AI Economy

At publication, the price of every asset is recorded along with its source and timestamp. That snapshot is the fixed starting point, and its fingerprint is written onchain so it cannot be altered afterwards.

asset return = (current price − publication price) ÷ publication price
basket return = Σ (weight × asset return)

The benchmark is basktr-core-10-v1, an equal weight of 10 liquid assets (BTC, ETH, SOL, AAPL, MSFT, NVDA, AMZN, TSLA, SPY, QQQ), measured over exactly the same window. It is a fixed published list, not whatever the app happens to offer, so adding new assets can never change what a score means. Its prices are frozen into the snapshot at publication, so it cannot drift later either.

outperformance = basket return − benchmark return

Tokenized equities carry the issuer's share-per-token multiplier. If a corporate action changes it while a basket runs, the token price moves by the same factor without anyone gaining anything, so both ends are restated in the terms they were published under and the correction is disclosed on the basket.

Outperformance is the headline because raw return mostly reflects when a basket was published. Measuring against the same window answers the only question that matters: did the weighting beat simply holding everything?

No score, rating or projection is applied. If any price is unavailable, the result is reported as unavailable rather than estimated.