Spaces:
Running
Running
File size: 8,981 Bytes
1cd56b6 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 | import sys, os
import numpy as np, pandas as pd
from scipy.stats import spearmanr
import random
import warnings; warnings.filterwarnings('ignore')
sys.path.insert(0, os.path.dirname(__file__))
from backtesting.engines.v30_causal_engine import get_data, evaluate_slice, CAP, V30_PARAMS
from backtesting.engines.v36_research_engine import SECTOR_MAP, SECTORS
from backtesting.audits.v53_causal_audit import run_v53_causal
def test_1_1_ic_analysis(dc, vf):
print("--- Test 1.1: Information Coefficient (IC) Analysis ---")
mom_long, mom_short = 175, 21
m_sig = (dc[vf].shift(mom_short) / dc[vf].shift(mom_long)) - 1
fwd = dc[vf].pct_change(60).shift(-60)
ic_vals = []
for i in range(mom_long+10, len(dc)-60, 60):
sig = m_sig.iloc[i].dropna()
z = pd.Series(index=sig.index, dtype=float)
for sector in SECTORS:
stks = [t for t in sig.index if SECTOR_MAP.get(t) == sector]
if len(stks) > 1:
mu, sigma = sig[stks].mean(), sig[stks].std()
z[stks] = (sig[stks] - mu) / (sigma if sigma > 1e-8 else 1e-8)
common = z.dropna().index.intersection(fwd.iloc[i].dropna().index)
if len(common) >= 15:
top = z[common].nlargest(15)
corr, _ = spearmanr(top.values, fwd.iloc[i][top.index].values)
if not np.isnan(corr): ic_vals.append(corr)
ic_mean = np.mean(ic_vals)
ic_std = np.std(ic_vals)
n = len(ic_vals)
t_stat = ic_mean / (ic_std / np.sqrt(n))
print(f"Mean IC: {ic_mean:.4f}")
print(f"Std IC: {ic_std:.4f}")
print(f"t-stat: {t_stat:.2f}")
if t_stat > 2.0:
print("Result: PASS (Statistically significant edge)")
else:
print("Result: FAIL/WEAK (Check signal power)")
return t_stat
def test_1_2_regime_ic_decay(dc, spy, vf):
print("\n--- Test 1.2: Regime-Conditioned IC Decay Analysis ---")
mom_long, mom_short = 175, 21
m_sig = (dc[vf].shift(mom_short) / dc[vf].shift(mom_long)) - 1
fwd = dc[vf].pct_change(60).shift(-60)
sma = spy.rolling(200).mean()
periods = {
"Period 1 (2008-2012)": ("2008-01-01", "2012-12-31"),
"Period 2 (2013-2018)": ("2013-01-01", "2018-12-31"),
"Period 3 (2019-2025)": ("2019-01-01", "2025-12-31")
}
# Calculate daily IC for top 15
ic_series = []
regimes = []
dates = []
for i in range(mom_long+10, len(dc)-60, 60):
sig = m_sig.iloc[i].dropna()
z = pd.Series(index=sig.index, dtype=float)
for sector in SECTORS:
stks = [t for t in sig.index if SECTOR_MAP.get(t) == sector]
if len(stks) > 1:
mu, sigma = sig[stks].mean(), sig[stks].std()
z[stks] = (sig[stks] - mu) / (sigma if sigma > 1e-8 else 1e-8)
common = z.dropna().index.intersection(fwd.iloc[i].dropna().index)
if len(common) >= 15:
top = z[common].nlargest(15)
corr, _ = spearmanr(top.values, fwd.iloc[i][top.index].values)
if not np.isnan(corr):
dt = dc.index[i]
ic_series.append(corr)
dates.append(dt)
regimes.append("ON" if spy.iloc[i] > sma.iloc[i] else "OFF")
ic_df = pd.DataFrame({'IC': ic_series, 'Regime': regimes}, index=dates)
for name, (start, end) in periods.items():
mask = (ic_df.index >= start) & (ic_df.index <= end)
sub = ic_df[mask]['IC']
if len(sub) > 0:
t_stat = sub.mean() / (sub.std() / np.sqrt(len(sub)))
print(f"{name}: t-stat = {t_stat:.2f} (n={len(sub)})")
print("\nBy Regime:")
for reg in ["ON", "OFF"]:
sub = ic_df[ic_df['Regime'] == reg]['IC']
if len(sub) > 0:
t_stat = sub.mean() / (sub.std() / np.sqrt(len(sub)))
print(f"Risk-{reg} days: t-stat = {t_stat:.2f} (n={len(sub)})")
def run_variant(dc, spy, vf, daily_ret, mode='random'):
rebal_days, vol_target, riskoff_haircut = 60, 0.18, 0.50
sma_lookback, mom_long, mom_short, top_n = 200, 175, 21, 15
txn_frac = 20 / 10000.0
price_mom = (dc[vf].shift(mom_short) / dc[vf].shift(mom_long)) - 1
sma = spy.rolling(sma_lookback).mean()
nav, paper_nav = CAP, CAP
peak_paper_nav, trough_paper_nav = CAP, CAP
stop_active = False
pick_tks = []
current_weights = pd.Series(dtype=float)
port_rets, hist = [], []
days = 0
for i in range(1, len(dc)):
if len(port_rets) >= 21:
w_window = port_rets[-60:] if len(port_rets) >= 60 else port_rets[-21:]
vs = vol_target / (np.std(w_window)*np.sqrt(252)+1e-8)
else: vs = 0.5
sp, sm = spy.values[i-1], sma.values[i-1]
if pd.isna(sm) or sp <= sm: vs *= riskoff_haircut
vs = float(np.clip(vs, 0.05, 1.0))
day_ret = 0.0
if pick_tks:
lr = daily_ret.iloc[i][[t for t in pick_tks if t in daily_ret.columns]].dropna()
if not lr.empty:
wt = current_weights.reindex(lr.index).fillna(0)
if wt.sum() > 0: wt = wt / wt.sum()
day_ret = (lr * wt).sum() * vs
paper_nav *= (1 + day_ret)
paper_nav -= paper_nav * txn_frac * 2 / rebal_days * vs
paper_nav = max(paper_nav, 0.01)
if not stop_active:
nav *= (1 + day_ret)
nav -= nav * txn_frac * 2 / rebal_days * vs
nav = max(nav, 0.01)
port_rets.append(day_ret)
hist.append(nav)
peak_paper_nav = max(peak_paper_nav, paper_nav)
paper_dd = (paper_nav / peak_paper_nav) - 1.0
if not stop_active:
if paper_dd <= -0.15:
stop_active = True
trough_paper_nav = paper_nav
nav -= nav * txn_frac
else:
trough_paper_nav = min(trough_paper_nav, paper_nav)
if paper_nav >= trough_paper_nav * 1.05:
stop_active = False
peak_paper_nav = paper_nav
nav -= nav * txn_frac
days += 1
if days >= rebal_days:
days = 0
avail = price_mom.iloc[i].dropna().index.tolist()
if not avail: continue
if mode == 'random':
new_picks = random.sample(avail, min(top_n, len(avail)))
current_weights = pd.Series(1.0/len(new_picks), index=new_picks)
elif mode == 'equal':
# For EW, we could buy everything, but we will pick top N just for EW equivalent or all?
# Framework says "Equal-weight all stocks in universe"
new_picks = avail
current_weights = pd.Series(1.0/len(new_picks), index=new_picks)
else:
new_picks = avail
pick_tks = new_picks
return pd.Series(hist, index=dc.index[1:len(hist)+1])
def test_1_3_null_hypothesis(dc, spy, vf, daily_ret):
print("\n--- Test 1.3: Null Hypothesis Randomization Test ---")
# Original
c_orig = run_v53_causal(dc, spy, vf, daily_ret, **V30_PARAMS)
m_orig = evaluate_slice(c_orig, "2008-01-01", "2025-12-31")
orig_sharpe = m_orig['sharpe']
print(f"Original Strategy Sharpe: {orig_sharpe:.4f}")
# Equal Weight
c_ew = run_variant(dc, spy, vf, daily_ret, mode='equal')
m_ew = evaluate_slice(c_ew, "2008-01-01", "2025-12-31")
ew_sharpe = m_ew['sharpe']
print(f"Equal-Weight Universe Sharpe: {ew_sharpe:.4f}")
# Random (Run 100 iterations as per framework)
rand_sharpes = []
print("Running 100 Random iterations...", end="", flush=True)
for _ in range(100):
c_rand = run_variant(dc, spy, vf, daily_ret, mode='random')
m_rand = evaluate_slice(c_rand, "2008-01-01", "2025-12-31")
rand_sharpes.append(m_rand['sharpe'])
if _ % 10 == 0: print(".", end="", flush=True)
print()
rand_median = np.median(rand_sharpes)
print(f"Random Picks Median Sharpe: {rand_median:.4f}")
print(f"Random Picks Max Sharpe: {np.max(rand_sharpes):.4f}")
excess = orig_sharpe - rand_median
print(f"\nExcess Sharpe vs Random: +{excess:.4f}")
if excess > 0.15:
print("Result: PASS (Signal adds strong value beyond risk management)")
elif excess > 0.10:
print("Result: WEAK PASS (Signal adds marginal value)")
else:
print("Result: FAIL (Signal has no edge over random/risk management)")
if __name__ == "__main__":
print("========================================")
print(" V53 FRAMEWORK VALIDATION - PHASE 1")
print("========================================")
dc, spy, vf, daily_ret = get_data()
test_1_1_ic_analysis(dc, vf)
test_1_2_regime_ic_decay(dc, spy, vf)
test_1_3_null_hypothesis(dc, spy, vf, daily_ret)
|