Point-in-time data captures a value exactly as it was known on a given historical date, with no later restatements applied — essential for honest backtesting because it prevents look-ahead bias.
Most databases overwrite history: when a company restates earnings or a 13F is amended, the old value disappears. Point-in-time storage keeps the original alongside an "as-of" timestamp, so you can reconstruct exactly what an investor could have known on any past date.
Without it, backtests silently cheat — using restated or revised figures that were not available at the time — and overstate a strategy's performance. With it, an as_of query is just another parameter.
Asking "what did this fund's 13F show as of the day it was first filed" — before any amendment — requires point-in-time storage.
A point-in-time archive of filings is one of Arkolith's core moats: it turns "history" from a single mutable snapshot into a queryable timeline you can backtest against.
Arkolith turns this into live, sourced data your agent can query — SEC filings, insider activity, and market data behind one key, every datapoint traceable to its origin.