Forensic

Week 7 — Cash with no bot name

15 July 2026AmsterdamJordi Lluis

This is the wallet report I keep. It replaces “$18 to $221” as an observed path.

July 15 cash forensic
Start$19.979317 USDC.e observed trading collateral.
Liquid peak$210.215219 on 2 May 18:41:54 UTC.
Settled trading P&L−$13.700515. A separate $6.25 left on contracts not tagged TRADE/REDEEM; still unresolved, not relabelled as trading loss.
Dust$0.038307 stablecoins at the audit.
Public CSVDifferent math: assumed $18 start, $221.40 peak. Not liquid cash.

Before the peak: 347 markets, 497 fill rows, +$203.40 reconstructed. After: 790 markets, 1,093 fill rows, −$208.93 reconstructed, more frequent and larger. Take out the top ten pre-peak winners and the pre-peak result is −$29.19. 90.06% of that reconstructed run-up profit has no reliable single-bot label. After the peak, the next 17 sub-$0.10 markets went 0-for-17 (−$44.11).

There was a real early gain and a real later loss. The files do not show a durable named strategy. Lifetime settled trading cash is negative.

Case study · Evidence map

← Blog