Blend V3 Proposal

[royalf00l]

Blend V3 Proposal

Overview

Templar Protocol proposes to take the lead on active development, maintenance, and operational stewardship of Blend Protocol v3. Our goal is to make Blend an institution grade money market and unlock the functionality the ecosystem needs to grow in v3 and beyond.

Updates

RWA focus

• Institutional Onboarding: Assets, Vault Curators & Managers, Risk Analysis, Liquidations
• Oracleizing off-chain NAV, transparent book to market
• Fixed Rate and Fixed Duration Loans
• Permissioned pools: compliance-ready configurations for RWA issuers & regulated entities using Predicate or similar service for pre-action hooks

Composability & Vaults

• CDP Strategy vaults enshrining the asset manager for 1click yield
• PT/YT token support on pool + oracle - Spectra Integration
• Structured products with senior/junior tranches for institutional and RWA use cases. Requires changes to the backstop

Risk Management

• Hybrid market + governance approach: configurable guardrails for pool creators
• Improvements to backstop (separate discussion TODO link)

Multichain

• EVM, Solana, passkey, privy wallet support for interacting with Blend. Public good that could be used across Stellar ecosystem
• Cross-chain collateral support via Templar’s chain abstraction tech: e.g. borrow USDC on Stellar against BTC on Blend or use BTC in backstop

Security

• Hypernative Monitoring Integration for all v3 pools and vaults
• Continuous invariant testing, fuzzing, security checks for all core contracts
• Granular per-function pause capabilities, more granular parameterization of pools

Breaking Protocol Changes

1. Permissioned Pre-Action Hooks
2. Backstop modifications (potentially including tranche formalization)
3. Fixed Rate and Fixed Duration Loans
4. PT/YT token support: Maturity timestamp needs to be tracked and parameterization around maturity needs to be supported.

[levinson]

First off, thank you taking the time to think about the future of Blend and come up with ideas for v3 protocol enhancements. It would be helpful if we could break out a separate discussion threads for each of the proposed concrete changes.

For each concrete change I suggest the following format:

• Summary of the issue: What does this change seek to address?
• Current Design: High-level summary
  • Detailed description: What's the current technical design/behavior in Blend v2?
  • Limitations: What practical limitations does the current design impose?
• Proposed Design: High-level summary
  • Detailed description: What technical changes are required to support the proposed design?
  • Backwards compatibility: How would existing v2 pools be supported in v3?
  • Pros/Cons: List out all advantages/disadvantages to the proposed change
• Alternative Designs: What other designs were considered?
  • Detailed description
  • Backwards compatibility
  • Pros/Cons

This is just a proposed format. The most important parts are that each concrete change has it's own thread with detailed description of the current design and proposed design. Make these as granular as possible -- do not bundle changes together. For example "permissioned pools" should have it's own thread, explaining the existing permissionless design of Blend v2 pools, practical limitations it imposes for pool issuers, proposed solution with high level summary and detailed technical description.

While the proposed design changes may enable several RWA use cases, the protocol enhancements should not be limited to or specific to RWAs (after all RWAs are just Stellar tokens). They also should not be specific to supporting Templar integrations.