Run a strategy with qkt

Install the engine, re-run the backtest on your own machine, then paper-trade the strategy before it touches a broker.

Video: Run a strategy with qkt

A recorded walkthrough of this guide is coming soon.

1. Install qkt

qkt needs Java 21 or newer. On Linux or macOS (Windows users: follow the Windows install guide):

git clone https://github.com/elitekaycy/qkt
cd qkt
./gradlew installDist
export PATH="$PWD/build/install/qkt/bin:$PATH"
qkt --version

2. Download your strategy

In My library, choose Download strategy file. You get a single .qkt file. Its first line records the licence to your account; the rest is the strategy exactly as it was tested. Its symbols start with BACKTEST:, which means “no broker, simulate”.

3. Re-run the backtest yourself

qkt downloads the tick data it needs on first use. Run the same window as the listing and open the report:

qkt backtest gold-ema-pullback-v1.qkt \
    --from 2021-01-04 \
    --to 2025-12-31 \
    --instruments instruments.yaml \
    --config labs-backtest-100000.yaml \
    --broker mt5-sim \
    --execution mt5-realistic \
    --starting-balance 100000 \
    --report-dir ./out
# then open out/report.html in a browser

These are the settings the research pipeline uses; instruments.yaml (the pinned cost profile) and labs-backtest-100000.yaml (the pinned qkt settings) are on the strategy’s Code tab. Figures will be close to the listing but not identical, because the listing was tested on broker ticks while a fresh install downloads a public tick feed.

4. Paper-trade it

The daemon runs strategies continuously against live prices, with simulated fills:

qkt daemon &
qkt deploy gold-ema-pullback-v1.qkt --as gold
qkt status gold
qkt logs gold --follow
qkt stop gold

Leave it running for a few weeks and compare what it does with the research. Then go on to trading it on MetaTrader 5. Full details are in the qkt quickstart and the paper trading guide.