reports, blogs, and insightful discussions across social platforms, with links included throughout.

Derivatives and Volatility

Learning to Trade I - Greek, Parameter and Statistical Hedging (Buehler)

These lecture notes cover the evolution of derivatives hedging from traditional Greek-based methods to statistical hedging. Their central message is that hedging decisions can be learned from market data while accounting for transaction costs, real-world market dynamics, and tail risk.

Visual Recovery (Hu and Wu)

Investors often reduce option markets to implied volatility, skew, or term structure. This paper shows that the entire implied volatility surface contains additional information. Using the full surface to predict stock returns generates a long-short equity strategy with a 1.76% monthly return spread. Key takeaway: Preserve the full options surface instead of compressing it into a few metrics.

The Efficacy of Zero-Days-to-Expiration (0DTE) Options Strategies An Empirical Study of 3,909 Option Alpha Backtests (Kam)

0DTE premium selling can be profitable, but this study argues that curated "top" backtests materially overstate the true edge due to overfitting and data snooping. Key takeaway: Be sceptical of spectacular backtests; robustness matters more than headline returns.

Equities

In good and in bad times? The relation between anomaly returns and market states (Muller and Preissler)

This paper studies 133 equity anomalies across 56 markets and finds that they generate significantly higher long-short returns during unfavorable market states, with roughly 75% of the performance difference attributable to the short side. Key takeaway: When anomaly returns strengthen in bad times, mispricing appears to explain the evidence better than conventional risk compensation.

Sentiment-driven factor timing: applying news sentiment to Fama-French factors (Uhl and Mugrauer)

News sentiment can be aggregated at the factor level to dynamically time Fama-French factors. The strategy is shown to outperform an equal-weighted benchmark and generate statistically significant alpha. Key takeaway: News can improve factor timing, not just stock selection.

When Does Risk-Managed Momentum Add Value? An Out-of-Sample, Regime-Conditional, and Statistically Tested Study of a Volatility-Targeted Dual-Momentum Strategy (Goyal)

Investors often ask whether risk-managed momentum beats buy-and-hold. This paper argues that's the wrong question. Its value is regime-dependent: It meaningfully reduced losses in the slow 2022 bear market but struggled during the rapid COVID crash because monthly momentum reacted too slowly. Key takeaway: The effectiveness of momentum-based risk management depends on how quickly markets fall.

Diversification all the Way Down: Multifactor Investing Within Factors, Across Factors and Across Time (Renshaw, Gu, Candar, and Ang)

Most investors diversify across factors. This paper argues that's only the beginning. It shows diversification should also happen within factors (using multiple signals) and across timing models, not just assets. The biggest edge comes from more robust factor construction, while tactical tilts provide only a modest improvement. Key takeaway: The strongest portfolios diversify at multiple levels, not just across assets.

Fixed Income

MuniVIX: Implied Volatility and Stress Regimes in the Municipal Bond Market (Marlowe)

This paper builds MuniVIX, an implied volatility index from municipal bond ETF options, then strips out Treasury-rate volatility to isolate municipal-specific stress. In historical municipal-driven stress episodes, this residual often led broader rate volatility by 1–6 weeks and predicted ETF outflows. Key takeaway: Looking beyond headline volatility indices can reveal early signs of market-specific stress.

Machine Learning and Large Language Models

The Validation Bottleneck: Alpha Discovery When Hypotheses Are Free (Wuebben)

Investment research is entering a new era. AI can now produce an endless stream of plausible trading ideas, but the supply of credible evidence has not changed. This paper argues that while hypotheses scale almost without limit, convincing out-of-sample proof still arrives only with time. Key takeaway: As idea generation becomes abundant, rigorous validation becomes increasingly valuable.

Order Flow and Cryptocurrency Returns: A Machine Learning Approach with Out-of-Sample Validation (Saluja, Sreenu, Srivastava, and Srivastava)

Order flow has strong predictive power for crypto returns. Nonlinear machine learning models extract this signal more effectively than traditional linear methods, producing long-short portfolios with up to 0.86% daily alpha and an annual Sharpe ratio of 3.97. Key takeaway: Order flow provides incremental information beyond what is already reflected in crypto prices.

Trend Following

The Science and Practice of Trend-Following Systems (Sepp and Lucic)

This rigorous paper develops a unified theory of trend-following systems, deriving closed-form results for expected returns, Sharpe ratios, turnover, trading costs, and optimal lookback horizons. Key takeaway: The effectiveness of trend following depends as much on implementation as on the trend signal itself.

Blogs

Shorting Leveraged ETFs: A Free Lunch or Just an Expensive One? (Quantseeker)

Portfolio optimization with macro factors and neural networks (Macrosynergy)

The Self-Driving Portfolio: Promise, Pitfalls, and the Practitioner Gap (CFA Institute)

Machine Learning Improves Return Forecasting (CFA Institute)

Podcasts

Ex-Balyasny PM: Quant is like Blackjack, Fundamental is like Poker (Odds on Open)

The Next Evolution of Trend Following ft. Nick Baltas (Top Traders Unplugged)

Kieran Duff - Trading for a Living is Easier Now (The Algorithmic Advantage)

Social Media & Industry Research

Hynix Hijinks (Acadian Asset Management, Owen Lamont)

Flying below the radar: Insider trading by executives below the top (Alpha Architect)

Fed Views: The Case for July (Citadel)

Last Week’s Most Popular Links

Bimodality Everywhere: International Evidence of Deep Momentum (Han and Qin)

When Do Anomalies Deserve Capital? Macro-Regime Memory and Dynamic Cross-Sectional Factor Allocation (Taheri Hosseinkhani)

When Simplicity Beats Optimization: Evidence from Factor Timing, Volatility Management, and the 1/N Benchmark (Feng)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Shorting Leveraged ETFs: A Free Lunch or Just an Expensive One?

QuantSeeker — Sun, 26 Jul 2026 21:12:47 GMT

Leveraged ETFs are known to suffer from volatility decay, particularly in turbulent markets. That suggests an intuitive trade: Short leveraged ETFs while hedging their market exposure. With directional exposure largely neutralized, the portfolio should be positioned to harvest the decay.

This week, I examine recent research on this trade and several related implementations, extend the evidence through July 2026, and test what survives key implementation frictions. Most versions become considerably less attractive, but one appears more robust.

Weekly Research Recap

QuantSeeker — Tue, 21 Jul 2026 23:06:43 GMT

Each Tuesday, I share the most interesting market and investing insights I came across during the past seven days, including new research papers, blogs, and podcasts. Links are included throughout for readers who want to explore the ideas in more detail.

Crypto

Demand-Side Fee Flows and Return Predictability on Ethereum (Dunnes and Eckberg)

Ethereum may have a fundamental valuation signal after all. This paper builds a valuation metric from Ethereum’s protocol fees relative to token supply and finds that periods of unusually strong fee activity predict higher ETH returns over the next 1–2 months. The effect appears only after the 2024 Dencun upgrade, with out-of-sample R² reaching 14.4% at a 45-day horizon. Key takeaway: Protocol fundamentals seem to contain predictive information for crypto returns.

The “Actual Retail Price” of Crypto Trades (Rösch, Shohfi, Stanco, and Walz)

Retail crypto trading may be far more expensive than most investors realize. This paper executed nearly 500 real trades across Coinbase, Kraken, Crypto.com, and Robinhood, finding all-in trading costs of 253 to 834 bps on major crypto exchanges, while Robinhood and crypto ETFs often provided much cheaper execution. Key takeaway: For retail crypto investors, choosing the right platform can save more than picking the right coin.

Equities

War Risk and Time-Varying Return Predictability of the Dividend-Price Ratio (Yu, Xiao, Zhou)

The dividend-price ratio predicts future U.S. stock returns primarily during periods of active war. Outside high-war periods, it adds little forecasting value. The authors argue that war speeds the correction of market mispricing, making valuation signals more informative. Key takeaway: Even classic valuation signals aren’t universally useful, their predictive power depends on the market regime.

When Do Anomalies Deserve Capital? Macro-Regime Memory and Dynamic Cross-Sectional Factor Allocation (Taheri Hosseinkhani)

Macro timing is one of investing’s oldest ideas. This paper applies it in a different way. Instead of timing the market, it dynamically allocates capital across factor families like value, momentum, quality, and liquidity using macro regimes. Across 212 equity signals, the approach achieves an out-of-sample Sharpe ratio of 1.51. Key takeaway: In factor investing, the timing of factor allocation can matter as much as the factors themselves.

On Volatility-managed Portfolios: The Role of Conditional Factor Correlations (Barroso, Maio, and Wang)

Only 3 of 9 individual equity factors deliver robust gains from volatility scaling. However, volatility scaling remains effective for diversified factor portfolios. It’s not just about timing volatility, it automatically reduces exposure when factor correlations rise, and the paper shows much of its edge comes from this mechanism. Key takeaway: In multi-factor investing, volatility scaling helps manage the impact of rising factor correlations, not just volatility itself.

Fixed Income

Understanding Corporate Bond Excess Returns (Cui, Lu, and Song)

The excess returns on corporate bonds aren’t as credit-driven as many investors assume. This paper matches every corporate bond with a synthetic Treasury having identical cash flows, isolating returns unique to credit markets. It finds that roughly 41% of the excess return measured over T-bills is actually compensation for interest-rate risk. Key takeaway: Separate duration from corporate-bond-specific risk before evaluating corporate bond strategies.

FX

The Shape of Carry: Construction, Leverage, and Crash Risk Across Currency and Bond Markets (Gouws)

Carry delivers a modest 0.5 Sharpe across currencies and bonds but comes with severe negative skew and crash risk. Proper portfolio construction makes the premium more investable: Diversify carry sources and size positions by risk. Key takeaway: Proper risk management is key to harvesting the carry premium.

Machine Learning and Large Language Models

Bimodality Everywhere: International Evidence of Deep Momentum (Han and Qin)

This paper enhances traditional momentum by predicting the entire return distribution rather than a single outcome, then ranking stocks by their expected returns. Across 45 countries, this improved the average annualized Sharpe ratio from 0.98 for traditional momentum to 2.49. Key takeaway: Rather than relying only on the most likely outcome, use the entire predicted probability distribution when constructing momentum portfolios.

Portfolio Construction

When Simplicity Beats Optimization: Evidence from Factor Timing, Volatility Management, and the 1/N Benchmark (Feng)

Investors often assume sophisticated optimization beats simple diversification. This paper suggests otherwise. Across nine equity factors (1976–2025), recursive mean-variance optimization and dynamic factor timing generally failed to outperform a simple equal-weighted (1/N) portfolio out of sample. Key takeaway: Simple diversification remains remarkably difficult to beat once estimation risk is taken into account.

Risk Management

A Brief History of Financial Risk (Hartzmark and Solomon)

For most of market history, investors didn’t think about risk the way we do today. This paper shows that before Markowitz, risk wasn’t volatility, it was largely the chance of losing money over the long run, while stocks were often viewed as speculation rather than investments. Modern portfolio theory fundamentally changed how investors think about risk. Key takeaway: Today's definition of risk is relatively recent, and it fundamentally changed investing.

Blogs

Beyond Diversification: Factor Risk in Sovereign Portfolios (CFA Institute)

Can AI Do Financial Research? (Quantpedia)

Podcasts

~100% Returns in 2025, No Losing Year Since 2008: Erik Smolinski on Edge for Retail Trader Edge (Odds on Open)

The Creator of Claude Code on The Hottest Piece of Software in the World (Odd Lots)

Why Trend Following Is Changing Faster Than Ever ft. Yoav Git (Top Traders Unplugged)

Social Media & Industry Research

The Intramonth Momentum Cycle (Alpha Architect)

Two Accounting Anomalies: One May Be Risk, the Other Is Mispricing (Alpha Architect)

A More Fragile World (Citadel)

Last Week’s Most Popular Links

Winners Glide, Losers Stumble: A Behavioral Reversal Tilt to Momentum (Wen and Feng)

Betting on the Leverage Effect (Doshi, Jacobs, and Sert)

Do AIs Make Good Traders, and Do They Make Good Traders Better? (Bell, Haghani, and White)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Weekly Research Recap

QuantSeeker — Tue, 14 Jul 2026 20:52:38 GMT

Welcome to this week’s recap, your curated roundup of the most actionable investing insights from the past seven days, drawing from academic research, industry reports, blogs, and social media, with links to everything.

Equities

Betting on the Leverage Effect (Doshi, Jacobs, and Sert)

Stocks with the strongest negative return-volatility correlation (”leverage effect”) earn significantly higher future returns, and the signal outperforms traditional skewness measures. The premium comes almost entirely from increases in volatility, not volatility declines. Key takeaway: Measure asymmetric risk through return-volatility dynamics, not skewness alone.

Winners Glide, Losers Stumble: A Behavioral Reversal Tilt to Momentum (Wen and Feng)

Most momentum research tries to improve the winner side. This paper argues that the real opportunity lies on the losing side. Among past losers, recent rebounds often fade, so overweighting those names on the short side improves momentum and reduces momentum crash risk. Key takeaway: Momentum has more to gain from better short selection than from better long selection.

Redeeming Value (Penman and Reggiani)

Everyone says value investing is about buying low P/B stocks. This paper argues that it’s too simplistic. Book-to-price only becomes informative when combined with earnings. Together, they help reveal the market’s implied expected return, or the growth expectations embedded in today’s price. Key takeaway: Value investing isn’t about buying “cheap” stocks. It’s about identifying mispriced growth expectations.

FX

Rethinking currency factors: The case for mean-variance optimisation (Fan, Kearney, Li, and Liu)

Instead of equal-weighting positions in carry, value, and momentum portfolios, this paper shows that optimizing constituent weights with Bayesian mean-variance optimization boosts the carry factor’s out-of-sample Sharpe ratio from 0.69 to 0.76 and the value factor’s from 0.19 to 0.50. The edge comes from better portfolio construction, not better forecasting. Key takeaway: In factor investing, portfolio construction can be as important as the signal itself.

Skewness Risk Premia and the Cross Section of Currency Returns (Li, Sarno, and Zinna)

Currency investors are potentially overlooking skewness risk premia. Currencies with the highest skewness risk premia subsequently earn the highest excess returns. The signal outperforms other option-based measures and helps explain returns beyond the classic carry trade. Key takeaway: Skewness risk is a distinct, priced source of currency risk.

Machine Learning and Large Language Models

Do AIs Make Good Traders, and Do They Make Good Traders Better? (Bell, Haghani, and White)

Instead of racing to build better market forecasts with AI, this paper suggests the bigger edge lies elsewhere. Claude and ChatGPT matched elite macro traders surprisingly well at inferring market direction from macro news. Their biggest weakness wasn’t prediction; it was position sizing. When researchers explicitly improved the AI’s risk-sizing framework, performance improved without better forecasts. Key takeaway: Correct sizing often matters more than signal.

More Than Words: Visual Sentiment in Social Media Videos and Stock Returns (Jiang, Jin, and Ma)

Using facial expressions from finance influencers’ Douyin (TikTok) videos, the authors build a visual sentiment index that predicts next-day market returns, even after controlling for what the influencers actually say. The effect is strongest during trading hours and in retail-driven stocks. Key takeaway: Nonverbal cues can contain information that text alone misses.

Options

Here Today, Gone Today: First Evidence on European Zero-Day Options (Wilkens)

This paper offers a systematic study of European 0DTE options on the EURO STOXX 50 and DAX and finds that put-call parity holds almost perfectly at mid quotes. However, once you cross the bid-ask spread, round-trip execution costs consume 10 to 13% of an ATM straddle on average and up to 40–70% in the final 30 minutes before expiry. Key takeaway: In 0DTE trading, execution quality can matter more than pricing efficiency.

Textual Risk Option Momentum (Chen, Vasquez, and Yang)

Option momentum spreads across firms with similar risk disclosures. Options on companies whose textual risk peers had strong past delta-hedged returns earn higher future delta-hedged returns, even after controlling for own-option momentum and other peer networks. Key takeaway: Risk similarity creates hidden links through which option-market information diffuses.

Portfolio Construction

An Automated Black-Litterman Framework for Global Fixed Income (Carver, Harris, and Lin)

Traditional bond allocation struggles because expected returns are noisy. This paper combines factor forecasts with an automated Black-Litterman framework that adjusts confidence based on each model's predictive power. Across 13 global bond indices, Sharpe improves from 0.31 to 0.45 with only a modest increase in drawdown. Key takeaway: Portfolio construction isn't just about better forecasts; it's about knowing how much to trust them.

When and Why Naïve Diversification Works: A Simple Diagnostic Strategy (Feng, Huang, Wang, and Zhang)

Equal weighting often beats optimization out of sample, not because optimization is wrong, but because estimation error is costly. This paper argues that equal weighting becomes nearly optimal when the forecast-error covariance structure is close to a specific eigenvector condition. As the data move away from that condition, optimized weights regain their edge. Key takeaway: Before optimizing, ask whether your data justify it.

Blogs

Bitcoin’s Fading Diversification Benefit (Quantseeker)

Economic surprises and global asset class returns (Macrosynergy)

Quantitative Value (Concretum Group)

Podcasts

Has Trend Following Changed Forever? ft. Alan Dunne (Top Traders Unplugged)

From Skid Row Showers to Short Selling Master · David Capablanca (Chat with Traders)

Social Media & Industry Research

After the Reset: Time to Focus on Fundamentals (Citadel)

Bonds Seem to Not Diversify Anymore. Now What? (Alpha Architect)

The Toolkit Expands (Citadel)

Last Week’s Most Popular Links

Does Momentum Time Risk Rather Than Returns? (Zakamulin)

Is Trend Still Your Friend?: A Microstructural Account of the Demise of Short-Term Trend-Following (Kurth, Eisler, Rej, Bouchaud)

Liquidity Premium and Investment Horizons (Aldridge)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Bitcoin's Fading Diversification Benefit

QuantSeeker — Mon, 13 Jul 2026 23:01:52 GMT

Bitcoin has long been viewed as a portfolio diversifier, an asset that moves independently of stocks and cushions equity drawdowns. However, things have changed. Its correlation with equities has increased; its market beta has risen from essentially zero to roughly one; and the Sharpe ratio benefit of adding a 10% Bitcoin allocation to a portfolio has fallen sharply, although the magnitude depends on the underlying asset mix.

The Changing Correlation and Market Beta

The rolling 3-month correlation between BTC and SPY daily returns tells the basic story. From 2013–2019, it oscillated between roughly -0.3 and +0.3 with no clear trend, an asset trading on its own idiosyncratic cycle, largely detached from equities. Since 2020, the whole distribution has shifted up, spending most of its time between 0.3 and 0.7.

The full-sample correlation between Bitcoin and the S&P 500 is modest (0.16), but it obscures a pronounced regime change. Before 2020, the correlation was essentially zero (-0.02); since 2020, it has averaged 0.37. Investors relying on the full-sample estimate would therefore materially understate Bitcoin's current relationship with equities.

The shift is even clearer when looking at market beta. Before 2020, Bitcoin's market beta was -0.12, implying essentially no systematic equity exposure. Since 2020, it has increased to 1.07, meaning a dollar invested in Bitcoin now carries roughly the same market risk as a dollar invested in the S&P 500. That's not a subtle change. Bitcoin went from behaving largely as an unrelated asset to behaving much more like an equity position.

What might explain this shift? The timing is consistent with a growing literature on Bitcoin’s financialization. An IMF working paper (Che et al., 2023) identifies a common “crypto factor” whose correlation with global equities rose sharply after 2020 as institutional participation increased and the marginal crypto and equity investor became increasingly similar. Conlon and McGee (2020) showed that Bitcoin failed to act as a safe haven during the March 2020 market crash, while Nguyen (2022) found that S&P 500 returns began exerting a stronger influence on Bitcoin returns during periods of high uncertainty.

The rolling correlation also exhibits a second upward shift in early 2024, coinciding with the launch of US spot Bitcoin ETFs on January 11. Around that event, the 30-day rolling correlation increased from roughly 0.02 immediately before the launch to around 0.28 immediately afterwards. While this evidence is descriptive rather than causal, it is consistent with the hypothesis that spot ETFs further integrated Bitcoin into mainstream investment portfolios, as discussed in Hong et al. (2025).

Downside Risk: Where the Real Damage Shows Up

Weekly Research Recap

QuantSeeker — Tue, 07 Jul 2026 19:20:53 GMT

This week’s Tuesday Roundup features the most valuable investing insights I came across over the last seven days, spanning new academic and industry research, as well as high-quality blog posts. Links are included throughout for further reading.

Equities

Lost in the Multiverse: Methodological Uncertainty in Studying Global Equity Returns (Cakici, Fieberg, Neszveda, Piljak, and Zaremba)

Most country-level investing “anomalies” may be less robust than they appear. This paper shows that signals like momentum, beta, and idiosyncratic risk often fail to remain robust across alternative research designs. The most reliable predictors were market size, dividend yield, sovereign risk, and short-term momentum. Key takeaway: Always ask whether an investing edge survives different implementations.

Does Momentum Time Risk Rather Than Returns? (Zakamulin)

This paper challenges the standard view of time-series equity momentum. In the U.S. equity market, its edge comes primarily from identifying future volatility rather than predicting future returns. Long-only momentum improves risk-adjusted performance largely by avoiding high-volatility regimes, not by forecasting higher returns. Key takeaway: For equity momentum, risk timing matters more than return timing.

Liquidity Premium and Investment Horizons (Aldridge)

Not all trading volume is equally informative. Signed order flow is shown to predict the cross-section of next month’s stock returns, while raw volume adds far less information. Key takeaway: Net buying and selling carry more information than raw trading volume.

Options

Hearing and Seeing the Fed: Multimodal Signals Beyond Text in Intraday Option Markets (Filippou, Zhao, Zhou, and Zhou)

Markets don’t just react to what the Fed says; they react to how it’s said. Jerome Powell’s tone of voice and facial expressions are shown to contain incremental information beyond the transcript. Those multimodal signals improved intraday return forecasts and generated profitable 0DTE option trades even after full bid-ask spreads. Key takeaway: How the Fed communicates matters alongside what it communicates.

Portfolio Construction

Bond Portfolio Optimization (Ben Slimane, Cherief, Roncalli, and Xu)

Most portfolio optimization assumes historical risk is informative about future risk. The authors argue this assumption breaks down for individual bonds. As bonds age, duration falls, and credit risk evolves, making historical covariance matrices progressively less informative. Key takeaway: Fixed income requires dynamic risk models, not equity models with different inputs.

Hidden Factors in Portfolio Risk Models: A Finite-Sample Approach to Residual PCA (Kolm and Ritter)

The paper argues that residual PCA is a sensible way to search for missing risk factors, but the resulting factor should only be used if it is strong enough to be distinguished from estimation noise and its estimated loading direction can be reliably recovered. Key takeaway: Validate residual PCA factors before incorporating them into portfolio construction.

Prediction Markets

Settlement Manipulation in Prediction Markets (Dai, Jia, and Yu)

Prediction markets are often praised for improving price discovery. This paper argues that Polymarket's 5-minute Bitcoin prediction market can do the opposite. The authors show that it creates incentives to manipulate the underlying into settlement, degrading price discovery and transferring wealth from retail traders to manipulators. Key takeaway: Ultra-short asset-price prediction markets can create incentives that distort the underlying market rather than improve price discovery.

Trend Following

Is Trend Still Your Friend?: A Microstructural Account of the Demise of Short-Term Trend-Following (Kurth, Eisler, Rej, Bouchaud)

Why did short-term trend following stop working after 2009? This paper argues that the culprit wasn't overcrowding or capacity constraints. Instead, fast trend largely disappeared in small-tick futures contracts, while remaining surprisingly resilient in large-tick contracts, consistent with HFT-driven changes in market microstructure. Key takeaway: The viability of fast trend following depends on contract microstructure, not just the signal itself.

Blogs

Volatility Indicators for Predicting S&P 500 Drawdowns (Quantseeker)

Jumping back in the pool(ing): testing pooling by asset class and portfolio weight distance (Rob Carver)

Silicon vs. Satoshi: Tactical Asset Rotation Between NASDAQ-100 and Bitcoin (Quantpedia)

Podcasts

Ben Wellington – Complex Feature Engineering at Two Sigma (Flirting with Models)

The Bridge Ep. 11: Expensive Isn't a Bubble (iCapital, featuring Cliff Asness)

Ex-Two Sigma Quant: Why You Should Bet Against Conviction (Odds on Open)

Social Media & Industry Research

Learning to Replicate Expert Judgment in Financial Tasks (Mira Murati, Thinking Machines)

When Fear Spikes, Should You Buy? (Victor Haghani, Elm Wealth)

From Energy to Metals: Why to Still Diversify Into Commodities (Goldman Sachs)

The Disappearing Overnight Drift (Boyarchenko, Larsen, and Whelan; New York Fed)

Cruel Summer for Fixed Income (Citadel)

Last Week’s Most Popular Links

Skewness Managed Portfolios (Gong, Lynch, and Ogden)

Oil–VIX States as Conditional Signals for Cross-Asset Allocation (Aylsworth, Poechhacker, Schwaiger, and Werbach)

Systematic FX trading with regression learning and transaction cost analysis (Macrosynergy)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Volatility Indicators for Predicting S&P 500 Drawdowns

QuantSeeker — Mon, 06 Jul 2026 21:50:41 GMT

In a recent post, Which Macro Indicators Actually Predict Market Drawdowns?, I discussed and tested new research on forecasting S&P 500 drawdowns using macro-financial indicators.

In this post, I revisit that topic and test whether two volatility indicators that we have previously shown to be useful for forecasting realized volatility and timing VX futures also help predict future S&P 500 drawdowns. The findings suggest there is additional information embedded in option markets that can help predict near-term market risk.

Finally, I apply the framework to today's market and discuss the current drawdown outlook.

A quick note: Beginning Wednesday, the price of a paid subscription will increase from $39/month to $49/month. Annual subscriptions will remain unchanged. Existing paid subscribers will keep their current pricing for as long as they remain subscribed. The price increase therefore applies only to new subscribers. Thank you for your continued support.

Weekly Research Recap

QuantSeeker — Tue, 30 Jun 2026 19:29:34 GMT

Here’s this week’s Tuesday roundup, a selection of the most practical investing ideas from recent academic studies, industry research, blogs, and insightful discussions across social platforms, with links included throughout.

Asset Allocation

Oil–VIX States as Conditional Signals for Cross-Asset Allocation (Aylsworth, Poechhacker, Schwaiger, and Werbach)

Oil prices alone are an incomplete allocation signal. This paper argues that it's the combination of oil and market volatility (VIX) that matters. High oil with low volatility resembles a late-cycle expansion, while high oil with high volatility signals stress. Key takeaway: The same oil price can imply very different portfolio risks.

Crypto

The Party’s Over: How the Bitcoin ETF Killed Crypto’s Cool Factor (Krause)

This paper argues that Bitcoin ETFs didn't just institutionalize crypto; they changed its character. After the ETF launch, Bitcoin's volatility and equity beta declined, while Dogecoin's Sharpe ratio fell from 0.30 to 0.01 and Google search interest dropped 63%. Key takeaway: As Bitcoin becomes institutionalized, the retail speculation that once fueled crypto booms is likely to become less dominant.

Equities

Returns to “Do-Nothing” Portfolios (Bessembinder)

The last decade rewarded concentration in mega-cap stocks. However, history tells a different story. From 1971 to 2025, concentrated portfolios generally underperformed broader portfolios, while randomly selected narrow portfolios lagged the index more often than they beat it. Key takeaway: Recent mega-cap dominance is the exception, not the rule.

The CAPE that Cried Wolf (Palazzo)

CAPE may not have failed; its denominator changed. The author argues that accounting changes have depressed reported earnings, making traditional CAPE overstate valuations. Adjusting for these distortions restores much of its long-run predictive power. Key takeaway: Before dismissing CAPE, consider whether the accounting changed.

Main Street in Wall Street (Ghezzi)

Wall Street may know more about Main Street than we think. Filtering out high-frequency market noise, stock prices can be used to recover investors’ real-time expectations of economic activity, often before traditional macro data catches up. Key takeaway: Markets are a leading macro indicator, not just a pricing mechanism.

Skewness Managed Portfolios (Gong, Lynch, and Ogden)

Many classic factor premiums are driven by just a handful of extreme winners. The authors show that tilting anomaly portfolios toward stocks with higher expected skewness on the long side and lower expected skewness on the short side improves returns by 5.5% annually across 18 anomalies while also boosting Sharpe ratios. Key takeaway: Expected skewness appears to add value beyond traditional factor signals.

AI Premium (Borri, Liu, and Tsyvinski)

AI adoption may now be priced as a systematic risk factor. Using 380 trillion AI tokens across 400+ LLMs, the authors construct an AI factor from realized AI consumption. Firms with the highest AI exposure subsequently outperformed the least exposed by 64 bps per week, even after standard factor adjustments. Key takeaway: Markets seem to price AI exposure as a distinct source of risk.

Machine Learning and Large Language Models

What NLP Sentiment Can’t Hear -How Vocal Delivery Improves Earnings-Call NLP Sentiment (Pope)

Adding vocal delivery to NLP sentiment improves return predictability. The strongest effect is on the downside: Negative language delivered with vocal strain underperformed similar negative language delivered with greater control by about 1% over the next 20 trading days. Key takeaway: Voice appears to make text-based sentiment more informative, not by replacing NLP, but by helping investors interpret it.

Portfolio Optimization

Adapting the Black-Litterman Framework to Fixed-Income Portfolios (Tandle)

The author adapts the Black–Litterman framework to fixed-income investing by combining market-implied returns with macroeconomic views. While the model improved returns in some risk regimes, it generally failed to beat a simple equal-weight portfolio on a risk-adjusted basis and required much higher turnover. Key takeaway: Sophisticated optimization isn't automatically superior to simple diversification.

Blogs

Buying the Dip Isn’t Free (Quantseeker)

Rolling, rolling, rolling.... updating statistical estimates yes or no (Rob Carver)

One of These Things (Is Not Like the Others). Or is it? Pooling rule p&l estimates across instruments. (Rob Carver)

Systematic FX trading with regression learning and transaction cost analysis (Macrosynergy)

Guardrails Make the Researcher: What an AI Agent Got Right (And Wrong) Replicating Nine Equity Anomalies (Quantpedia)

Podcasts

Peter Hecht – Portable Alpha: Solving the Funding Problem of Alternatives (Flirting with Models)

Ex-Citadel Quant on Trading the Most Asymmetric Market - Neel Somani (Odds on Open)

Social Media & Industry Research

Beta-arbitrage returns and stock market bubbles (Ptuomov)

Reads you may have missed (Man Group)

Catching Angels Before They Fall (Man Group)

Credit’s Systematic Shift (Acadian Asset Management)

Last Week’s Most Popular Links

Boundaries of Time Series Momentum (Suominen and Hjalmarsson)

Beyond Growth Rates: Macro Trends and Price Cycles (Favero, Melone, Myers, and Tamoni)

To Cluster Or Not To Cluster That is the Question... (Rob Carver)

Paid subscriptions may not be available in all jurisdictions and may change without notice.

Buying the Dip Isn't Free

QuantSeeker — Thu, 25 Jun 2026 22:03:26 GMT

Buying the dip is one of the oldest rules in investing.

Markets sell off, prices become cheaper, and the instinctive reaction is obvious: This is the opportunity to buy.

The logic seems compelling. If you liked the market a month ago, shouldn’t you like it even more now that it’s cheaper?

Not necessarily.

A market decline can mean one of two things. Prices may have temporarily overreacted and will eventually recover, or they may be reflecting a genuine deterioration in fundamentals or macro conditions. In real time, you don’t know which.

Even if the decline eventually proves temporary, another question remains: Does buying the dip actually improve your portfolio?

The answer is less obvious than most investors assume.

Buying after a sell-off isn’t free. Every purchase must be financed by selling another asset or by holding cash in reserve. If the money comes from bonds, your portfolio becomes progressively more equity-heavy as markets fall. If it comes from cash, that cash carries an opportunity cost while waiting for the next decline.

The question, then, isn’t whether buying at lower prices feels intuitive. It’s whether the additional equity exposure created during market declines is rewarded with sufficiently higher returns to justify the extra risk.

Weekly Research Recap

QuantSeeker — Tue, 23 Jun 2026 23:39:21 GMT

This week’s Tuesday Roundup brings together the most useful investing ideas I found recently, from fresh academic studies and market research to thoughtful blog posts. Links are included throughout so you can explore further.

Equities

High-end IPO prices (Mueller and Verwijmeren)

This paper finds a surprisingly strong signal in IPOs priced exactly at the top of the range. Across 9,152 U.S. IPOs, high-end priced deals generated significantly higher first-day returns than IPOs priced just below or above that level. Key takeaway: The most underpriced IPOs may not be those that exceed expectations, but those that stop exactly at them.

Short-term Reversal in Chinese A-Shares: Statistical Predictability, Economic Tradability, and Multiple Testing (Fang and Wang)

Short-term reversal in Chinese A-shares appears real. A strategy that buys recent losers and sells recent winners earned 1.72% per month and remained significant after controlling for common risk factors. Yet the investment case weakens considerably once transaction costs, short-sale constraints, and multiple testing are considered. Key takeaway: Statistical predictability does not necessarily translate into investable alpha.

Explaining Two Prominent Accounting Pricing Anomalies: The Accrual Anomaly and the Post-Earnings-Announcement Drift (Penman and Zhu)

Two of accounting’s most famous anomalies may have very different origins. This paper argues that the accrual anomaly largely reflects rational pricing of risk, while post-earnings-announcement drift (PEAD) looks more like genuine mispricing. Key takeaway: Investors should be careful not to treat every anomaly as evidence of market inefficiency.

Which Portfolios? The Construction Dependence of Factor Model Performance (Shin)

Which factor model is best? This paper suggests the answer depends as much on the test portfolios as on the model itself. Using thousands of randomly generated portfolios, the author finds that rankings shift materially with portfolio construction, weighting, and rebalancing choices. Key takeaway: Many debates about the “best” factor model may really be debates about test-asset design.

Macro

Beyond Growth Rates: Macro Trends and Price Cycles (Favero, Melone, Myers, and Tamoni)

Stocks and consumption growth barely move together. Yet this paper finds that stock prices remain anchored to consumption over the long run. When prices stray too far from that anchor, future returns become more predictable. Key takeaway: Consumption contains more information about expected returns than investors realize.

Momentum

Boundaries of Time Series Momentum (Suominen and Hjalmarsson)

Time-series momentum is one of the most persistent effects in finance. Using nearly 100 years of U.S. data and evidence from 20 international markets, the authors find that trend-following works best when valuations are in a normal range. Near extreme levels of CAPE, dividend yield, or the yield curve slope, trends are more likely to reverse, and momentum breaks down. Key takeaway: Valuation extremes are where trend-following strategies are most vulnerable.

Volatility

Volatility Disagreement in the Options Market (Bali, Kelly, and Moerke)

Investors often focus on volatility itself. This paper suggests disagreement about future volatility may matter more. Using a machine-learning measure of volatility disagreement, the authors find that high-disagreement stocks subsequently exhibit much lower delta-hedged straddle returns. Key takeaway: Volatility disagreement appears linked to option overpricing.

Blogs

Testing an AI-Assisted Research Workflow for Multi-Asset Pullback Strategy Discovery (Quantpedia)

To Cluster Or Not To Cluster That is the Question... (Rob Carver)

Breaking Badly: finding the structural breaks in parameter estimates (Rob Carver)

Recession Risk Through a Real-Economy Lens (CFA Institute)

Podcasts

LTCM Co-founder Victor Hagani: “Taking Risk Is Always a Negative.” (Odds on Open)

What Separates Market Wizards From All Other Traders (TraderLion)

Why Most Trend Following Improvements Should Fail ft. Rob Carver (Top Traders Unplugged)

Social Media & Industry Research

Systematic Strategies & Quant Trading (HedgeNordic)

Chasing Trends or Chasing Performance? (Quantica Capital)

Last Week’s Most Popular Links

Regime-Aware Asset Allocation with Dual-Regime Signals and Regime-Dependent Asset Selection (Luo and Mulvey)

A new decomposition approach to modeling financial returns: Conditioning sign on magnitude (Brou and Luger)

Order flow and cryptocurrency returns (Anastasopoulos, Gradojevic, Liu, Maynard, and Tsiakas)

Paid subscriptions may not be available in all jurisdictions and may change without notice.