Minute-bar trend trading strategy for slower entries

For those of us not watching ticks. Same structure, using the one-minute values so the logic reasons about bars rather than every print.

1. The rule

# BUY if Price > Sma(60) and ChangeAngle(60) > 20: if BuyVol_1m > SellVol_1m: Buy()# SELL if ProfitPct >= 3.0 or ProfitPct <= -1.5 or HoldTime >= 3600: Sell()The _1m suffix is the whole trick. If your reasoning happens on bars, your conditions should read bars. Mismatching the two is how you end up with a strategy that behaves differently from how you think about it.

2. Why a minute version at all

Per-second factors react to every tick, which is appropriate for a scalp and wrong for a trend idea that intends to hold for twenty minutes. The _1m suffixed factors aggregate the same underlying data over the completed minute, so the entry stops flickering. Note that the minute in progress is included as it builds. This is not a completed-bar join. That matters because using only completed bars would mean your live strategy and your backtest disagree about what was known at each moment.

3. When it fails

It enters later than the per-second version by construction. In a fast move that is a real cost, and no amount of tuning recovers it — you chose a slower clock deliberately.

4. What to change first

The averaging window, not the timeframe. Sma(60) on minute data is a very different thing from Sma(60) on seconds. Check that the window you wrote is the window you meant.

5. What the round trip asks of it

Slower entries are partly a cost decision, not only a signal one. The 3.0% target against a 1.5% stop needs 33% before costs and 50% after — the same seventeen point toll as any other 2:1 structure at this target size. But a minute-bar strategy takes far fewer trades than a one-second one, and the toll is charged per trade. That is the underrated argument for slowing down: not that minute bars see something ticks do not, but that you pay the round trip a tenth as often. See what day trading a US stock really costs.