would have done during those signal months is particularly useful:

Trigger Months Avg SP5 return during avoided month
+9% day only 64 +0.48%
−10%/60d only 6 −1.19%
Both conditions 4 −6.10%

The 60-day decline signal is rare, but when it coincides with the +9% euphoria signal, it is extremely useful. Those four months compounded to roughly a −24% loss if SP5 had been held.

The +9% signal by itself is much more nuanced: those 64 months actually averaged slightly positive. It works because its return distribution contains enough nasty periods that avoiding them improves the long-run path—not because every +9% warning predicts a losing month.

That also explains why the rule can reduce MDD dramatically without being a perfect “negative month” predictor.

The crisis-removal test was encouraging

I removed each major crisis entirely and recalculated the remaining history.

If the Canary were nothing more than a lucky 2000–02 rule, its advantage should disappear when we remove that period. It doesn't.

Without 2000–02, untimed SP5 was about 34.3% CAGR, Canary about 33.2%; Canary still had a much lower MDD, roughly −24% versus −45%.

Without 2008–09, Canary still had roughly 31.5% CAGR / −29.7% MDD, versus untimed at about 31.8% / −53.7%.

Removing COVID or 2022 likewise did not destroy the Canary result.

So no single crisis accounts for the whole drawdown advantage. That is important.

The economic-regime result is also interesting

I used the FRED real-retail-sales and industrial-production series you supplied earlier, with the previous month's year-over-year readings so we aren't using the future month's economic number.

When both real retail sales and industrial production were negative, untimed SP5's conditional annualized return was only about 8.0%. Canary was about 21.2%.

When both were positive, the two were much closer:

Untimed ~34.3%; Canary ~36.2%.

And one regime actually hurt the Canary: real retail positive / industrial production negative produced about 29.3% for untimed SP5 versus 18.0% for Canary.

That again argues against a magical all-weather rule. It seems most valuable in systemic contraction/stress, while it can unnecessarily sideline good momentum during some narrower industrial slowdowns.

My conclusion The rule does not simply improve SP5 everywhere.
Its major benefit occurs when market momentum is weak, volatility is high, or economic activity is broadly contracting—exactly when momentum strategies are most vulnerable to violent leadership reversals.

And it survives removing any single famous crisis.

The one caution I would emphasize is the 2020–2026 result. Canary gave up a lot of return there: about 26% versus 42% CAGR for untimed SP5. That does not invalidate it—the whole purpose is to create a much lower-drawdown S&P strategy—but it shows that the rule can be costly during extremely powerful modern momentum runs.