Point-in-time storage
Every record is timestamped at publish moment. Backtests automatically lock the window — no look-ahead bias.
About Horsorion
Hong Kong racing data, built for professionals
Horsorion builds data infrastructure for syndicates, quant teams, and racing analytics groups working on Hong Kong racing.
Our story
Why we built Horsorion
Hong Kong horse racing is one of Asia's most mature betting markets, with 700+ races each season and total annual turnover in the hundreds of billions of HKD. Yet for syndicates and quant teams, clean, standardised historical data has long been the single biggest bottleneck.
Horsorion was founded by a Hong Kong-based quantitative modelling team with one goal: deliver HK racing data truly suitable for modelling — not raw scraper output. We built the full pipeline in-house, cross-validating against public sources and storing every record under point-in-time principles so your backtests stay clean of look-ahead bias.
We deliberately keep a low profile. Client identities, team composition, and pricing are all handled under NDA. Reach out via email or WhatsApp and we will engage on terms you are comfortable with.
Headquarters
Hong Kong SAR
Legal status
HK Business Registration
Data history
1980 – Today
Update cadence
Race-day T+0
Serving regions
APAC · Europe · Americas
Engagement
NDA-backed commercial terms
Global reach · Since 2015
40 years of race records.
Serving syndicates across APAC, Europe, and the Americas.
Global reach
Clients across four continents
From Asia-Pacific to Europe and the Americas, quant teams and racing analysts use Horsorion data to power their models and betting decisions.
- United Kingdom
- United States
- China
- Hong Kong
- Australia
* Active regions shown only — client identities and counts withheld under NDA.
What sets us apart
How we engineer the data
Schema stability promise
Historical columns never change. New columns ship as additive versions, so your pipeline never breaks on upgrade.
Multi-source validation
Every race is reconciled against multiple independent sources plus manual spot-checks before release.
Ready to talk?
Request schemas, samples, or a tailored proposal — we reply within 24 hours.