← All tasks
T10 · Mean Reversion

Short-term Momentum

Mamdouh Medhat and Maik Schmeling, 2022 · The Review of Financial Studies

3%
pass@1 across 16 systems
2/16
systems with ≥1 successful attempt
1/5
1/5 attempts passed for the best system
80 attempts · 16 systems × 5
Claude Code + Opus 5
0%
Claude Code + Opus 4.8
0%
Claude Code + Opus 4.7
20%
Codex + GPT-5.6-sol
0%
Codex + GPT-5.5
20%
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.

Double-sorting on the previous month's return and share turnover separates short-term reversal (low-turnover D1) from short-term momentum (high-turnover D10); long-short sleeves on each side are combined 50/50 monthly into a single STMOM+STREV* portfolio.

Type
technical
Implementation
pandas
Rebalance
monthly
Long-short
yes
Declared task setup
Data
One-command fetcher for WRDS-sourced task data
Operational scope
Implement high-turnover short-term momentum and a tradable low-turnover reversal sleeve, then combine their returns with equal weight.
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

18 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 (8)
  • A1
    no backfill
    77/79
  • A2
    no global normalize
    69/79
  • A3
    no future data access
    30/79
  • A4
    causal consistency
    30/79
  • A5
    no survivorship bias
    76/79
  • A8
    transaction timing
    30/79
  • A9
    signal delay test
    30/79
  • A10
    e2e metric consistency
    2/79
TYPE B · PAPER-SPECIFIC (10)
  • B1
    turnover calc
    §§1.1 “we deflate the total volume of trades by the number of shares outstanding to obtain share turnover (TO_{1,0})”
    provenance · paper-explicit
    78/79
  • B2
    nasdaq volume adjustment
    §footnote 6 “Prior to February 2001, we divide NASDAQ volume by 2.0. From February 2001 to December 2001, we divide by 1.8. From January 2002 to December 2003, we divide by 1.6. From January 2004 and onward, NASDAQ volume no longer differs from NYSE and AMEX volume, and we apply no adjustment.”
    provenance · paper-explicit
    73/79
  • B3
    nyse breakpoints return
    §§1.2 “conditional decile sorts based on NYSE breakpoints”
    provenance · paper-explicit
    72/79
  • B4
    nyse breakpoints turnover
    §§1.2 “conditional decile sorts based on NYSE breakpoints, first on r_{1,0} and then on TO_{1,0} within r_{1,0} deciles”
    provenance · paper-explicit
    62/79
  • B5
    conditional sort order
    §§1.2 “We use conditional decile sorts based on NYSE breakpoints, first on r_{1,0} and then on TO_{1,0} within r_{1,0} deciles”
    provenance · paper-explicit
    70/79
  • B6
    stmom turnover selection
    §§1.2, Figure 1 “short-term momentum trades only in the highest-turnover decile”
    provenance · paper-explicit
    44/79
  • B7
    strev turnover selection
    §§1.2, Figure 1 “short-term reversal* trades only in the lowest-turnover decile”
    provenance · paper-explicit
    44/79
  • B8
    stmom long short signal
    §§1.2, Figure 1 “short-term momentum strategy, which buys the previous month's winners and shorts the losers within the highest turnover decile”
    provenance · paper-explicit
    75/79
  • B9
    strev long short signal
    §§1.2, Figure 1 “short-term reversal* strategy, which buys the previous month's winners and shorts the losers within the lowest turnover decile, generates a negative and significant average return”
    provenance · paper-explicit
    5/79
  • B10
    universe exchanges
    §§1.1 “Our main sample is all NYSE/AMEX/NASDAQ stocks”
    provenance · paper-explicit
    68/79
What this task reveals

The lowest-passing observed requirement is e2e metric consistency: 2 of 79 observed checker verdicts pass. The full property list above shows whether errors concentrate in domain-level validity or the paper-specific strategy logic.

Open paper → All tasks
← Previous in Mean Reversion
T09 · Deep Learning Statistical Arbitrage