The model where losers support winners may face its ultimate test in the future.

Written by: Maher, Foresight News

An on-chain casino with up to 1000 times leverage was exploited on its first day, with $128 million lifted using a cost of $52,000.

This incident occurred on Papertrade.

That evening, a professional player first purchased about 6,060 ETH, raising the price by 0.12%; the collateral for the thousand-fold long position doubled instantly; an hour later, they opened 30 short positions nominally worth $297 million, then sold off 6,590 ETH to crash the market. The cost on the Hyperliquid side was about $52,000, while the withdrawal from Papertrade was about $128 million. This trader had nearly two years of trading records on Hyperliquid, and this transaction went through normal trading channels.

Winners draw blood, losers provide blood

Papertrade is a synthetic perpetual contract platform built on Hyperliquid HyperEVM, launched on October 10: there is no order book, no market makers, no funding rates; all users bet against a shared USDC pool, with prices directly reading the mid-point of the best buy and sell prices from the Hyperliquid order book.

After this incident, the community gradually began to realize that the protocol's "losses equal mining" sustainability faced significant uncertainty.

The core and most contentious design of Papertrade is its relationship binding the liquidity pool (LP) with the PAPER token.

This pool was initially empty and completely filled through losses or liquidations realized by traders. When users make profits, their profits come from the pool; if the pool temporarily lacks funds to cover, excess profits go into a queue to await payout after subsequent losses come in. Meanwhile, users who incur losses will receive PAPER tokens.

According to the documentation, when the cash balance in the pool is below $2 million, every $1 loss mints 100 PAPER; above this threshold, the minting ratio gradually decreases according to the high water mark. PAPER can be staked to share protocol revenues (including a share of profits from profitable trades and excess portions after the pool exceeds the set limit).

This mechanism is described by the team as "turning liquidation into ownership": losers receive a share of the platform's future profits while incurring losses, equivalent to gradually becoming joint owners of the house. Theoretically, as long as a majority of highly leveraged traders continue to incur losses, the pool will continue to be filled, and PAPER holders will be able to receive dividends, allowing the system to continue.

However, the point of contention remains whether this model can be stable in a confrontational environment, as it relies on one premise: the number of losers must exceed the number of winners, and the speed of losses must cover the payouts of profits. Once there are concentrated large profits (for example, triggered by pushing external quotes to achieve targeted withdrawals), the pool can be rapidly depleted.

Additionally, the former is human nature; in high-leverage markets, most participants are destined to incur losses, making this a high probability in the short term. Initial data from the launch period also partially validated this point. Data showed that on the first day, traders' net realized losses were estimated to be in the range of $18 million to $30 million, while stakers received rewards amounting to about $10.3 million to $12.6 million. The official later stated that approximately $25 million had been distributed to staking users, and the liquidity pool size once reached about $5 million.

However, maintaining a continuous inflow of losses is fraught with uncertainty: it requires losses to be a perpetual cash flow rather than a stagnant amount. But the market will not always generously supply new losers; trends, professional players, and targeted withdrawals can turn inflows into outflows at any given moment.

Prior to the protocol's launch, Delphi Digital analyst Tempest had publicly warned about this: the long-short squeezing will dilute the real losses that get the airdrop, and also inflate the apparent data of "loss inflow." Therefore, its loss inflow data may not be as much as officially disclosed.

The price fluctuation involved in this attack was only 0.12%.

If one day BTC or ETH experiences a unidirectional market of 1% to 2%, with positions in the same direction collectively profiting, the $5 million-sized pool will face a nominal exposure in the tens of billions, and the payout queue will truly undergo its test.

Aside from the ability to maintain payouts, the protocol’s queueing has also faced criticism from the community,

Payouts must also queue

Since the liquidity pool started from zero and fully relies on subsequent losses for filling, when users close a profitable position, the protocol will immediately return their initial margin, but if the pool currently lacks sufficient funds to cover all profits, the excess portion will enter a queue, waiting to be paid out after subsequent losing traders’ losses fill in.

The documentation clearly states: the pool allows negative balances to occur under specific circumstances, and the unpaid profits become public debts rather than immediately withdrawable balances.

This mechanism was magnified at the early stages of launch. The platform expected numerous users to trade simultaneously, thus employing a queuing priority arrangement—officials acknowledged that, due to technical constraints, small nominal value orders could not get the same priority as large orders; otherwise, the front end might face a lot of junk requests leading to congestion.

The team claims the existing queuing mechanism has been improved for fairness under limiting conditions, but community feedback shows that some users experienced notable confirmation delays and priority differences on the first day of launch.

Some complained, "Winners have to queue FIFO for the next loser to get liquidated," believing this resembles using paper profits as IOUs rather than instant settlement in an exchange.

The queuing issues intersect with price manipulation controversies. In response, officials mentioned that due to anticipated congestion at the initial stage, they set the position limit somewhat loosely. The queue itself, due to the uncertain transaction confirmation time, also naturally guards against manipulation. As traffic has decreased, the team has begun tightening position restrictions on-chain and at the relay level.

However, from the user’s perspective, the queuing uncertainty has twofold impact: first, profits cannot be withdrawn in a timely manner, affecting capital turnover; second, in a confrontational environment, if targeted withdrawals continue to occur, the queue may lengthen, further reducing new users’ willingness to enter and trade continuously.

From a mechanism design standpoint, queuing is the inevitable outcome of "bootstrapping the pool from zero": without external market makers or pre-injected funds, profit payouts can only rely on the inflow pace of subsequent losses. While this can theoretically sustain itself, in actual operation, confirmation delays, priority differences, and potential queue backlogs will increase user friction.

Currently, public data has not provided a complete time series for queuing debts, but compared to centralized exchanges and other DEXs, it remains a significant pain point. For a platform that boasts high leverage and zero slippage, the uncertainty of settlement could weaken its experience advantages compared to traditional perpetual exchanges. If the queue persists for long or repeatedly lengthens during fluctuations, it could impact the platform's trade retention and the liquidity pool's filling efficiency.

If the pool runs low on funds, winners will have to queue; as queuing times extend, some users may choose to leave, reducing new loss inflows, making it harder for the pool to recover, creating a vicious cycle.

If one day, the first large market movement that causes the payout queue surfaces, this "losers support winners" assembly line will face its ultimate test.

Arbitrage by opportunists, 88% drop in open contracts

From grand narrative to a mess on the ground, all it takes is a few days.

After its official launch, the platform's cumulative nominal trading volume once broke the staggering mark of $600 billion, while open interest (OI) skyrocketed to above $3.3 billion.

However, according to the latest data on the official website, its OI has now plummeted to around $400 million, a reduction of approximately 88%. This is far from just market cooling, but a typical "locust invasion"—once the underlying mechanism flaws were thoroughly understood, professional studios and opportunists completed a wild extraction of the liquidity pool and quickly retreated, leaving behind a devastating liquidity vacuum.

What supported this false prosperity of tens of billions in scale were the risk-free "hedged crypto" arbitrage operations. In this zero-sum game of "losses equal mining" with a thousand times leverage, real retail investors have become the minority, while opportunists are indifferent to the long or short direction. They utilize dual accounts to open positions in neutral hedges at same price and amount. Under extreme leverage, tiny price fluctuations can cause one side of the position to crash rapidly, triggering the payout mechanism to mint PAPER tokens at very low cost. Meanwhile, the profitable position on the other end exits risk-free, directly withdrawing real USDC from the LP pool.

This structural scheme of "transferring from one hand to the other" effectively turns the protocol-backed liquidity pool into an unconditional ATM. When opportunists have drained the initial pool liquidity, leading to endless queuing for payouts, the breakeven point of this "cryptocurrency play" is broken.

Currently, PaperTrade is built on HyperEVM, with prices directly reading from Hyperliquid's BBO. The trading activity itself consumes Hyperliquid's block space and liquidity attention, but if PaperTrade continues to attract high-leverage trading demand, it may divert some extreme leverage users originally headed for Hyperliquid. Conversely, if the protocol's model and risk control prove repeatedly fragile, it might become a marginalized niche product.