Every figure InvestIQ shows comes from a named source and is computed by code. This page lists those sources, how fresh the data is, and what the AI models receive and write. Your personal data is covered by the privacy policy.
1Market and financial data — Financial Modeling Prep
Most of the data comes from Financial Modeling Prep, a market-data provider. InvestIQ uses it for:
- quotes and price history, daily and intraday;
- financial statements (annual and quarterly), ratios, key metrics and enterprise values;
- the discounted-cash-flow value and the financial-health scores (Piotroski F-Score, Altman Z-Score);
- analyst estimates, ratings and price-target consensus;
- results history and the results calendar;
- insider trades and the trades disclosed by members of the US Congress;
- company news;
- macroeconomic series (US Treasury rates, inflation, the federal funds rate, unemployment) and the economic calendar;
- sector performance and sector valuation multiples;
- company profiles, peers, dividends, ETF composition, index constituents and exchange rates;
- earnings-call transcripts, where available.
Institutional holdings (13F filings) are not covered by our current data plan: the app says so rather than showing an estimate.
2Company filings — SEC EDGAR
For US companies, InvestIQ also reads the filings published on EDGAR, the public database of the US Securities and Exchange Commission: the list of filings, results announcements (form 8-K) and the financial data companies report in XBRL format, over up to ten years. This data is public and lets a figure be traced back to the company’s own filing.
3The web, as a last resort
The Copilot and the War Room can run a web search, through OpenAI’s search tool, when a question needs something the data above does not contain, such as a very recent event. The War Room is limited to 8 searches per debate. Figures found on the web never appear in the Copilot’s data cards, which accept only figures from InvestIQ’s own data.
4How fresh the data is
InvestIQ fetches each figure once and shares it across users for a set period. It is refreshed at most this often:
| Data | Refreshed at most every |
|---|---|
| Quotes | 1 minute |
| Price history | 15 minutes |
| Insider trades | 2 hours |
| Analyst estimates, results | 6 hours |
| Financial statements, ratios, DCF value, Congress trades | 24 hours |
| Scores of the ~600-stock universe | 7 days (rescored in rotation every night) |
A stock you open in Stock Analysis is scored at that moment, with the data above. Depending on the exchange, the provider’s own quote can be delayed.
5What InvestIQ computes
Scores and sub-scores, ratios built from the statements, valuations (discounted cash flows, reverse DCF, sensitivities), portfolio measures (weights, concentration, volatility, drawdown) and ranks within the universe are computed by InvestIQ’s own deterministic code. The rules of the score are in the methodology.
6AI models: what they receive and what they write
InvestIQ uses language models from OpenAI and DeepSeek by default. In the Copilot you can also choose Anthropic’s Claude, and in the War Room models from Anthropic, Google, xAI and Mistral.
The models receive your question and the results computed for it — scores, figures, excerpts of filings or news. They write explanations, summaries and debates; they do not compute figures. The checks applied to their text are described in the editorial policy.
The score and the figures of Stock Analysis involve no AI model at all.
7Errors and corrections
Providers make mistakes, and so can we. If a figure looks wrong, write to contact@getinvestiq.com with the stock and the page: we check it against the company’s filings and correct what is ours to correct.