This analysis uses real executed historical CryptoPunk sales in strict time order to test whether PunkPredictor decisions improved risk-adjusted returns versus the Buy-All benchmark.
Dataset scope: only sales that occurred while PunkPredictor was live, using the latest publicly shown prediction available before each sale timestamp.
P&L measurement scope: only punks with at least two observed sales in this dataset, so each result is a real buy-to-next-sale outcome.
Enter only when predicted sale price (valuation) is above the current sale price.
Enter every eligible sale with no prediction filter, then compare outcomes side by side.
| Metric | PunkPredictor | Buy-All Benchmark |
|---|
Trade construction: Buy Date/Price is the sale where the strategy would enter, and Sell Date/Price is the next observed sale for that same punk. Rows only exist for punks with at least two observed sales.
| Image | Punk ID | Buy Date | Sell Date | Buy Price | Predicted Selling Price | Sell Price | Buy-All Takes? | PunkPredictor Takes? | Profit | Realized Return % |
|---|
