Do Trading Indicators Actually Work? We Ran 502,988 Backtests

The short version

We ran 502,988 backtests — every one of 358 indicators on 741 assets, out-of-sample, with realistic costs. Here's the uncomfortable truth: 57% of assets had some indicator that beat buy-and-hold — but any single indicator beat buy-and-hold on only about 8–14% of assets.

Both things are true

That gap is the whole story. With 358 indicators to choose from, you can almost always find one that fit an asset's past. That's not the same as indicators 'working' — it's selection. The real question isn't 'do indicators work,' it's 'does this indicator work on this asset, out-of-sample?'

When we hold each indicator to that standard, most lose to simply holding — especially on trending assets like the major indices, where nothing reliably beats buy-and-hold.

So why does anyone use them?

Because on the right asset, a real edge exists — and because indicators do something holding doesn't: they cut risk. Even when an indicator trails buy-and-hold on raw return, it often sidesteps the worst drawdowns. That's a legitimate reason to use one — just not the reason most courses sell you.

How to actually use this

Don't ask 'what's the best indicator.' Ask 'what's the best indicator for the thing I trade' — and check it held up out-of-sample. That's the entire point of this site: look up your asset and see what actually worked, honestly, including when the answer is 'just hold.'