Concepts

How it works

What gets measured, what gets published, and what each guarantee means.

Percentages, never amounts

The public record contains ratios only. There is no balance, position, trade or exchange identifier in it, so someone reading your profile cannot tell whether you trade with a hundred dollars or a hundred thousand.

One combined record

All your connected accounts are added into a single curve, counted from when you connected them. You cannot pick the flattering account or hide the weak one. That is what makes the record hard to curate. Your one choice is whether the record is public at all, through privacy mode.

What is measured

MetricMeaningNeeds
ReturnHow much the record grew since you connected, ignoring deposits and withdrawals.None
Max drawdownThe biggest fall from a peak, as a percentage.None
Sharpe ratioReturn relative to how much it swung day to day (risk-free rate taken as zero).30 days
Win rateThe share of closed trades that ended in profit.A closed trade

Return is time-weighted: the period is split at every deposit or withdrawal and the pieces are chained, so adding money neither creates nor hides performance. Crypto is valued in a stablecoin unit, which is not the same as a dollar, and the unit is always stated.

When there is not enough history, a metric is shown as unavailable. That is different from zero.

Proofs and signatures

Two things can back a statement. Your signature shows that you authorized exactly that statement. A zero-knowledge proof shows that a result was computed correctly from data that stays hidden. They answer different questions, and Linvesther never presents one as the other.

Proofs of performance exist today for individual exchange accounts. Proofs for the combined record and for claims are planned. See Security and limits.

Whose data it is

A proof shows a calculation is right. It cannot show the data it ran over was honest, because whoever supplies the data can sign anything. So every proof records which collector signed the data, and a verifier compares that against the collectors it trusts.

  • Trusted collector. The signer is on your list, and the proof covers a period the collector is trusted for.
  • Self-attested. The signer is not on your list. The figures are only as trustworthy as whoever holds that key. This is what you get from an instance someone runs for themselves.
  • Revoked or out of date. The collector was on the list but is no longer trusted, or the period is outside the dates it was trusted for.

Public profiles and claims show which of these applies. The list is yours to keep: see Verify a proof.

Five independent guarantees

There is no single trust badge. A result is judged on five separate questions:

GuaranteeThe question
OriginWho supplied the data. A collector reads your exchange with a read-only key and signs what it saw; proofs name that collector. That is not the exchange signing the data.
CoverageWhether the data covers the period without unexplained gaps.
CalculationWhether the result comes with proof that it follows from the data by the published method.
RegistryWhether the result is recorded on the public blockchain.
AvailabilityWhether the evidence needed to check it can still be reached.

Being strong on one says nothing about another. A perfectly proven calculation over false data is still a false result, which is why origin and coverage are shown on their own.

How it works