An on-chain claim, and the confidence it supports

Graduate · enrolled learners

This lesson opens with Mango Markets, October 2022.

What happened

On 11 October 2022 Avraham Eisenberg drained about 110 million dollars from Mango Markets, a Solana lending and trading protocol. He funded two accounts with about 5 million dollars, took a large leveraged position in an MNGO perpetual swap, and bought the thinly-traded MNGO token on the exchanges feeding Mango's price oracle, so the oracle price jumped over 13-fold in 30 minutes, then used the inflated value of his MNGO position as collateral to borrow and withdraw almost everything else in the protocol. He then publicly admitted the conduct while asserting it was legal, returned about 67 million dollars under a DAO vote while keeping about 47 million, was charged in a criminal complaint unsealed in December 2022, and was convicted by a jury in April 2024 of commodities fraud, commodities manipulation and wire fraud; on 23 May 2025 the judge set all three convictions aside, for improper venue and, on wire fraud, for want of any material misrepresentation, while finding the evidence of a manipulative device sufficient. Two very different claims sit in this case. The factual claim, that these specific on-chain actions moved the price and drained the protocol, and that Eisenberg controlled the accounts, was supported at very high confidence by the chain and his own admissions. The characterization, that this was illegal manipulation rather than permitted trading, was contested, found by a jury and then set aside by the judge, and in none of it did the chain decide.

The decision point

A research report is a claim about the chain, and the discipline that separates an analyst from a rumor is attaching to each part of a claim only the confidence the evidence supports. The Mango case shows why this must be granular. The who and the what happened, Eisenberg's accounts moved the price and took the funds, were nearly certain, grounded in on-chain fact and his public statements. The what it means, that this was illegal manipulation, was a legal characterization the chain could not settle and that Eisenberg actively disputed. An analyst who reports both at the same high confidence overstates the second; one who hedges the first because the second is contested understates the fact. So a claim is not one label but a set: this part is on-chain fact at high confidence, this part is a supported inference, this part is a contested interpretation, and stating each at its true confidence is the foundation of a report that survives scrutiny.

Recorded loss

$110,000,000

Orientation and Year One are open: anyone can read them without an account. From Year Two onward the lessons are for enrolled learners, because progress through the later years only means anything if it is tracked against a record.

It is free. We do not sell the list and there is nothing to buy at the end of it.