How severe drops actually recover
Relationship between SMRS and post-event abnormal returns vs. SPY at 10, 30, and 60 trading days. Tests whether market reaction predicts the path back.
MRS × abnormal-return correlation
Pearson r between SMRS and the post-event abnormal return at each window.
MRS vs. 30-day abnormal recovery
tier-colouredAwaiting more recovery data.
Recovery by SMRS tier
Average market-adjusted return at 10, 30, and 60 trading days after the information shock.
Awaiting tier-level data.
Backtest — long high-MRS / short low-MRS
Hypothetical pair: buy MRS 8–10 (overreaction), short MRS 4–6 (fundamental decline). Educational only — no transaction costs, slippage, or borrow.
All scored events (0)
Every scored disclosure with its drop, MRS, abnormal-return windows and sector. Click headers to sort.
How these numbers are computed
Abnormal returns are measured against SPY over the same window, anchored to the disclosure-day close. A <window>-day abnormal return is stock_return − spy_return over the trading days following the event. Recovery values are persisted on each scored event keyed by (ticker, disclosure_date).
Tier buckets follow the SMRS tier definitions: Mild (0–4), Moderate (4–6), Severe (6–8), Extreme (8–10). The strategy backtest pairs the Extreme tier (long) against the Moderate tier (short); spread = long_avg − short_avg. Sample-size warnings fire below a 20-event threshold per leg and are first-class output, not chrome.
Pearson correlation is computed on the joined dataset (events with both an MRS and a recovery measurement). Coverage updates on a 12-hour cron — events become eligible 90 calendar days post-disclosure, so a fresh case appears in this view roughly a quarter after it surfaces in the catalog.
Research disclaimer. This analysis examines whether the intensity of an information shock (as measured by SMRS) correlates with subsequent stock-price recovery. The backtest is hypothetical and excludes transaction costs, slippage, borrowing costs, liquidity constraints, and execution risk. Past market behavior does not predict future results. Not investment advice.