{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1e495ff0",
   "metadata": {},
   "source": [
    "---\n",
    "title: Momentum + Value — a Strategy That Survives Out of Sample\n",
    "summary: Short-horizon momentum combined with value, in a wide dollar-neutral book — a real edge that holds up out of sample (most of the time).\n",
    "tags: [momentum, value, multifactor, out-of-sample]\n",
    "---\n",
    "\n",
    "# Momentum + Value — a Strategy That Survives Out of Sample\n",
    "\n",
    "Most of the classic single factors look great in a backtest and then evaporate out of sample (see the\n",
    "other examples). This one is different: **short-horizon momentum** (last month's winners keep winning,\n",
    "briefly) blended with **value** (cheap beats expensive) is one of the most durable combinations in\n",
    "equities. We hold a *wide* dollar-neutral book and let weekly rebalancing keep turnover in check — and\n",
    "we check that the edge **survives out of sample across many markets**, not just one lucky draw."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ee44e66e",
   "metadata": {},
   "outputs": [],
   "source": [
    "try:\n",
    "    import convexpi.lab  # noqa\n",
    "except ImportError:\n",
    "    import subprocess, sys\n",
    "    subprocess.run([sys.executable, \"-m\", \"pip\", \"install\", \"-q\", \"convexpi-lab\"])\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "print(\"ready\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2606fef3",
   "metadata": {},
   "source": [
    "## The strategy\n",
    "\n",
    "Combine two real edges into one score and trade the spread. We weight short-horizon momentum\n",
    "(`mom_1m`) a bit more than value (`val_bm`), and go long the top half / short the bottom half — a wide\n",
    "book diversifies away single-name noise, and weekly rebalancing (the grader's default) keeps the\n",
    "high-turnover momentum leg from being eaten by costs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "668193d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from convexpi.lab import Strategy\n",
    "\n",
    "class MyStrategy(Strategy):\n",
    "    \"\"\"Short-horizon momentum + value, dollar-neutral, wide book (top/bottom half).\"\"\"\n",
    "    weights = {\"mom_1m\": 1.0, \"val_bm\": 0.5}\n",
    "    frac = 0.5\n",
    "\n",
    "    def on_day(self, day, features, prices, portfolio):\n",
    "        n = len(prices)\n",
    "        score = np.zeros(n)\n",
    "        for name, wt in self.weights.items():\n",
    "            score += wt * np.nan_to_num(features.get(name, np.zeros(n)))\n",
    "        k = max(1, int(n * self.frac))\n",
    "        order = np.argsort(score)\n",
    "        w = np.zeros(n)\n",
    "        w[order[-k:]] = 1.0 / k     # long the strongest\n",
    "        w[order[:k]] = -1.0 / k     # short the weakest\n",
    "        return w"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "964f44f9",
   "metadata": {},
   "source": [
    "## Does it survive? Check across many markets\n",
    "\n",
    "A single backtest can be lucky. The honest test is **robustness**: evaluate out of sample on many\n",
    "independent synthetic markets and look at the *distribution* of OOS Sharpe — not one number."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a1d95c3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from convexpi.lab import SyntheticMarket, Grader\n",
    "\n",
    "oos = []\n",
    "for seed in range(1, 21):\n",
    "    r = Grader(SyntheticMarket(n_stocks=200, n_days=1260, seed=seed)).evaluate(MyStrategy())\n",
    "    oos.append(r.oos_sharpe)\n",
    "oos = np.array(oos)\n",
    "\n",
    "print(f\"out-of-sample Sharpe across 20 markets:\")\n",
    "print(f\"  mean      : {oos.mean():+.2f}\")\n",
    "print(f\"  win rate  : {np.mean(oos > 0):.0%}  (fraction of markets with positive OOS)\")\n",
    "print(f\"  range     : [{oos.min():+.2f}, {oos.max():+.2f}]\")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8, 3))\n",
    "ax.hist(oos, bins=10, color=\"steelblue\"); ax.axvline(0, color=\"grey\", lw=1)\n",
    "ax.axvline(oos.mean(), color=\"darkorange\", lw=2, label=f\"mean {oos.mean():+.2f}\")\n",
    "ax.set_title(\"OOS Sharpe distribution (20 markets)\"); ax.set_xlabel(\"OOS Sharpe\"); ax.legend()\n",
    "plt.tight_layout(); plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20eb3628",
   "metadata": {},
   "source": [
    "## The honest part\n",
    "\n",
    "Positive **on average and most of the time** — but look at the left tail: some markets still hand it a\n",
    "losing draw. That's the truth about even good strategies: an edge is a *tilt in the distribution*, not\n",
    "a guarantee. The permanent leaderboard grades this on one hidden market, so the badge you see is one\n",
    "draw from that histogram.\n",
    "\n",
    "## What I'd try next\n",
    "\n",
    "- Add a third lightly-weighted factor and see if it tightens the distribution (lifts the left tail).\n",
    "- Vary the book width (`frac`) and rebalance horizon to trade off noise vs. turnover cost.\n",
    "- Compare against the single-factor examples — diversification is what turns a coin-flip into an edge."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}