Monolithic vs. Modular Lending

01 · Concept — what problem does it solve?

Every lending protocol must answer one structural question: do all assets share one pool, or does each market stand alone? Monolithic designs (Aave v3, Compound) pool everything together — deep liquidity, maximum capital efficiency, but a single shared risk surface. Modular designs (Morpho Blue, Euler v2) deploy many minimal, isolated markets — risk is ring-fenced, listing is permissionless, but liquidity fragments. The whole industry is drifting modular; understanding why is understanding the core trade-off in DeFi credit.

02 · Mechanics

• Monolithic — one pool: all lenders' capital co-mingles; USDC suppliers earn from every borrower at once; governance must vote to list each asset and set its parameters.
• Monolithic upside: deep, passive-friendly liquidity and high utilization — one big book serving ETH, BTC, stables, and LSTs simultaneously.
• Monolithic downside: a bad-debt event in one collateral can be socialized across all depositors; new-asset listing is a slow governance bottleneck.
• Modular — many markets: each market fixes one collateral, one loan asset, one oracle, one LTV at deploy time. Anyone can create one permissionlessly for a few hundred dollars of gas.
• Modular upside: total risk isolation (bad debt can't cross markets) and instant long-tail/RWA listing without governance.
• Modular downside: liquidity is fragmented (every new market starts at zero) and passive lenders need a curation layer (vaults) to pick markets for them.

03 · Formulas

// both price money the same way — utilization → rate
U = borrowed / supplied        // monolithic: pool-wide;  modular: per-market

// the difference is the blast radius of a failure
monolithic:  bad_debt → socialized across the whole pool
modular:     bad_debt → capped at that single isolated market

04 · Edge cases & risks

• The CRV bad-debt lesson (Aave, 2022–23): an oversized CRV borrow left ~$1.6M bad debt socialized across Aave's shared USDC suppliers, and a later $63M position threatened far more. Monolithic contagion, made concrete — and the reason borrow caps exist.
• Modular's cold-start problem: a well-designed isolated market with no liquidity is useless; bootstrapping needs incentives or a curator's allocation.
• Curation moves the trust, not removes it: modular shifts risk assessment from protocol governance to vault curators (Gauntlet, Steakhouse). You now trust them to pick safe markets and oracles.
• The convergence: Aave v4 adds a modular hub-and-spoke layer; Morpho aggregates isolated markets under vaults. Both are reaching for "deep liquidity and contained risk" from opposite directions.