Grid Parameter Optimizer
Direct answer
The Grid Parameter Optimizer backtests eight common grid counts from 10 to 100 over the same history and lays out each one's historical net, return, max drawdown and fees; results are historical only.
The Grid Parameter Optimizer automatically backtests grid counts from 10 to 100 over the same history and lays them side by side: realized net, total return, max drawdown and fees for each.
Read the output as a historical parameter comparison: a grid count that performed best in the past is not a promise for the future. This is a common research step, not the conclusion itself.
Grid Parameter Optimizer
Result
All figures follow the assumptions on this page and are for reference only.
- Bars
- —
- Best grid count (historical, ref.)
- —
- Its realized net (historical)
- —
Scenarios / details
| Grids | Sells | Realized (USDT) | Total return | Max DD | Fees (USDT) | Final equity (USDT) |
|---|
Key data & facts
- Grid counts swept
- 10 / 20 / 30 / 40 / 50 / 60 / 80 / 100Eight counts, each running a full historical backtest
- Only variable
- grid countRange, capital and fee stay fixed
- Very dense grids
- capital sliced thinRaises the fee share
- Example: 50 vs 10 grids
- historical realized netHigher net at 50 grids means the sample favors medium density
What is the Grid Parameter Optimizer?
Grid count is one of the most tuned parameters, yet its effect on return, drawdown and fees depends on market structure.
This tool runs eight common counts over the same sample and lines the results up, exposing parameter sensitivity instead of predicting markets.
How to use it
- Pick the data source (prefer your real history), then enter range, capital and fee.
- Click Calculate and compare realized net, drawdown and fees row by row.
- Treat the “highest historical net” count as a reference for that sample only.
- For a candidate count, open the backtest page to inspect its equity curve.
Method
Grids ∈ {10, 20, 30, 40, 50, 60, 80, 100}
Each count runs a full historical backtest; metrics are compared.
Worked example
If 50 grids clearly beat 10 grids on the demo series, the sample's oscillation favors medium density; if 10 grids wins, price crosses the range rarely and fewer grids mean less fee drag.
Repeat the same exercise on your real data before trusting the conclusion on your sample.
Practical tips
- Read the whole table, not just the top row. A count with similar net but smaller drawdown and fees is often the more robust choice.
- Beware overfitting. A “best” found on one historical sample can fail on another — cross-validate across time windows.
FAQ
Why avoid the word “best”?
Because these are comparisons on one historical sample with no future guarantee. “Historical reference” is more honest and less misleading.
Is more grids better?
Tables usually show medium counts balancing return and drawdown: too few under-trade, too many bleed fees. It depends on your sample.
Can I copy the optimized setting straight to live trading?
No. Historical winners may reflect one market regime; confirm across samples and against live constraints (minimum order, fee tiers) first.
Sources & references
- What Is Backtesting?Investopedia
- What Is the Sharpe Ratio?Investopedia