ML-QuantSubscribe

SSRNRisk, Credit & Banking

Forecasting Distress with Statements That Did Not Yet Exist: Look-Ahead Leakage and Data Vintages in Early-Warning Models

Using statements and database values not yet available when forecasts are made inflates distress-prediction accuracy by up to 49%, a leakage concentrating in profitability ratios.

Featured in No. 140 on 20 Nov 2026 · 2 days after release

Released
18 Nov 2026
First featured
No. 140 · 20 Nov 2026
Published in
Not yet, as far as Semantic Scholar knows
Fanfare
3 of 5
Identifier
SSRN 7709638
Authors
Jinhyung Bae

Citations and venue from Semantic Scholar (ODC-BY), refreshed weekly. Summary: Quant Letter (CC BY 4.0).

    Type to search. Try rough volatility, LLM agents or FinGPT.

    ↑↓ move↵ openesc closeFull search page