A filter always sounds good. "Only trade in an uptrend." "Only trade in high volatility." "Only trade in altcoin season." Each of these rules promises to cut the bad trades and keep the good ones. The uncomfortable truth: the same filter that rescues one strategy can strangle the next. A regime gate is not a magic switch.
We didn't claim this — we measured it. 189 filter-and-strategy combinations on crypto, each filter in isolation. Here's what the data says.
The 200 WMA: broadly helpful, with an asterisk
The 200-week filter is simple: only trade when price is above its 200-week average — stay out of deep bear markets. Across many strategies it lifts the median return sharply: Supertrend from −5.1% to +33.4% (about 38 points), Ichimoku from −31.4% to +34.3% (about 66 points), EMA Cross about 31 points. Avoiding downtrends is a large part of the edge.
But it isn't universal. On some strategies the filter is roughly flat (Bollinger+RSI, Smoothed Heiken Ashi), and on Stochastic RSI it slightly hurts. And the asterisk: we also score every combination with a Deflated Sharpe test, which corrects for the fact that if you try enough filters, some shine by chance. Most of the "helps" verdicts don't clear that stricter bar. The median lift is real and useful — but "lifts the median" is not "statistically bulletproof."
ATR volatility: a matching problem, not a switch
The ATR filter lets you choose: trade only in calm, only in volatile, or only in expanding conditions. There's no universally right answer. For trend-followers, the low-ATR filter (calm entries before the move) helped: Golden Cross +14.8, Supertrend +13.9, EMA Cross +12.4, RSI/SMA +10.5 points. Apply that same filter to Bollinger+RSI — a mean-reversion strategy — and it hurt (−3.9). What strengthens a trend-follower starves a mean-reverter of exactly the setups it needs.
ATR isn't a "better" switch. It's a matching problem: trend strategies often want the calm-before or the expansion; mean-reversion wants the chop the filter removes.
Altcoin Season: the most polarizing filter
One filter stands out because it acts to both extremes. The Altcoin-Season gate — only trade alts during altcoin season — lifted Ichimoku by about 43 points, Supertrend by 26, Fear&Greed Cadence by 13, EMA Cross by 12 (over 450 runs). On BTC Signal, by contrast, it demolished performance: from +10.9% to −12.5%, a drop of about 23 points. On Williams Alligator, −14; on Smoothed Heiken Ashi, −8.
Why the split? BTC Signal already gates on Bitcoin's trend. Adding an altcoin-season gate on top double-filters it into silence — it sits in cash through the very moves it should catch. That's the lesson filters teach loudest: never apply one blindly.
The Bullmarket gauge — and the one that passed
The strongest filter in aggregate is the Bullmarket gauge on strict mode: only trade when the bull-market signal is fully confirmed. On Ichimoku it added about 64 points, EMA Cross 34, Supertrend 31, MACD 18. Impressive numbers — and exactly where overfitting hides.
Hence the Deflated Sharpe test. Of about 189 combinations, exactly one passed: EMA Cross + Bullmarket-strict. Everything else — including the +64-point Ichimoku result — is real in the median but doesn't clear the strictest bar. We don't hide that; we show you the honest confidence level, not just the pretty number. And even the strongest gauge cuts both ways: on BTC Signal, the strict filter hurt (−9 points).
What this means for you
A filter isn't an upgrade you bolt on. It's a bet that a particular regime is the right one for a particular strategy. Sometimes the bet pays off dramatically; sometimes it costs the whole edge. The only honest method is to check the delta per strategy — and put the overfitting test right next to it. The platform shows you both.
(One special case: the Min Profit Guard isn't a regime gate but a per-trade loss cap — a different tool for a different problem, giving profits back.)
Study the Past — Improve your Future. 🥋