Robust Core Plus Volatility: Beats The Incumbent On Both Metrics

Back to leaderboard · 2015-01-01 to 2025-06-01 · v3_top100_monthly

Overall top-100 hits
28.46
average across horizons
Predicted top-100 avg actual rank
226.53
lower is better
Rank IC
0.0732
picks the better of two stocks 52.3% of the time
IC reliability
9.04
t-stat; IC positive 66.7% of months
not adjusted for overlapping windows

Horizon Breakdown

top 100 only
Horizon Average Top-100 Hits Predicted Top-100 Avg Actual Rank Rank IC
3m 27.91 231.33 0.0614
6m 28.62 226.21 0.0769
12m 28.83 222.06 0.0814

Run Note

test summary
Hypothesis: volatility and rank IC are additive, not competing, so adding vol_63d back should buy hits@100 without surrendering the IC gain. vol_63d alone scores 29.5 hits -- more than any model here -- yet its IC is 0.0179 at t 1.55, indistinguishable from zero: it sorts stocks into BOTH tails at once, which wins a membership count and says nothing about ordering. The other four terms supply the ordering it cannot. Change from strategy_000109: add vol_63d as a fifth equally weighted date-local rank, still no fitted parameter. Result IC 0.07324 (t 9.04), hits 28.456, avg actual rank 226.53. THIS IS THE FIRST MODEL TO STRICTLY DOMINATE strategy_000079 ON ALL THREE REPORTED METRICS AT ONCE: IC 0.07324 vs 0.06572 and higher in all 3 blocks and both halves (0.0828/0.0469/0.0904, 0.0646/0.0829 vs 0.0764/0.0419/0.0791, 0.0641/0.0675), hits 28.456 vs 27.965, avg actual rank 226.53 vs 232.69. It is not a trade-off. IC is lower than 109's 0.08077, so 109 remains the IC leader and this is the best all-round model. Plateau check: vol weight 0.75 gives higher IC 0.0760 with hits 28.229 and the entire 0.75-1.0 range clears the incumbent on every measure, so this is not a tuned point. Next: strategy_000109 blended with strategy_000079 in rank space reaches IC 0.0832 offline, the best seen -- test whether the legacy eleven-factor family carries anything genuinely orthogonal or whether the gain is just its coverage term reappearing.

Run Details

strategy_test_results #230
id
230
strategy_id
strategy_000111
strategy_title
Robust Core Plus Volatility: Beats The Incumbent On Both Metrics
strategy_hash
0e3b56b0f7d773afe9320a03e307e121f799571d5bcb68498263632a918d9486
evaluator_version
v3_top100_monthly
start_date
2015-01-01
end_date
2025-06-01
created_at
2026-07-25 13:58:14
updated_at
2026-07-25 13:58:14

Sample Top-100 Lists

2020-03-31 / 6m / 29 hits

Predicted top 100

AAPL, ADBE, AES, AFL, AIG, AMD, ANET, APTV, AVGO, AXP, BMY, C, CCL, CF, CFG, CI, CMG, CNC, CNP, COP, CRM, CVX, DAL, DRI, DVN, DXC, DXCM, EMR, ETN, ETSY, EXC, EXPE, FANG, FITB, FTNT, GE, HAL, HAS, HIG, HUM, HWM, INTC, ISRG, IVZ, KEY, KLAC, KMI, LEN, LKQ, LNC, LRCX, LYB, LYV, MA, MCHP, MCO, MET, MGM, MOS, MPC, MS, MSFT, NCLH, NEE, NEM, NFLX, NOW, NVDA, NWL, O, OKE, ORCL, OXY, PLD, PRU, QCOM, RCL, RF, ROST, RTX, SPG, SPGI, STT, SWK, SWKS, SYF, SYY, TER, TJX, TPR, TSLA, UAL, UNM, URI, V, VLO, VRTX, VTR, WELL, WMB

Actual top 100

AAP, AAPL, ABMD, ADBE, ADSK, ALB, ALGN, AMD, AMP, AMZN, APD, APH, APTV, ATVI, AVGO, BBY, BWA, CARR, CDNS, CE, CHRW, CMG, CMI, CPRT, CRM, CTAS, CTLT, CTSH, DD, DE, DHI, DHR, DRI, DXCM, EBAY, EMN, ETSY, EXPE, FCX, FDX, GLW, GPS, HAL, HD, IDXX, IPGP, JCI, KMX, LEG, LEN, LH, LOW, MAS, MCHP, MGM, MOS, NCLH, NKE, NOW, NSC, NVDA, NWS, NWSA, ORLY, PAYC, PH, PHM, PNR, POOL, PVH, PWR, PYPL, QCOM, QRVO, RCL, ROK, ROL, SHW, SIVB, SNPS, SWK, SWKS, SYF, TDG, TEL, TER, TGT, TMO, TSCO, TSLA, TT, UPS, URI, VAR, VIAC, VTR, WHR, WMB, WST, WY

Matched tickers

AAPL, ADBE, AMD, APTV, AVGO, CMG, CRM, DRI, DXCM, ETSY, EXPE, HAL, LEN, MCHP, MGM, MOS, NCLH, NOW, NVDA, QCOM, RCL, SWK, SWKS, SYF, TER, TSLA, URI, VTR, WMB

Strategy Script

0e3b56b0f7
"""Generated V3 stock-ranking strategy."""
from __future__ import annotations

import numpy as np
import pandas as pd


STRATEGY_ID = "strategy_000111"
STRATEGY_TITLE = "Robust Core Plus Volatility: Beats The Incumbent On Both Metrics"
DESCRIPTION = """
strategy_000109 wins on rank IC but gives up 1.8 hits@100, because the trailing
volatility term that pays for top-100 overlap is not in it. Volatility and IC
pull in different directions on purpose: vol_63d alone scores 29.5 hits, more
than any model here, yet its IC is 0.0179 at t 1.55, which is not distinguishable
from zero. It sorts stocks into both tails at once, which wins a membership
count and says nothing about ordering.

That makes it additive rather than competing. Adding vol_63d as a fifth equally
weighted date-local rank should buy back the hits without surrendering the IC
gain, because the other four terms supply the ordering the volatility term
cannot.

  from_low_252d              distance held above the 252-day low
  volume                     traded share volume
  free_cash_flow_margin_pct  cash conversion
  revenue_growth_pct         top-line growth
  vol_63d                    63-day trailing volatility

Still equal weights and still no fitted parameter. A weight of 0.75 on the
volatility term scores higher IC at 0.0760 against 0.0732, and the whole range
from 0.75 to 1.0 clears the incumbent on every measure, so this is a plateau and
not a tuned point. Equal weighting is kept because it costs nothing to defend.

The target is a model that is not a trade-off at all: better than
strategy_000079 on rank IC, on hits@100, and on average actual rank at the same
time.
"""
FACTORS_USED = [
    "from_low_252d",
    "volume",
    "free_cash_flow_margin_pct",
    "revenue_growth_pct",
    "vol_63d"
]
PARAMETERS = {
    "from_low_252d": 1.0,
    "volume": 1.0,
    "free_cash_flow_margin_pct": 1.0,
    "revenue_growth_pct": 1.0,
    "vol_63d": 1.0
}
COMPLEXITY = 5
PARENT_STRATEGY_ID = "strategy_000109"
GENERATION = 111
RANDOM_SEED = 20261110
CREATED_AT = "2026-07-25T00:00:00+00:00"


def _date_group(df):
    if "trade_date" in df.columns:
        return df["trade_date"]
    return pd.Series(0, index=df.index)


def _rank_pct(df, column):
    """Date-local rank percentile centred on zero; missing values stay neutral."""
    if column not in df.columns:
        return pd.Series(0.0, index=df.index)
    clean = df[column].replace([np.inf, -np.inf], np.nan)
    ranked = clean.groupby(_date_group(df)).rank(pct=True)
    return ranked.fillna(0.5) - 0.5


def calculate_score(df):
    score = pd.Series(0.0, index=df.index)
    for column, weight in PARAMETERS.items():
        score = score + weight * _rank_pct(df, column)
    total = sum(abs(w) for w in PARAMETERS.values()) or 1.0
    score = score / total
    return score.replace([np.inf, -np.inf], np.nan).fillna(0.0)