Rewrite as a probabilistic model with walk-forward evaluation.

Replace the 2024 model (model.py, ~2000 lines) with the btcmodel package, the
baseline for future work:

- Forecasts are quantiles of log price at each horizon, scored with CRPS in a
  walk-forward backtest (origins every 30 days from 2014, horizons 1 month to
  4 years). Skill is relative to a zero-drift random walk, with circular
  block-bootstrap intervals and a count of independent windows.
- Development data stops at 2024-11-26, the last day the 2024 model saw.
  Later outcomes are a holdout, scored only by `backtest --holdout`.
- Models: random_walk, drift_rw, and cycle (the 2024 model's cycle-position
  drift, now kernel-smoothed and recency-weighted). On development data
  nothing beats the random walk with confidence; cycle loses at every horizon.
- Prices: the Investing.com archive moves to data/ (cut at 2024-11-26; its
  last row was intraday) and is extended with Coinbase daily closes by
  `update`.

Also: Nix flake dev shell (Python 3.13, pandas 3), ruff in place of black,
pytest suite, and a rewritten README. NOTES.md is removed as inaccurate, and
poetry is dropped.

btcmodel/models/__init__.py

@@ -0,0 +1,15 @@
+"""
+Candidate models.
+
+A model is any object with a `name` and a
+`forecast(history: pd.DataFrame, horizons: np.ndarray) -> Forecast` method.
+`history` holds every row up to and including the forecast origin and nothing
+after it; the harness guarantees that, so models can use all of it freely.
+"""
+
+from .baselines import DriftRandomWalk, RandomWalk
+from .cycle import CycleModel
+
+# Order is fixed: it sets each model's colour in every chart.
+MODELS = {m.name: m for m in (RandomWalk(), DriftRandomWalk(), CycleModel())}
+BASELINE = RandomWalk.name

btcmodel/models/baselines.py

@@ -0,0 +1,47 @@
+"""Reference forecasts every other model has to beat."""
+
+from dataclasses import dataclass
+from typing import ClassVar
+
+import numpy as np
+import pandas as pd
+
+from ..data import log_returns
+from ..forecast import Forecast
+
+
+@dataclass(frozen=True)
+class RandomWalk:
+    """
+    Zero-drift random walk in log price: "it stays about here, give or take".
+
+    Volatility is the trailing standard deviation of daily log returns.
+    """
+
+    name: ClassVar[str] = "random_walk"
+    vol_window: int = 365
+
+    def forecast(self, history: pd.DataFrame, horizons: np.ndarray) -> Forecast:
+        sigma = log_returns(history).iloc[-self.vol_window :].std()
+        mean = np.log(history["close"].iloc[-1])
+        return Forecast.normal(history.index[-1], horizons, mean, sigma * np.sqrt(horizons))
+
+
+@dataclass(frozen=True)
+class DriftRandomWalk:
+    """
+    Random walk whose drift is the mean daily log return over the trailing
+    `drift_window` days (one halving cycle by default): "it keeps doing what it
+    did last cycle".
+    """
+
+    name: ClassVar[str] = "drift_rw"
+    drift_window: int = 1460
+    vol_window: int = 365
+
+    def forecast(self, history: pd.DataFrame, horizons: np.ndarray) -> Forecast:
+        returns = log_returns(history)
+        mu = returns.iloc[-self.drift_window :].mean()
+        sigma = returns.iloc[-self.vol_window :].std()
+        mean = np.log(history["close"].iloc[-1]) + mu * horizons
+        return Forecast.normal(history.index[-1], horizons, mean, sigma * np.sqrt(horizons))

btcmodel/models/cycle.py

@@ -0,0 +1,73 @@
+"""The 2024 model, distilled."""
+
+from dataclasses import dataclass
+from typing import ClassVar
+
+import numpy as np
+import pandas as pd
+
+from ..data import log_returns
+from ..forecast import Forecast
+from ..halving import cycle_position
+
+# Longer than any cycle so far (the longest, cycle 0, is 1425 days).
+MAX_CYCLE_DAYS = 1500
+
+
+@dataclass(frozen=True)
+class CycleModel:
+    """
+    Expected return depends on how many days it has been since the last halving.
+
+    The drift for day d of the cycle is a weighted mean of the daily log returns
+    observed around day d of every past cycle. Of the 2024 model's ~2000 lines,
+    this idea did all the work. It differs from that model in two ways:
+
+    - Neighbouring cycle days are pooled with a Gaussian kernel. The 2024 model
+      averaged each day separately and then took a rolling mean.
+    - Past cycles are down-weighted, halving each `recency_half_life` cycles, so
+      the 10-100x cycles of 2011-2017 don't set the level. The 2024 model
+      averaged all cycles equally, then scaled by ~0.7; it overshot the
+      2025 peak by ~60%. (This is the idea on the old `tuning-b` branch.)
+
+    Where the data is thin, the drift shrinks toward the overall weighted mean,
+    as if `prior_days` extra observations sat at that value. Noise is a
+    constant-volatility random walk, so the distribution is closed-form.
+    """
+
+    name: ClassVar[str] = "cycle"
+    bandwidth_days: float = 30.0
+    recency_half_life: float = 1.0
+    prior_days: float = 10.0
+    vol_window: int = 365
+
+    def drift_by_cycle_day(self, history: pd.DataFrame) -> np.ndarray:
+        """Expected daily log return for each day of the cycle, shape (MAX_CYCLE_DAYS,)."""
+        returns = log_returns(history)
+        cycle, day = cycle_position(returns.index)
+        current_cycle = cycle_position(history.index[-1:])[0][0]
+        weight = 0.5 ** ((current_cycle - cycle) / self.recency_half_life)
+
+        sum_wr = np.bincount(day, weights=weight * returns.values, minlength=MAX_CYCLE_DAYS)
+        sum_w = np.bincount(day, weights=weight, minlength=MAX_CYCLE_DAYS)
+
+        # Peak-1 kernel, so smoothed weights count (recency-weighted) days of data.
+        half_width = int(np.ceil(4 * self.bandwidth_days))
+        offsets = np.arange(-half_width, half_width + 1)
+        kernel = np.exp(-0.5 * (offsets / self.bandwidth_days) ** 2)
+        smooth_wr = np.convolve(sum_wr, kernel, mode="same")
+        smooth_w = np.convolve(sum_w, kernel, mode="same")
+
+        overall = sum_wr.sum() / sum_w.sum()
+        return (smooth_wr + self.prior_days * overall) / (smooth_w + self.prior_days)
+
+    def forecast(self, history: pd.DataFrame, horizons: np.ndarray) -> Forecast:
+        drift = self.drift_by_cycle_day(history)
+        origin = history.index[-1]
+        future = origin + pd.to_timedelta(np.arange(1, horizons.max() + 1), unit="D")
+        _, future_day = cycle_position(future)
+        cumulative = np.cumsum(drift[future_day])
+
+        sigma = log_returns(history).iloc[-self.vol_window :].std()
+        mean = np.log(history["close"].iloc[-1]) + cumulative[horizons - 1]
+        return Forecast.normal(origin, horizons, mean, sigma * np.sqrt(horizons))