10 crypto strategies, paper-traded live
Each starts with $1,000 in paper money on BTC, ETH and SOL. The rules come from CoinQuant's public strategy library. Results are net of realistic exchange costs, and the benchmark is plain buy-and-hold.
Leaderboard
Return on the $1,000 start, closed trades plus open positions marked at the latest price (selling costs included).
small sample fewer than 10 closed trades, so the numbers say little yet. vs buy-and-hold is the return minus buy-and-hold over the same period, in percentage points. Backtest is the same rules from 2022 to 2026-09-27, averaged over the three pairs, at Kraken costs.
What the bots are watching
How far each strategy is from its next buy (or, while holding, its next sell), as of the last closed bar. Nearest to a trigger first. Tap a cell for the full rule.
Equity curves
Return on $1,000 since each strategy started, as simulated with Kraken costs. Showing the top 3 and buy-and-hold; tap a name to show or hide it.
Recent trades 0
Newest first, across all strategies. P&L is after the simulated fees and spread.
- Loading…
▸ What the backtest said (2022 → 2026-09-27) 0 / 10 passed
At realistic exchange fees none of the 10 passed. 5 of 10 made money at CoinQuant's flat 0.1% fee, but only 1 did at Kraken's costs, so costs, not signals, are the main problem. We're running them live anyway to see.
| Strategy | At Kraken costs | At CoinQuant 0.1% | Trades | vs buy-and-hold | Profit factor | Gate |
|---|---|---|---|---|---|---|
| Buy and Hold (benchmark) | +7% | +8% | 3 | — | 1.35 | BENCHMARK |
| Ichimoku Cloud 4H | +3% | +155% | 185 | -4 pp | 1.01 | FAIL |
| EMA 21/55 4H | -5% | +226% | 266 | -12 pp | 0.98 | FAIL |
| MTF RSI 4H + 1D SMA200 filter | -35% | -7% | 67 | -42 pp | 0.60 | FAIL |
| RSI(14) Mean Reversion 4H | -71% | -41% | 133 | -77 pp | 0.50 | FAIL |
| MACD + RSI>50 + SMA200 4H | -76% | +104% | 449 | -83 pp | 0.65 | FAIL |
| Donchian(20) Breakout 4H | -91% | +31% | 548 | -97 pp | 0.60 | FAIL |
| MACD(12,26,9) + RSI>50 4H | -93% | +92% | 685 | -100 pp | 0.59 | FAIL |
| 2.5% below SMA20 4H | -97% | -65% | 605 | -104 pp | 0.30 | FAIL |
| Supertrend(10,3) 1H | -100% | -47% | 1419 | -106 pp | 0.39 | FAIL |
| 1.25% below SMA20 1H | -100% | -80% | 2407 | -106 pp | 0.14 | FAIL |
Returns are the average over BTC, ETH and SOL, each from $1,000 with 100% of equity per trade, long only, filled at the bar close. Kraken costs are 0.82% per side (fee plus half-spread) vs CoinQuant's flat 0.10%. Gate: pooled profit factor of at least 1.1 at Kraken costs and 20+ trades on every pair. Past and simulated results don't predict future ones.
The strategies
Rules as published by CoinQuant. "CoinQuant reported" is their own backtest with their own dates and fees, so it isn't comparable to our numbers. Each strategy page has the full detail.
How it works
- A bot checks every strategy once a minute and acts only when a 1-hour or 4-hour bar closes. It reads the last closed bar and buys or sells at market. It trades long only, with no leverage.
- Each strategy has its own $1,000 paper account split across BTC-USD, ETH-USD and SOL-USD.
- Fills are simulated against Coinbase's public price feed. We charge Kraken's entry-tier taker fee plus an estimated half-spread on every side. The toggle above re-prices the same trades under other exchanges' costs.
- Buy-and-hold buys all three coins on day one and never sells. A strategy has to beat that to be worth the effort.
- If a strategy's rules change, its scoreboard restarts. The old record is kept, not merged.
Also see our stock playbook paper portfolios.
Full risk disclaimer