Financial data API

Index-level fundamentals, calculated on the aggregate: total market capitalisation of an index over the total earnings, book value or EBITDA of its constituents, rather than an average of the constituents' own ratios. The same method is applied to every market and every period.

Readings are point-in-time: each value reflects only what was public on or before its own date, and a figure is never restated when a company later revises its accounts. That is what makes a backtest against this data honest rather than hindsight-assisted.

The data methodology (PDF) sets out the full construction — what counts as earnings, which items are normalised out, and how missing figures are handled.

No sign-up, no key, no client library. These URLs work in a browser, in curl, and in anything that can fetch a URL.

https://siblisresearch.com/api/v1/indices
https://siblisresearch.com/api/v1/NDX/pe-trailing
https://siblisresearch.com/api/v1/UKX/pe-trailing
https://siblisresearch.com/api/v1/DAX/pe-forward
https://siblisresearch.com/api/v1/N225/cape

Python

import requests

url = "https://siblisresearch.com/api/v1/NDX/pe-forward"
for row in requests.get(url).json()["data"]:
    print(row["trading_day (EOD)"], row["value"])

R

library(httr); library(jsonlite)

url <- "https://siblisresearch.com/api/v1/NDX/pe-forward"
df  <- fromJSON(content(GET(url), "text"))$data

Google Sheets — a live cell formula, no add-on:

=IMPORTDATA("https://siblisresearch.com/api/v1/NDX/pe-forward")

The response

{
  "ticker":     "NDX",
  "index_name": "NASDAQ 100",
  "frequency":  "daily",
  "ratio":      "pe_forward",
  "data": [
    { "trading_day (EOD)": "2024-12-31", "value": 26.41 },
    { "trading_day (EOD)": "2025-06-30", "value": 27.62 }
  ]
}

The free tier carries a subset of the ticker universe and a handful of period-end snapshots — /v1/dates lists exactly which. The subscriber API returns the same shape at daily frequency for every index, with date-range filtering and a statistical aggregates endpoint. Because the shape is identical, code written against the free tier keeps working when a key is added.

Full endpoint list, authentication, rate limits and error codes are in the API reference, and there is an API guide (PDF) covering the same ground offline.

The API is a plain REST interface with stable URLs and no authentication on the free tier, which makes it usable directly as a tool by an AI agent or an LLM workflow: the agent fetches a URL and receives labelled JSON. There is nothing to wrap and no integration work to do.

Open-source examples for Python, R, Excel and Google Sheets are on GitHub, ready to run against the free tier.

47 country and regional indices and 44 sector indices, covering every major developed and emerging market alongside the world, developed, emerging and European aggregates, and the eleven equity sectors calculated separately for the United States, global, and global excluding the United States.

The longest series begin in 1970 and the majority of the universe carries twenty years or more, built on one consistent method throughout. A valuation reading only means something against a reference period, and a reference period that spans several complete cycles is a different thing from one that starts in the 1990s.

Quantitative researchers

Decades of index-level fundamental data into Python or R in one request. Point-in-time history means a backtest sees what was known on each date rather than what was restated afterwards.

Portfolio managers and analysts

Valuation dashboards in Excel, Power BI or Tableau that refresh on their own. The CAPE, P/E and price-to-book series are the inputs for valuation-based allocation work.

Academic researchers

A citable source with a published methodology. Siblis valuation data has been used in peer-reviewed journals, central bank working papers and government research — see where it has been cited.

Anyone benchmarking a single stock

A company multiple means nothing on its own. The index and sector ratios here are the denominator in expensive compared with what?

This is not a company fundamentals API. There is no per-ticker balance sheet, income statement or cash flow feed here, and no financial statement data for individual companies. If that is what you need, providers such as Finnhub, Alpha Vantage and FinancialData.net cover it.

What this API provides is the market-level view — aggregated valuation and fundamental metrics for whole equity indices, built from the same underlying company filings. That serves two questions a per-company feed answers badly:

• Top-down research. Is the world market expensive? Is European technology trading at a premium to American technology? Is the emerging markets index at a historically low CAPE? Those are index-level questions, and answering them by summing a per-company feed means rebuilding index membership, weighting and normalisation yourself, for every date in the history.
• Benchmarks for bottom-up work. When a single stock is analysed from its own filings, the market and sector multiples are what turn those figures into a judgement.

The financial statement data underneath every ratio is collected in-house, from primary sources: regulatory filings including SEC EDGAR and its international equivalents, official corporate disclosures, and company reports. There is no third-party valuation feed behind these numbers.

Collection

Crawling tools built and refined over more than a decade of daily operation collect earnings, income statement and balance sheet figures, analyst estimates and dividend records. Filing delays, restatements and corporate actions — splits, spin-offs, index changes — are handled automatically, and edge cases are flagged for a person rather than resolved by a default.

Normalisation and aggregation

Reported figures are normalised to remove non-recurring items — impairments, restructuring and litigation charges, M&A costs, gains and losses on discontinued operations — under one set of rules applied across GAAP, IFRS and local standards. Constituent figures are then aggregated to the index level by market capitalisation, and per-share figures are adjusted at every constituent change so an EPS series stays continuous through membership changes.

Automated validation

Every newly calculated value is cross-checked against independent sources and tested for statistical outliers against its own history and against peer indices before publication. Values outside the thresholds are held back rather than published and corrected later.

Review and benchmarking

Flagged values are reviewed against primary sources by the research team, and each release is benchmarked against the major commercial data providers. Where a discrepancy turns out to be ours, the correction is applied to the full historical series and subscribers are told.

Do I need an API key to start?

Not for the free tier. Those endpoints need no key and no account. A key is issued on subscription and unlocks daily frequency, the full index universe and the complete history.

What does point-in-time mean here?

A reading reflects the information that was public on or before its own date. When a company restates its results, the historic reading is left as first published. The only exception is a calculation error on our side, which is corrected and disclosed.

How far back does the history go?

To 1970 on the longest series. Coverage varies by index and by ratio; the sample dashboard shows exactly what exists for each one.

Can I use it with AI agents or LLM tools?

Yes. Plain REST, stable URLs, labelled JSON, no authentication on the free tier.

Is there a rate limit?

Subscriptions carry a daily request quota — 1,000 on Pro and 50,000 on Enterprise. The API reference has the detail.

Can I redistribute the data?

Free-tier figures may be reused with attribution. Subscriber data is licensed to the subscriber and may be used in published research with a citation, but not redistributed as a dataset. The terms of use set out both.