Evidence the signal can generate before exponential economic decay exhausts its useful life.
Higher initial Sharpe and longer half-life expand the available information budget.
Enter its initial Sharpe, economic half-life and how many candidates you screened. CAC compares the total evidence the signal can ever generate with the evidence your certification standard requires.
Above the red curve, lifetime information meets the declared evidence requirement.
To reach the frontier, keep SR 2.00 and raise half-life to 7.24 years, or keep the current half-life and raise Sharpe to 2.69.
A Sharpe-2 signal selected from 100 candidates needs a 7.24-year half-life to reach the default certification frontier.Canonical Gaussian contract: 5% family error budget, 90% power, Bonferroni correction.
The browser exposes the closed-form canonical calculation. Richer unknown-scale and data-driven certificates remain in the AlphaValue Python package.
Evidence the signal can generate before exponential economic decay exhausts its useful life.
Higher initial Sharpe and longer half-life expand the available information budget.
Evidence demanded by reliability, power and the search multiplicity declared before deployment.
Screening more candidates tightens the per-candidate error budget and moves the frontier upward.
Compare the economic information budget with the information required by the deployment contract.
Below one, the declared lifetime is too short under the chosen canonical contract.
A significance test asks whether observed performance clears a statistical bar. A track-record calculation asks how long a stable signal needs. CAC asks whether a decaying signal has enough lifetime information to clear the bar at all.
Ask whether the performance already observed is distinguishable from luck, possibly after accounting for selection.
Ask how much observation time a signal of constant strength needs to reach the chosen evidence standard.
Ask whether the entire remaining economic life of a decaying signal contains enough information to reach the evidence requirement at all.
Canonical scope. The interactive threshold uses the paper's Gaussian, known-volatility, exponential-decay, fixed-sample contract. Reaching the frontier means certification is feasible under that contract; it is not a profitability claim.
Mathematical guarantees, numerical constructions and descriptive evidence answer different questions.
A reliability frontier in a specified statistical experiment.
The canonical Gaussian model puts statistical learning and opportunity decay on a common information clock. Its feasibility frontier determines whether the declared reliability contract can be met.
The guarantees require the likelihood and observation assumptions in the paper. Cash value also depends on the payoff map and its uniform regularity conditions.
A descriptive empirical illustration.
The funding panel examines how a lifetime score relates to persistence in future signed funding, including information beyond contemporaneous funding magnitude.
This association does not establish executable hedged-trade returns. No confirmatory trading-profit validation is claimed.
Residual information and cash payoffs have distinct asymptotic scales.
The calculator is one surface of a theorem-driven release with source, tests and a reproducibility record.