Comparing Bitcoin Covenant Proposals for Vaults

The recent report on Bitcoin vault constructions provides a comprehensive analysis of various mechanisms such as presigned transactions, CTV (CHECKTEMPLATEVERIFY), APO/APOAS (Anyprevout/Anyprevout Anysighash), TXHASH, CCV (Covenants by consensus verification), and the CAT-based Purrfect Vault construction. It delves into critical aspects like partial withdrawals, timing for committing withdrawal addresses, in-transaction fee management flexibility, operational complexity, and the associated on-chain costs.

The findings highlight that CTV is particularly effective for straightforward vault setups with precomputed outputs due to its simplicity. In contrast, CCV offers enhanced capabilities for vaults that necessitate partial withdrawals or need the flexibility to choose a withdrawal address at the time of trigger. TXHASH is noted for providing greater flexibility in commitment but requires more input and vigilance from the designer of the vault. Furthermore, the use of APOAS and OP_CAT is deemed advantageous not only for their primary functions within vault applications but also for their potential in broader non-vault related applications.

This research was aimed at a wide technical audience and originated from a semester-long project under the guidance of Michael Maurer and Neha Narula during fall 2025. The motivation behind this study stemmed from the observation that discussions around vault constructions were dispersed among various BIPs, implementations, and forum discussions. This report aims to serve as a consolidating resource that brings together disparate analyses of vault technologies. Feedback and corrections are welcomed, especially regarding interpretations of the APOAS and CCV structures, the methodology used for comparisons, and other opcode combinations or properties of vaults.