Quantitative research
Research you can reason about.
Tiger Prism is a causal-AI research and risk platform for quantitative investment teams. Licensed software — your data, your parameters, your conclusions.
- Causal, not correlational
- Every signal inspectable
- Your data, your parameters
Correlation tells you what moved together
Most quantitative platforms are optimisation engines built on correlation. They are good at finding what has moved together and projecting that forward. They are silent on why.
That gap matters most exactly when it costs most. Relationships that held through a decade of one regime stop holding when the regime changes, and a model built on correlation gives no warning, because it never encoded a mechanism to break.
Why this is hard to fix
Adding explainability to a correlational model usually means attribution after the fact — which factor loaded, how much it contributed. That tells you what the model did. It doesn't tell you whether the relationship it found is causal, spurious, or the product of a confounder neither of you has named.
Mapping direction, not just association
Tiger Prism is built around causal discovery. Rather than measuring how variables move together, it maps directed relationships — which variable acts on which, and through what — across macro factors, rates, volatility indices, cross-asset signals and instrument returns.
The output is a graph you can read. Every analytical signal carries the chain that produced it, so a researcher can interrogate the mechanism rather than accepting a number.
What that enables
Because the relationships are directed, you can ask intervention questions rather than only historical ones. What happens downstream if real yields move fifty basis points and credit spreads hold? A correlational model answers by analogy to periods that looked similar. A causal model answers by propagating the change through the structure it has learned, and shows you the path it took.
And what it doesn't claim
A causal graph is a model, not the world. Tiger Prism surfaces the assumptions behind each edge and the strength of the evidence for it, because a structure presented without its caveats is just a more confident kind of guess.
What the platform produces
Illustrative only. Clients load their own instruments, set their own parameters and draw their own conclusions. Quantius does not provide investment advice.
From discovery to decision support
Factor attribution
Decompose returns into contributions from the factors that drove them, with the causal path for each rather than a regression loading in isolation.
Scenario analysis
Define an intervention, hold what you choose to hold, and model the downstream path across your instruments. Assumptions are stated alongside every result.
Risk decomposition
See where portfolio variance actually originates, by instrument and by underlying factor, and how that concentration shifts under scenario.
Exportable research
Reports and risk summaries are generated by the platform from your data and your configured parameters, ready to export for your own internal use.
Human in the loop, by design
Every model output is visible in full. There are no hidden scores and no black-box recommendations, because a research tool whose reasoning you cannot inspect is not a research tool.
All decisions remain entirely with the client.
What Quantius is, and is not
Quantius LLC is a software company. Tiger Prism is a licensed platform. Quantius does not manage, advise on, control, or have any access to client investments, portfolios, capital or financial accounts at any time.
Clients load their own portfolio data, strategies and risk parameters, run their own research, and reach their own conclusions. Reports are produced by the software from client-defined inputs — Quantius does not prepare, author or contribute analytical content to any report.
Quantius LLC is not a Registered Investment Adviser, broker-dealer, commodity trading adviser, or any other form of regulated financial services provider under US law.
Who we work with
Quantitative hedge funds
Systematic teams that want causal structure alongside existing signal research, without building the infrastructure in-house.
Family offices
Sophisticated allocators with real analytical needs and no appetite for a standing quant team.
Boutique asset managers
Firms that want institutional-grade quantitative capability proportionate to their size.
Registered investment advisers
Teams needing defensible, explainable analytics they can stand behind with clients and committees.
Start with your own data
Bring your instruments and your parameters, and we'll set up a working environment for your team to evaluate against research you've already done.