PM market making

• ✓When your view of the odds differs from the market's pricing
• ⚠Liquidity thins near resolution; binary outcomes can go to zero

Binary prediction-market simulation (YES/NO, settles YES after 180 steps) — not a promise of future returns.

"""事件做市（Prediction-market making）—— 订单簿价差 + 库存偏移。

思路：在二元市场对 YES token 做市：以滚动中间价为参考，价格低于参考（偏便宜）时增持库存、高于参考时
减仓，等价于围绕中间价的双边报价在成交后把库存拉回中性，赚取价差与短期均值回归。价差过窄（低于
min_spread）时不报价——做市无利可图。本 demo 为长仓做市（库存 0..max_inventory）。

适用性：instrument="event"、venues=["POLYMARKET"]。
"""
from __future__ import annotations

from ..base import StrategyBase
from ..context import StrategyContext
from ..indicators import clamp01, sma
from ..registry import register


@register("事件做市")
class PmMarketMaking(StrategyBase):
    description = "预测市场对 YES 做市：围绕滚动中间价按库存偏移持仓，价差过窄则不报价。"
    params = {"ref_window": 20, "max_inventory": 30.0, "spread": 0.06, "min_spread": 0.005}
    venues = ["POLYMARKET"]
    symbols: list = []
    instrument = "event"

    def __init__(self, ref_window: int = 20, max_inventory: float = 30.0,
                 spread: float = 0.06, min_spread: float = 0.005) -> None:
        self.ref_window = int(ref_window)
        self.max_inventory = float(max_inventory)
        self.spread = float(spread)
        self.min_spread = float(min_spread)

    async def on_tick(self, ctx: StrategyContext) -> None:
        sym = ctx.conn_symbol()
        bid, ask, _depth = await ctx.orderbook(sym)
        if ask - bid < self.min_spread:
            return  # 价差过窄，做市无利可图
        hist = await ctx.history(sym, self.ref_window + 1)
        ref = sma(hist, self.ref_window)
        if ref is None or ref == 0:
            return
        dev = (hist[-1] - ref) / self.spread  # >0 偏贵、<0 偏便宜（以 spread 归一）
        frac = clamp01(0.5 - dev * 0.5)  # 便宜 → 库存大
        await ctx.target(sym, frac * self.max_inventory, band=self.max_inventory * 0.08)

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