DeFi is optimizing for gas, not for markets

Decentralized finance (DeFi) offers a transparent alternative to traditional finance, but its current architecture is largely constrained by the computational limitations of gas fees, hindering financial robustness and resilience, according to João Garcia, DevReal lead at Cartesi. This focus on computational minimalism, while seemingly a technical detail, is increasingly defining the boundaries of what DeFi can achieve, often prioritizing efficiency over market resilience.

Execution Environments and Market Stress

DeFi protocols, including exchanges, lending markets, and stablecoins, are tightly bound by their execution environments. Risk parameters often remain static, with collateral thresholds adjusting slowly through governance rather than automatic recalibration. Liquidation engines typically rely on fixed formulas, unable to adapt to shifting volatility or correlations due to the absence or emulation of floating-point arithmetic and the high cost of iterative simulations on chains like Ethereum. This forces financial logic into simplified, deterministic forms, which perform adequately in stable conditions but falter during market stress.

Historical events highlight these limitations. During MakerDAO’s “Black Thursday” in March 2020, vaults were liquidated at near-zero bids due to collapsing prices and network congestion. In subsequent downturns, protocols like Aave and Compound used mass liquidations triggered by fixed collateral ratios. The destabilization of Curve's pools in 2023, following a smart contract exploit, demonstrated how treating LP tokens as static collateral in lending protocols amplified systemic risk. In these instances, the issue was not decentralization itself, but rigid financial logic operating within an execution layer incapable of continuous risk recomputation.

In contrast, traditional markets utilize extensive stress scenario simulations and dynamic margin requirements, supported by substantial computational infrastructure. Public blockchains, however, were not initially designed for such iterative financial processing.

The Illusion of Simplicity and Externalized Intelligence

While simplifying protocols at the base layer reduces certain attack surfaces, it pushes complexity elsewhere. When risk cannot be modeled and recomputed transparently on-chain, it migrates off-chain to dashboards, analytics teams, and discretionary parameter adjustments. This reliance on rapid human coordination and the disproportionate influence of oracles and large token holders during volatility spikes creates a less transparent operational reality, despite a decentralized base.

DeFi's convergence on simplified finance is not due to the superiority of static ratios but because richer computational models were prohibitively expensive. As markets deepen and instruments become more interdependent, this compromise becomes increasingly problematic, turning safeguards like fixed thresholds into potential amplifiers of stress.

Computation as a Foundational Primitive

The core constraint facing DeFi is not decentralization but execution design. If verifiable execution environments can approximate general-purpose computing, the financial design space expands significantly. Native floating-point support, iterative algorithms, and access to numerical libraries would allow complex financial models to be expressed directly on-chain, rather than approximated. This could enable lending protocols to incorporate scenario-based stress testing, adjust margin requirements dynamically based on observed volatility, and recompute multivariable risk scores transparently.

The goal is to keep financial intelligence visible and enforceable within the protocol, rather than externalizing it into less auditable operational layers. The limitations of DeFi are largely architectural choices, not inherent inevitabilities of decentralization.

Navigating DeFi's Future

DeFi faces a structural crossroads. One path maintains gas-optimized minimalism, pushing sophisticated financial logic off-chain, which limits responsible scaling. The alternative is to embrace more capable execution environments, treating computation as a first-class primitive to allow systems to adapt, recompute, and stress-test transparently. Without on-chain complex risk logic, DeFi will continue to rely on discretion in practice, despite its code-level simplicity.

Markets will not adapt to virtual machine constraints. For decentralized finance to operate at a meaningful scale, its computational foundations must evolve to match the financial ambitions built upon them.