Day Trading Setups
9/20 EMA Cross Scalp
A fast EMA crossing a slower one is one of the most common scalp triggers there is. It also generates a lot of trades, and a lot of trades is exactly where a small, boring cost per round trip goes from ignorable to the whole story.
Same signal, two ledgers
Both EMAs below are computed for real, bar by bar, and every crossover is booked mechanically the instant it happens, no discretion. Drag the friction slider and every trade's cost comes straight out of a per-share cents figure, the kind of number that's easy to wave off in isolation and easy to forget adds up across dozens of scalps. Gross tracks the signal in a world with no costs, Net tracks the same signal paying them.
How it works
- Both lines are real exponential moving averages. Each one weights recent closes more heavily than a simple average does, computed with the standard smoothing formula, seeded from a simple average of the first period's bars.
- A trade books the instant the fast EMA crosses the slow one. No confirmation bar, no filter, the system flips long the moment fast crosses above slow and flips short the moment it crosses back below, entering and exiting at that bar's close.
- Friction is charged once per round trip, in cents per share. That figure gets converted into a percentage of the entry price for each individual trade, since the same 3 cents matters a lot more on a $10 stock than a $150 one.
- Gross and Net run the identical trade log. Every entry, exit, and direction is exactly the same for both lines, the only difference is whether the cost gets subtracted, which is what isolates exactly how much of the edge the friction is eating.
Where this breaks
A fast/slow cross is a whipsaw generator in anything but a clean trend
Tighten the EMA periods or widen the friction slider and watch how quickly a system that looked fine gross can go flat or negative net, that gap is not a bug in the simulation, it's the actual mechanism. A 9/20 cross fires constantly in a choppy, directionless stretch, each cross individually tiny and each one still paying the full round-trip cost. The gross line can survive a lot of that noise because small losses and small gains roughly cancel before costs, the net line usually can't, because every single one of those small trades bleeds a fixed amount regardless of whether it wins or loses.