Persistence of the Three-Factor Structure as Liquidity Deepens

Determine whether the cross-sectional factor structure of Bittensor subnet alpha token returns—captured by a three-factor model comprising the market factor, the size factor (small-minus-big sorted on market capitalization), and momentum—persists as automated market maker pool depths (TAO reserves) increase and as the Bittensor subnet token market matures.

Background

The paper documents that returns of Bittensor subnet alpha tokens are well described by a three-factor model including market, size, and momentum, with the size premium derived structurally from constant-product AMM mechanics. Empirical tests, including a natural experiment around the December 2025 halving, support the theory and show the size premium scales with emissions.

However, transaction cost analysis reveals severe capacity constraints due to deterministic AMM slippage in thin pools, and the authors note that the market is young and segmented. As the ecosystem evolves—specifically, as AMM pools deepen (greater TAO reserves) and the market matures—the stability of the identified factor structure is uncertain, motivating the explicit open question about its persistence under changing liquidity and market conditions.

References

Whether the factor structure persists as pool depths increase and the market matures is an open question for future work.

Common Risk Factors in Decentralized AI Subnets  (2603.29751 - Maymin, 31 Mar 2026) in Conclusion (Section: Conclusion, label sec:conclusion), final paragraph