run_phase7.py

"""
run_phase7.py  (Phase 7e report)
================================
End-to-end Phase 7 report:
  1. Build the real SX5E universe (BL data layer) + archetype ESG profiles
  2. Show the PAB exclusion outcome on the real benchmark
  3. Run the three compliance scenarios (cost of each path)
  4. Compute and plot the cost-of-green frontier

Run:  python3 run_phase7.py   (pulls live yfinance data; needs network)
Outputs: console report + cost_of_green_frontier.png
"""
from __future__ import annotations

import numpy as np
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt

from phase7.esg_data import build_universe
from phase7.esg_optimiser import (run_scenarios, cost_of_green_frontier,
                                  benchmark_waci)
from sfdr2.exclusions import screen_portfolio

MAX_W = 0.12   # UCITS-style per-name cap; also regularises the optimiser
YEAR = 2026


def main():
    print("Building real SX5E universe via the BL data layer...")
    b = build_universe(years=3)
    tickers = b["tickers"]

    print("=" * 70)
    print("PHASE 7 - ESG-TILTED MINIMUM-TRACKING-ERROR OPTIMISER")
    print("=" * 70)

    # --- exclusion outcome on the real benchmark --------------------------
    scr = screen_portfolio(b["holdings"])
    print("\n[1] PAB exclusions on the real SX5E benchmark")
    print("-" * 70)
    for r in scr["results"]:
        if r.excluded:
            why = ", ".join(x["screen"] for x in r.breaches)
            print(f"   EXCLUDED  {r.issuer:<20} ({why})")
    s = scr["summary"]
    print(f"   {s['n_excluded']} excluded ({s['weight_excluded_pct']:.1f}% of weight); "
          f"{s['weight_clear_pct']:.1f}% clear")

    # --- three compliance scenarios ---------------------------------------
    sc = run_scenarios(b, year=YEAR, max_weight=MAX_W)
    print(f"\n[2] Compliance-path cost (carbon target = 1.5C pathway "
          f"{sc['pathway_target']:.1f} tCO2e/EURm; {int(MAX_W*100)}% name cap)")
    print("-" * 70)
    print(f"   Parent benchmark WACI: {sc['parent_waci']:.1f} tCO2e/EURm")
    print(f"   {'Scenario':<24}{'TE (bps)':>10}{'WACI':>8}{'Qual wt':>9}{'Feasible':>10}")
    for k, r in sc["results"].items():
        print(f"   {r['label']:<24}{r['te_bps']:>10.1f}{r['waci']:>8.1f}"
              f"{r['qualifying_weight']*100:>8.0f}%{('Yes' if r['success'] else 'No'):>10}")

    # --- cost-of-green frontier -------------------------------------------
    print("\n[3] Cost-of-green frontier")
    print("-" * 70)
    df = cost_of_green_frontier(b, year=YEAR, n_points=25, floor=None,
                                max_weight=MAX_W)
    feas = df[df.feasible]
    print(f"   Parent WACI {df.attrs['parent_waci']:.1f} -> "
          f"pathway target {df.attrs['pathway_target']:.1f} tCO2e/EURm")
    print(f"   {len(feas)} feasible points; "
          f"TE {feas.tracking_error_bps.min():.0f}-{feas.tracking_error_bps.max():.0f} bps")
    print(f"   {'Intensity cut %':>16}{'Achieved WACI':>15}{'TE (bps)':>11}")
    for _, row in feas.iloc[::3].iterrows():
        print(f"   {row['intensity_cut_pct']:>16.1f}{row['achieved_waci']:>15.1f}"
              f"{row['tracking_error_bps']:>11.1f}")

    # --- plot --------------------------------------------------------------
    fig, ax = plt.subplots(figsize=(8.5, 5.2))
    ax.plot(feas["intensity_cut_pct"], feas["tracking_error_bps"],
            "o-", color="#1E2761", lw=2, ms=5, label="Min-TE frontier")
    # mark the pathway-aligned point
    parent = df.attrs["parent_waci"]
    tgt = df.attrs["pathway_target"]
    pathway_cut = 100.0 * (parent - tgt) / parent
    ax.axvline(pathway_cut, color="#C47D00", ls="--", lw=1.5,
               label=f"1.5C pathway ({pathway_cut:.0f}% cut)")
    ax.set_xlabel("Carbon-intensity reduction vs parent benchmark (%)")
    ax.set_ylabel("Tracking error (bps p.a.)")
    ax.set_title("The cost of green: active risk vs decarbonisation\n"
                 "SX5E universe, minimum tracking error, 12% name cap",
                 fontsize=12, fontweight="bold")
    ax.grid(alpha=0.25)
    ax.legend()
    fig.tight_layout()
    fig.savefig("cost_of_green_frontier.png", dpi=130)
    print("\n   Saved chart: cost_of_green_frontier.png")
    print("-" * 70)


if __name__ == "__main__":
    main()