Mean reversion

• ✓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.

"""均值回归（Mean reversion）—— 布林带 / z-score。

思路：价格相对滚动均值显著偏低（z-score ≤ -k）时买入，回到均值（z-score ≥ 0）时离场。适合震荡、
无趋势的行情；趋势行情会连续亏损，需配合风控（max_position / 止损，本 demo 用仓位上限）。
参数：window（均值窗口）、k（进场偏离倍数）、size（目标数量）。
"""
from __future__ import annotations

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


@register("均值回归")
class MeanReversion(StrategyBase):
    description = "z-score 偏离均值买入、回归离场（震荡行情）。"
    params = {"window": 20, "k": 1.0, "size": 5.0}

    def __init__(self, window: int = 20, k: float = 1.0, size: float = 5.0) -> None:
        self.window = int(window)
        self.k = float(k)
        self.size = float(size)

    async def on_tick(self, ctx: StrategyContext) -> None:
        sym = ctx.conn_symbol()
        hist = await ctx.history(sym, self.window + 1)
        z = zscore(hist, self.window)
        if z is None:
            return
        if z <= -self.k:
            await ctx.target(sym, self.size, band=self.size * 0.1)  # cheap vs mean → accumulate
        elif z >= 0:
            await ctx.target(sym, 0.0, band=self.size * 0.1)        # reverted → exit

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