The Test Bench › gate

Does Risk Ratios (SPY:TLT + 5 sector RS) actually work?

DEAD

No edge left once tested on a fair footing.

The claim

A chart-based signal that times the market — StockCharts / sector-rotation chartists

What we did

Tested against Buy & hold SPY, same window (2002+), 2002-2026, prior-day signal, OOS 2014+.

The result

Result: 0.05 dSharpe (vs buy-hold), OOS

SPY:TLT actively hurts (OOS -0.39); sector RS gates +0.13/+0.18 in-sample fade to ~0 OOS; composite +0.05

Verdict

The popular 'Risk On / Risk Off' ratio panel -- SPY:TLT plus five sector relative-strength ratios -- run as 50-day-EMA crossover gates on SPY. The famous SPY:TLT is the WORST: as a timer it underperforms buy-and-hold (dSharpe -0.16, -0.39 out-of-sample). The sector gates (XLY, XLI) flatter in the rear-view (+0.13, +0.18) and fade to roughly zero out-of-sample. The >=4-of-6 composite cuts drawdown (-55%->-22%) but only matches buy-and-hold on risk-adjusted terms (+0.05 OOS) while handing back half the return. An intuitive panel with no out-of-sample timing 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.

← Recession Monitor (3M/10Y + 2Y/10Y inversion) All findings SKEW<145 →

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.