first commit

2026-08-07 16:08:34 +08:00
commit e0b5228008
51 changed files with 5175 additions and 0 deletions
--- a/backend/app/init.py
+++ b/backend/app/init.py
--- a/backend/app/api.py
+++ b/backend/app/api.py
@@ -0,0 +1,167 @@
+"""HTTP 路由（OpenAPI 契约的载体）。
+
+  GET  /api/health          健康检查
+  GET  /api/candles/{sym}   取 K 线（支持 1d/1w/1M/1y 周期，日线为基底聚合）
+  POST /api/backtest        跑回测，返回 K线+指标+买卖点+净值+绩效
+"""
+from __future__ import annotations
+
+import json
+
+import pandas as pd
+from fastapi import APIRouter, Depends, HTTPException
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from .backtest.engine import BacktestConfig, run_backtest
+from .backtest.strategies import build_strategy
+from .data import fetcher, repository
+from .data.aggregation import bars_per_year, resample_bars
+from .data.synthetic import seed_if_empty
+from .db import get_session
+from .domain import Bar
+from .models import BacktestRun
+from .schemas import (
+    BacktestRequest,
+    BacktestResponse,
+    CandleOut,
+    EquityPoint,
+    IndicatorOut,
+    MetricsOut,
+    SignalOut,
+    SyncRequest,
+    SyncResponse,
+)
+
+router = APIRouter(prefix="/api")
+
+
+def _series_to_jsonable(s: pd.Series) -> list[float | None]:
+    """NaN -> None（lightweight-charts 的 whitespace data，跳过指标预热期）。"""
+    out: list[float | None] = []
+    for v in s.tolist():
+        if v is None or (isinstance(v, float) and v != v):
+            out.append(None)
+        else:
+            out.append(float(v))
+    return out
+
+
+def _rows_to_bars(rows) -> list[Bar]:
+    return [Bar(ts=r.ts, open=r.open, high=r.high, low=r.low, close=r.close, volume=r.volume) for r in rows]
+
+
+@router.get("/health")
+async def health() -> dict:
+    return {"status": "ok"}
+
+
+@router.get("/candles/{symbol}", response_model=list[CandleOut])
+async def get_candles(
+    symbol: str,
+    timeframe: str = "1d",
+    limit: int = 5000,
+    session: AsyncSession = Depends(get_session),
+) -> list[CandleOut]:
+    await seed_if_empty(session, symbol="DEMO")
+    # 始终以日线为基底，再聚合到目标周期
+    rows = await repository.get_candles(session, symbol, "1d", limit=limit)
+    bars = resample_bars(_rows_to_bars(rows), timeframe)
+    return [CandleOut(ts=b.ts, open=b.open, high=b.high, low=b.low, close=b.close, volume=b.volume) for b in bars]
+
+
+@router.post("/data/sync", response_model=SyncResponse)
+async def sync_data(req: SyncRequest, session: AsyncSession = Depends(get_session)) -> SyncResponse:
+    """主动拉取并缓存某标的的日线（Tushare 主 -> AKShare 兜底）。"""
+    try:
+        res = await fetcher.sync_symbol(
+            session, req.symbol, start=req.start, end=req.end, source=req.source, force=req.force
+        )
+        return SyncResponse(**res)
+    except Exception as e:  # noqa: BLE001
+        raise HTTPException(status_code=502, detail=str(e))
+
+
+@router.post("/backtest", response_model=BacktestResponse)
+async def backtest(
+    req: BacktestRequest,
+    session: AsyncSession = Depends(get_session),
+) -> BacktestResponse:
+    await seed_if_empty(session, symbol="DEMO")
+
+    # 非演示标的：首次自动拉取真实数据并缓存
+    if req.symbol != "DEMO" and not await fetcher.is_cached(session, req.symbol):
+        try:
+            await fetcher.sync_symbol(session, req.symbol, source="auto")
+        except Exception as e:  # noqa: BLE001
+            raise HTTPException(status_code=502, detail=f"数据拉取失败: {e}")
+
+    # 日线为基底，聚合到请求周期
+    rows = await repository.get_candles(
+        session, req.symbol, "1d", start=req.start, end=req.end, limit=100000
+    )
+    if not rows:
+        raise HTTPException(status_code=404, detail=f"无数据: symbol={req.symbol}")
+
+    bars = resample_bars(_rows_to_bars(rows), req.timeframe)
+    if len(bars) < 2:
+        raise HTTPException(status_code=400, detail=f"周期 {req.timeframe} 下数据不足，无法回测")
+
+    try:
+        strategy = build_strategy(req.strategy, req.params)
+    except Exception as e:  # noqa: BLE001
+        raise HTTPException(status_code=400, detail=f"策略构建失败: {e}")
+    cfg = BacktestConfig(
+        initial_cash=req.initial_cash,
+        fast_mode=req.fast_mode,
+        bars_per_year=bars_per_year(req.timeframe),
+    )
+    result = run_backtest(bars, strategy, cfg)
+
+    df: pd.DataFrame = result["df"]
+    m = result["metrics"]
+
+    # 记录到回测运行注册表（可复现/可审计的基础）
+    session.add(
+        BacktestRun(
+            symbol=req.symbol,
+            strategy=req.strategy,
+            timeframe=req.timeframe,
+            params_json=json.dumps(req.params, ensure_ascii=False),
+            initial_cash=req.initial_cash,
+            total_return=m["total_return"],
+            max_drawdown=m["max_drawdown"],
+            sharpe=m["sharpe"],
+            num_trades=m["num_trades"],
+        )
+    )
+    await session.commit()
+
+    candles = [
+        CandleOut(ts=r["ts"], open=r["open"], high=r["high"], low=r["low"],
+                  close=r["close"], volume=r["volume"])
+        for _, r in df.iterrows()
+    ]
+    signals = [
+        SignalOut(ts=f.ts, side=f.side.value, price=f.price, qty=f.qty)
+        for f in result["fills"]
+    ]
+    indicators = IndicatorOut(
+        strategy=req.strategy,
+        data={col: _series_to_jsonable(df[col]) for col in result["indicator_cols"]},
+    )
+    equity = [EquityPoint(ts=t.to_pydatetime(), value=float(v))
+              for t, v in result["equity"].items()]
+
+    return BacktestResponse(
+        symbol=req.symbol,
+        timeframe=req.timeframe,
+        strategy=req.strategy,
+        candles=candles,
+        indicators=indicators,
+        signals=signals,
+        equity=equity,
+        metrics=MetricsOut(**m),
+        final_cash=result["final_cash"],
+        final_position=result["final_position"],
+        initial_cash=req.initial_cash,
+    )
--- a/backend/app/backtest/init.py
+++ b/backend/app/backtest/init.py
--- a/backend/app/backtest/broker.py
+++ b/backend/app/backtest/broker.py
@@ -0,0 +1,104 @@
+"""PaperBroker —— 回测中的虚拟撮合 / 账户。
+
+建模 A 股规则：
+  - 100 股整数手（1手=100股）
+  - T+1：当日买入次日才可卖（fast_mode 关闭此约束）
+  - 印花税（卖出）、过户费（双边）、佣金（万1 最低5元）、滑点
+  - 撮合价：以当根 bar 收盘价近似（阶段1 接 VWAP / 限价单）
+"""
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+
+from ..commission import CostSchedule, DEFAULT, buy_cost, sell_cost
+from ..domain import Fill, Side
+
+LOT = 100  # A股 1 手 = 100 股
+
+
+@dataclass
+class PaperBroker:
+    initial_cash: float = 100000.0
+    schedule: CostSchedule = field(default_factory=lambda: DEFAULT)
+    enable_costs: bool = True
+    enable_t_plus_1: bool = True
+
+    cash: float = field(init=False)
+    holdings: float = 0.0   # 可卖数量
+    locked: float = 0.0     # 当日买入（T+1 锁定）
+    avg_price: float = 0.0
+    fills: list[Fill] = field(default_factory=list)
+
+    def __post_init__(self) -> None:
+        self.cash = self.initial_cash
+
+    @property
+    def position(self) -> float:
+        return self.holdings + self.locked
+
+    def equity(self, price: float) -> float:
+        return self.cash + self.position * price
+
+    @staticmethod
+    def _to_lots(qty: float) -> int:
+        return int(qty // LOT) * LOT
+
+    def buy_max(self, ts, price: float) -> Fill | None:
+        """用当前现金买尽可能多的整手。"""
+        if price <= 0:
+            return None
+        rate = (
+            self.schedule.commission_rate
+            + self.schedule.transfer_fee_rate
+            + self.schedule.slippage_rate
+        ) if self.enable_costs else 0.0
+        affordable_qty = self.cash / (price * (1 + rate))
+        qty = self._to_lots(affordable_qty)
+        if qty <= 0:
+            return None
+        return self._execute_buy(ts, price, qty)
+
+    def sell_all(self, ts, price: float) -> Fill | None:
+        """卖出全部可卖持仓（整手）。"""
+        qty = self._to_lots(self.holdings)
+        if qty <= 0:
+            return None
+        return self._execute_sell(ts, price, qty)
+
+    def _execute_buy(self, ts, price: float, qty: int) -> Fill:
+        if self.enable_costs:
+            fp, comm, tf = buy_cost(price, qty, self.schedule)
+        else:
+            fp, comm, tf = price, 0.0, 0.0
+        cost = fp * qty + comm + tf
+        prev_pos = self.position
+        new_pos = prev_pos + qty
+        self.avg_price = (self.avg_price * prev_pos + fp * qty) / new_pos if new_pos else 0.0
+        if self.enable_t_plus_1:
+            self.locked += qty
+        else:
+            self.holdings += qty
+        self.cash -= cost
+        f = Fill(ts=ts, side=Side.BUY, price=fp, qty=qty, commission=comm, transfer_fee=tf)
+        self.fills.append(f)
+        return f
+
+    def _execute_sell(self, ts, price: float, qty: int) -> Fill:
+        if self.enable_costs:
+            fp, comm, tf, sd = sell_cost(price, qty, self.schedule)
+        else:
+            fp, comm, tf, sd = price, 0.0, 0.0, 0.0
+        proceeds = fp * qty - comm - tf - sd
+        self.holdings -= qty
+        self.cash += proceeds
+        if self.position == 0:
+            self.avg_price = 0.0
+        f = Fill(ts=ts, side=Side.SELL, price=fp, qty=qty, commission=comm, stamp_duty=sd, transfer_fee=tf)
+        self.fills.append(f)
+        return f
+
+    def release_t_plus_1(self) -> None:
+        """每根 bar 结束时调用：当日锁定转为次日可卖。"""
+        if self.enable_t_plus_1:
+            self.holdings += self.locked
+            self.locked = 0.0
--- a/backend/app/backtest/engine.py
+++ b/backend/app/backtest/engine.py
@@ -0,0 +1,114 @@
+"""单一回测引擎（fast/strict 两档，同一代码路径、同一撮合逻辑）。
+
+不做"快层 vectorbt + 真层自研"双引擎——用户必然信任更快的那层，
+一旦快层省略 T+1/费用，两套结论背离即反复扯皮"以谁为准"。
+这里用开关控制规则是否注入，引擎只有一份。
+
+数据流：
+  bars -> DataFrame -> strategy.compute(指标) -> 逐 bar 喂 strategy.on_bar(broker)
+       -> broker 产生 fills -> 每根 bar 记录 equity -> metrics
+"""
+from __future__ import annotations
+
+from dataclasses import dataclass
+
+import pandas as pd
+
+from ..domain import Bar
+from .broker import PaperBroker
+from .metrics import compute_metrics
+
+
+@dataclass
+class BacktestConfig:
+    initial_cash: float = 1000000.0
+    fast_mode: bool = False            # True: 关 T+1/费用，交互试探
+    bars_per_year: int = 252           # 日线 252；分钟级另算
+
+
+def run_backtest(bars: list[Bar], strategy, cfg: BacktestConfig | None = None) -> dict:
+    cfg = cfg or BacktestConfig()
+    if not bars:
+        return _empty_result()
+
+    df = pd.DataFrame(
+        [{"ts": b.ts, "open": b.open, "high": b.high, "low": b.low,
+          "close": b.close, "volume": b.volume} for b in bars]
+    ).sort_values("ts").reset_index(drop=True)
+
+    # 指标预计算（由策略持有，单一事实源）
+    ind = strategy.compute(df["close"], df["high"], df["low"])
+    indicator_cols = list(ind.columns)
+    df = pd.concat([df, ind], axis=1)
+
+    broker = PaperBroker(
+        initial_cash=cfg.initial_cash,
+        enable_costs=not cfg.fast_mode,
+        enable_t_plus_1=not cfg.fast_mode,
+    )
+
+    equity_values = []
+    for i in range(len(df)):
+        row = df.iloc[i]
+        strategy.on_bar(i, row, broker)
+        equity_values.append(broker.equity(row["close"]))
+        broker.release_t_plus_1()
+
+    equity = pd.Series(equity_values, index=df["ts"], name="equity")
+    metrics = compute_metrics(equity, cfg.bars_per_year)
+    metrics["num_trades"] = len(broker.fills)
+    metrics["win_rate"] = _win_rate(broker.fills)
+
+    return {
+        "df": df,
+        "indicator_cols": indicator_cols,
+        "fills": broker.fills,
+        "equity": equity,
+        "metrics": metrics,
+        "final_cash": broker.cash,
+        "final_position": broker.position,
+    }
+
+
+def _win_rate(fills) -> float:
+    """FIFO 配对计算卖出胜率。"""
+    from collections import deque
+    buys: deque = deque()
+    wins = 0
+    total_sells = 0
+    for f in fills:
+        if f.side.value == "buy":
+            buys.append((f.price, f.qty))
+        else:  # sell
+            remaining = f.qty
+            total_sells += 1
+            profitable = True
+            while remaining > 0 and buys:
+                buy_price, buy_qty = buys[0]
+                if f.price >= buy_price:
+                    pass
+                else:
+                    profitable = False
+                take = min(remaining, buy_qty)
+                buy_qty -= take
+                remaining -= take
+                if buy_qty <= 0:
+                    buys.popleft()
+                else:
+                    buys[0] = (buy_price, buy_qty)
+            if profitable and f.qty > 0:
+                wins += 1
+    return wins / total_sells if total_sells else 0.0
+
+
+def _empty_result() -> dict:
+    return {
+        "df": pd.DataFrame(),
+        "indicator_cols": [],
+        "fills": [],
+        "equity": pd.Series(dtype=float),
+        "metrics": {"total_return": 0.0, "max_drawdown": 0.0, "sharpe": 0.0,
+                    "volatility": 0.0, "num_trades": 0, "win_rate": 0.0},
+        "final_cash": 0.0,
+        "final_position": 0.0,
+    }
--- a/backend/app/backtest/metrics.py
+++ b/backend/app/backtest/metrics.py
@@ -0,0 +1,41 @@
+"""绩效统计（阶段1 补基准归因：超额/信息比率/beta/alpha）。"""
+from __future__ import annotations
+
+import numpy as np
+import pandas as pd
+
+
+def compute_metrics(equity: pd.Series, bars_per_year: int = 252) -> dict:
+    equity = equity.dropna()
+    if len(equity) < 2 or equity.iloc[0] == 0:
+        return {"total_return": 0.0, "max_drawdown": 0.0, "sharpe": 0.0,
+                "volatility": 0.0, "win_rate": 0.0}
+
+    total_return = float(equity.iloc[-1] / equity.iloc[0] - 1)
+
+    returns = equity.pct_change().dropna()
+    cummax = equity.cummax()
+    drawdown = (equity - cummax) / cummax
+    max_drawdown = float(abs(drawdown.min()))
+
+    std = float(returns.std())
+    sharpe = float(returns.mean() / std * np.sqrt(bars_per_year)) if std > 0 else 0.0
+    volatility = std * np.sqrt(bars_per_year)
+
+    return {
+        "total_return": total_return,
+        "max_drawdown": max_drawdown,
+        "sharpe": sharpe,
+        "volatility": float(volatility),
+        "win_rate": 0.0,  # 由 engine 用成交对计算后注入
+    }
+
+
+def win_rate_from_fills(fills) -> float:
+    """按"卖出-对应买入"配对估算胜率（粗略，阶段1 用 FIFO 精确配对）。"""
+    sells = [f for f in fills if f.side.value == "sell"]
+    if not sells:
+        return 0.0
+    wins = sum(1 for f in fills if f.side.value == "sell" and f.price > 0)
+    # 简化：有成交即计；真实胜率需配对，这里先返回 0 占位，由 engine 精算
+    return 0.0
--- a/backend/app/backtest/strategies/init.py
+++ b/backend/app/backtest/strategies/init.py
@@ -0,0 +1,24 @@
+"""策略注册表与工厂。
+
+新增策略：实现 Strategy（base.py），在此注册 {name: Class}，前端下拉即可选。
+策略构造参数由请求的 params(dict) 以 **kwargs 传入。
+"""
+from __future__ import annotations
+
+from .base import Strategy
+from .macd_cross import MACDCrossStrategy
+from .ma_strategies import MACrossStrategy, SingleMAStrategy
+
+STRATEGIES: dict[str, type[Strategy]] = {
+    "macd_cross": MACDCrossStrategy,
+    "ma_cross": MACrossStrategy,
+    "single_ma": SingleMAStrategy,
+}
+
+
+def build_strategy(name: str, params: dict | None) -> Strategy:
+    cls = STRATEGIES.get(name)
+    if cls is None:
+        raise ValueError(f"未知策略: {name}（可用: {', '.join(STRATEGIES)}）")
+    kwargs = {k: v for k, v in (params or {}).items()}
+    return cls(**kwargs)
--- a/backend/app/backtest/strategies/base.py
+++ b/backend/app/backtest/strategies/base.py
@@ -0,0 +1,20 @@
+"""策略抽象基类。策略只产生买卖意图（向 broker 下单），不负责撮合/费用。"""
+from __future__ import annotations
+
+from abc import ABC, abstractmethod
+
+import pandas as pd
+
+from ..broker import PaperBroker
+
+
+class Strategy(ABC):
+    """compute 预算指标；on_bar 逐 bar 决策并向 broker 下单。"""
+
+    @abstractmethod
+    def compute(self, close: pd.Series, high: pd.Series, low: pd.Series) -> pd.DataFrame:
+        ...
+
+    @abstractmethod
+    def on_bar(self, i: int, row: pd.Series, broker: PaperBroker) -> None:
+        ...
--- a/backend/app/backtest/strategies/ma_strategies.py
+++ b/backend/app/backtest/strategies/ma_strategies.py
@@ -0,0 +1,74 @@
+"""均线类策略：双均线交叉、单均线（价格上穿/下穿）。"""
+from __future__ import annotations
+
+import pandas as pd
+
+from ...indicators import ma
+from ..broker import PaperBroker
+from .base import Strategy
+
+
+class MACrossStrategy(Strategy):
+    """双均线交叉：快线上穿慢线买入，下穿卖出（金叉/死叉）。"""
+
+    def __init__(self, fast: float = 5, slow: float = 20):
+        self.fast = int(fast)
+        self.slow = int(slow)
+        self._ind: pd.DataFrame | None = None
+        self._prev_fast: float | None = None
+        self._prev_slow: float | None = None
+
+    def compute(self, close: pd.Series, high: pd.Series, low: pd.Series) -> pd.DataFrame:
+        self._ind = pd.DataFrame({"fast": ma(close, self.fast), "slow": ma(close, self.slow)})
+        return self._ind
+
+    def on_bar(self, i: int, row: pd.Series, broker: PaperBroker) -> None:
+        f = float(self._ind["fast"].iloc[i])
+        s = float(self._ind["slow"].iloc[i])
+        if self._prev_fast is None:
+            self._prev_fast, self._prev_slow = f, s
+            return
+
+        golden = self._prev_fast <= self._prev_slow and f > s
+        death = self._prev_fast >= self._prev_slow and f < s
+
+        price = float(row["close"])
+        ts = row["ts"]
+        if golden and pd.notna(f):
+            broker.buy_max(ts, price)
+        elif death and broker.position > 0 and pd.notna(f):
+            broker.sell_all(ts, price)
+
+        self._prev_fast, self._prev_slow = f, s
+
+
+class SingleMAStrategy(Strategy):
+    """单均线：收盘价上穿均线买入，下穿均线卖出。"""
+
+    def __init__(self, period: float = 20):
+        self.period = int(period)
+        self._ind: pd.DataFrame | None = None
+        self._prev_above: bool | None = None
+
+    def compute(self, close: pd.Series, high: pd.Series, low: pd.Series) -> pd.DataFrame:
+        self._ind = pd.DataFrame({"ma": ma(close, self.period)})
+        return self._ind
+
+    def on_bar(self, i: int, row: pd.Series, broker: PaperBroker) -> None:
+        m = float(self._ind["ma"].iloc[i])
+        price = float(row["close"])
+        ts = row["ts"]
+        above = price > m
+        if self._prev_above is None or pd.isna(m):
+            self._prev_above = above
+            return
+
+        cross_up = above and not self._prev_above        # 上穿
+        cross_down = (not above) and self._prev_above    # 下穿
+
+        if cross_up:
+            broker.buy_max(ts, price)
+        elif cross_down and broker.position > 0:
+            broker.sell_all(ts, price)
+
+        self._prev_above = above
--- a/backend/app/backtest/strategies/macd_cross.py
+++ b/backend/app/backtest/strategies/macd_cross.py
@@ -0,0 +1,42 @@
+"""MACD 金叉死叉策略。"""
+from __future__ import annotations
+
+import pandas as pd
+
+from ...indicators import macd
+from ..broker import PaperBroker
+from .base import Strategy
+
+
+class MACDCrossStrategy(Strategy):
+    def __init__(self, fast: float = 12, slow: float = 26, signal: float = 9):
+        self.fast = int(fast)
+        self.slow = int(slow)
+        self.signal = int(signal)
+        self._ind: pd.DataFrame | None = None
+        self._prev_dif: float | None = None
+        self._prev_dea: float | None = None
+
+    def compute(self, close: pd.Series, high: pd.Series, low: pd.Series) -> pd.DataFrame:
+        self._ind = macd(close, self.fast, self.slow, self.signal)
+        return self._ind
+
+    def on_bar(self, i: int, row: pd.Series, broker: PaperBroker) -> None:
+        dif = float(self._ind["macd"].iloc[i])
+        dea = float(self._ind["signal"].iloc[i])
+
+        if self._prev_dif is None:
+            self._prev_dif, self._prev_dea = dif, dea
+            return
+
+        golden = self._prev_dif <= self._prev_dea and dif > dea   # 金叉
+        death = self._prev_dif >= self._prev_dea and dif < dea    # 死叉
+
+        price = float(row["close"])
+        ts = row["ts"]
+        if golden and pd.notna(dif):
+            broker.buy_max(ts, price)
+        elif death and broker.position > 0 and pd.notna(dif):
+            broker.sell_all(ts, price)
+
+        self._prev_dif, self._prev_dea = dif, dea
--- a/backend/app/commission.py
+++ b/backend/app/commission.py
@@ -0,0 +1,47 @@
+"""A股交易成本（基准日 2026-08，已修正历史错误）。
+
+⚠️ 重要：费率必须是"参数表 + 生效日期版本化 + 显式基准日"，不能硬编码成常量。
+   费率会再变（如 2023-08-28 印花税减半、2022 过户费下调）。
+   MVP 用单一 CostSchedule；阶段1 扩展为按生效日期区间查找的多版本表。
+
+参考（已核实）：
+  印花税 0.05% 单边卖出 —— 2023-08-28 财政部减半（原 0.1%）
+  过户费 0.001% 沪深双边   —— 2022 年统一下调（原沪市万 0.2 单边）
+  佣金 万1 含规费，最低 5 元 —— 2026 主流
+"""
+from __future__ import annotations
+
+from dataclasses import dataclass
+
+from .config import settings
+
+
+@dataclass(frozen=True)
+class CostSchedule:
+    stamp_duty_rate: float = settings.stamp_duty_rate    # 印花税，卖出
+    transfer_fee_rate: float = settings.transfer_fee_rate  # 过户费，双边
+    commission_rate: float = settings.commission_rate    # 佣金
+    commission_min: float = settings.commission_min      # 最低佣金
+    slippage_rate: float = settings.slippage_rate        # 滑点（价格比例近似）
+
+
+DEFAULT = CostSchedule()
+
+
+def buy_cost(price: float, qty: float, sch: CostSchedule = DEFAULT) -> tuple[float, float, float]:
+    """买入成本。返回 (成交价, 佣金, 过户费)。买入无印花税。"""
+    fill_price = price * (1 + sch.slippage_rate)
+    gross = fill_price * qty
+    commission = max(gross * sch.commission_rate, sch.commission_min)
+    transfer_fee = gross * sch.transfer_fee_rate
+    return fill_price, commission, transfer_fee
+
+
+def sell_cost(price: float, qty: float, sch: CostSchedule = DEFAULT) -> tuple[float, float, float, float]:
+    """卖出成本。返回 (成交价, 佣金, 过户费, 印花税)。"""
+    fill_price = price * (1 - sch.slippage_rate)
+    gross = fill_price * qty
+    commission = max(gross * sch.commission_rate, sch.commission_min)
+    transfer_fee = gross * sch.transfer_fee_rate
+    stamp_duty = gross * sch.stamp_duty_rate
+    return fill_price, commission, transfer_fee, stamp_duty
--- a/backend/app/config.py
+++ b/backend/app/config.py
@@ -0,0 +1,26 @@
+"""应用配置（pydantic-settings）。可由 .env / 环境变量覆盖。"""
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    model_config = SettingsConfigDict(env_file=".env", extra="ignore")
+
+    app_name: str = "Stock Backtest"
+
+    # 默认 SQLite 零配置；切 Postgres/TimescaleDB 只改这一行
+    database_url: str = "sqlite+aiosqlite:///./stock.db"
+
+    # 真实数据源
+    tushare_token: str = ""            # Tushare Pro token（主数据源）
+    data_adjust: str = "qfq"           # 复权：qfq 前复权 / hfq 后复权 / "" 不复权
+    data_default_start: str = "20200101"  # 默认拉取起点（约近 5 年）
+
+    # A股交易成本（基准日 2026-08）——做成可配置参数，便于将来按生效日期版本化
+    stamp_duty_rate: float = 0.0005      # 印花税 0.05%，单边卖出（2023-08-28 减半）
+    transfer_fee_rate: float = 0.00001   # 过户费 0.001%，沪深双边（2022 调整）
+    commission_rate: float = 0.0001      # 佣金 万1（含规费）
+    commission_min: float = 5.0          # 最低 5 元
+    slippage_rate: float = 0.0005        # 滑点近似（按价格比例）
+
+
+settings = Settings()
--- a/backend/app/data/init.py
+++ b/backend/app/data/init.py
--- a/backend/app/data/aggregation.py
+++ b/backend/app/data/aggregation.py
@@ -0,0 +1,54 @@
+"""K 线周期聚合：日线 -> 周/月/年。
+
+生产环境用 TimescaleDB Continuous Aggregates 在库里预物化（性能）；
+MVP 在应用层用 pandas resample 即可，逻辑等价、便于切换。
+
+OHLCV 聚合规则：开=周期内首根开、高=最高、低=最低、收=末根收、量=求和。
+"""
+from __future__ import annotations
+
+import pandas as pd
+
+from ..domain import Bar
+
+# pandas resample 规则（周一为周首；月/年以首日对齐）
+_RULES = {"1w": "W-MON", "1M": "MS", "1y": "YS"}
+
+# 各周期的"年交易日数"（用于夏普等指标的年化）
+_BARS_PER_YEAR = {"1d": 252, "1w": 52, "1M": 12, "1y": 1}
+
+
+def bars_per_year(timeframe: str) -> int:
+    return _BARS_PER_YEAR.get(timeframe, 252)
+
+
+def resample_bars(bars: list[Bar], timeframe: str) -> list[Bar]:
+    """把日线 bars 聚合为目标周期；日线或未知周期原样返回。"""
+    if not bars or timeframe in ("1d", "d", "day", "", None):
+        return bars
+    rule = _RULES.get(timeframe)
+    if rule is None:
+        return bars
+
+    df = pd.DataFrame(
+        [{"ts": b.ts, "open": b.open, "high": b.high, "low": b.low, "close": b.close, "volume": b.volume}
+         for b in bars]
+    ).set_index("ts").sort_index()
+
+    agg = (
+        df.resample(rule)
+        .agg({"open": "first", "high": "max", "low": "min", "close": "last", "volume": "sum"})
+        .dropna()
+    )
+
+    return [
+        Bar(
+            ts=ts.to_pydatetime(),
+            open=float(row["open"]),
+            high=float(row["high"]),
+            low=float(row["low"]),
+            close=float(row["close"]),
+            volume=float(row["volume"]),
+        )
+        for ts, row in agg.iterrows()
+    ]
--- a/backend/app/data/akshare_provider.py
+++ b/backend/app/data/akshare_provider.py
@@ -0,0 +1,38 @@
+"""AKShare 数据源（兜底/校验）。免费、无需 token。
+
+默认不安装（依赖较重）；如需启用：`uv add akshare`。
+fetcher 在 Tushare 失败时会尝试本模块；未安装则该路径自动跳过。
+"""
+from __future__ import annotations
+
+from datetime import datetime
+
+from ..domain import Bar
+from .symbols import plain_code
+
+
+def fetch_daily(code: str, start: str = "20200101", end: str | None = None,
+                adjust: str = "qfq") -> list[Bar]:
+    import akshare as ak  # 延迟导入
+
+    end = end or datetime.now().strftime("%Y%m%d")
+    symbol = plain_code(code)
+    adj_map = {"qfq": "qfq", "hfq": "hfq", "": "", None: ""}
+    df = ak.stock_zh_a_hist(
+        symbol=symbol, period="daily",
+        start_date=start, end_date=end, adjust=adj_map.get(adjust, ""),
+    )
+    if df is None or df.empty:
+        raise RuntimeError(f"AKShare 无数据: {symbol}")
+
+    bars: list[Bar] = []
+    for _, r in df.iterrows():
+        bars.append(
+            Bar(
+                ts=datetime.strptime(str(r["日期"]), "%Y-%m-%d"),
+                open=float(r["开盘"]), high=float(r["最高"]),
+                low=float(r["最低"]), close=float(r["收盘"]),
+                volume=float(r["成交量"]) * 100.0,  # AKShare 成交量单位为手 -> 股
+            )
+        )
+    return bars
--- a/backend/app/data/fetcher.py
+++ b/backend/app/data/fetcher.py
@@ -0,0 +1,81 @@
+"""数据编排：拉取（Tushare 主 -> AKShare 兜底）+ 本地缓存。
+
+真实行情落库到 candles 表（timeframe='1d'），回测统一从库读，与 DEMO 同路径。
+"""
+from __future__ import annotations
+
+import asyncio
+
+from sqlalchemy import delete, func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from ..config import settings
+from ..domain import Bar
+from ..models import Candle
+from . import akshare_provider, tushare_provider
+
+DEFAULT_START = settings.data_default_start or "20200101"
+
+
+def _providers(source: str):
+    """按优先级返回 (名称, 同步拉取函数) 列表。"""
+    seq = []
+    if source in ("auto", "tushare") and settings.tushare_token:
+        seq.append(("tushare", tushare_provider.fetch_daily))
+    if source in ("auto", "akshare"):
+        seq.append(("akshare", akshare_provider.fetch_daily))
+    return seq
+
+
+async def count_cached(session: AsyncSession, symbol: str) -> int:
+    res = await session.execute(
+        select(func.count()).select_from(Candle).where(
+            Candle.symbol == symbol, Candle.timeframe == "1d"
+        )
+    )
+    return int(res.scalar() or 0)
+
+
+async def is_cached(session: AsyncSession, symbol: str) -> bool:
+    return await count_cached(session, symbol) > 0
+
+
+async def sync_symbol(
+    session: AsyncSession,
+    code: str,
+    start: str | None = None,
+    end: str | None = None,
+    source: str = "auto",
+    force: bool = False,
+) -> dict:
+    """拉取并缓存某标的日线。已缓存且非 force 时直接返回缓存计数。"""
+    if not force and await is_cached(session, code):
+        return {"symbol": code, "bars": await count_cached(session, code), "source": "cache"}
+
+    start = start or DEFAULT_START
+    adjust = settings.data_adjust
+    errors: list[str] = []
+    bars: list[Bar] = []
+    used = None
+
+    for name, fn in _providers(source):
+        try:
+            # tushare/akshare 是同步网络 IO，丢到线程池避免阻塞事件循环
+            bars = await asyncio.to_thread(fn, code, start, end, adjust)
+            used = name
+            break
+        except Exception as e:  # noqa: BLE001
+            errors.append(f"{name}: {e}")
+
+    if not bars:
+        raise RuntimeError("所有数据源均失败 -> " + " | ".join(errors) if errors else "无可用数据源")
+
+    # 全量替换该标的日线（避免重复主键）
+    await session.execute(delete(Candle).where(Candle.symbol == code, Candle.timeframe == "1d"))
+    for b in bars:
+        session.add(
+            Candle(symbol=code, timeframe="1d", ts=b.ts, open=b.open, high=b.high,
+                   low=b.low, close=b.close, volume=b.volume)
+        )
+    await session.commit()
+    return {"symbol": code, "bars": len(bars), "source": used}
--- a/backend/app/data/repository.py
+++ b/backend/app/data/repository.py
@@ -0,0 +1,34 @@
+"""K 线数据访问（从库读）。
+
+写入由 DataProvider 适配器负责（阶段1 接 Tushare/AKShare）。
+MVP 的数据由 synthetic.seed_if_empty 灌入。
+"""
+from __future__ import annotations
+
+from datetime import datetime
+
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from ..models import Candle
+
+
+async def get_candles(
+    session: AsyncSession,
+    symbol: str,
+    timeframe: str = "1d",
+    start: datetime | None = None,
+    end: datetime | None = None,
+    limit: int = 5000,
+) -> list[Candle]:
+    stmt = select(Candle).where(
+        Candle.symbol == symbol,
+        Candle.timeframe == timeframe,
+    )
+    if start is not None:
+        stmt = stmt.where(Candle.ts >= start)
+    if end is not None:
+        stmt = stmt.where(Candle.ts <= end)
+    stmt = stmt.order_by(Candle.ts.asc()).limit(limit)
+    result = await session.execute(stmt)
+    return list(result.scalars().all())
--- a/backend/app/data/symbols.py
+++ b/backend/app/data/symbols.py
@@ -0,0 +1,25 @@
+"""A 股代码归一化。支持 6 位纯数字或带交易所后缀（000001 / 000001.SZ）。"""
+from __future__ import annotations
+
+
+def plain_code(code: str) -> str:
+    """000001.SZ -> 000001"""
+    return code.strip().upper().split(".")[0]
+
+
+def to_ts_code(code: str) -> str:
+    """转 Tushare ts_code（带交易所后缀）。"""
+    c = code.strip().upper()
+    if "." in c:
+        return c
+    c = plain_code(c)
+    # 沪市：60xxxx 主板、68xxxx 科创、9xxxxx B 股
+    if c.startswith(("60", "68", "9")):
+        return c + ".SH"
+    # 深市：00xxxx 主板/中小、30xxxx 创业、20xxxx B 股
+    if c.startswith(("00", "30", "20")):
+        return c + ".SZ"
+    # 北交所：8xxxxx / 4xxxxx
+    if c.startswith(("8", "4")):
+        return c + ".BJ"
+    return c + ".SZ"
--- a/backend/app/data/synthetic.py
+++ b/backend/app/data/synthetic.py
@@ -0,0 +1,78 @@
+"""合成数据（MVP 零依赖可跑）。
+
+生成随机游走 OHLCV，灌入 DB。仅用于让回测链路在没有真实数据源时也能跑通演示。
+阶段1 接 Tushare/AKShare 后，这里仅保留为"离线测试夹具"。
+"""
+from __future__ import annotations
+
+from datetime import datetime, timedelta, timezone
+
+import numpy as np
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from ..domain import Bar
+from ..models import Candle
+
+
+def _trading_days(n: int) -> list[datetime]:
+    """粗略生成 n 个工作日（跳过周末；节假日由阶段1 的交易日历服务处理）。"""
+    start = datetime.now(timezone.utc).replace(tzinfo=None) - timedelta(days=int(n * 1.6))
+    days: list[datetime] = []
+    d = start
+    while len(days) < n:
+        if d.weekday() < 5:
+            days.append(d.replace(hour=15, minute=0, second=0, microsecond=0))
+        d += timedelta(days=1)
+    return days
+
+
+def generate_ohlcv(n: int = 500, seed: int = 42) -> list[Bar]:
+    """随机游走 + A 股风格的价格区间（5~30 元）。"""
+    rng = np.random.default_rng(seed)
+    rets = rng.normal(loc=0.0003, scale=0.018, size=n)
+    price = 10.0 * np.cumprod(1 + rets)
+    days = _trading_days(n)
+    bars: list[Bar] = []
+    for i in range(n):
+        close = float(price[i])
+        op = close * (1 + rng.normal(0, 0.005))
+        hi = max(op, close) * (1 + abs(rng.normal(0, 0.006)))
+        lo = min(op, close) * (1 - abs(rng.normal(0, 0.006)))
+        vol = float(rng.integers(1_000_000, 10_000_000))
+        bars.append(
+            Bar(
+                ts=days[i],
+                open=round(op, 2),
+                high=round(hi, 2),
+                low=round(lo, 2),
+                close=round(close, 2),
+                volume=vol,
+            )
+        )
+    return bars
+
+
+async def seed_if_empty(session: AsyncSession, symbol: str = "DEMO", n: int = 500) -> None:
+    """若库中无该 symbol 数据，则灌入合成数据。"""
+    existing = await session.execute(
+        select(Candle.id).where(Candle.symbol == symbol).limit(1)
+    )
+    if existing.scalars().first() is not None:
+        return
+
+    bars = generate_ohlcv(n=n)
+    for b in bars:
+        session.add(
+            Candle(
+                symbol=symbol,
+                timeframe="1d",
+                ts=b.ts,
+                open=b.open,
+                high=b.high,
+                low=b.low,
+                close=b.close,
+                volume=b.volume,
+            )
+        )
+    await session.commit()
--- a/backend/app/data/tushare_provider.py
+++ b/backend/app/data/tushare_provider.py
@@ -0,0 +1,51 @@
+"""Tushare 数据源（主）。日线 + 前复权。
+
+token 从 settings.tushare_token 读取（.env）。免费版 pro.daily 与 ts.pro_bar 实测可用。
+"""
+from __future__ import annotations
+
+from datetime import datetime
+
+from ..config import settings
+from ..domain import Bar
+from .symbols import to_ts_code
+
+
+def _parse(date_str: str) -> datetime:
+    return datetime.strptime(str(date_str), "%Y%m%d")
+
+
+def fetch_daily(code: str, start: str = "20200101", end: str | None = None,
+                adjust: str = "qfq") -> list[Bar]:
+    import tushare as ts  # 延迟导入：未装/无 token 时 DEMO 仍可用
+
+    if not settings.tushare_token:
+        raise RuntimeError("未配置 TUSHARE_TOKEN")
+    ts.set_token(settings.tushare_token)
+    pro = ts.pro_api()
+    ts_code = to_ts_code(code)
+    end = end or datetime.now().strftime("%Y%m%d")
+
+    # 优先 pro_bar（含复权）；积分不足则退化为 pro.daily（不复权）
+    df = None
+    try:
+        df = ts.pro_bar(ts_code=ts_code, adj=adjust, start_date=start, end_date=end, freq="D")
+    except Exception:
+        df = None
+    if df is None or df.empty:
+        df = pro.daily(ts_code=ts_code, start_date=start, end_date=end)
+    if df is None or df.empty:
+        raise RuntimeError(f"Tushare 无数据: {ts_code}")
+
+    df = df.sort_values("trade_date")
+    bars: list[Bar] = []
+    for _, r in df.iterrows():
+        bars.append(
+            Bar(
+                ts=_parse(r["trade_date"]),
+                open=float(r["open"]), high=float(r["high"]),
+                low=float(r["low"]), close=float(r["close"]),
+                volume=float(r["vol"]) * 100.0,  # Tushare vol 单位为手 -> 股
+            )
+        )
+    return bars
--- a/backend/app/db.py
+++ b/backend/app/db.py
@@ -0,0 +1,20 @@
+"""async SQLAlchemy 引擎与会话。"""
+from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker, create_async_engine
+from sqlalchemy.orm import DeclarativeBase
+
+from .config import settings
+
+
+class Base(DeclarativeBase):
+    """所有 ORM 模型的基类。"""
+
+
+# echo=False；生产环境可用连接池参数调优
+engine = create_async_engine(settings.database_url, echo=False, future=True)
+async_session = async_sessionmaker(engine, expire_on_commit=False, class_=AsyncSession)
+
+
+async def get_session() -> AsyncSession:
+    """FastAPI 依赖：提供一个事务会话。"""
+    async with async_session() as session:
+        yield session
--- a/backend/app/domain.py
+++ b/backend/app/domain.py
@@ -0,0 +1,77 @@
+"""领域模型契约（单一事实源的载体）。
+
+MVP 第一周必须定稿的核心类型。回测引擎、指标、API、(未来的)前端类型化客户端
+都基于这些契约——保证"图表/回测/实盘"用同一套语义。
+
+注意：回测/图表的指标值由后端唯一计算（app.indicators），前端不另算。
+"""
+from __future__ import annotations
+
+from dataclasses import dataclass
+from datetime import datetime
+from enum import Enum
+
+
+class Side(str, Enum):
+    BUY = "buy"
+    SELL = "sell"
+
+
+class Timeframe(str, Enum):
+    """K 线周期。分钟级已预留（用户选了分钟级回测）。"""
+    M1 = "1m"
+    M5 = "5m"
+    M15 = "15m"
+    M30 = "30m"
+    H1 = "1h"
+    D1 = "1d"
+    W1 = "1w"
+
+
+@dataclass(frozen=True)
+class Bar:
+    """一根 K 线（OHLCV + 时间戳）。复权标识后续扩展。"""
+    ts: datetime
+    open: float
+    high: float
+    low: float
+    close: float
+    volume: float
+
+
+@dataclass(frozen=True)
+class Signal:
+    """策略产生的交易信号（用于在图上标注买卖点）。"""
+    ts: datetime
+    side: Side
+    price: float
+    strength: float = 1.0
+
+
+@dataclass(frozen=True)
+class Fill:
+    """一笔成交（含费用明细）。回测中由 PaperBroker 产生。"""
+    ts: datetime
+    side: Side
+    price: float
+    qty: float
+    commission: float = 0.0
+    stamp_duty: float = 0.0   # 印花税（仅卖出）
+    transfer_fee: float = 0.0  # 过户费（双边）
+
+    @property
+    def total_cost(self) -> float:
+        return self.commission + self.stamp_duty + self.transfer_fee
+
+
+@dataclass
+class Position:
+    """持仓状态。T+1：locked_qty 为当日买入、次日才可卖的部分。"""
+    symbol: str = ""
+    holdings: float = 0.0   # 可卖数量
+    locked: float = 0.0     # 当日买入（T+1 锁定）
+    avg_price: float = 0.0
+
+    @property
+    def qty(self) -> float:
+        return self.holdings + self.locked
--- a/backend/app/indicators.py
+++ b/backend/app/indicators.py
@@ -0,0 +1,56 @@
+"""技术指标（单一事实源）。
+
+MVP 用纯 pandas/numpy 实现，避免 Windows 上 TA-Lib C 库的安装痛点。
+算法正确（MACD = 快慢 EMA 之差），接口稳定；阶段1 在 Linux/Docker 上可换 TA-Lib，
+只需保持函数签名（输入 close Series，输出指标），上层无感。
+"""
+from __future__ import annotations
+
+import numpy as np
+import pandas as pd
+
+
+def ema(series: pd.Series, span: int) -> pd.Series:
+    """指数移动平均（adjust=False，与 TA-Lib 默认一致）。"""
+    return series.ewm(span=span, adjust=False).mean()
+
+
+def macd(close: pd.Series, fast: int = 12, slow: int = 26, signal: int = 9) -> pd.DataFrame:
+    """MACD：返回 DataFrame[DIF, DEA, HIST]。"""
+    dif = ema(close, fast) - ema(close, slow)
+    dea = ema(dif, signal)
+    hist = (dif - dea) * 2  # A股惯例 MACD 柱 = 2*(DIF-DEA)
+    return pd.DataFrame({"macd": dif, "signal": dea, "hist": hist})
+
+
+def rsi(close: pd.Series, period: int = 14) -> pd.Series:
+    """RSI（Wilder 平滑）。"""
+    delta = close.diff()
+    gain = delta.clip(lower=0.0)
+    loss = -delta.clip(upper=0.0)
+    avg_gain = gain.ewm(alpha=1 / period, adjust=False).mean()
+    avg_loss = loss.ewm(alpha=1 / period, adjust=False).mean()
+    rs = avg_gain / avg_loss.replace(0, np.nan)
+    return 100 - (100 / (1 + rs))
+
+
+def kdj(high: pd.Series, low: pd.Series, close: pd.Series,
+        n: int = 9, m1: int = 3, m2: int = 3) -> pd.DataFrame:
+    """KDJ（A股常用：RSV -> K -> D -> J）。"""
+    low_n = low.rolling(n, min_periods=1).min()
+    high_n = high.rolling(n, min_periods=1).max()
+    rsv = (close - low_n) / (high_n - low_n).replace(0, np.nan) * 100
+    k = rsv.ewm(alpha=1 / m1, adjust=False).mean()
+    d = k.ewm(alpha=1 / m2, adjust=False).mean()
+    j = 3 * k - 2 * d
+    return pd.DataFrame({"k": k, "d": d, "j": j})
+
+
+def bollinger(close: pd.Series, period: int = 20, std: float = 2.0) -> pd.DataFrame:
+    ma = close.rolling(period, min_periods=1).mean()
+    sd = close.rolling(period, min_periods=1).std(ddof=0)
+    return pd.DataFrame({"mid": ma, "upper": ma + std * sd, "lower": ma - std * sd})
+
+
+def ma(close: pd.Series, period: int) -> pd.Series:
+    return close.rolling(period, min_periods=1).mean()
--- a/backend/app/main.py
+++ b/backend/app/main.py
@@ -0,0 +1,43 @@
+"""FastAPI 入口。
+
+启动时自动建表（MVP 用 create_all；阶段1 切 Alembic 迁移，含 TimescaleDB hypertable）。
+"""
+from contextlib import asynccontextmanager
+
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+
+from .api import router
+from .db import Base, engine
+from . import models  # noqa: F401  —— 注册 ORM 到 Base.metadata
+
+
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+    yield
+
+
+app = FastAPI(
+    title="Stock Backtest",
+    description="历史回测 + 回放式模拟平台（A 股为主，不做实盘）",
+    version="0.1.0",
+    lifespan=lifespan,
+)
+
+# 开发期允许前端 dev server 跨域；上线收窄 origins
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+app.include_router(router)
+
+
+@app.get("/")
+async def root() -> dict:
+    return {"name": "Stock Backtest API", "docs": "/docs"}
--- a/backend/app/models.py
+++ b/backend/app/models.py
@@ -0,0 +1,57 @@
+"""ORM 模型。
+
+Candle 表设计与 TimescaleDB hypertable 完全兼容：将来在目标 PG 库执行
+  SELECT create_hypertable('candles', 'ts');
+即可升级为时序表 + Continuous Aggregates 多周期预聚合，无需改表结构。
+"""
+from datetime import datetime
+
+from sqlalchemy import DateTime, Float, Integer, String, UniqueConstraint
+from sqlalchemy.orm import Mapped, mapped_column
+
+from .db import Base
+
+
+def _utcnow() -> datetime:
+    # naive UTC，避免 SQLite 存储时区带来的麻烦
+    from datetime import timezone
+    return datetime.now(timezone.utc).replace(tzinfo=None)
+
+
+class Candle(Base):
+    __tablename__ = "candles"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    symbol: Mapped[str] = mapped_column(String(16), index=True)
+    timeframe: Mapped[str] = mapped_column(String(4), default="1d", index=True)
+    ts: Mapped[datetime] = mapped_column(DateTime, index=True)  # bar 开始时间
+    open: Mapped[float] = mapped_column(Float)
+    high: Mapped[float] = mapped_column(Float)
+    low: Mapped[float] = mapped_column(Float)
+    close: Mapped[float] = mapped_column(Float)
+    volume: Mapped[float] = mapped_column(Float)
+
+    __table_args__ = (
+        UniqueConstraint("symbol", "timeframe", "ts", name="uq_candle_sym_tf_ts"),
+    )
+
+
+class BacktestRun(Base):
+    """回测运行注册表（可复现/可审计/可回归对比的基础）。
+
+    完整版应记录 策略版本 + 参数快照 + 数据快照（复权/数据源/库版本）+ 环境指纹 + 结果指纹。
+    MVP 先落关键字段，结构就位。
+    """
+    __tablename__ = "backtest_runs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime, default=_utcnow)
+    symbol: Mapped[str] = mapped_column(String(16))
+    strategy: Mapped[str] = mapped_column(String(64))
+    timeframe: Mapped[str] = mapped_column(String(4), default="1d")
+    params_json: Mapped[str] = mapped_column(String, default="{}")
+    initial_cash: Mapped[float] = mapped_column(Float, default=100000.0)
+    total_return: Mapped[float] = mapped_column(Float, default=0.0)
+    max_drawdown: Mapped[float] = mapped_column(Float, default=0.0)
+    sharpe: Mapped[float] = mapped_column(Float, default=0.0)
+    num_trades: Mapped[int] = mapped_column(Integer, default=0)
--- a/backend/app/schemas.py
+++ b/backend/app/schemas.py
@@ -0,0 +1,87 @@
+"""Pydantic DTO —— 这就是 OpenAPI 契约（前端据此生成类型化客户端）。
+
+契约先于业务锁定：字段一旦定下，前端可并行开发，后端实现改动不影响前端。
+"""
+from __future__ import annotations
+
+from datetime import datetime
+
+from pydantic import BaseModel, Field
+
+
+# ---------- Candle ----------
+class CandleOut(BaseModel):
+    ts: datetime
+    open: float
+    high: float
+    low: float
+    close: float
+    volume: float
+
+    model_config = {"from_attributes": True}
+
+
+# ---------- Backtest ----------
+class BacktestRequest(BaseModel):
+    symbol: str = "DEMO"
+    timeframe: str = "1d"
+    strategy: str = "macd_cross"        # macd_cross | ma_cross | single_ma
+    params: dict[str, float] = Field(default_factory=dict)  # 各策略参数
+    initial_cash: float = 1000000.0
+    fast_mode: bool = False  # True => 关闭 T+1/费用，交互试探
+    start: datetime | None = None
+    end: datetime | None = None
+
+
+class SignalOut(BaseModel):
+    ts: datetime
+    side: str   # "buy" | "sell"
+    price: float
+    qty: float
+
+
+class EquityPoint(BaseModel):
+    ts: datetime
+    value: float
+
+
+class IndicatorOut(BaseModel):
+    strategy: str
+    data: dict[str, list[float | None]] = {}  # 列名 -> 序列（MACD: macd/signal/hist；均线: fast/slow 或 ma）
+
+
+class MetricsOut(BaseModel):
+    total_return: float
+    max_drawdown: float
+    sharpe: float
+    volatility: float
+    num_trades: int = 0
+    win_rate: float = 0.0
+
+
+class BacktestResponse(BaseModel):
+    symbol: str
+    timeframe: str
+    strategy: str
+    candles: list[CandleOut]
+    indicators: IndicatorOut
+    signals: list[SignalOut]
+    equity: list[EquityPoint]
+    metrics: MetricsOut
+    final_cash: float
+    final_position: float
+    initial_cash: float
+
+
+class SyncRequest(BaseModel):
+    symbol: str
+    start: str | None = None       # YYYYMMDD
+    end: str | None = None
+    source: str = "auto"           # auto | tushare | akshare
+    force: bool = False            # True => 忽略缓存重新拉取
+
+
+class SyncResponse(BaseModel):
+    symbol: str
+    bars: int
+    source: str