Arbitrage Theory Gap Exposed: Polymarket's Negative-Risk Markets
A new research paper draws a critical distinction between theoretical and executable arbitrage in prediction markets, showing that apparent mispricings in Polymarket's negative-risk binary markets may be structurally unexploitable until settlement. Meanwhile, a star-farming spam campaign is actively polluting the Polymarket bot ecosystem on GitHub and Telegram.
Payoff-Space vs. Protocol-Space Arbitrage: The Gap That Costs You Money
The most operationally significant finding today comes from academia, not GitHub. A new paper formalizes something experienced Polymarket operators have felt but rarely articulated: **not all arbitrage that exists on paper can be executed in practice**. The distinction is between *payoff-space* no-arbitrage — where deterministic outcome identities imply a mispricing — and *protocol-executable* no-arbitrage, where position transformation constraints prevent you from capturing that spread until settlement. Polymarket's negative-risk markets, which link binary contracts for mutually exclusive outcomes, are the clearest observable case of this. If you have been sizing arb positions based on raw price identities without accounting for settlement timing and position structure, you have been overestimating your edge. Read the paper before building another arb bot.
A second paper complements this by formalizing arbitrage across multi-condition prediction markets, providing mathematical frameworks for identifying mispriced outcome combinations and constructing portfolios that satisfy exhaustive, mutually exclusive resolution requirements. Together, these two papers are the most useful theoretical output for Polymarket quants in recent weeks.
---
New Bots: 7 Found, 3 Worth Your Time
**`winning-wallet-finder`** — Worth cloning. This is the cleanest statistical approach seen this week. It computes z-scores over trader performance history to separate wallets with a genuine edge from noise, runs a backtest of a copy-trading strategy against that filtered set, and then fires live alerts. The z-score framing is the right move — raw win rate alone is nearly useless for wallet screening at Polymarket's volume.
**`polymarket-btc-updown-jury-bot`** — Worth cloning. Targets BTC 5-minute Up/Down markets with a jury-based consensus model that aggregates signals across multiple sub-models before committing to a position. Crucially, it includes orderbook data in the backtest, which is rare and meaningful. BTC Up/Down markets are high-churn and difficult; the jury architecture at least addresses the signal instability problem.
**`polymarketfuturesbot`** — Monitor only. Scalping strategy on US Futures prediction markets, likely price-action and order-flow driven. No public backtest results. Watch for updates but do not deploy capital based on current documentation.
**`TG-Polymarket-bot`** — Monitor only. Whale copytrading via on-chain proxy wallet monitoring with Telegram alerts. Concept is sound but execution quality is unknown. The Telegram whale alert space is currently heavily polluted (see Security section below).
**`polymarket-insider-bot`** — Monitor only. No README, no description, name suggests a scraper or execution layer. Treat as unknown until source review is complete.
---
Security: Spam Campaign Active, Stay Disciplined
No confirmed malware today, but an active campaign warrants attention. Three related alerts describe a coordinated **star-farming operation** targeting Polymarket whale alert bots on both GitHub and Telegram. Duplicate Telegram bots with garbled-username GitHub repos are being seeded across the ecosystem, likely to manufacture social proof for low-quality or honeypot tools.
Operational guidance: do not judge a Polymarket bot by its GitHub star count this week. Verify repo age, commit history, and whether the strategy description is internally coherent before cloning anything from the whale-alert or arbitrage-scalping category. This campaign is specifically targeting those two categories.
---
HN Signal Worth Noting
The "Nothing Ever Happens" bot — which simply buys No on all non-sports markets — hit 471 points on Hacker News. Ignore the meme framing. The underlying strategy observation is not trivial: non-sports markets systematically overestimate the probability of discrete events occurring within a given window. If you do not have a counter-thesis to this, it should affect your prior calibration on event markets.
The Spain regulatory block on Polymarket and Kalshi is a jurisdictional development to track for operator exposure, not an immediate alpha event.
---
What to Watch
- **Negative-risk market pricing** across linked binary pairs — the arbitrage theory paper makes this the highest-priority structure to audit this week.
- **`winning-wallet-finder` deployment results** — the z-score wallet screening methodology is worth stress-testing against recent Polymarket data.
- **Star-farming campaign expansion** — if it moves from Telegram/GitHub into Discord or prediction market forums, alert surface widens significantly.