Methodology and data sources
This page describes where the numbers on Alysly originate: which sources sit behind them, how often they are refreshed, why our values may differ from another provider's, and where AI does part of the work.
How an individual metric is calculated is not here but in the data notes. They hold one entry per kind of data, linked as a footnote under the block it belongs to. This page answers the question of origin, that one the question of method.
What we cover
| Coverage | Count |
|---|---|
| Stocks with a detail page | 4,334 |
| ETFs | 99 |
| of those with holdings data | 95 |
| Indices behind those ETFs | 44 |
| Cryptocurrencies | 370 |
| Countries with indicators | 45 |
| Macro time series | 15 |
As of 21 September 2026. The table is recounted on every build.
Where the data comes from
- Prices. Daily closing prices from a commercial market data provider. One point per trading day, no intraday data, no real-time quotes.
- Company financials. SEC EDGAR, the XBRL annual filings of companies that report to the SEC. Public domain.
- ETF holdings and master data. From commercial providers, based on each fund's latest published holdings list. Funds publish on their own schedules, which is why the as-of date sits directly at the holdings table.
- Rates, inflation and macro series. European Central Bank, Eurostat and the Federal Reserve Bank of St. Louis (FRED). All three permit reuse with attribution; the full attribution including license sits in the structured data block of each series.
- Country indicators. World Bank Open Data, licensed CC BY 4.0.
- Crypto. Prices and market capitalization from public exchange and market data interfaces.
We do not name the market data provider. It has no bearing on whether our calculations can be followed, and the license does not require attribution. Where a source is part of the statement itself or its license requires naming, it is written out. That applies to SEC EDGAR, the World Bank, the ECB, Eurostat and FRED.
How often it is refreshed
The entire dataset is rebuilt automatically every trading day: fetch prices, compute metrics, generate pages, publish. There is no manual step in between at which individual figures could be adjusted after the fact.
Not every part is equally fresh. Prices are one trading day old, company financials a quarter or a year, ETF holdings weeks depending on the fund company. The relevant as-of date sits at each data block, and the data notes explain for every block how it was produced.
Calculations and limits
How an individual metric is produced and where it stops being reliable sits where the metric is shown: as a footnote under the block and in full on the data notes page. That is also where the uncomfortable parts are. Fund holdings are older than the prices beside them. Index compositions are derived from a proxy ETF. Commodity series show the front-month futures contract. Since September 2024, their returns are chained contract by contract across roll dates. Contract data for earlier roll dates is not available, so the series can jump at those contract changes. Company financials exist only for SEC filers.
Two decisions apply across the whole site and therefore sit here. Returns are price returns: excluding dividends, taxes and costs, annualized from one year upwards. And we deliberately do not show a price-to-earnings ratio. It would need a current price for every stock that has financials, and we only have that for a small share of them. A metric missing from most pages causes more confusion than it resolves.
What ETF coverage says and what it does not
Every stock page states how many ETFs hold the stock and how that compares with all other stocks. Both refer only to our own fund coverage, never to the whole market. That coverage is unevenly distributed across providers. A claim such as "held in 3 % of all ETFs worldwide" could not be derived from it, so we do not make one.
How the percentile is built, and why high coverage says nothing about the quality or valuation of a company, is set out in the data note on ETF coverage.
Why our figures may differ from other providers
Two sites can report the same metric differently without either of them calculating wrongly. The most common reasons:
- A different universe. "Held in 26 ETFs" depends entirely on which ETFs are counted in the first place.
- A different as-of date. Closing price against intraday, yesterday against today, differing fund disclosure dates.
- Dividends. Our returns are price returns. Anyone reporting total return will come out higher for every dividend payer.
- Currency. The same stock in euros and in dollars produces different annual returns.
- Definition. Volatility over 30, 90 or 250 days is the same metric with a different result.
Use of AI
Analysis and presentation are partly AI-assisted: market reports, summaries and chart commentary are generated automatically from the same underlying data. The calculation rules and text templates are reviewed by hand, not every single report, and an automated gate checks the data before each release. The figures are computed, not generated. No price, no metric and no percentile on this site comes out of a language model.
Found an error?
We would rather hear about a wrong number than have it sit there unnoticed: mail@alysly.com. Who is behind Alysly is on the About Us page.
Not investment advice. Alysly provides data and tools, not recommendations. All information without guarantee.