← All tasks
T03 · Momentum

Fundamentally, Momentum Is Fundamental Momentum

Robert Novy-Marx, 2015 · National Bureau of Economic Research

0%
pass@1 across 16 systems
0/16
systems with ≥1 successful attempt
0/5
no system produced a successful attempt
80 attempts · 16 systems × 5
Claude Code + Opus 5
0%
Claude Code + Opus 4.8
0%
Claude Code + Opus 4.7
0%
Codex + GPT-5.6-sol
0%
Codex + GPT-5.5
0%
OpenCode + Kimi-K3
0%
OpenCode + Opus 4.7
0%
OpenCode + GPT-5.5
0%
OpenCode + Sonnet 4.6
0%
OpenCode + Qwen3.7-Max
0%
OpenCode + GLM-5.2
0%
OpenCode + DeepSeek V4-Pro
0%
OpenCode + GLM-5.1
0%
OpenCode + Qwen 3.6 Plus
0%
OpenCode + MiniMax-M3
0%
OpenCode + MiniMax M2.7
0%
all checkers pass Type-A clean, Type-B fail fail checking timeout no result / incomplete
Strategy

What the paper asks for.

A Fama-French-style 2x3 SUE (Standardized Unexpected Earnings) factor captures fundamental-momentum: independent monthly sorts on NYSE-median market equity (size) and NYSE 30/70 SUE breakpoints generate six value-weighted portfolios, and the long-short factor = 0.5 * (LARGE_HIGH - LARGE_LOW) + 0.5 * (SMALL_HIGH - SMALL_LOW).

Type
fundamental
Implementation
pandas
Rebalance
monthly
Long-short
yes
Declared task setup
Data
One-command fetcher for WRDS-sourced task data
Operational scope
The paper-defined methodological core is retained; the fixed instruction instantiates only the data mappings, interfaces, and evaluation conventions needed for execution.
Six-stage quantitative strategy pipeline
  1. 01 Data loading
  2. 02 Feature engineering
  3. 03 Model fitting
  4. 04 Signal generation
  5. 05 Portfolio construction
  6. 06 Execution
Checker suite

19 properties · 80 attempts.

Each row is one requirement the strategy must satisfy. The bar reports the share of observed checker verdicts that pass; missing checker output is not converted into a property verdict.

TYPE A · DOMAIN-LEVEL (9)
  • A1
    no backfill
    §§2.1 “SUE is defined as the most recent year-over-year change in earnings per share, scaled by the standard deviation of the earnings innovations over the last eight announcements, subject to a requirement of at least six observed announcements over the two year window.”
    stage · data loading
    63/77
  • A2
    no global normalize
    §§2.1 “scaled by the standard deviation of the earnings innovations over the last eight announcements, subject to a requirement of at least six observed announcements over the two year window.”
    stage · feature eng
    61/77
  • A3
    no future data access
    §§2.1 “SUE is defined as the most recent year-over-year change in earnings per share”
    stage · feature eng
    60/77
  • A4
    causal consistency
    §§3 “Portfolios are rebalanced monthly, and returns are value weighted.”
    stage · signal gen
    54/77
  • A5
    no survivorship bias
    §§3 (Table 2 scope) “The full sample covers January 1975 through December 2012, dates determined by the data requirements for making the SUE and CAR3 strategies.”
    stage · portfolio
    48/77
  • A7
    data frequency alignment
    §§3 “these factors are constructed from underlying portfolios that are formed monthly, as the intersection of two size and three earnings momentum portfolios.”
    stage · data loading
    68/77
  • A8
    transaction timing
    §§3 “Portfolios are rebalanced monthly, and returns are value weighted.”
    stage · execution
    65/77
  • A9
    signal delay test
    §§3 “Portfolios are rebalanced monthly, and returns are value weighted.”
    stage · signal gen
    66/77
  • A10
    e2e metric consistency
    §Table 2 Panel B “Panel B: y D SUE α 0.59 0.70 0.38 0.71 0.41 0.46 0.34 [7.14] [8.68] [5.31] [7.20] [4.18] [3.53] [3.36]”
    0/77
TYPE B · PAPER-SPECIFIC (10)
  • B1
    sue min observations required
    §§2.1 “subject to a requirement of at least six observed announcements over the two year window”
    provenance · paper-explicit
    stage · feature eng
    62/77
  • B2
    sue eps proxy field
    §§2.1 “For earnings per share I use Compustat quarterly data item EPSPXQ (Earnings Per Share (Basic) / Excluding Extraordinary Items).”
    provenance · paper-explicit
    stage · data loading
    64/77
  • B3
    sue observability via rdq
    §§2.1 “Earnings announcement dates are Compustat quarterly data item RDQ.”
    provenance · paper-explicit
    stage · feature eng
    55/77
  • B4
    size split at nyse median
    §§3 “The size portfolios divide stocks into large or small cap universes, based on NYSE median market capitalization.”
    provenance · paper-explicit
    stage · signal gen
    71/77
  • B5
    sue split at nyse 30 70
    §§3 “The earnings momentum portfolios divide the world into three portfolios divided at the 30th and 70th percentiles, using NYSE breaks, of earnings surprises”
    provenance · paper-explicit
    stage · signal gen
    69/77
  • B6
    portfolio weighting value weighted within
    §§3 “Portfolios are rebalanced monthly, and returns are value weighted.”
    provenance · paper-explicit
    stage · portfolio
    61/77
  • B7
    factor is ew of size legs
    §§3 “The earnings momentum factors are each formed as an equal weighted average of value weighted large cap and small cap earnings momentum strategies, which buy the upper tertile and short the bottom tertile of the earnings surprises portfolios based on the corresponding measure of earnings surprises.”
    provenance · paper-explicit
    stage · portfolio
    57/77
  • B8
    long short direction
    §§3 “which buy the upper tertile and short the bottom tertile of the earnings surprises portfolios”
    provenance · paper-explicit
    stage · signal gen
    57/77
  • B9
    delay bars zero
    §§2.1 + §3 “performance measured over the preceding year, skipping the most recent month to avoid diluting price momentum with short term reversals (r2;12).”
    provenance · paper-inferred
    stage · execution
    61/77
  • B10
    mid tertile not used in factor
    §§3 “which buy the upper tertile and short the bottom tertile of the earnings surprises portfolios”
    provenance · paper-explicit
    stage · portfolio
    60/77
What this task reveals

The lowest-passing observed requirement is e2e metric consistency: 0 of 77 observed checker verdicts pass. The full property list above shows whether errors concentrate in domain-level validity or the paper-specific strategy logic.