Alternative Data

134 datasets from 175 publishers in more than 50 countries, point-in-time, mapped to stocks and tested for lead.

Alternative data only gives you an edge when it can be trusted, and when you know it leads. ORTEX collects from 175 publishers across more than fifty countries, archives every figure as it first appeared, cleans and normalizes it, maps it to the right stock, and then tests it against each company's reported results, so you know which datasets actually lead. Car registrations, airline traffic, bank books, drug approvals and power output arrive as one consistent dataset.

At a glance

  • 134 datasets
  • 814 series
  • 175 publishers in 50+ countries
  • 9,900+ securities mapped

44 days median lead on reported revenue across the 54 statistically significant, out-of-sample leads found on a company's results.

ORTEX lead measurement, tested on the date each figure was published.

What you can do with it

  • Nowcasting. Read revenue drivers weeks before the company reports.
  • Systematic research. Point-in-time series with two clocks, ready to backtest.
  • Thematic work. Sector and economy series on the same platform.

Clean, point-in-time, ready to trade

The questions every fund asks first, answered.

  • Cleaned and normalized. Hundreds of formats reshaped into one, with the same units, periods and naming. Periods still filling in are flagged, not headlined.
  • Truly point-in-time. Every observation carries the period it describes and the moment it first became public, never the moment it was processed.
  • History you cannot rebuild. Every figure archived as it first appeared. Many publishers overwrite their own history; ORTEX keeps it.
  • Mapped to the right stock. Every company matched through a curated, reviewed cross-reference. An ambiguous name is never guessed.
  • Tested for lead. Every company series tested out of sample against reported results, and marked as a measured lead, no lead, or not yet tested.
  • Signals included. Z-scores against a company's own history, percentiles across peers, year-on-year change and trend on every series.

Vehicles, 41 datasets

One of the broadest vehicle registration and sales datasets available anywhere.

  • Europe, 20 datasets. New registrations by brand from national registers in the UK, Germany, Italy, Spain, the Netherlands, Belgium, Austria, Poland, Ireland, Czechia, Finland, Greece, Switzerland, Luxembourg, the Baltics and Ukraine, with electric splits and brand mix for every EU country.
  • North America, 11 datasets. Carmakers' own monthly US sales within days, registrations by state and fuel, EV fleets by make, California zero-emission sales, Canadian sales from 1946 and Mexico.
  • Asia-Pacific, 9 datasets. Japan, Korea and India carmaker sales, India's national register, Hong Kong, Singapore, Malaysia, Australia and New Zealand.
  • Daily where it matters. Spain and Greece every day, US carmaker sales within one to three days of month end, Germany ten days after.
  • Electric, by brand. Battery-electric and plug-in splits by brand across Europe, the US states and Asia.
  • Recalls and complaints. US recalls, complaints and investigations daily, from 1995.

Transport, financials and filings

Airlines, airports, bank books, consumer credit and company filings: 55 datasets.

  • Airlines, 12 datasets. Passengers, capacity, load factors and punctuality by airline for the US, UK, France, Canada, Mexico, Brazil, Peru, India, Korea, Taiwan, Japan and Australia.
  • Airports, 11 datasets. Daily flight counts for more than 300 European airports since 2018, EU passengers and movements, US airports and the major operators in Amsterdam, Copenhagen, Paris, Zurich, Germany, Spain, Turkey and the Nordics.
  • Deep history. French airports from 1990, Canadian air traffic from 1981, Brazil from 2000.
  • Road and rail. US trucking capacity and Hong Kong public transport by operator.
  • Financials, 16 datasets. US, Australian and Canadian bank books, card and auto loan performance weeks before lenders report, bank enforcement actions, consumer complaints daily, stock-loan and options volume, payments and crypto flows.
  • Corporate filings, 14 datasets. Executed buybacks in the US, Japan and Korea, tender offers, insider clusters, filing red flags, private raises, federal contract awards daily from 2007, Taiwan monthly revenue, Brazilian operating previews and layoff notices.

Healthcare, energy, consumer and macro

Approvals, power, production, gaming, travel and the macro prints that move markets: 38 datasets.

  • Healthcare, 7 datasets. Clinical trial status daily, FDA decisions and letters, device clearances and recalls, Medicare, Medicaid drug utilization and weekly generic drug prices.
  • Energy and power, 8 datasets. Daily output for every US nuclear reactor, plant generation by operator, power prices, degree days, Gulf of Mexico production from 1996 and Norwegian fields from 1979.
  • Industry and trade, 6 datasets. Korean exports including early 10- and 20-day prints, Taiwan export orders, Japanese machine tool orders, Chilean copper by mine, US mine inspections and Brazilian mining royalties.
  • Consumer and leisure, 11 datasets. Macau gaming, US sports betting by operator, Steam players captured hourly, China game licenses, Texas store openings, Hong Kong and Singapore arrivals, and telecom subscribers.
  • Macro, 6 datasets. US payrolls, CPI and weekly jobless claims from 1986; UK, EU and Canadian labor and prices; and the daily US Treasury cash balance.
  • Market-level context. Sector and economy series alongside the company-level ones, on the same point-in-time clock.

Facts

Coverage134 datasets, 814 series, 50+ countries
Mapping9,900+ securities
HistoryDecades on many series; Canadian vehicle sales from 1946
FrequencyHourly to quarterly, collected as published
APIAll 134 datasets, raw values and derived signals
DeliveryAPI, CSV, ORTEX platform
EvaluationSample files and a data dictionary on request

Why ORTEX

Most alternative data is sold as raw files with the lead left for you to find. ORTEX archives every figure as first published, maps it, and tells you which series actually led a company's results, and by how much.

Measured leads

Selected series with a statistically significant, out-of-sample lead on a company's reported revenue.

Where the data led the results

SeriesCompanyLead on revenueDirection right
New car sales, JapanNissan65 days100%
New car sales, JapanHonda35 days100%
New car registrations, GermanyPolestar80 days100%
New car registrations, GermanyMazda30 days90%
New car registrations, GermanyGeneral Motors15 days82%
New car registrations, NetherlandsMercedes-Benz26 days77%
New car registrations, IndiaMahindra & Mahindra38 days62%

Lead is measured from the date each figure was published to the company's report. Direction is how often the series called the direction of revenue growth.

Tested481 company and series pairs with enough history
Significant54 measured leads, median 44 days
HonestSeries with no lead are marked as such

Knowing which series lead, and which do not, is what turns a dataset into a signal.

Documents and API

Coverage, history and figures are as at November 2026. Provided for information purposes only; not investment advice. Past performance is not a guide to future results. Trial access, pricing and licensing: [email protected].


Did this page help you?