New work shows PE’s outperformance largely boils down to leverage and appraisal smoothing—replicable with public small-cap + leverage. Meanwhile, funds with the lowest holdings overlap beat peers by ~1.2% a year.

One paper shows why low-cost index investing wins over time—the triumph of compounding and belief. Another reveals how VC returns can be replicated (and beaten) with staged leverage on the Nasdaq-100.

New research highlights ROIIC—not ROIC—as the forward signal for multiple expansion, and shows how sector-specific factors (like in REITs) sharpen attribution and cut false alpha.

New research shows that framing stock selection as a classification problem (winner vs. loser) outperforms return forecasting, while López de Prado argues AI fails in investing because it chases correlations instead of causal mechanisms.

Columbia, Cornell, and Sussex researchers challenge core assumptions—showing how to train models for trading, why “good” loan terms signal weakness, and how price-weighting beats cap-weighting.

Columbia research shows that when you apply Kelly to actual return series, optimal leverage shrinks 22–36% versus theory—especially in crisis periods and with less frequent rebalancing.

New research shows wealthy individuals now hold $1–2T in PE assets—and their returns rival institutions. The real driver? Advisor quality, not wealth.

New research shows AI-powered replication models track hedge fund performance with 89% accuracy—raising questions about fees, alpha, and the future of “2 and 20”.

From earnings call analysis to compliance stress tests, AI is reshaping hedge funds. But the real edge comes from building systems, not plugging tools.

Bubbles aren’t just noise. A new model shows how to price speculation, exploit the upside, and manage exits for better risk-adjusted returns.