You’ll learn: an EMA(9)/EMA(21) crossover entry, a
$500 stop and $1,000
target, and how to read the exit reasons. Time: about five minutes.The idea, in one sentence
When the fast moving average crosses up through the slow one, momentum has turned up — go long. Then risk$500 to make $1,000 on each trade: a 2-to-1
reward-to-risk shape.
Step 1 — Describe it
Open a new chat and type the idea exactly as you’d say it to another trader:Buy ES on an EMA(9)/EMA(21) cross with a 1,000 target.That’s the whole prompt. You don’t pick bar sizes, date ranges, or directions — AskFutures fills in sensible defaults and shows you what it chose.
Because you didn’t say otherwise, AskFutures applies its silent defaults: the
last 1 year of data, Day Trading on 1-minute bars, and an
end-of-day exit so no position is carried overnight. It also makes the
strategy both directions unless you ask for long-only. Want longs only?
Just add “longs only” to the prompt.
Step 2 — Read the strategy card
AskFutures builds the strategy and shows you a card. Skim it before you backtest — this is where you confirm it understood you.Market
ES (E-mini S&P 500). See choosing a symbol if
you’d rather trade the micro, MES.Entry rule
Go long when the 9-period EMA crosses above the 21-period EMA. Because the
strategy is both-directions by default, it also goes short on the opposite
cross.
Exit rules
A 1,000 profit target, and an end-of-day close —
whichever comes first. These are structured exits; see
risk & trade management.
Parameters
The tunable numbers are surfaced for you — typically
ema_fast = 9,
ema_slow = 21, stop_loss = 500, and target = 1000. These are exactly
what the optimizer can sweep later.Step 3 — Run the backtest
Ask for it in plain English:Backtest it.A fixed, deterministic engine replays a year of real ES prices one bar at a time, applying your exact rules. It reports P&L, win rate, drawdown, and a full trade list — each trade tagged with why it closed (
stop, target, or
session close). The same rules on the same data always produce the same
numbers; the AI never invents results.
Step 4 — Read the exit mix
The single most useful thing on the results panel for this strategy is how the trades ended. Sort or scan the trade list by exit reason:- Lots of
targetexits means the 2-to-1 shape is doing its job. - Lots of
stopexits means the entry is catching too many false crosses. - Lots of
session closeexits means trades are still open at the bell — the stop and target are spaced wide relative to a single day’s range.
Step 5 — Iterate
Keep chatting. Each edit becomes a new saved version you can compare side by side.Tighten or widen the risk
Tighten or widen the risk
Change the stop to 900.A tighter stop usually means more
stop exits but smaller losers — watch
whether win rate falls faster than the average loss shrinks.Make it long-only
Make it long-only
Only take long trades.Useful when you only want to trade with an up-cross and ignore the shorts.
Add a trend filter
Add a trend filter
Only go long when the 50-period EMA is rising.Thinning out counter-trend crosses is the classic next move. See signals & indicators.
Let the optimizer search
Let the optimizer search
Optimize the fast and slow EMA periods and the stop and target.The optimizer sweeps the parameters across a range and ranks the combinations — no manual trial and error.
Next steps
Trend pullback with adaptive risk
Step up to ATR-based stops, a daily-trend filter, and an RSI trigger.
Risk & trade management
Every exit type explained: stops, targets, trailing, max-time, EOD.
Optimize a strategy
Sweep the EMA periods, stop, and target instead of guessing.
Iterate & refine
How to chat your way from a rough idea to a tighter strategy.