What it is
Entries on a higher-timeframe signal (a pullback to the 20 EMA in an uptrend, a close above a weekly high), stops measured in ATR, holds from two days to a few weeks. Often run as a basket across several coins to smooth the equity curve.
swing trading
Swing trading: hold for days, pay funding, survive the gap.
Swing strategies ride multi-day moves on 4-hour and daily charts: pullbacks in a trend, range breaks, momentum rotations. Fewer trades than scalping, so fees matter less and funding, overnight moves and position sizing matter more.
Official Binance broker · Non-custodial · Live only when you confirm · Updated
What the backtest must get right
Funding on perpetuals, paid every 8 hours for the whole hold, which can erase a month of gains in a crowded trade; stop fills on the 1-minute path through a violent candle, not at the candle's close; and a long enough history to include at least one bear market.
A prompt that builds one
Name the trend filter, the entry, the stop and the symbols.
“Swing long BTC, ETH and SOL on 4h: trend up when close is above the 200 EMA, buy a pullback to the 20 EMA with a 2.5 ATR stop, trail it, equal capital split, 2021 to now.”Read the basket honestly
A basket run splits capital equally and the honesty grade is the worst symbol's, not the best. The overfit check counts every variant you tried across the basket, so adding a coin to fix the curve is counted too.
faq
Swing trading strategy FAQ
Does the backtest include funding for a 10-day hold?
Yes. OptAlgo charges the exact funding rate at every funding timestamp of the hold, from the exchange's funding history.
Can I swing trade spot and perps in one strategy?
Hosted strategies can run mixed spot and perpetual legs on the same parent bot when the account mode allows it; the backtest reports each leg's costs.
Do I need to code this?
No. Describe the rules to Claude or ChatGPT through the OptAlgo connector and the assistant writes the strategy against the OptAlgo SDK, validates it for free and backtests it. You can read and download the code.
How do I know the backtest is not overfit?
Every run returns an overfit verdict based on how many variants you tried. Then ask for a walk-forward or a holdout: the strategy must keep its edge on data it never saw before it is worth paper trading.