Trend Bots: Growth and Steady ← hub
Two versions of the same daily trend-following bot, backtested on 7 years of data and now running as live paper accounts. Nothing here places real orders. Not financial advice.
Read this first. These are simulated results on today's surviving coins (survivorship flatters every crypto backtest), with fees of 0.14% per round trip included and no look-ahead. Real fills will differ. Expect about 4 in 10 (Growth) or 5 in 10 (Steady) trades to win, with long stretches of small losses between big winners.
The two bots
| Growth | Steady |
|---|
| Buy | A daily close above the previous 20-day high, only when Bitcoin is in an uptrend (BTC above its rising 100-day average). Buy at the next daily open (03:00 your time). |
| Size | Risk 0.5% of the account per trade (distance from entry to the previous 10-day low), never more than 20% of the account in one coin, at most 10 positions. |
| Exit | Everything is sold at the next open after a daily close below the previous 10-day low. | Sell half at +1R (limit order), move the stop on the rest to your entry (break-even), then exit the rest on the same 10-day-low rule. |
| Character | Higher return, bigger drawdowns, about 4 in 10 trades win | Higher win rate, smaller drawdowns, lower return, about 5 in 10 trades win |
Backtest: headline
Growth of $10,000 (log scale)
■ Growth ■ Steady ■ Bitcoin buy-and-hold
Return by year
Stress tests: does it survive being hurt?
Parameter sensitivity and walk-forward
Entry high (rows) and exit low (columns). Cell = return per year / worst drawdown. If the settings were curve-fitted, only one cell would look good.