ClickHouse says UK investment manager iSAM Funds cut its options-market data footprint by 35 to 50 times after moving from Postgres, while increasing single-core ingestion from 8,000 to 800,000 rows per second. The case study shows how faster analytics can let researchers work with more history while removing some of the data-processing scaffolding around it.
ClickHouse Watch analysis
What happened
In a case study published on 9 October, ClickHouse describes how iSAM’s options desk built a research platform on ClickHouse Cloud. The company says the team tested ClickHouse alongside Postgres on the same hardware, storing data in time-series order. It reports 35 to 50 times the compression of the Postgres footprint and a jump in single-core ingestion from 8,000 to 800,000 rows per second.
ClickHouse says the platform keeps full options-market history available for queries, helping researchers test ideas across a wider data set. The desk stopped maintaining some pre-computed rollups and materialised views because, according to its head of options development, queries ran quickly enough on the columnar database. The researchers use the desk’s own systems to access the data rather than querying ClickHouse directly.
Why it matters
Those figures are from a customer case study, not an independently controlled benchmark, but they make the practical trade-off unusually clear. Faster ingestion and compression can make it feasible to retain and query more market history, while reducing the extra pipelines built to prepare that data in advance. In this account, the payoff is a shorter path from a researcher’s idea to a test.
The story also shows why the database is not the whole deployment decision. iSAM chose ClickHouse Cloud rather than running the service itself, and the case study describes AWS connectivity, enterprise authentication and backups to the firm’s own storage buckets as part of its setup.
Our read
The most interesting result is not the eye-catching ingestion figure on its own. It is the combination of keeping more history online and simplifying the route from raw data to a query. That is a compelling case for this particular workload, not a promise that every Postgres system will see the same gains. If you are weighing a migration, treat these numbers as a reason to run your own representative test, not as a substitute for one.
What to watch
- Whether iSAM or ClickHouse shares independent measurements or workload details behind the reported gains.
- How the platform performs as iSAM adds more options markets and reconstructs larger historical data sets.
- Whether the reported reduction in pre-computation holds for other workloads, not just this options research platform.
Discussion spark: When a database migration promises both far faster ingestion and fewer pre-computed data pipelines, which would matter more in your own workload: the speed gain or the simpler architecture?
Sources and evidence
- Source update (9 October 2026, 14:01 UTC)
not affiliated with or endorsed by ClickHouse