Quantitative research & consulting
You bring the hypothesis. We test it properly, against out-of-sample data, realistic costs and the ways a good-looking backtest fools people. You get the honest picture, and the decision stays entirely yours.
The boundary, stated first
We are engineers and researchers, not advisors. We do not design strategies to sell, we do not supply signals or tips, and we do not make recommendations about what you should trade. The idea originates with you; we bring the rigour to test and refine it. Every investment decision, and the responsibility for it, stays with you. If you want someone to hand you a money-making system, that isn’t us, and it’s worth being wary of anyone who says it’s them.
What we actually do
- Validate a hypothesis you bring us. We turn a trading idea into a testable specification and put it through honest out-of-sample and walk-forward testing, not a single curve fitted to the past.
- Test it against reality. Realistic transaction costs, slippage, spreads and market impact, so the result reflects what would have happened, not an idealised fill.
- Check it isn’t fooling you. Overfitting and look-ahead bias, parameter sensitivity, Monte Carlo and regime analysis, the checks that separate a real effect from a lucky one.
- Optimise and implement. Where a hypothesis holds up, we refine its implementation and build it into working software, the custom system that runs it.
How we report
You get the evidence, not a verdict dressed up as a promise. That means the assumptions we made, the conditions under which the idea did and didn’t hold, and where the risks sit, including the uncomfortable findings. We don’t publish or imply returns, and we don’t claim a strategy will make money. Honest research sometimes concludes “not this idea, not yet”, and that conclusion is worth more than a flattering chart.
For a concrete example of that discipline in our own research, see our case study on backtesting a linear-regression support strategy, where a change is only kept if it still holds up on data the tuning never touched.
For licensed managers and firms
Where the client is a licensed asset manager or firm, we work purely as a research and technology vendor: you hold the licence, the mandate and every decision; we run the research and the engineering under contract, for a fixed fee, with no share of outcomes and no performance claims. There’s a dedicated page on that arrangement:institutional research & operations.
Common questions
- Do you give me a strategy to trade?
- No. You bring the hypothesis, an idea about how a market behaves, and we test it rigorously. We’re researchers and engineers, not advisors: we don’t supply strategies, signals or recommendations, and we never tell you what to trade.
- What if the research shows my idea doesn’t hold up?
- Then we tell you plainly, with the evidence. A hypothesis that fails an honest test has saved you from risking real money on it. That’s a useful result, not a failed engagement. We report what the data shows, not what you hoped to hear.
- Can you work under an NDA for a fund or manager?
- Yes. We work as a vendor to licensed managers and firms under NDA. You hold the licence and every investment decision; we run the research and the build. We make no performance claims and take no share of outcomes.