Most pattern guides ignore timeframe — as if a hammer means the same thing on a 1-minute chart as on a daily. It doesn't. The 1-hour timeframe sits in the structural sweet spot: candle bodies are wide enough to outrun noise, but cycles are short enough to produce multiple setups per week. Here's which patterns thrive on the hourly chart and why we run our detector exclusively at this timeframe.
The lower the timeframe, the more candle shapes look like patterns that aren't really patterns. A 1-minute hammer can form because a single market-maker pulled bids briefly. A 5-minute bullish engulfing can be triggered by an algo that's now done. Anyone watching short timeframes ends up trading noise.
The higher the timeframe, the slower the cycle. Daily-chart reversal patterns are real, but they fire once a week per ticker if you're lucky, and the stops are wide enough that small accounts can't size meaningfully. Daily charts are for swing traders with capital.
1-hour bars resolve the tension. Each candle represents a session worth of order flow — wide enough to make wick lengths statistically meaningful and bodies wide enough to clear typical bid-ask spreads after fees. A 1-hour reversal pattern produces roughly 3–8 candidate setups per week per ticker — that's detection cadence, not execution cadence.
Setup frequency does not equal trade frequency. On margin accounts under $25k, FINRA's Pattern Day Trader rule caps round-trip trades at 3 in any rolling 5-business-day window — Thesis enforces a stricter 1-equity-entry-per-business-day cap on small accounts to keep you well clear of that line. Crypto and cash accounts are not subject to PDT and trade with the full setup cadence. See our small-account guide for what this means for your specific account type.
We backtested both. The signal-to-noise ratio on 5-minute and 15-minute pattern setups in our universe is too low to overcome round-trip transaction costs at small position sizes. Patterns fire constantly; very few follow through; the ones that do work produce gains that don't clear fees by a comfortable margin.
On crypto specifically, the round-trip spread on most pairs consumes the entire small-gain target on intraday timeframes. Even when the pattern works directionally, the realized P&L is negative after fees. We confirmed this with hundreds of paper trades before disabling sub-1hr dispatch.
The 5m and 15m buffers still get filled in the app — they're useful for chart context — but only 1hr drives pattern evaluation and trade decisions.
Trading exclusively on 1-hour bars means setups don't fire every few minutes. You won't be glued to a screen. The trade-off is fewer, more selective setups with cleaner geometry — entries near meaningful pivots, stops outside reasonable noise, targets that leave room to be wrong on timing without being wrong on direction.
That fits the autopilot use case better than any shorter timeframe could. The bot watches the hourly close, evaluates the pattern, decides, and either trades or doesn't. Once a position is open, the in-memory ratchet manages stop and target on every live tick. See how the AI layer works for the decisioning side, or how it's tuned for small accounts for the risk side.
Free practice mode shows hourly pattern detections in real time, before connecting a broker.
Create free account →