The Stationarity Assumption Is Free: Regime-Conditional Factor Selection and the Robustness of the Unconditional SDF

Momentum and trend following over 1963–2024, studied by Glancszpigel in 2026. The abstract mentions a Sharpe ratio, profits, an out-of-sample test and multiple testing but not a benchmark, transaction costs or the size of the universe.

“The stationarity assumption—so often criticized on theoretical grounds—is effectively free at the Sharpe-ratio level: across our four specifications, the annualized return gap versus the static benchmark ranges from −15 to +18 basis points, statistically indistinguishable from zero under multiple-testing adjustment.”

Also stated: Long-short, Macro, out-of-sample test, multiple testing

Missing from abstract: benchmark, costs, universe