Browse working papers and research releases with direct links to supporting evidence, code, research tools, and companion quantitative methods where available.
Practitioner-oriented research on systematic investing, asset allocation, robustness, and portfolio decision design.
How a New Forecasting Model Changes Cross-Asset Allocation Decisions
A forecasting model can add portfolio value without winning the average prediction-error contest. This study evaluates Chronos-2 inside a matched cross-asset Forecast-to-Rank process and asks which portfolio decisions actually change, how integration changes the result, and whether those changes earn their keep.
Benchmark-Relative Evidence in Multi-Asset Allocation Using Historical-Vintage FRED-MD Data
Adding historical-vintage macro information to a matched price-based forecasting and ranking process does not produce a robust incremental portfolio benefit in this multi-asset setting. The paper evaluates that question relative to investable benchmarks and under explicit information-timing checks.
Long-Horizon Evidence on Downside Protection, Benchmark-Relative Drawdowns, and Persistent Equity Exposure
A defensive strategy can reduce its own drawdown and still trail equity for years. The analysis separates realized downside protection from the benchmark-relative path over long rolling holding periods.
A Practical Architecture for Dynamic Asset Allocation When Strategy Choice Is Uncertain
Historical performance and risk still guide strategy choice. ADAA adds a complementary check: whether the selected rules make genuinely different decisions and whether the portfolio depends too heavily on one historically best-performing rule or one precise set of weights.
Academic and technical research across macroeconomic forecasting, financial data science, AI/ML, and empirical methodology.
Statistical Distinguishability and Temporal Concentration in Data-Rich Macroeconomic Forecasting
Lower average forecast errors can coexist with limited statistical separation and performance concentrated in a relatively small number of months. The analysis evaluates average accuracy, formal inference, and temporal stability under a common FRED-MD forecasting design.
Practical explanations of statistical, econometric, portfolio, and machine-learning methods used across SlackQuant Research. Each article focuses on how a method works, how to implement it, and where it can fail, with hands-on examples where useful.
Explore Quantitative Methods →