1. From volatility harvesting to parameter design

In the previous chapter, the core idea of grid trading was stated very simply:

use price fluctuations to repeatedly buy low and sell high, and collect structural spread income.

This chapter asks the next practical question:

How do we quantify and design that structure?

2. Price range, number of grids, and spacing

A standard grid system can be abstracted as:

Parameter	Symbol	Meaning
Upper bound	$P_{\text{max}}$	highest price in the grid range
Lower bound	$P_{\text{min}}$	lowest price in the grid range
Number of grids	$N$	number of grid layers
Grid spacing	$\Delta P$	distance between adjacent levels

Arithmetic grid

For an arithmetic grid:

$$\Delta P = \frac{P_{\text{max}} - P_{\text{min}}}{N}$$

Geometric grid

For a geometric grid:

$$r = \left(\frac{P_{\text{max}}}{P_{\text{min}}}\right)^{1/N},\quad P_i=P_{\text{min}}r^i$$

Choosing the range

A useful rule of thumb is to let the range cover roughly 95% of the expected fluctuation zone.

• a range that is too narrow triggers often but is easy to break out of;
• a range that is too wide spreads capital too thin and dilutes return density.

3. Arithmetic vs geometric grids

An arithmetic grid is intuitive and easy to compute. A geometric grid is often better when volatility scales with price.

The choice is not purely mathematical. It reflects how the market actually moves.

4. Initial position building

Two common starting modes are:

• aggressive start: build a meaningful base position immediately;
• gradual start: let the system accumulate inventory step by step.

In practice, many systems use a hybrid form.

5. A simple return and drawdown view

If the grid spacing is $\Delta P$ and the trade size per execution is $q$, then the profit of one completed buy-sell cycle is approximately

$$\text{Profit}_1 = q \cdot \Delta P.$$

This makes the basic trade-off very clear:

tighter spacing improves trading frequency, but wider exposure increases path risk.

Final remarks

A basic grid system is not just a set of orders. It is a parameterized structure.

Its performance depends on whether the chosen range, spacing, layer count, and capital allocation are aligned with the volatility characteristics of the traded asset.

贡献者: Junyuan He