Best BTC grid bot settings: step size, levels and capital.
A grid bot has only a few settings, but each one changes how it behaves. Here's what each setting does, what we learned from backtesting them on real BTC prices, and a simple way to choose yours.
1. Step size: how far apart the orders are
The step is the gap between grid levels. With a 0.5% step, the bot buys 0.5% below the price and sells each buy 0.5% higher.
The rule that matters most: on Bybit and Binance spot, you pay about 0.1% on the buy and 0.1% on the sell, so 0.2% per round trip. Your step must be clearly bigger than that.
| Step | Profit per round trip, after fees | How often it trades |
|---|---|---|
| 0.35% (Active) | about 0.15% | very often |
| 0.5% (Balanced) | about 0.3% | often |
| 1.0% (Calm) | about 0.8% | less often |
Smaller steps trade more but keep less per trade. Bigger steps keep more but wait longer. Around 0.5% is the sweet spot for BTC's normal daily moves.
2. Levels: how many orders
Levels decide how many buy orders sit below the price. Step × levels = the price range the grid covers. 0.5% × 5 levels covers a 2.5% drop.
- More levels cover a deeper drop, so the bot keeps trading in dips. But your money is split into smaller orders, so each trade earns less.
- Fewer levels mean bigger orders and more profit per trade, but the bot runs out of buy orders sooner when BTC falls.
In our 6-month backtest with $5,000, 5 levels made +9.2% and 10 levels made +8.4%, even though 10 levels made almost twice as many trades. More trades is not the goal; net profit is.
3. Capital: how much to use
- Minimum: exchanges reject very small orders (about $5 each), so each level needs at least about $6. With 5 levels, that's about $30 minimum, but $300+ makes the profits easier to see.
- Profit scales with capital: the same settings with twice the money make roughly twice the dollar profit per trade.
- The real limit is you: without a stop loss, your whole budget can end up in BTC during a drop. Only use money you can leave in BTC for a long time.
4. Stop loss: on or off?
A stop loss sells everything if BTC falls a set amount (for example 8%) below the grid.
- With a stop loss: small, limited losses in a crash. But in a dip that recovers, it sells at the bottom and misses the recovery. In our 6-month test, it turned +9.2% into −1.1%.
- Without a stop loss: the bot holds and waits for BTC to come back, then sells at a profit. But you must accept temporary drops; ours reached −27% at the worst moment.
There's no perfect answer. It depends on whether you'd rather limit losses or wait for recovery.
5. Re-centering: following the price
If BTC rises far above your buy orders, they may never fill. A good bot moves (re-centers) its unfilled buy orders back near the current price. Daily Trade does this automatically using a model built on 8 years of BTC data, and only when it isn't holding BTC from earlier buys.
A simple way to choose
- Start with Balanced (0.5%, 5 levels).
- Run a free backtest with your amount on 30, 90 and 180 days. Look at the worst dip, not just the profit.
- Run it in paper mode with virtual money for a week or two.
- Go real with a small amount, and increase step by step only if the real results match.
See it with your own numbers
Run a free backtest on real BTC prices, or start the bot with $1,000 of virtual money. No risk, no card.
Free backtest → Start free