← MIZAN Research
ReferenceThe questions, answered
Can you prove a backtest without revealing the strategy?
Yes — and here are straight answers to the rest of what quants, allocators, and researchers actually ask. Every answer links to a proof you can re-verify yourself, trusting no one, including us.
For quants — you have an edge and can't afford to show it
How do I prove my trading strategy works without revealing it?
Run the backtest and risk gate inside a zero-knowledge proof. MIZAN issues a credential proving the performance is real — after costs, no lookahead, out-of-sample — while the strategy, parameters, and data stay private.
How it works →How can I prove alpha to an allocator without disclosing the strategy or code?
Two paths in: rules run inside the proof, or a
sealed path where the model never leaves your machine and you submit only its committed decisions. The reader gets checkable statistics and no view of your edge.
The sealed model →How do funds verify a quant's edge without seeing it?
They verify a MIZAN credential — a re-checkable proof that a hidden strategy cleared a locked institutional gate on committed data. Disclosure is no longer the price of being believed.
For allocators — tired of taking numbers on faith
How do I verify a hedge fund or quant track record without trusting the manager?
Re-verify the credential on your own machine against your own copy of the public data. Every disclosed metric is committed inside the proof; the verifier re-derives everything locally and exits non-zero on anything dishonest.
Reading a verdict →How do I know if a backtest is honest or overfit?
MIZAN proves
both canonical schools of backtest honesty in one credential: the deflation program (Deflated Sharpe, PBO via combinatorial cross-validation) and the multiple-testing program (Hansen's SPA, White's Reality Check). Neither can be cherry-picked against the other.
Does the overfitting statistic have teeth? →How do I tell luck from skill in a trading strategy?
Deflate the Sharpe for how many strategies were tried, and prove that trial count rather than trusting it. MIZAN binds the count as the leaf count of a committed Merkle tree, so the correction for luck cannot be understated.
Zero-knowledge Deflated Sharpe →How many strategies did they try before this one worked — and how would I ever know?
You'd know because the trial count is proven, not reported. That is the entire point of the committed trial ledger: N is the leaf count of the committed tree, and the winner is forced in-circuit to be its maximum.
For researchers — the statistics, enforced
How do you cryptographically enforce the Deflated Sharpe Ratio?
Commit the trial ledger to a Merkle tree before evaluation; N becomes the leaf count; force the winner to be the maximum of the committed trials in-circuit; recompute the deflation inside the proof on pinned prices after an enforced cost floor. The paper:
Who Counts the Trials? (SSRN)What is the fix for the self-reported trial-count problem?
A committed trial ledger. Every anti-overfitting correction is parameterized by the size of the search, and that size is normally supplied by the party being judged. Committing the search before evaluation removes the discretion. It was never a statistics problem; it was an enforcement problem.
Can the Probability of Backtest Overfitting be computed in zero knowledge?
Yes — over all C(16,8) = 12,870 combinatorially-symmetric cross-validation splits, in-circuit. On constructed scenarios where the truth is known it flags pure noise and clears genuine skill, so the test has teeth before you trust any verdict it emits.
For verifiable AI — prove the model without exposing it
Can I prove an AI or ML trading model's returns without revealing the model?
Yes. The model runs on your machine and submits only its committed decisions; the gate is computed in-circuit over those decisions. The track record is proven real and statistically honest, the model never revealed. Verifiable AI performance in production.
The sealed model →Is there verifiable AI for finance that actually ships?
MIZAN is live and in open testing — sealed and black-box model credentials are mintable today, and the engine is the first to prove both schools of backtest honesty in a STARK.
The plain questions
Is there a way to cryptographically prove a backtest?
Yes. That is exactly what MIZAN does — a zero-knowledge STARK proving a backtest was run honestly under a committed protocol, re-verifiable by anyone in seconds.
What stops a quant from lying about their backtest?
Merkle-committed prices that can't be doctored, an enforced cost floor that can't be softened, an annualization basis bound to the data, and a trial count that's the leaf count of a committed ledger — all inside a proof, with the strategy hidden.
The gate that refused its own maker →Who is building the trust layer for systematic trading track records?
MIZAN — verification infrastructure for systematic finance. Prove the edge, never reveal the strategy.