#!/usr/bin/env python3
# =============================================================================
# Author: Simon-Pierre Boucher
# Contact: contact@spboucher.ai
# =============================================================================
"""Step 09 — Portfolio sorts and economic significance.

Daily equal-weighted quintile sorts on each option-implied variable
(long-short Q5−Q1 performance), a 3×3 double sort on ATM IV × implied
skewness, and transaction-cost sensitivity of the implied-skewness strategy.

Inputs : data/processed/merged_options_rv.parquet
Outputs: results/portfolio_sort_results.csv, results/double_sort_iv_skew.csv,
         results/transaction_cost_analysis.csv
"""

import warnings

import numpy as np
import pandas as pd

import _bootstrap  # noqa: F401
from wp7 import config
from wp7.data_io import load_merged
from wp7.portfolio import double_sort, portfolio_sort

warnings.filterwarnings('ignore')


def main():
    print("=" * 70)
    print("PORTFOLIO SORTS & ECONOMIC SIGNIFICANCE")
    print("=" * 70)
    config.ensure_output_dirs()

    df = load_merged()
    stocks = df[~df['ticker'].isin(config.NON_STOCK_TICKERS)].copy()

    # ── Single sorts ──
    all_sort_results = []
    for sort_var, sort_label in config.SORT_VARIABLES.items():
        for ret_var, ret_label in [('ret_1d', '1-Day'), ('ret_5d', '5-Day')]:
            res = portfolio_sort(stocks, sort_var, ret_var, n_quantiles=5)
            if res is None:
                continue

            print(f"\n  {sort_label} → {ret_label} Returns:")
            print(f"  {'Q':>6} {'Mean(bps)':>10} {'Ann.Ret%':>10} {'Ann.Vol%':>10} "
                  f"{'Sharpe':>8} {'t-stat':>8} {'N':>6}")

            for q in [1, 2, 3, 4, 5, 'LS_5_1']:
                if q not in res:
                    continue
                r = res[q]
                q_label = f"Q{q}" if isinstance(q, int) else "L/S(5-1)"
                print(f"  {q_label:>6} {r['mean_daily']*10000:>10.2f} "
                      f"{r['annualized_return']*100:>10.2f} "
                      f"{r['annualized_vol']*100:>10.2f} {r['sharpe']:>8.3f} "
                      f"{r['t_stat']:>8.3f} {r['n_days']:>6}")

                all_sort_results.append({
                    'sort_variable': sort_label,
                    'return_horizon': ret_label,
                    'quintile': q_label,
                    'mean_daily_bps': r['mean_daily'] * 10000,
                    'annualized_return_pct': r['annualized_return'] * 100,
                    'annualized_vol_pct': r['annualized_vol'] * 100,
                    'sharpe_ratio': r['sharpe'],
                    't_statistic': r['t_stat'],
                    'n_days': r['n_days'],
                    'pct_positive': r['pct_positive'],
                    'max_drawdown_pct': r['max_drawdown'] * 100,
                })

    sort_df = pd.DataFrame(all_sort_results)
    sort_df.to_csv(config.RESULTS_DIR / "portfolio_sort_results.csv", index=False)

    print("\n" + "=" * 70)
    print("LONG-SHORT PORTFOLIO SUMMARY (Q5 - Q1)")
    print("=" * 70)
    ls = sort_df[sort_df['quintile'] == 'L/S(5-1)'].copy()
    print(ls[['sort_variable', 'return_horizon', 'mean_daily_bps',
              'annualized_return_pct', 'sharpe_ratio', 't_statistic']]
          .round(3).to_string(index=False))

    # ── Double sort: ATM IV × implied skewness → 5-day returns ──
    print("\n" + "=" * 70)
    print("DOUBLE SORT: IV_ATM x IMPLIED_SKEWNESS → 5-Day Returns")
    print("=" * 70)
    ds = stocks[['iv_atm_30d', 'implied_skewness', 'ret_5d', 'ticker', 'trade_date']].dropna()
    ds_results = double_sort(ds, 'iv_atm_30d', 'implied_skewness', 'ret_5d')

    ds_pivot = ds_results.pivot(index='q1', columns='q2', values='mean_bps')
    ds_pivot.index = ['Low IV', 'Med IV', 'High IV']
    ds_pivot.columns = ['Low Skew', 'Med Skew', 'High Skew']
    print(ds_pivot.round(2).to_string())

    ds_t = ds_results.pivot(index='q1', columns='q2', values='t_stat')
    ds_t.index = ['Low IV', 'Med IV', 'High IV']
    ds_t.columns = ['Low Skew', 'Med Skew', 'High Skew']
    print("\nt-statistics:")
    print(ds_t.round(3).to_string())

    ds_results.to_csv(config.RESULTS_DIR / "double_sort_iv_skew.csv", index=False)

    # ── Transaction-cost sensitivity (implied-skewness L/S, weekly) ──
    print("\n" + "=" * 70)
    print("TRANSACTION COST SENSITIVITY (Long-Short on Implied Skewness, 5D)")
    print("=" * 70)
    res_skew = portfolio_sort(stocks, 'implied_skewness', 'ret_5d', n_quantiles=5)
    if res_skew and 'LS_5_1' in res_skew:
        ls_gross = res_skew['LS_5_1']
        print(f"  {'TC (bps)':>10} {'Net Ret(bps)':>12} {'Ann.Ret%':>10} {'Sharpe':>8}")
        tc_results = []
        for tc_bps in [0, 5, 10, 15, 20, 30, 50]:
            # Weekly rebalance: full two-sided turnover spread over 5 days
            turnover_per_day = 2.0 / 5
            daily_tc = tc_bps / 10000 * turnover_per_day
            net_daily = ls_gross['mean_daily'] - daily_tc
            net_ann = net_daily * 52
            net_sharpe = (net_daily / ls_gross['std_daily'] * np.sqrt(52)) \
                if ls_gross['std_daily'] > 0 else 0
            print(f"  {tc_bps:>10} {net_daily*10000:>12.2f} "
                  f"{net_ann*100:>10.2f} {net_sharpe:>8.3f}")
            tc_results.append({'tc_bps': tc_bps, 'net_daily_bps': net_daily * 10000,
                               'net_ann_ret_pct': net_ann * 100, 'net_sharpe': net_sharpe})
        pd.DataFrame(tc_results).to_csv(
            config.RESULTS_DIR / "transaction_cost_analysis.csv", index=False)

    print("\nPORTFOLIO SORTS COMPLETE.")


if __name__ == "__main__":
    main()