A volatility-burst backtest can hide the drawdown between month-ends.
Our daily-data reconstruction exposes drawdown and execution-timing risks.
Curated by a 30-year hedge fund veteran
Read the paper.
Question the result.
Thirty years of hedge fund experience applied to academic finance. Human-led curation, practical investment judgment, and a close look at what survives contact with data.
Explore the research ↓Start here01 / MOMENTUM
Partial replication
Six days explain the momentum story. Can we trade it?
A compelling calendar effect meets a less compelling implementation.
Read the research brief ↗The research archive
Ideas worth examining.
House adaptation: execution timing unresolved
Predictable rebalancing does not guarantee a profitable front-run.
A plausible institutional flow meets weak futures returns and a stronger, imperfect bond proxy.
A backtest spends information every time we change the question.
The Axiomatic Trader connects model complexity, research search and position sizing through explicit assumptions.
A sparse neural model improves ranking. Size still changes the answer.
Feature structure reduces one model’s parameter burden, but the stronger portfolio claim depends on weighting and costs.
An earnings surprise can make investors miss the market news.
A study of earnings announcements finds delayed macro information in subsequent stock returns. Trading it requires precise timing.
A better volatility-smile model is not automatically a better hedge.
Higher-order surface dynamics help explain curvature. The improvement is much smaller for the whole smile.
The market’s correlation map can change without giving us a trade.
A reconfiguration index describes a volatility environment. Its premium measure is not a variance-swap payoff.
A steadier index tracker can still be a worse tracker.
Entropy regularization reduces some portfolio churn. The study’s tables show a trade-off, not universal superiority.
Leveraged ETF risk can arrive from the neighboring trade.
A feedback model finds a Korean transmission channel and U.S. nulls. Venue depth is part of the hypothesis.
Diversification starts with what stocks respond to.
MINGLE builds a graph from shared exposures. The gains are promising, but crisis protection remains limited.
Three matrices beat the index. The replication question is harder.
Rank dynamics offer a parsimonious portfolio design, but universe selection and uncertainty qualify the headline.
Less volatility drag does not make aggressive leverage safe.
A compact neural risk model improves the authors’ backtest. Its drawdowns and universe rules deserve equal billing.
Can the VIX curve tell us when protection is worth carrying?
An industry white paper uses curve disorder to switch between equity and volatility futures.
Missing prices can turn a risk model into a portfolio problem.
A neural covariance estimator reports lower risk, but the implementation deserves an equally careful audit.
Trend following can have a right tail without having an edge.
A useful distinction between the shape of returns and the source of expected profit.
A crash and a grinding bear market need different protection.
A framework for combining puts and trend without mistaking simulation weights for an allocation rule.
KellyBoost gets the objective right. That can still mean betting too much.
A direct route from features to portfolio weights exposes the cost of estimation error.
A better prediction target did not rescue our model.
We changed one training choice and kept the rest of the experiment fixed.
The strongest insider signal may be the one we cannot observe.
A Norwegian register study shows why access to the right data comes before a backtest.
When a news signal depends on which news you can see.
The disagreement measure is elegant. Substituting the event feed changed the research problem.
An insider trade is not automatically an information signal.
A classic classification rule met a much weaker result in our later US sample.
Six days explain the momentum story. Can we trade it?
A compelling calendar effect meets a less compelling implementation.
Go beyond the abstract
The detail is in the analysis.
Read the implementation questions, evidence, and full assessments on Substack.