← All tasks
T23 · Event-Driven Trading

Days to Cover and Stock Returns

Harrison Hong et al. (5 authors), 2015 · National Bureau of Economic Research

34%
pass@1 across 16 systems
12/16
systems with ≥1 successful attempt
5/5
5/5 attempts passed for the best system
80 attempts · 16 systems × 5
Claude Code + Opus 5
80%
Claude Code + Opus 4.8
20%
Claude Code + Opus 4.7
20%
Codex + GPT-5.6-sol
20%
Codex + GPT-5.5
100%
OpenCode + Kimi-K3
40%
OpenCode + Opus 4.7
20%
OpenCode + GPT-5.5
60%
OpenCode + Sonnet 4.6
40%
OpenCode + Qwen3.7-Max
80%
OpenCode + GLM-5.2
0%
OpenCode + DeepSeek V4-Pro
0%
OpenCode + GLM-5.1
20%
OpenCode + Qwen 3.6 Plus
0%
OpenCode + MiniMax-M3
40%
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.

Sort common stocks each month by days-to-cover, computed as short interest divided by average daily share turnover, then hold an equal-weight long-low DTC / short-high DTC portfolio for the following month.

Type
fundamental
Implementation
pandas
Rebalance
monthly
Long-short
yes
Declared task setup
Data
One-command fetcher for WRDS-sourced task data
Operational scope
Fix WRDS/CRSP short-interest, shares, turnover, split, unit, and monthly-alignment conventions before evaluation.
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
    stage · data loading
    77/77
  • A2
    no global normalize
    stage · feature eng
    76/77
  • A3
    no future data access
    stage · signal gen
    68/77
  • A4
    causal consistency
    stage · signal gen
    66/77
  • A5
    no survivorship bias
    stage · data loading
    75/77
  • A7
    data frequency alignment
    stage · feature eng
    75/77
  • A8
    transaction timing
    stage · execution
    65/77
  • A9
    signal delay test
    stage · signal gen
    68/77
  • A10
    e2e metric consistency
    31/77
TYPE B · PAPER-SPECIFIC (10)
  • B1
    dtc formula
    §§II.A line 1252-1257 “Our baseline DTC measure divides the SR by the daily turnover ratio averaged within a month”
    provenance · paper-explicit
    stage · feature eng
    77/77
  • B2
    average daily turnover formula
    §§II.A line 1255 “The daily turnover ratio is averaged within a month to get a monthly variable”
    provenance · paper-explicit
    stage · feature eng
    72/77
  • B3
    decile sort n quantiles
    §Table 3 “decile portfolios sorted on DTC”
    provenance · paper-explicit
    stage · signal gen
    73/77
  • B4
    cross sectional sort per month
    §Table 3 caption “Each month, we sort all stocks into decile portfolios based on DTC”
    provenance · paper-explicit
    stage · signal gen
    75/77
  • B5
    long short assignment
    §Table 3 “long-short = decile 1 − decile 10”
    provenance · paper-explicit
    stage · signal gen
    65/77
  • B6
    equal weight within leg
    §Table 3 “equal-weighted DTC long-short = 1.19% per month”
    provenance · paper-explicit
    stage · portfolio
    65/77
  • B7
    one month lag signal to return
    §Table 3 footnote “monthly long-short returns formed from sorts on prior-month DTC”
    provenance · paper-explicit
    stage · signal gen
    65/77
  • B8
    universe filter exchanges and sharecodes
    §§II.A line 1255 “We require stocks to be listed on NYSE, AMEX and NASDAQ and common stocks (i.e. share type code equals to 10 or 11)”
    provenance · paper-explicit
    stage · data loading
    77/77
  • B9
    shortint uses split adjusted
    §§II.A line 1207-1228 “Short interest is measured as a fraction of shares outstanding”
    provenance · benchmark-specified
    stage · feature eng
    67/77
  • B10
    long short dollar neutral
    §Table 3 “decile 1 minus decile 10”
    provenance · paper-explicit
    stage · portfolio
    70/77
What this task reveals

The lowest-passing observed requirement is e2e metric consistency: 31 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.

Open paper → All tasks
← Previous in Event-Driven Trading
T22 · The Pre-FOMC Announcement Drift
Next in Event-Driven Trading →
T24 · Earnings Extrapolation and Predictable Stock Market Returns