Skip to main content
DjinnDJINN

Education & Research

Research on Bittensor, dTAO, and decentralized AI economics. External educational resources on sports analytics and prediction markets.

PaperDjinn Research

The Pricing of Options on AMM-Native Tokens: Theory and Evidence from Bittensor

Proves that a token whose only market is a constant-product AMM follows a constant-elasticity-of-variance (CEV) price process, with the exponent fixed at 1/2 by the bonding curve. Tests the prediction across 92 Bittensor subnets (February 2025 to April 2026): the realized-variance elasticity has a median of -0.96, matching the theoretical -1, with every subnet showing a negative slope. Yields a structural benchmark for option pricing, hedging, and the implied-volatility skew on AMM-native tokens.

Academic paper · PDF · April 2026
PaperDjinn Research

Mechanical Factor Premia in Automated Market Makers: Evidence from Bittensor Subnet Tokens

Shows that the constant-product AMM pricing Bittensor subnet tokens makes the cross-section of returns mechanically predictable: smaller pools take larger percentage price impact, so smaller subnets earn higher returns by arithmetic rather than risk. Documents a size premium of 1.0% per day across 128 subnet tokens from February 2025 through June 2026, identifies it as mechanical rather than risk-based, and shows AMM slippage bounds the strategy to roughly $10,000 in AUM, which is why it persists.

Academic paper · PDF · June 2026
PaperGershteyn & Zevelev

Market Microstructure and Emission Policy in Decentralized Incentive Networks

A flow-based simulation of how Bittensor’s $1.2M/day emission policy across 128 subnets interacts with short selling. Shows signal smoothing and short-selling corrections are substitutes, not complements: a short-selling correction improves allocation accuracy by 16.5% under the price-proportional rule but degrades it by 11.5% under the flow-smoothed rule.

Academic paper · PDF · April 2026
PaperHarry Crane

Estimating U.S. User Activity in Offshore Prediction Markets

The first rigorous estimate of how much offshore prediction-market volume (Polymarket, Opinion, Limitless, Predict, Myriad) comes from U.S. users, who are formally barred from these venues yet participate widely. Because on-chain wallets carry no demographics and geo-IP is defeated by VPNs, it triangulates the U.S. share from two behavioral proxies, the mix of sports traded and the hour-of-day shape of trading, each benchmarked against U.S.-only anchors (Kalshi and state sportsbook handle) and international ones. Over the trailing twelve months (May 2025 through April 2026) it attributes roughly 30% of Polymarket’s $55.6 billion volume, and $11 to $34 billion of all offshore volume, to U.S. users.

Research report · June 2026
EducationAnalytics.Bet

Analytics.Bet

Education platform for sports betting and prediction markets. Home of the Master of Sports Betting program with courses from world-class experts and professional bettors. Also offering a live AI in Sports Betting & Prediction Markets course.

Courses · Live & self-paced