Back to collection
Research

A portfolio backtest and a task-credit balance answer different questions

SwopLabs-referenced allocation claims concern a simulation; a fixed task-credit conversion concerns a payment liability.

Two women inspect a closed parcel on a workshop table.
Technical illustration

A SwopLabs-referenced account says an AI allocation agent outperformed a 60/40 portfolio in a twenty-year backtest, but supplies no model specification, exact sample dates, cost assumptions or size of the advantage. That is a simulation claim, not a live investment record. The comparison with JumpTask’s stated conversion of 1,000 credits to one dollar concerns a different obligation: a credit balance needs a redemption rule and payment capacity, while an allocation model needs reproducible testing and sensitivity to assumptions. Neither a simulated ranking nor a fixed conversion establishes custody quality or dependable live execution.

References