Research
Hypotheses, not hunches. Every idea arrives with a mechanism, a data set and a falsifiable prediction — and most of them fail, which is the point.
Read morePrivate investment manager · Founded 2011
Evidence,
at scale.
We build models from data, test them against markets, and run them with the discipline of an engineering firm. No house view. No narrative. Just systems that survive contact with reality.
01 — In brief
A research firm that happens to invest.
We are organised like a laboratory: small teams, long horizons, and a shared conviction that most market behaviour is measurable if you are patient enough to look.
2011
Founded
41
Researchers
& engineers
& engineers
$2.4bn
Assets under
management
management
3
Offices
Figures as at 30 September 2026.
02 — How we work
Three disciplines, one loop.
Research proposes. Risk disposes. Execution proves whether either of them was right. The loop never closes.
Risk
Position sizing is the strategy. We size from the distribution of outcomes rather than from conviction, and we assume our correlations are wrong.
Read moreExecution
Microstructure treated as a first-class problem. Slippage is a research output, not an operational afterthought.
Read more03 — The pipeline
From raw data to live capital.
One path, no shortcuts. A model that cannot survive every stage is not a model we trade.
Ingest
Point-in-time data, cleaned and versioned. No revisions, no look-ahead.
Hypothesise
A stated economic or behavioural mechanism, written down before testing.
Validate
Out-of-sample, cost-aware, stress-tested across regimes and liquidity.
Deploy
Small at first, monitored daily, retired without sentiment when it decays.
04 — Standard
If we cannot explain why it works, we do not trade it.
Statistical significance is a floor, not an argument. Every live model carries a written thesis and an owner.
-
No black boxes
Models can be complex; their justification cannot be. A strategy we cannot describe in a paragraph does not get capital.
-
Capacity before scale
We size to what a market can absorb, not to what a pitch deck promises. Closing to new capital is a feature.
-
Disagreement is required
Researchers are expected to attack each other's work in the open. Ideas that cannot survive that room do not reach a portfolio.
Careers
We hire slowly.
We keep people a long time.
If you like being wrong quickly, in good company, and then being right for a decade — we would like to meet you.
See open roles