FoxAlgo → Research → Academic Audit → G-Score Investing

Academic audit

Failedequity XS fundamental

G-Score Investing

The three-gate gauntlet · genuine only if it clears all three and survives adversarial refutation
Gate 1
Survivorship-free
free
clean panel
Eliminated here
Gate 2
Placebo ≥ P95
P1
outranked ~2 of 200 baskets
Gate 3
Cost-aware net
RF -0.91
net-negative after costs
Failed
Worst 12-month leg (RF)-0.84
−1.00 floor0
Every strategy here — winners included — loses in its worst 12 months. Depth is honest context, not the verdict.
Rejected at the luck gate — its net ranked no better than random baskets (below the P95 skill line).

The G-score is a fundamentals-based growth-quality score from Mohanram (2005) that ranks stocks on profitability, earnings stability, and growth signals, buying high-scoring names and shorting low-scoring ones.

What we found

Rebuilt as a survivorship-free long/short and run through realistic costs and delisting outcomes, the G-score showed no growth-quality edge. Its risk-adjusted RF is negative (-0.91) and its worst year is also negative (-0.84), so it neither ranked stocks profitably overall nor held up in a bad year. The placebo test placed the real result at the 1.0 percentile, meaning random baskets did about as well or better, so what remained after costs and delisted names is consistent with noise rather than a factor.

How we tested it
2005–2026 test windowmodelled liquidity-aware costssurvivorship free
  • Data: survivorship-free 1077-name US common-stock panel, 2005-2026. Realistic modelled costs.
  • Placebo / robustness test: real result vs random baskets or shuffled signals (real vs the 95th percentile of random)
Source: Mohanram (2005), "Separating Winners from Losers... the G-score", Review of Accounting Studies
Read the paper ↗
← The Academic Audit — all 54 studies

Research, not investment advice. “Validated” factor-legs are market-neutral diversifying building blocks with a losing worst year — none is a standalone tradeable strategy. Metrics are cost-aware and modelled (not live fills); the 2005–2026 test window is out-of-sample versus the source paper. Dollar figures are not returns and are omitted by design.