Will OpenAI or Anthropic IPO first?

What this archive page covers

DepthFeed's market index identifies this event as KXOAIANTH-40 in the financials category. It reports 20,348 stored order-book observations across 43 distinct days, from 2026-06-30 through 2026-08-11. That is an average of about 473 indexed observations per captured day; the timestamps in the returned rows, not the average, determine the actual spacing.

The page is a coverage record for historical market data. It does not assert who won, whether the event resolved, or what an official settlement source reported. Use the event title and identifier to locate the exact contracts, then verify contract rules and resolution with the venue before interpreting a probability path.

Research uses for this financials market

For financial contracts, compare probability and liquidity before and after company or market announcements. The archive shows the displayed book; issuer filings, news, and the contract's official settlement source remain separate evidence.

A realistic replay should walk the recorded bid/ask ladder at the decision timestamp, apply the modeled order size, and preserve the observed spread. A midpoint-only backtest can create fills that were never available. Compare results across multiple observations and sizes instead of treating one displayed quote as a stable execution price.

From event coverage to API rows

1. Use market discovery to resolve event KXOAIANTH-40 into the exact contract ticker or outcome identifier your study needs.
2. Request stored snapshots for that exact market and keep the response timestamps, cursor, native identifiers, and order-book arrays together.
3. Model fills against displayed levels, then join external news, polling, filings, or settlement evidence with an explicit as-of rule.
4. Record the coverage dates and gaps used by the test so another researcher can reproduce the sample instead of assuming the archive is continuous outside these boundaries.

Review the API documentation for current discovery and snapshot contracts, or start with the historical-data overview to compare history windows and access limits.

Limits to preserve in the model

• A stored observation describes displayed liquidity at that time, not hidden size or queue priority.
• Activity between two observations is unknown; do not interpolate a book state that was not captured.
• A paper or simulated fill does not consume live liquidity and can understate impact for large orders.
• The market title is not the settlement rule. Preserve the venue's exact contract and resolution source.
• Coverage is bounded by the dates shown above and by the exact markets returned through discovery.