Corporate Default Event Database for Equity Research

A corporate default event database supports equity research when it preserves the public disclosure date, the underlying default date, the filing evidence, and nullable severity fields as separate parts of each event.

Alphanume Corporate Default Events is a point-in-time feed of public-company defaults labeled from SEC filing text. The route GET /v1/corporate-default-events returns only rows flagged is_default = 1. Each record carries the issuer ticker and filing source, plus structured obligation and severity details when the filing explicitly supports them.

For equity research, the first job is creating a dated panel of what investors learned. event_date is the date of the filing that disclosed the event. default_date is the earlier contractual date stated inside the filing when available. A backtest of public information should anchor on the disclosure, while a study of reporting lag can compare both clocks without pretending the earlier default was already known.

Read the event as a filing record

The enriched fields include obligation name and type, creditor, principal outstanding, accelerated amount, missed payment amount, grace-period end, and acceleration state. Any can be null when the filing does not state the fact. Keep those nulls because replacing an unstated amount with zero reverses the meaning of the source.

A default is a broken debt obligation. A restructuring changes terms, a bankruptcy is a court process, and a cure or waiver can resolve an earlier default. Those events can occur in sequence, though the default row does not make them synonyms or imply zero equity value.

Return events by first-known date

The API accepts exact date or range filters and returns results ordered by event_date descending. Exact date cannot be mixed with date_gte, date_lte, date_gt, or date_lt. The current route has no ticker filter and no pagination, so a range request returns all matching rows allowed by the account tier.

import os
import requests
import pandas as pd

response = requests.get(
    "https://api.alphanume.com/v1/corporate-default-events",
    headers={"X-API-Key": os.environ["ALPHANUME_API_KEY"]},
    params={"date_gte": "2020-01-01", "date_lte": "2025-12-31"},
    timeout=30,
)
response.raise_for_status()
payload = response.json()
events = pd.DataFrame(payload["data"])

assert len(events) == payload["count"]
assert events["event_date"].notna().all()
events["first_known_date"] = pd.to_datetime(events["event_date"])
events.to_parquet("corporate_default_first_known.parquet", index=False)

Save the raw response before grouping. A single issuer can disclose several default stages or obligations, and treating every row as an independent company outcome overstates the denominator. Retain the all-event table, then create a separate first-event-per-ticker view under a written rule.

Build the equity event panel

The event endpoint does not include prices, filing acceptance time, delisting returns, or a complete later-resolution history. Those are separate inputs. A daily study can use the first eligible session strictly after event_date. A finer rule requires a separately sourced EDGAR acceptance timestamp whose accession and URL remain in the audit. Begin every join from the event table and use left joins so missing outcomes remain visible. Starting with today's traded ticker universe removes distressed names that disappeared.

Event type can segment the panel into missed payments, covenant issues, acceleration, and other labeled defaults. Report the number of null types and evidence fields within each slice before comparing returns. A smaller clean subset can answer a narrower question, but it should not be presented as the entire default population.

Watch the main failure modes

• Disclosure lag. Measuring from default_date can use a fact the market learned only on event_date.
• Repeated distress. Several rows from one issuer can dominate an event-type average.
• Survivorship. Delisted and bankrupt names need retained identifiers and explicit terminal outcomes.
• Null severity. Missing dollar values mean the filing did not support them, not that the obligation was small.
• Lifecycle confusion. Default, restructuring, bankruptcy, cure, and recovery require separate dated labels.

Free access covers a trailing 20-trading-session window delayed by one session. A long equity panel requires historical access, and a restricted-range response should not be recorded as a period with no defaults.

Create one auditable year

Pull one calendar year and export the untouched response, all-event panel, first-event-per-ticker view, and missing-outcome audit. Check a sample of each event type against filing_url, preserve all null severity fields, and calculate returns only after the disclosure-date cohort is frozen.

Explore the field contract on the Corporate Default Events page and use the Corporate Default Events guide for the broader event-study workflow.