With everything going on in the macro right now I went back and checked how my signal engine actually performs during market stress.

Some context. I built a quantitative engine that detects institutional accumulation patterns through volume and price structure. Backtested across 234 US stocks over roughly 20 years. 2,375 independent signals after deduplication. Overall win rate 60.3% at a 20 day hold.

What I found was that signals fired when the broad market was below its 50 day moving average actually performed better not worse. About two thirds of those bear market signals were winners compared to the 60% baseline. The sample is smaller so I dont want to overstate the precision. But directionally it held across both halves of the dataset when I split it for walk-forward testing.

The logic makes sense to me. Fear shakes out weak holders. If a stock is still showing accumulation structure during a broad selloff the remaining buyers are more committed and the sellers are running out. Thats a better setup than the same pattern in a calm market where everyone is already long. Not saying go buy everything during a crash. But the idea that you should stop looking for setups during macro fear doesnt match what I see in 20 years of data. The engine doesnt stop working in bear markets. If anything it works a little better.

Anyone else tested their swing setups across different market regimes? Most backtests I see dont separate bull vs bear which I think is a blind spot.