Skip to content

EMA crossover strategy

Build a complete EMA-crossover momentum strategy, backtest it on Bitcoin, and look at the report. Five minutes end-to-end.

What you'll build

A strategy that goes long when the 9-period EMA crosses above the 21-period EMA, with a fixed bracket (50-point stop, 200-point target).

1. Write the strategy

Create strategies/ema-cross.qkt:

strategies/ema-cross.qkt
STRATEGY ema_cross VERSION 1

SYMBOLS
    btc = BACKTEST:BTCUSDT EVERY 1m

RULES
    WHEN ema(btc.close, 9) CROSSES ABOVE ema(btc.close, 21)
    THEN BUY btc SIZING 0.1
         BRACKET {
           STOP_LOSS BY 50 PCT,
           TAKE_PROFIT BY 200 PCT
         }
         LOG "long entry on cross above"

    WHEN ema(btc.close, 9) CROSSES BELOW ema(btc.close, 21)
    THEN CLOSE btc
         LOG "exit on cross below"

What's going on:

  • SYMBOLS declares one stream — the alias btc reads BTCUSDT from the BACKTEST broker on 1-minute candles
  • The first RULES block fires on the up-cross; BRACKET attaches a percent-based SL/TP atomically
  • The second fires on the down-cross to flatten before the bracket completes
  • LOG lines tag entries/exits so you can audit later

2. Get some data

If you don't have historical BTCUSDT yet, populate the local data store via the bundled Dukascopy script:

./scripts/fetch-dukascopy.sh BTCUSDT 2024-01-01 2024-02-01

This writes daily-partitioned gzipped CSVs at ~/.qkt/data/symbols/BTCUSDT/. Subsequent runs read from cache. A qkt fetch CLI wrapper is on the roadmap.

Don't have or want Dukascopy? The repo ships a tiny sample under data/sample/symbols/BTCUSD/ — pass --data-root data/sample to use it for a smoke run.

3. Backtest it

qkt backtest strategies/ema-cross.qkt --from 2024-01-01 --to 2024-02-01

Output (truncated):

Trades:           14
Final realized:   234.50
Win rate:         0.571
Sharpe (daily):   1.34
Max drawdown:     -180.25

Report: ./reports/ema_cross-20240301-103245.html

Open the HTML report — equity curve, drawdown periods, Monte Carlo fan, per-trade risk.

4. Tune it

To try a different fast/slow pair, edit the LET lines at the top of the strategy file:

LET fast = 12
LET slow = 26

LET clauses are how qkt strategies expose tunable parameters — they're substituted at compile time wherever they're referenced.

(A --param key=value CLI override is on the roadmap — see Planned features. For now, edit the file and re-run.)

5. Run a sweep

To find the best (fast, slow) pair, see Run a parameter sweep.

What just happened

This strategy uses three things every qkt strategy uses:

  1. One stream declaration — btc = BACKTEST:BTCUSDT EVERY 1m. Multi-stream strategies just add more aliases.
  2. A WHEN ... THEN ... rule — the condition is evaluated on every candle close; the action fires on the first tick after the rule transitions from false to true (edge-triggered).
  3. A bracket — the entry, stop-loss, and take-profit go in as one atomic group. The broker handles all three; the engine never sees orphaned legs.

The same file you just ran in backtest will paper-trade with qkt run strategies/ema-cross.qkt and live-trade with qkt deploy strategies/ema-cross.qkt. The parity contract guarantees the trades match.

See also