The Test Bench › gate

Does Recession Models (Sahm / claims / leading) actually work?

DEAD

No edge left once tested on a fair footing.

The claim

A macro / intermarket reading that times the market — Macro / recession trackers

What we did

Tested against Buy & hold SPY, same window (2012+), 2012-2026, both directions, vs +0.33 null floor.

The result

Result: 0.07 dSharpe (vs buy-hold)

Sahm rule dSharpe +0.07 (IS -0.35 -> OOS +0.35 = an OOS mirage); claims/leading all noise

Verdict

Recession indicators (Sahm rule, jobless-claims trend, OECD leading) are slow by design and noise as equity-timing gates. The Sahm rule nominally cleared the floor but on a negative in-sample = a data-mining artifact the discipline flags, not an edge.

What the test did and did not do

tested out of sample
point-in-time data
checked for look-ahead
net of real costs

Almost nobody publishing a backtest says which of these they did. We list the ones we failed as well as the ones we passed.

Tested 2026-06-23. The same treatment, applied to every other system, is on the test bench.

← RVX (Russell Implied Vol) All findings Recession Monitor (3M/10Y + 2Y/10Y inversion) →

Written with AI assistance from our own test database. The numbers are computed, not composed.
General market commentary for education, not personalised investment advice. Backtested results are hypothetical. Trading carries a real risk of loss.