Base rate

Combined signals: base rate from 382 out-of-sample tests

By Justin JaegerPublished Updated

Combined signals were tested 382 times across 28 idea variants. 30% beat their benchmark in 2015–2026 after costs, with a median Sharpe difference of −0.04. Verdict as of 2026-11-01: no consistent edge.

Key numbers

MeasureValue
Tests382
Idea variants28
Beat benchmark30.4%
Median Sharpe vs benchmark−0.04
Looked good before 201559
… and held after39.0%
VerdictNo consistent edge

By asset class

Asset classTestsBeat benchmarkMedian Sharpe vs bmVerdict
US equities23725%−0.06No consistent edge
international equities10236%−0.03No consistent edge
multi-asset portfolios2715%−0.14Dead end

Caveats

The 2015–2026 test window has been looked at many times during research, so we label it and do not call it a clean holdout. "Beat benchmark" compares each test with a fixed asset mix or buy-and-hold in the same window; a weak benchmark flatters the result. Many tests use US-listed ETFs.

Terms used: base rate, out-of-sample test, placebo test, dead end.

Before your next backtest

Let your AI check how ideas like yours did before it runs a single test. In our pilot, agents that checked first spent 1% of their backtests on known dead ends instead of 15%, and every backtest not wasted is time and tokens you keep. Cite this entry as ER-MIX-ALL, as of 2026-11-01.

Related evidence

Questions

Do combined signals work?

Combined signals were tested 382 times across 28 idea variants. 30% beat their benchmark in 2015–2026 after costs, with a median Sharpe difference of −0.04. Verdict as of 2026-11-01: no consistent edge.

Is this investment advice?

No. These are research statistics, not recommendations.