How to Use a Signal Screener Without Fooling Yourself
A screener shows you what's firing right now — not what has a proven edge. Here's the honest workflow.
The Screener's Built-In Trap
Every signal screener does the same thing: it scans across assets and surfaces the ones where an indicator just triggered. That sounds useful, but it has a structural flaw. Firing now is not the same as working consistently. When you see a list of RSI crosses or MACD hooks in real time, you're looking at coincidences that happened to align today — not evidence that any of them will repeat with positive expectancy.
The temptation is to pick the setup that looks cleanest and trade it. That's cherry-picking, and it's how screeners most often hurt people. You didn't filter for edge. You filtered for recency and visual appeal.
Start With Edge, Not Alerts
The honest workflow reverses the order. Before you look at what's firing, you need to know which indicator-asset combinations have a demonstrated out-of-sample edge. Across 660,005 backtests covering 903 assets and 382 indicators — tested on 1-Hour, 4-Hour, Daily, and Weekly timeframes — only 26% of all combinations beat a simple buy-and-hold baseline. That means roughly three in four screener alerts you see belong to combinations with no verified edge.
Run the screener after you've built your shortlist, not before. Your shortlist is the small set of indicators that have beaten buy-and-hold for a specific asset on a specific timeframe, verified out-of-sample. The screener then tells you whether one of those pre-approved setups is active right now. That's the only role it should play.
Win Rate Is a Trap
The most dangerous number a screener can surface is a high win rate. Some indicators show median in-sample win rates above 70%, which looks compelling on a dashboard. But win rate is calculated on historical bars the indicator was fitted to — it's an in-sample figure. When those same indicators run out-of-sample, most of them collapse.
From the backtest data: Murrey Math Lines carries a median win rate of 74.3% but beat buy-and-hold in only 11% of assets tested. Holy Grail Confluence sits at 73.3% win rate, 8% beat rate. RSI Mean-Reversion: 71.7% win rate, 10% beat rate. Money Flow Index: 72.2% win rate, 9% beat rate. If your screener is sorting by win rate, it's likely sorting you toward the indicators most prone to underperforming out-of-sample. Ignore that column.
Filter by Asset Class, Then Timeframe
Once you accept that edge is asset-specific and timeframe-specific, the screener becomes a narrower tool — which is exactly what you want. The top indicators differ substantially across asset classes. For stocks, Fibonacci Pivots, Projection Bands, Intraday Momentum Index, and Camarilla Pivots appear most often among the strongest combinations. For forex, Fisher Transform dominates. For crypto, MA Envelope and Fibonacci Pivots lead. For commodities, Keltner Mean-Reversion ranks first.
That divergence means a screener running the same indicator across all asset types is almost certainly generating noise in most of them. Narrow it to an asset class you actually trade, cross-reference with the verified top performers for that class, and only pay attention to alerts that match that filtered set.
What These Results Are — and Aren't
Everything cited in this article comes from hypothetical backtests with realistic transaction costs applied. A backtest tells you what would have happened if a rule had been followed mechanically in the past on historical data. It does not guarantee future results, and nothing here is financial advice. Markets change, and an indicator that showed edge historically may not maintain it going forward.
The value of the backtest data is not that it tells you what to trade. It's that it narrows the screener alert list from hundreds of random triggers down to the small subset worth examining at all. Think of it as a noise filter, not a signal generator.
Questions, answered
Why do screeners show so many signals if most don't have edge?
Screeners don't filter for edge — they filter for trigger conditions. Any indicator can produce a signal on any bar. The alert count tells you nothing about positive expectancy. Across 382 indicators and 903 assets tested, only 26% of all indicator-asset combinations beat buy-and-hold out-of-sample. A raw screener list is mostly noise by construction.
Should I use short signals from a screener?
Short signals are harder to find edge in. Across the full backtest dataset, only 17.4% of combinations showed edge on the short side. If your screener is generating short alerts, the prior probability that any given one has real edge is lower than it is for long setups — apply stricter filtering, not looser.
How do I find which combinations are worth watching?
Start at <a href="/assets">the asset pages</a> — each one lists the indicators that actually beat buy-and-hold for that specific asset, with the timeframe noted. Build your screener filters from that verified list, not from what looks active tonight.
Are these backtests a promise of future returns?
No. These are hypothetical results from historical data with transaction costs applied. Past backtest performance does not guarantee future results. Use the data to narrow your focus and reduce noise — not as a trading signal or financial advice.
Every figure here comes from our own out-of-sample backtests, costs included — not a course or a guess. Educational information only — not investment advice. Hypothetical backtested results; past performance does not guarantee future results. Trading involves risk of loss.
Keep reading
Get the weekly edge report
The best-performing indicator per asset, what changed this week, and the honest caveats — straight to your inbox.