Deflated Sharpe ratio calculator (PSR and DSR)
Deflated Sharpe calculator
Calculates the Probabilistic and Deflated Sharpe Ratio, exposing trial count, dispersion, sample length and non-normal return assumptions.
Observed, simulated, model-estimated and planned claims remain visibly separate.
Open research instruments7 calculators
Canli Capital, the research house that publishes ALPHAC, built each of these. Every one runs entirely in your browser, is free, requires no account or key, and is built from the same source-bound arithmetic this project's own published record is held to. Most started life under a house name; each now names itself for what it computes, and keeps the house name as a subtitle where one exists.
Deflated Sharpe calculator
Calculates the Probabilistic and Deflated Sharpe Ratio, exposing trial count, dispersion, sample length and non-normal return assumptions.
Runs Combinatorially Symmetric Cross-Validation on your own matrix of variant returns, the same core the validation API imports.
Selection Risk Lab
Searches a series with provably no edge and deflates your best result once the search is counted against it.
Execution Reality Lab
Prices one strategy under six fill, delay and impact assumptions, and measures which are costs and which are re-timings.
Breadth Lab
Computes how much a book (the set of sleeves run together) of N sleeves is worth, and the correlation ceiling that no amount of breadth can pass.
Trial accounting
Reconstructs the public search denominator: every hypothesis identity, legacy and prospective, that a deflated Sharpe here is corrected against.
Evidence chain
Recomputes the signed, append-only hash chain behind this record in your browser, so you do not have to trust the website that shows it to you.
A single Sharpe ratio says nothing about how many variants produced it. Run the deflated Sharpe calculator against your trial count, then check whether your in-sample winner actually predicts anything out of sample with the overfitting calculator.
The selection risk simulator lets you search a series that provably has no edge, so you can see a good-looking number turn into what it actually is before it happens to your own research.
The execution cost calculator separates costs, which reduce a result on every series, from re-timings, which only hurt on average.
The portfolio breadth calculator shows the correlation ceiling that no amount of added breadth can pass.
The same validation runs through a free keyed API: see the API quickstart for the deflated-Sharpe, overfitting and breadth routes. None of these tools writes to any ledger, artifact or published record; every input stays in your browser.
For the full walk-through with the API route beside each calculator, see how to validate a backtest.