Built to a standard that is public, and testable.
Opulence Alpha applies, end to end, the research standards an institutional desk is governed by: a promotion gate a model must clear before it serves a single prediction, point-in-time data with look-ahead excluded at the source, and an independent service that scores, after the fact, whether the beliefs were any good. Those standards are published. This platform is meant to be tested against them.
Every number on this platform is validated by whoever publishes it.
Most investment software is assembled from components that trust each other. A model produces a forecast, a sizer consumes it, a report describes how it went — and no part of that chain has any incentive or ability to contradict another part.
Opulence is built on the opposite convention. A service may publish a quantity only if it also validates that quantity. If nobody backtests it, nobody publishes it. That rule determined the service boundaries, and it is why the platform has a service whose only function is to say, after the fact, whether the beliefs were any good — reporting to nobody, able to promote nothing, and structurally unable to benefit from a flattering answer.
It is a slower way to build. It is the only way to build something a quantitative reader should believe.
For interrogating a view, not for being handed one.
Every trading day the platform publishes what it believes across a 1,266-symbol universe: a distribution rather than a point estimate, the regime state that estimate was formed under, and the evidence behind it. All of it is built to be argued with.
What it supplies is process. Separation between the service that forms a view and the service that decides what risk that view may take. A promotion standard a policy must clear before it is run. An independent measurement of whether last quarter's conviction was skill or luck. Institutional-quality data has been available for a decade; the process around it has not.
Stated plainly: it is for giving a decision you are already making a harder test to pass.
Five services. Five questions.
No service answers another's question, and no service acts on its own conclusion.
Four things this platform will not do.
Opulence is a research system. It does not connect to a broker. Every decision in your account is made by you.
The platform publishes research across a universe. It does not know your tax position, your liabilities, your other holdings or your objectives, and it will not pretend otherwise.
Every belief carries its dispersion, its direction probability and the reason it passed or failed its gate. Where the platform is uncertain, the interval says so.
Including in its own marketing.
Three audiences, and what each uses it for.
Independent evidence against a view they have already formed, and a calibrated read on how uncertain that view should make them.
Defensible, documented research behind client conversations, and standing analysis across many books at once.
A production-grade estate to interrogate — feature store, regime state, belief distributions and the evidence lattice behind them.
The standard this platform is held to.
One rule governs every service: a quantity may be published only if the service publishing it also validates it. Nothing reaches a screen because it was persuasive — a model clears deflated Sharpe, probability of backtest overfitting and a Harvey factor-zoo t-statistic in the promotion pipeline, or it serves nothing at all.
That rule is applied to this website too. Every figure on these pages is drawn from the same evidence layer the product runs on, and a claim that cannot be evidenced to that standard is withheld rather than rounded into something publishable. Correspondence about methodology, validation or the data foundation is answered directly, and technical objections are welcome.