Auto-Balancer: Harnessing idle network resources for enhanced market stability

TL;DR

The paper tackles inefficiencies in host-blockchain markets caused by external arbitrage traders and underutilised idle resources. It introduces Auto-Balancer, an endogenous mechanism embedded in block production that internalises emerging arbitrage via balancer transactions executed after user transactions, targeting price synchronization across venues while leveraging idle compute and avoiding inventory risk. The approach formalises a multi-objective optimization to minimize cross-market price discrepancies, $\min_{\mathbf{x} \in \mathcal{X}} \mathbb{E}[\sum_{i,j} \sum_{a} \Delta^a_{ij}(\mathbf{x})]$, while maximizing resource utilisation, $U(w)=w/C$, under $\mathbb{E}[\psi(w)] \le \delta$, balanced by weights $\lambda_1, \lambda_2$. It employs an epoch-based governance framework where searchers propose balancer transaction sets using $\mathbf{O}_s(\mathbb{E}[\mathbf{S}_e], \mathbb{E}[\mathbf{r}_e], c)$ and are evaluated on expected profit, gas costs, simulations, and credibility, with rewards distributed from the arbitrage pool to searchers, marketplaces, and the treasury via dynamic weights $\omega_x$, and gas-fee incentives for block producers with a slashing mechanism to deter misordering. The framework aims to improve liquidity, reduce price fragmentation, and promote sustainable ecosystem growth without introducing inventory risk.

Abstract

We propose a mechanism embedded into the foundational infrastructure of a blockchain network, designed to improve the utility of idle network resources, whilst enhancing market microstructure efficiency during block production by leveraging both network-owned and external capital. By systematically seeking to use idle network resources for internally capture arbitrageable inefficiencies, the mechanism mitigates extractable value leakage, reduces execution frictions, and improves price formation across venues. This framework optimises resource allocation by incentivising an ordered set of transactions to be identified and automatically executed at the end of each block, redirecting any realised arbitrage income - to marketplaces operating on the host blockchain network (and other stakeholders), which may have otherwise been extracted as rent by external actors. Crucially, this process operates without introducing additional inventory risk, ensuring that the network remains a neutral facilitator of price discovery. While the systematic framework governing the distribution of these internally captured returns is beyond the scope of this work, reinvesting them to support the ecosystem deployed on the host blockchain network is envisioned to endogenously enhance liquidity, strengthen transactional efficiency, and promote the organic adoption of the blockchain for end users. This mechanism is designed specifically for Supra's blockchain and seeks to maximally utilise its highly efficient automation framework to enhance the blockchain network's efficiency.