Grid

• ✓Buys low / sells high while price oscillates around a mean
• ⚠Averaging against a strong one-way trend can draw down hard

Backtest curve is an example on a deterministic simulated price path — not real returns.

"""网格（Grid trading）—— 区间等距加减仓。

思路：以起始价为参考，价格每下跌一个网格步长就加一格仓位、每回升一格就减一格，库存随下跌线性累积
（0..levels 格），在震荡区间内低买高卖。本 demo 为长仓网格；参数 grid_pct（格距%）、levels（格数）、
unit（每格数量）。适合区间震荡；单边下跌会满仓套牢，故设 levels 上限作为风控。
"""
from __future__ import annotations

from ..base import StrategyBase
from ..context import StrategyContext
from ..registry import register


@register("网格")
class Grid(StrategyBase):
    description = "区间等距网格：每跌一格加仓、每涨一格减仓（长仓）。"
    params = {"grid_pct": 0.01, "levels": 6, "unit": 1.0}

    def __init__(self, grid_pct: float = 0.01, levels: int = 6, unit: float = 1.0) -> None:
        self.grid_pct = float(grid_pct)
        self.levels = int(levels)
        self.unit = float(unit)
        self.ref: float | None = None

    async def on_start(self, ctx: StrategyContext) -> None:
        self.ref = await ctx.price(ctx.conn_symbol())

    async def on_tick(self, ctx: StrategyContext) -> None:
        sym = ctx.conn_symbol()
        px = await ctx.price(sym)
        if self.ref is None or self.ref == 0:
            self.ref = px
            return
        step = self.ref * self.grid_pct
        if step <= 0:
            return
        # how many grid steps below the reference (clamped to [0, levels])
        n = int(max(0, min(self.levels, (self.ref - px) // step + (1 if px < self.ref else 0))))
        await ctx.target(sym, n * self.unit, band=self.unit * 0.5)

The full SDK and all strategies are MIT-licensed open source — backtest, paper-trade, or fork them directly.