On Saturday evening, a trader pulled $1.28 million out of Papertrade by moving the price of ETH. Another wallet has since shown a quieter variant: 1,012 trades, $10.1 billion in volume, $979,000 in losses on its positions… and a positive final balance. That is what Journal du Coin reports, citing Lookonchain.
A platform that pays those who lose
Papertrade launched on Saturday 10 October on HyperEVM, Hyperliquid’s smart contract layer. It offers perpetuals, futures contracts with no expiry, on BTC and ETH, with leverage of up to 1,000x. There is no order book: a shared USDC pool acts as the sole counterparty, and prices are taken from Hyperliquid.
What sets it apart is the reward mechanism. Strictly speaking, it is not a volume bonus: according to Journal du Coin, every loss or liquidation paid into the pool triggers the issuance of PAPER tokens to the losing trader. This is “loss mining”. The schedule starts at a maximum of 100 PAPER per dollar lost as long as the pool holds less than $2 million, then decreases as it grows. Volume is therefore not rewarded as such; it serves to manufacture eligible losses.
PAPER can be staked. Staking entitles holders to a share of the platform’s revenue, paid in USDC, and the pool’s growth beyond $5 million also goes to stakers. The tokens are, however, not yet transferable: transfers will only open in a later phase, with no date announced so far.
The figures for wallet 0x35b2
According to Lookonchain, cited by Journal du Coin, wallet 0x35b29bdd racked up 1,012 trades for $10.1 billion in volume, or nearly $10 million per trade on average. Its net loss came to $979,000. In return, it received 252 million PAPER and collected around $997,000 in USDC through staking. The balance works out to a profit of around $17,800, not counting the PAPER still held and, for now, non-transferable.
At the most generous rate, 252 million PAPER implies at least $2.52 million in eligible losses. Journal du Coin infers from this that PAPER is calculated losing position by losing position, with winning trades offsetting the losses without deducting anything from the rewards.
The flaw described before launch
The research firm Delphi Digital had described the problem even before launch, Journal du Coin reports. The scheme: open a long position and a short position of the same size simultaneously. At 1,000x, a $10,000 margin is enough to control $10 million of ETH, and a 0.1% move wipes out that margin. One leg is liquidated, its $10,000 goes to the pool and triggers the issuance of PAPER; the other gains roughly as much, paid by that same pool. For the trader, the net result comes down to fees.
Lookonchain does not detail the strategy of wallet 0x35b2. Journal du Coin merely considers that its profile, with tickets close to $10 million and a tiny net loss relative to the volume, fits this scenario.
Two wallets, two separate flaws
The two cases should not be confused. The $1.28 million trader operated from the address 0x2025…840e, identified by The Block. He attacked the price source by moving ETH by $3 on Hyperliquid; his liquidated shorts had cost him $450,000 for 17.6 million PAPER, again according to The Block. Wallet 0x35b2, for its part, exploited the reward schedule without touching the market, and no public data currently links the two addresses.
What this says about incentives in DeFi
The case illustrates a flaw in incentive programmes: as soon as a metric is rewarded, it can be produced artificially. By seeking to compensate losers, Papertrade created an incentive to manufacture losses covered by gains on the other side. According to The Defiant, cited by Journal du Coin, traders had lost a net $18.2 million on the platform as of 10 October, and stakers had received $12.6 million. Every PAPER obtained on the cheap dilutes other stakers and captures a share of the losses suffered by those who genuinely lose.
On top of this comes the dependence on Hyperliquid’s prices, whose HLP vault had already taken $4 million in losses in March 2025 on a single ETH position. As for wallet 0x35b2’s profit, it remains thin relative to the sums committed and rests on a temporary launch schedule. Nothing in these figures suggests a reproducible or profitable strategy.
