← All tasks
T02 · Momentum
Momentum Crashes
Kent Daniel and Tobias J. Moskowitz, 2016 · Journal of Financial Economics
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.
Dynamic momentum (Daniel & Moskowitz 2016, Eq. 6 / Table 7 "Dynamic"): scale a monthly value-weighted WML decile portfolio by an optimal risk-managed weight w*_t = (1/2lambda)(mu_{t-1}/sigma^2_{WML,t-1}), where the conditional mean is estimated by an expanding-window OLS on a bear-market x market-variance interaction and the conditional variance is a linear combination of a GJR-GARCH(1,1) forecast and a 126-day realized vol; weight is allowed to flip negative.
- Type
- technical
- 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
- 01 Data loading
- 02 Feature engineering
- 03 Model fitting
- 04 Signal generation
- 05 Portfolio construction
- 06 Execution
Checker suite
17 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 (6)
- A3no future data access§Section 2.1 “we first rank stocks based on their cumulative returns from 12 months before to one month before the formation date (i.e., the t-12 to t-2-month returns)”stage · feature eng73/76
- A4causal consistency§Section 2.1 “The value-weighted holding period returns of the decile portfolios are computed, in which portfolio membership does not change within a month except in the case of delisting.”stage · signal gen16/76
- A5no survivorship bias§Section 2.1 “The universe starts with all firms listed on NYSE, Amex, or Nasdaq as of the formation date”stage · portfolio66/76
- A7data frequency alignment§Table 7 caption “the WML returns each month are scaled by the realized volatility of the daily WML returns over the preceding 126 trading days”stage · data loading46/76
- A9signal delay test§Section 2.1 “we use a one-month gap between the end of the ranking period and the start of the holding period to avoid the short-term reversals”stage · signal gen54/76
- A10e2e metric consistency§Benchmark evaluation contract “The auxiliary A10 check compares end-to-end strategy behavior under the released differential tests and metric tolerances.”0/76
TYPE B · PAPER-SPECIFIC (11)
- B1formation period§Section 2.1 “We first rank stocks based on their cumulative returns from 12 months before to one month before the formation date (i.e., the t-12 to t-2-month returns).”provenance · paper-explicitstage · signal gen54/76
- B2decile sort§Section 2.1 “All firms meeting the data requirements are then placed into one of ten decile portfolios based on this ranking.”provenance · paper-explicitstage · signal gen66/76
- B3value weighted§Section 2.1 “The value-weighted (VW) holding period returns of the decile portfolios are computed.”provenance · paper-explicitstage · portfolio66/76
- B4bear lookback§Section 3.2 “The ex ante bear market indicator equals one if the cumulative CRSP VW index return in the past 24 months is negative and is zero otherwise.”provenance · paper-explicitstage · feature eng76/76
- B5market var window§Section 3.5, Equation 4 “Market variance is measured over the 126 days preceding the start of month t.”provenance · paper-explicitstage · feature eng40/76
- B6wml var window§Section 4.1 “To estimate the month t WML variance we use the 126-day WML variance estimated through the last day of month t-1.”provenance · paper-explicitstage · feature eng73/76
- B7regression interaction§Section 3.5, Equation 4 “The fitted regression uses the interaction between the bear market indicator and market variance over the preceding six months.”provenance · paper-explicitstage · model fit56/76
- B8expanding window ols§Figure 7 caption “The regression is run from the start of our sample (1927:07) up through month t-1.”provenance · paper-explicitstage · model fit13/76
- B9dynamic weight formula§Section 4 and Appendix C “The weight placed on the risky asset should be proportional to the expected excess return over the next period and inversely proportional to the conditional variance.”provenance · paper-explicitstage · portfolio70/76
- B10negative weight§Section 4 “The dynamic strategy weight is negative in 82 of the months in our sample, necessarily in months when the forecast return of the WML strategy is negative.”provenance · paper-explicitstage · portfolio72/76
- B11gjr garch§Section 4 and Appendix D “We form a linear combination of the forecast of future volatility from the fitted GJR-GARCH process with the realized standard deviation of the 126 daily returns preceding the current month.”provenance · paper-explicitstage · model fit18/76
What this task reveals
The lowest-passing observed requirement is e2e metric consistency: 0 of 76 observed checker verdicts pass. The full property list above shows whether errors concentrate in domain-level validity or the paper-specific strategy logic.