Research

DV-PMI program

The three working papers behind the indices on this site.

1. Paper 1 Who Profits from Prediction? Execution, not Information
   Della Vedova, J. (2026) · Working paper · SSRN
   Abstract

   Retail traders pick winners yet lose money. In 222 million Polymarket trades, profit reflects two nearly independent skills; execution, not forecasting, determines who profits. Terminal-payoff identification removes the benchmark contamination biasing conventional decompositions. Automated traders at coin-flip accuracy (49.9%) earn the only positive return, $133 million; more accurate retail (51.3%) lose $79 million. The negative cross-skill correlation those decompositions report is the contamination; the bias derived in the paper matches the empirical estimate to within one percent. Orthogonality appears in 14.1 million CBOE options; the inversion does not. Retail traders lose not because they are wrong, but because they are late.

2. Paper 2 Detecting Informed Trading in Prediction Markets: One Event at a Time
   Della Vedova, J. (2026) · Working paper · Revised June 2026 · SSRN abstract 6567238
   Abstract

   Trader-aggregated statistics cannot identify episodic informed trading: pooled averages dilute the signal, and the best-episode rule inflates with activity at the extreme-value rate. Identification requires the trader-event unit. We construct a per-event joint-accuracy test with multiplicity correction across a trader's events and calibrate its false-positive rate on a real-data placebo class where episodic private information is implausible — the corrected test yields zero discoveries at every threshold. Applied to the full Polymarket history (237 million resolved trades), the test shows statistical patterns consistent with informed trading for the trader charged in CFTC v. Spagnuolo (2026), where every trader-aggregated statistic does not.

3. Paper 4 Probability Weighting from Prediction Markets
   Della Vedova, J. and Grant, A. · Work in progress · View live index
   Abstract

   Prediction markets provide a model-free measurement of probability weighting: the price-probability gap in binary contracts is the Prelec weighting function, with no need to specify utility, estimate physical density, or extract risk-neutral density. Pooled across 233M Polymarket non-bot trades the fit is α = 0.664 (R² = 0.987), matching the Tversky-Kahneman (1992) experimental estimate of 0.65 to within standard error. We then test whether the dashboard's weekly prediction-market-implied weighting index predicts cross-sectional equity anomalies (lottery premium, IVOL spread, beta spread) and JKP factor returns.

Peer-reviewed publications

1. Investor Behavior at the 52-Week High
   Della Vedova, J., Grant, A. R., and Westerholm, P. J. · Journal of Financial and Quantitative Analysis (2023), 58(7), 2852-2889 ABS 4 · ABDC A* · FT50
   Abstract

   Household investors amplify the 52-week high anomaly through their trading behavior. Using daily household and institutional trading data, we show that households sharply increase their selling — particularly with limit orders — at the 52-week high price. This uninformed selling at a psychological price milestone leads to a doubling of unconditional 52-week-high anomaly returns, with institutions serving as the profitable counterparty.

2. Who Drives Momentum Returns? The Role Reversal of Trend-Seeking Households and Contrarian Institutions
   Della Vedova, J., Grant, A. R., and Westerholm, P. J. · European Financial Management (2025) ABS 3 · ABDC A
   Abstract

   Cross-sectional momentum is conventionally attributed to slow-moving institutional capital. Using investor-class flow data we show the opposite: households are the trend-followers buying recent winners, while institutions are the contrarians providing liquidity to those flows. The role reversal helps reconcile competing behavioral and intermediary-based explanations for the momentum premium.

3. Financial Uncertainty and the Cross-Section of Cryptocurrency Returns
   Colak, G., Della Vedova, J., Foley, S., and Mai, S. T. · Journal of Banking and Finance (2026), 188, 107717 ABS 3 · ABDC A*
   Abstract

   We identify financial uncertainty — the unforecastable component of multiple financial indicators — as the primary driver of cryptocurrency return premia. Coins with greater exposure to financial uncertainty earn higher subsequent returns; this premium is distinct from macro, real, and policy uncertainty, and is not captured by VIX alone. Portfolio analysis yields a financial uncertainty premium of approximately 21% per year.

4. The 52-Week High Liquidity Mirage: How Behavioral Convergence Impairs Price Discovery
   Della Vedova, J., Gao, M., Grant, A. R., Westerholm, P. J., and Bliss, B. A. · Journal of Banking and Finance (forthcoming) ABS 3 · ABDC A*
   Abstract

   Do trading patterns at the 52-week high reflect investor disagreement or behavioral convergence? Using StockTwits sentiment, TAQ microstructure data, and machine-learning informed trading measures for S&P 500 stocks, we document agreement rather than disagreement. Belief dispersion falls while both retail and institutional investors reduce trading and shift toward selling. Anchoring to the 52-week high generates a liquidity mirage: effective spreads narrow and depth increases even as informed-trading intensity declines. Apparent liquidity masks weaker information incorporation, and stocks underperform momentum benchmarks by 20 to 30 basis points over five days.

5. The 1% Problem: LLM Orchestration for Rare Event Detection in Manufacturing Supply Chains
   Dang, J., Della Vedova, J., and Yang, Y. · International Journal of Production Economics (2026), 300, 110119 ABS 3 · ABDC A
   Abstract

   Manufacturing quality failures occur at around 1% prevalence yet carry outsized operational consequences. Extreme class imbalance defeats traditional statistical methods, and machine learning deployment requires expertise many organizations lack. We position large language models as orchestration agents that guide feature engineering, clustering, ensemble modeling, and SHAP-to-action translation while keeping every computation deterministic, as regulated environments require. On 622,924 aerospace manufacturing transactions, F1 scores improve by 5 to 18 percentage points across quality escapes, component defects, and delivery delays, and the gains hold across three locally deployable open-weight LLMs.

6. Equity Borrowing Constraints and the Informed Trading Strategies of Short Sellers
   Blau, B. M., Della Vedova, J., Fox, C., and Smith, J. M. · Market Microstructure and Liquidity (forthcoming) ABDC B
   Abstract

   We examine how equity borrowing constraints shape informed short-selling behavior. When borrowing constraints are tighter, informed short sellers concentrate their activity in unusually large trades: the most return-predictive short sales are predominantly large short sales in stocks with higher borrowing constraints. The result suggests that constraints lead informed investors to trade more aggressively rather than withdraw.

7. Fostering Reverse Innovation with Value Chain Co-Creation
   Kortmann, S., Bliss, B. A., Zimmermann, C., and Della Vedova, J. · IEEE Transactions on Engineering Management (2025), 72, 768-785 ABS 3 · ABDC A
   Abstract

   Reverse innovation originates when products are first developed in emerging economies before diffusing to advanced economies. Using structural equation modeling and qualitative comparative analysis (fsQCA), we examine the antecedents and configurations that enable reverse innovation, finding that value-chain co-creation between parent firm and subsidiary is a key condition for successful reversal.

Other working papers

1. The Wrong Reference Price: Purchase Path and the Disposition Effect
   Della Vedova, J. (2026) · Working paper · SSRN
   Abstract

   The disposition effect requires a reference price to classify a position as a gain or a loss, and for multi-purchase positions the literature defaults to the volume-weighted average price (VWAP). Using population-level data on household investors, we show the relevant reference depends on how the position was built: investors who buy below their running cost anchor on the last purchase price, while investors who buy above anchor on the VWAP, consistent with Thaler's hedonic editing. The correction reverses two canonical findings: VWAP masks a 249-basis-point welfare cost, and experienced investors retain rather than shed their bias.

2. The Value of Openness
   Della Vedova, J., Siegel, S., and Warachka, M. · Working paper · SSRN
   Abstract

   Cities differ in their openness to adopting new innovative products, and that variation shapes local firm value creation. We construct a proxy for city-level openness from the likelihood that new music is first played by local radio stations. Openness exhibits persistent cross-sectional variation across U.S. cities traceable more than a century, and during 2000 to 2019 it explains variation in the success of new ventures and new product introductions.