Trade-to-earn is the mechanism design mistake of paying your own token for trading volume. It looks like growth spending: subsidise activity, bootstrap liquidity, take share from the incumbent. What it buys, reliably and provably, is wash trading, because volume is the single easiest statistic in finance for one actor to manufacture alone.
The claim is not theoretical. My paper The Economics of Wash Trading measures it: NFT wash trading concentrated overwhelmingly on venues with token-based incentives, tracked the value of those incentives, and showed no sign of achieving anything else, with no significant relationship between wash volumes and genuine future volumes.
The arithmetic of the exploit
A volume reward creates a mechanical profitability condition. For a self-trade:
- Cost: marketplace fees plus gas on the trades.
- Revenue: your volume's pro-rata share of the emission schedule, at the token's expected sale price.
- Risk: none at the moment of trade; both sides are you.
When revenue exceeds cost, the reward pool is an open faucet, and every rational actor scales until crowding or emission decay closes the margin. The venue is not buying liquidity; it is auctioning its token to whoever can print volume cheapest. On LooksRare the equilibrium arrived within days of launch.
Second-order damage
- The statistics die first. Volume, rankings, and market-share charts become unusable, poisoning analysis long after the programme ends, as covered on fake trading volume.
- Token holders fund it. Emissions sold by farmers are a wealth transfer from holders to wash traders, laundered through a growth budget.
- Real users mislearn. Participants who arrive for rewards behave like rewards, not like customers; retention collapses when emissions do.
- Regulatory surface grows. Manufactured activity is manipulation in regulated markets and increasingly treated as fraud in crypto, per the legality map.
Designing rewards that buy something real
Principles that survive contact with the data:
- Never pay raw volume. Pay outcomes that require independent counterparties: realised spread capture, filled resting depth, retained active buyers.
- Filter by funding, not identity count. Wallets are free; funded independence is not. Cluster by capital origin before crediting anything, using the methods on the detection page.
- Vest against the future. Rewards that unlock only if activity persists after emissions stop convert farmers into either users or non-claimants.
- Price the attack pre-launch. If a self-trade clears the profitability condition on day one, the programme is a wash trading subsidy with a marketing budget.
Trade-to-earn is the cleanest modern illustration of the paper's thesis: wash trading is not primarily a deception story, it is an incentive story. Fix the incentive and the fake volume stops being printed; leave it and no amount of tape-watching helps. The wider market context is on NFT markets.