The Negative Fee Mirage: HTX 'Trade to Earn' and the Arithmetic of Subsidized Liquidity
The ledger remembers what the headline forgets. HTX, the exchange formerly known as Huobi, closed the first phase of its "Trade to Earn" campaign with a set of numbers designed for a press release: up to 110% fee rebates on selected perpetual contracts, a daily prize pool of 6,000 USDT, a reported 63.37 million USDT in trading volume, and a ceremonial burn of 1.8 billion HTX tokens. The narrative is that this is a new chapter in exchange economics, where traders are paid to trade and the platform token becomes more scarce. The arithmetic suggests something else. A rebate above 100% is not a revenue share; it is a cost. A burn that is financed by a promotional budget is not deflation; it is marketing. A campaign that rewards trading volume without requiring the exchange to earn fees is not a sustainable flywheel; it is a combustion engine running on treasury fuel. The price of the token may rise for a week because of the noise. But the chain does not read press releases. It indexes cash flows, and the cash flow here is negative.
This is not moral outrage. Outrage is a luxury I abandoned after the 2017 Tezos audit, when publishing a 40-page vulnerability report taught me that most projects do not want clarity; they want cover. What follows is a forensic read of the HTX campaign, not as a headline but as a system. I want to answer one question: if every user gets 110% of their fees back, and the exchange also funds a prize pool, then who pays for the exchange? The answer is not mystical. It is the same answer that has appeared in every subsidized liquidity scheme since the first crypto exchange discovered that volume can be bought. Retail pays, market makers collect, and the platform monetizes attention while it lasts.
This campaign must be placed in context, because context is a form of evidence. HTX is not a new protocol with a novel idea. It is an aging centralized exchange with a history of turbulence. Huobi was once a top-tier venue in the Asia-Pacific region, but after the departure of its founder, after an acquisition that masked a series of capital structure questions, and after the tightening of regulatory pressure in mainland China, Huobi became HTX under the orbit of Justin Sun. The migration was not about technology. It was about branding, jurisdiction, and survival. An exchange in that position does not typically run a charity. It runs an acquisition campaign. And when a Venue with a declining user base offers to refund more than 100% of fees, the first question is not generosity; it is desperation.
"Trade to Earn" is a rebrand of an older mechanism. The industry called it transaction mining in 2018, when certain exchanges rewarded users with platform tokens for generating volume. The pattern was simple: inflate token supply, distribute it as a reward for trading, watch volume spike, and call the resulting price appreciation a tokenomics victory. The flaws were also simple. The rewards were funded by dilution. The volume was often wash trading. The users were not customers; they were extractors. What HTX has done is add a traditional finance wrapper to that old model. The underlying instruments are perpetual contracts on QQQ, NVDA, and MSFT. The marketing language is "TradFi meets DeFi," which sounds like convergence but is really a regulatory end-run. Stock perpetuals on a crypto exchange in Seychelles are not a bridge to traditional finance. They are a synthetic derivative product that exists in a jurisdiction where retail protections are thin. The only new technology in this campaign is the wording of the announcement.
Let me dissect the mechanics first, because the mechanics are where the truth leaks through. In a standard perp market, the exchange charges a taker fee, typically a few basis points, and the maker receives a rebate for adding liquidity. In the HTX promotion, the exchange rebates up to 110% of fees. That means for every one unit of fee collected, the exchange returns 1.1 units. The extra 0.1 does not appear from air. It appears from the exchange's treasury, from its marketing budget, or from the future sale of HTX tokens to someone who has not yet arrived. The 6,000 USDT daily prize pool is an additional expense. If the campaign runs for 30 days, that is 180,000 USDT of direct cash expense before computing the negative fee subsidy. This is not a revenue-generating business line; it is an expense line. The exchange is selling a dollar of trading infrastructure for 90 cents, and then handing out lottery tickets on top. No auditor can call that income. It is customer acquisition cost. The ledger labels it as such, even when the marketing department labels it as innovation.
The deeper problem is with the token burn. HTX announced the burn of 1.8 billion HTX as evidence that real value is being returned to holders. On its face, a buyback-and-burn is a transfer of protocol profits to token holders. But this burn exists in a circular relationship with the reward pool. The campaign distributes HTX as rewards. The campaign generates fees on trading volume. The exchange uses those fees, or its own balance sheet, to repurchase and burn HTX. The question that no announcement answers is whether the burn is larger than the sum of rewards distributed into circulation. If the exchange pays out 1.8 billion HTX in rewards and then burns 1.8 billion HTX, the net supply change is zero. That is not deflation. If the exchange pays out 2.5 billion in rewards and burns only 1.8 billion, the net supply change is positive, meaning the burn is a narrative decoration placed on top of silent inflation. Without a public wallet-level reconciliation of the reward pool source and the burn address source, the claim of scarcity cannot be verified. From my audit experience, whenever a project announces a burn to justify a token's value, the first question I ask is: where did the burned tokens come from, and where did the rewarded tokens come from? If both answers point to the same wallet, the ceremony is a zero-sum illusion. Silence in the code speaks louder than the pitch.
Now let me discuss the real balance sheet of Trade to Earn, because the sustainability math is the core of the teardown. Consider a professional market maker entering the HTX order book during the campaign. The market maker places quotes on both sides. It earns maker rebates, which the campaign may amplify. It captures the spread. It also receives the 110% fee rebate on any taker volume it generates. Its cost basis becomes negative. The prize pool is a pure lottery option, but at sufficient volume, the expected value of the lottery is positive. A quantitative trading firm can treat this as a guaranteed return, constrained only by position limits and the exchange's willingness to pay. Retail traders see the same numbers, but they do not have the same execution tools. They enter with directional bias. They see negative fees and assume that margin for error has expanded. In reality, the market maker is not the retail trader's counterparty in some collaborative profit machine; it is the opposing side of the trade. The negative fee does not reduce the market maker's edge. It funds that edge. Retail, meanwhile, pays the spread, pays the funding rate when the crowd is on the wrong side, and earns a reward that is calibrated to be too small to compensate for adverse selection. The network of subsidy flow has a simple topology: treasury money flows to the market maker, the market maker extracts spread from retail, and the exchange converts that temporary volume into a quarterly report that it can present to future investors. The campaign is not a return on investment. It is a return of marketing expense to a professional class.
The question of who benefits is not idle curiosity; it is a risk model. Every bug is a footprint left in haste, and this campaign is a footprint of an exchange trying to reverse an outflow of users. The exchange's center of gravity has shifted toward aggressive promotions because its organic liquidity has weakened. A genuinely liquid venue does not need to pay 110% of fees to attract volume. Binance, OKX, and Bybit do not offer permanent negative fees on blue-chip stock perpetuals because their order books already have sufficient depth. HTX is buying a market share that it cannot retain, and the market share it buys is disproportionately composed of arbitrageurs and extractors. The users who arrive because of the subsidy will leave when the subsidy ends. The cost remains. The exchange has spent cash to acquire volume that never intended to remain loyal. In accounting terms, this is a failed retention investment. In game-theoretic terms, it is a classic adverse selection result: the people most likely to respond to an excessive incentive are the people least likely to behave like long-term customers.
The regulatory component is where this campaign stops being merely reckless and becomes dangerous. HTX is offering perpetual contracts on individual company stocks and an index. A perpetual contract on NVDA is, in substance, a synthetic equity derivative. In the United States, retail access to such products through an unregistered offshore exchange violates the spirit of the Commodity Exchange Act and the Securities Exchange Act, and in many interpretations the letter as well. The SEC has repeatedly pursued unregistered securities offerings and platforms that facilitate them. The CFTC has a long history of bringing actions against offshore platforms that offer leveraged retail commodity trading to U.S. persons. The EU's MiCA regime imposes clear rules on crypto-asset service providers, and many EU member states maintain local prohibitions on CFD products for retail investors. The fact that the contract is settled in USDT does not change the underlying asset class. It is a 75x levered bet on Nvidia's share price, offered by a Seychelles entity to a global audience. That is not innovation; it is regulatory arbitrage. I have testified to the view that on-chain surveillance can distinguish legitimate activity from evasion, but even perfect surveillance cannot make an illegal retail derivative legal. The exchange is not merely testing the boundaries of tokenomics; it is testing the boundaries of securities law with real user funds.
This is not the first time I have seen a project ignore a structural warning. In 2020, when I analyzed Yearn.finance's yield aggregation strategies, the advertised APYs did not account for impermanent loss, and my report "The Illusion of Infinite Yield" showed that the marginal retail depositor was earning negative expected value after fees and slippage. The response from the community was not gratitude; it was hostility. In 2022, when I reconstructed the UST de-pegging timeline for Terraform Labs, I found internal risk warnings that had been ignored for six months because the founders believed the market would keep subsidizing their growth. The Luna collapse was not a black swan; it was a chronicle of incentives that could not balance. HTX's Trade to Earn campaign has the same chronological signature: a high-cost incentive, a narrative that declares the cost a value mechanism, and a reliance on continuous inflow to prevent the ledger from revealing the loss. History is not written; it is indexed. And the index for this campaign is already negative.
Let me now examine the competitive dynamic, because the market context matters. The exchange sector is brutally efficient. If HTX runs a negative-fee promotion, Binance, OKX, or Bybit can run a deeper one within days. There is no technical moat defending this campaign. The only defense is balance sheet size, and HTX's balance sheet is not the strongest in the industry. A subsidy war for trading volume is a war of attrition, and the exchange with the most durable capital wins. HTX does not have the capital base of Binance. It does not have the regulatory track record of Coinbase. It does not have the derivatives reputation of OKX. What it has is a willingness to spend aggressively and an asset, HTX, that can be printed from treasury to fund the burn ceremony. That is a war chest, but it is a war chest filled with the same asset it is trying to support. Using HTX rewards to attract volume, then buying and burning HTX to reduce supply, is a closed loop. It is not a revenue engine. It is a token velocity machine that generates the impression of activity while consuming external cash for the small portion of rewards paid in USDT.
I want to be precise about the distinction between marketing and law. There is nothing inherently fraudulent about a promotional campaign that loses money. Retail promotions often lose money. A restaurant can sell a dollar burger at a loss to attract customers. But the restaurant does not simultaneously announce that the dollar burger loss is proof of a new economic paradigm. The restaurant also does not claim that the resulting traffic justifies issuing new ownership stakes to those customers. HTX is doing all three: losing money on fees, labeling the loss a token burn, and rewarding users with the token that is supposed to become more scarce. This conflation of expense and income is the core of the deception. The industry has a word for a system that pays early participants with future inflows while promising that the token's price will rise because of the payment. That word is Ponzi-like. I am not accusing HTX of criminal fraud. I am stating that the incentive structure bears the formal signature of an unsustainable subsidy scheme, and that the formal signature is visible in the public materials. I have audited enough token models to know that the question is never whether the campaign works in week one. The question is whether it works in week fifty-two. For HTX, the only thing that can last fifty-two weeks is the burn of capital.
There is also an oft-ignored operational risk embedded in this campaign: the centralization of asset custody and reward distribution. Everything depends on the exchange's word. The actual volume, the actual negative fee rebates, the actual prize pool winners, the actual burn address, and the actual treasury balance are all controlled by a centralized entity with a contractual domicile in a low-regulation jurisdiction. Users cannot verify the exchange's accounting from chain data because the trading engine is off-chain. The ledger of the exchange is a private database, not a public hash. This is why I insist on a distinction between noise and signal. Pics are noise; the hash is the identity. In this context, the absence of a publicly verifiable audit trail for the campaign is not a minor omission. It is the defining feature of the risk. The user is asked to trust a centralized entity that has both a history of top-level disruption and a financial incentive to overstate the campaign's success. Trust is not a security model. It is a liability.
The narrative layer of the campaign deserves a closer look as well. The official framing uses the term "positive flywheel" to describe a system where more trading volume leads to more burn, which leads to a higher token price, which attracts more users, which creates more volume. In systems theory, a flywheel is only positive if each revolution stores energy. In this campaign, each revolution spends energy. The fee rebate is a cash cost. The prize pool is a cash cost. The burn is a balance sheet transaction that does not generate new profits. A flywheel that consumes fuel rather than accumulating it is not a flywheel; it is a turbine running on a fuel line. When the fuel line closes, the wheel stops. There is no price level at which the system becomes self-sustaining unless the exchange develops a revenue stream independent of the subsidy. Until that happens, the token price is a sentiment indicator, not a valuation. The market will treat it as a store of value until the market realizes that the store has an cash register at the front door and an exit at the back.
The final structural weakness is the governance vacuum. HTX is not a DAO. It is a private company. Its board does not publish minutes. Its treasury does not publish a custody attestation. Its token holders have no meaningful vote over the terms of the campaign. This matters because the campaign is a decision to spend shareholder or treasury capital. Under normal corporate governance, such a decision would require a board analysis of the expected return on customer acquisition. Under HTX's governance model, a single controlling figure can direct the campaign based on an assessment that is invisible to the market. That asymmetry is not just a governance flaw; it is a source of asymmetric information. The market is trading against an entity that sees the full balance sheet while the market sees only the press releases. If the campaign is performing worse than expected internally, the public might not learn until the subsidy is abruptly changed or the exchange freezes withdrawals. I do not say this because I believe a freeze is imminent. I say this because the absence of disclosure makes a freeze impossible to anticipate. The only rational position is to treat the exchange as a high-risk counterparty.
Now let me address the contrarian case, because an honest analyst must acknowledge what the bulls got right. First, the campaign is likely to generate a genuine short-term increase in trading volume. Negative fees are a proven attention mechanism. The 63.37 million USDT reported for the first phase is not nothing, and the second phase may be larger. Second, the campaign may create a real, if temporary, price catalyst for HTX. If the second phase expands the prize pool, extends the timeline, or adds more high-profile TradFi assets, speculators will respond. The token is sufficiently illiquid that a modest influx of demand can move its price. Third, the campaign is an effective piece of marketing in a crowded market. HTX has been absent from the top-tier narrative for years, and this promotion has succeeded in putting the brand back into the conversation. Fourth, there is a small subset of sophisticated traders who can extract positive expected value from the campaign. The market maker class will do well. High-frequency traders with low-latency infrastructure will do well. A disciplined individual who treats the negative fee as a rebate, not as a license to gamble, and who adheres to strict position sizing, can also profit. None of these observations change the sustainability analysis, but they explain why the campaign can feel like a golden age while it lasts. The edge exists only because the exchange is willing to overpay for participation.
The bulls are also correct that a buyback-and-burn, when funded by genuine revenue, is a valid use of profits. I have defended such mechanisms in protocol design documents. The problem with HTX is not the burn mechanism; it is the funding source. A burn funded by trading fees is deflation. A burn funded by treasury stock sold or by newly issued tokens is a category error. The first phase of the campaign did not release an audited accounting statement that would allow an observer to determine which category this burn falls into. Without that statement, the only defensible assumption is the worst case: the burn is funded from the same token inventory that is being used for rewards. If that assumption is false, HTX can prove it with a public wallet report. Until then, the claim of scarcity remains unverifiable. This is the precise point where the bulls' optimism collides with the auditors' caution. The map is not the territory; the chain is both. But the chain only carries the token transfer, not the internal accounting of the campaign.
The second phase of the campaign will be the true audit. The first phase proves only that a subsidy can attract volume. The second phase proves whether the subsidy is being reduced or expanded, whether the burn is accelerated or decelerated, and whether the exchange has found any organic revenue to replace the subsidy. My framework for evaluating phase two is simple. First, look at the reward pool size relative to the first phase. An increase is not a sign of health; it is a sign that the exchange has concluded it needs to buy more volume. Second, look at the burn address. If the burn rate drops while the reward rate rises, the narrative of scarcity is dead. Third, look at the exchange's overall spot derivatives volume across all markets, not just the promoted instruments. If total volume rises but promoted volume dominates, the campaign is cannibalizing organic activity. Fourth, look at the USDT reserve. A decline in exchange reserves indicates that the campaign is consuming liquid capital that could otherwise support withdrawals. The threshold for panic is not a number; it is a pattern of decline combined with declining transparency.
I have a particular interest in the interaction between on-chain transparency and regulation, because my 2025 work on an open-source surveillance framework for 12 blockchains taught me that enforcement can be automated. If the CFTC or the SEC decides to examine HTX's TradFi perpetuals, a chain-based reconstruction of the campaign's flows is not difficult. The exchange is centralized, but many of its USDT flows pass through public blockchains. The prize pool payments, the fee rebates, and the burn are all visible, at least at the wallet level. This means that the campaign leaves a record that can be subpoenaed, analyzed, and used as evidence. Every bug is a footprint left in haste; every incentive structure is a trail left for the regulator. I expect that the offshore jurisdiction will not protect HTX if a major market decides to act. The history of offshore crypto enforcement shows that exchanges can be indicted, that users can be barred, and that tokens can be delisted by global market participants. The campaign is not only financially unsustainable; it is legally exposed in a way that its marketing copy cannot conceal.
Let me also address the collateral damage to the broader industry. When an exchange offers retail users in the United States and Europe access to leveraged NVDA perpetuals, and when the exchange is incorporated in a jurisdiction that does not enforce local financial consumer protections, it normalizes a practice that other exchanges may be tempted to copy. The industry already struggles with the perception that it is a casino. A campaign that deliberately marries that casino machinery to regulated equity references strengthens the case for aggressive government intervention. This is not a minor reputational issue. It is a policy magnet. Every time a promoter extends a risky product farther than the local law allows, the responsible actors in the jurisdiction then have to defend their own market against the fallout. The SEC does not distinguish between a compliant exchange and an offshore exchange that is simply more visible. It distinguishes by enforcement priority. HTX's campaign raises that priority. That is an externality that no HTX token holder priced into the trade.
The comparison to the NFT metadata problem is instructive. In 2021, I examined BAYC and found that 80% of the collection's value was tied to off-chain metadata hosted on a centralized server. The community insisted that the art was forever. The infrastructure said otherwise. Three projects that followed the same centralized storage model lost all value when their servers failed. The lesson is that infrastructure fragility always outlasts narrative euphoria. HTX's Trade to Earn campaign has the same structural fragility, but its fragility is financial rather than technical. The infrastructure of the campaign is the exchange's willingness to incur losses. That willingness is a function of the exchange's balance sheet, the controlling individual's mood, and the pressure from regulators. All three are variable. None are disclosed in real time. This is what I mean when I say silence in the code speaks louder than the pitch: the absence of a verifiable reserve report is not neutrality; it is evidence.
One more technical point must be made about the mechanics of the negative fee in the context of perpetual funding. In a normal perpetual contract, the funding rate is the mechanism that keeps the contract price anchored to the spot price. Longs pay shorts when funding is positive, and shorts pay longs when funding is negative. A negative fee rebate interacts with this mechanism in a strange way. A trader can open and close positions simultaneously, generating crossed volume, paying almost no net fees, and capturing the funding rate differential. This is the classic manufacturing play. The exchange claims volume; the trader claims rebates; the funding rate is redistributed between them. Neither party is engaged in price discovery. Both are engaged in subsidy extraction. This is why raw volume alone is a worthless metric for measuring the health of the exchange. The quality of volume matters. In the first phase of Trade to Earn, a meaningful portion of the reported 63.37 million USDT may have been self-generated by professionals gaming the fee schedule. The exchange does not want to disclose that split, because it would reveal that the campaign is less an acquisition engine and more a transfer payment to quant funds.
I should also note the absence of any insurance or protection mechanism for users. On a centralized exchange, the user's position is only as safe as the exchange's solvency. During a campaign with negative fees and a daily prize pool, the exchange is spending its own capital to subsidize users. That spending increases the probability, however slightly, that the exchange will face a liquidity shortfall in a future stress event. This is not an allegation of fraud; it is a statement of probability theory. A firm that spends 110 basis points to earn 100 basis points is reducing its capital buffer. If a sharp market move triggers a cascade of liquidations, the exchange may be forced to pay socialized losses. The history of centralized exchanges is full of events where promotions and reserve shortfalls coincided. The ledger is indifferent to the marketing calendar. The question is not whether the campaign is profitable. The question is whether the users are the last to be repaid. With an opaque balance sheet, they will never know until the withdrawal queue freezes.
The Takeaway from this analysis is not that HTX is destined to collapse tomorrow. It is that the Trade to Earn campaign is a market event with a finite lifespan, a negative expected cash flow, and a regulatory time bomb hidden inside its underlying instruments. The second phase will determine whether the exchange can convert the subsidy into durable habit, or whether it will need to escalate the subsidy to maintain the same volume. If the latter, the campaign is a form of addiction, not growth. The only meaningful validation would be a phase two that reduces the incentive while maintaining volume. That would prove that users came for the product, not the payout. I expect the opposite. I expect the incentive to grow, the volume to fluctuate, and the burn to be celebrated regardless of the net supply effect. When the incentive stops, the volume will leave. That is not a forecast from a mystic. It is a statement about the price elasticity of extrinsically motivated traders. History is not written; it is indexed. And the index of this campaign, when the subsidy is stripped away, will show a line of volume that decays to its pre-campaign baseline.
The market is asking the wrong question. It asks whether the campaign will raise the price of HTX. The correct question is whether the campaign can survive without a permanent external subsidy. If the answer is no, then any price appreciation is a gift from the late entrants to the early extractors. I have seen this pattern in the Yearn yield analysis, in the Luna collapse, and in the NFT metadata post-mortems. The details change. The ledger does not. In every case, the participants who understood the economics of the incentive left before the subsidy was exhausted. The participants who believed the narrative stayed until the ledger reversed. Precision is the only apology the chain accepts. It does not accept press releases. It does not accept influencers. It accepts the difference between the amount paid and the amount earned. In the case of HTX's Trade to Earn, the difference is negative, and no burn ceremony can make a negative cash flow positive.
My final recommendation is not a call to short the token, nor is it an invitation to ignore the campaign. It is a call to demand a higher standard of disclosure. If HTX wants the market to view this as a legitimate token economics improvement, it should publish a full wallet-level accounting of the reward pool, the fee rebate outflow, the burn address balance, and the treasury inventory before and after the campaign. It should also specify the jurisdiction in which the TradFi perpetuals are legal to offer and identify the entity that holds the offsetting equity exposure. Without those disclosures, the campaign remains what it appears to be: a subsidy. The fact that the industry is willing to celebrate a subsidy as a structural breakthrough is the most dangerous part of this event. It tells me that the market is still willing to substitute narrative for arithmetic. The chain is a better accountant than the influencer. It remembers every transfer, every rebate, and every burn. It does not remember the headline, because the headline is noise. The hash is the identity. And the hash of this campaign is a long string of expenses with very little revenue at the end.