Input Output Global (IOG), the research and engineering firm behind the Cardano blockchain, published a design philosophy piece on March 10, 2026 arguing that the instability plaguing decentralized finance is not a feature of crypto itself but a product of specific, reversible design choices. The piece, authored by Rusty Shapiro, a DeFi product consultant at IOG, proposes what he calls a "DeFi Kernel": a minimal three-part architecture intended to support productive economic activity rather than speculative trading. The post arrives approximately one month after the Cardano Africa Tech Summit and within two weeks of the USDCx stablecoin launch on Cardano, a sequence that suggests a coordinated narrative push rather than a standalone thought leadership exercise, and readers should weigh the piece with that context in mind.

The Core Argument: Volatility Is Engineered In

Shapiro's central claim is that the two dominant building blocks of current DeFi, margin lending and automated market makers (AMMs), reinforce each other in ways that manufacture volatility rather than reduce it. Margin lending forces borrowers to post collateral and get liquidated when asset prices fall by a set threshold. Liquidators then sell those assets into a market that is already declining. AMMs, which price assets using a mathematical formula based on the ratio of two tokens in a liquidity pool, amplify those price moves because large trades shift the ratio and drive the price further in the same direction. "Current DeFi's volatility and speculation are designed in, not inherent to crypto," Shapiro writes.

This critique applies not only to protocols on Ethereum and Solana but also to AMM-based decentralized exchanges already live on Cardano itself, including Minswap, SundaeSwap, and WingRiders. Readers already familiar with those platforms should note that Shapiro's argument implicitly encompasses them.

AMM slippage at scale is a documented problem. Independent research suggests a $10 million trade on a constant-product AMM can incur slippage above 5%, while the same trade on a traditional limit order book might cost around 0.1%. Shapiro adds in his blog post that holding AMM slippage to 1% on a given trade requires a liquidity pool holding roughly 100 times the trade size in assets.

Three Primitives, One Kernel

The proposed DeFi Kernel is built from three components. The first is payments, which Shapiro considers a solved problem at the base blockchain layer. The second is a non-margin credit market: loans backed by collateral but governed by fixed terms, where liquidation only occurs on actual default rather than on price movement. Under this model, collateral stays locked for the duration of the loan, which removes it from circulating supply and reduces sell pressure. The third is an order book exchange, where buyers and sellers post limit orders at defined prices instead of trading against a liquidity pool formula. A working implementation of this third primitive already exists on Cardano: Cardano-Swaps (GitHub: fallen-icarus/cardano-swaps) is a peer-to-peer distributed order book DEX that demonstrates the model is not purely theoretical on this chain.

The blog post ties these primitives explicitly to Cardano's technical architecture. Cardano uses an Extended Unspent Transaction Output model (eUTxO), which assigns local, self-contained state to individual outputs rather than to transactions as a whole. A single transaction may consume and produce multiple UTXOs, each carrying its own datum. This makes transaction outcomes deterministic: fees are knowable before execution, and smart contracts behave predictably. Shapiro argues this architecture supports the credit reputation systems that a non-margin lending market would require over time.

On the question of network speed, Shapiro takes a deliberately contrarian position: "The base layer should not be high-throughput. Resilience and anti-fragility require deliberate slowness at the foundation." The argument is that historical data integrity, necessary for building borrower credit records on-chain, depends on a base layer that does not sacrifice completeness for speed. IOG's Hydra layer-2 protocol is positioned to handle throughput demands above that foundation. This framing is likely to draw pushback from developers working on Solana and Ethereum layer-2 systems, where high throughput is treated as a primary design goal rather than a concern to be delegated to a separate layer. That tension is not addressed in the IOG piece itself and represents the most contested claim in the design philosophy.

Cardano's Current DeFi Baseline

The blog post arrives at a specific moment in Cardano's development. Total value locked (TVL) on Cardano sits at approximately $141.9 million as of early March 2026, according to DefiLlama. That figure remains modest relative to Ethereum ($53.67 billion) and Solana ($6.55 billion), but recent on-chain flows show an unusual pattern. Stablecoin capital is entering the ecosystem while TVL in yield strategies has been declining. USDCx, a USDC-backed stablecoin launched on Cardano in late February in partnership with Circle, captured 37.16% of Cardano's stablecoin market within its first week, representing $17.73 million. Total stablecoin market cap on Cardano has reached $47.7 million, pushing the stablecoin-to-TVL ratio up 33%.

This is consistent with how AMBCrypto described the current trend: capital entering Cardano "in a more conservative, dollar-denominated form," oriented toward payments and treasury management rather than speculative rotation.

Regional Implications

The non-margin credit primitive carries the most direct relevance for emerging market users. Approximately 65% of Cardano activity is concentrated in emerging markets, according to AInvest, though that figure warrants caution given the platform's reliance on AI-generated financial content. More concrete evidence of Cardano's footprint in this space comes from the Atala PRISM student identity project in Ethiopia, which covers 5 million students and stands as one of the largest blockchain deployments globally. That project is the primary on-the-ground evidence for the "significant existing footprint" the IOG piece claims in Africa.

For small businesses and agricultural cooperatives in African markets such as Nigeria and Kenya, margin-based DeFi lending is structurally hostile. A price dip can trigger liquidation before a loan has had time to generate returns. A fixed-term, default-only liquidation model is structurally compatible with business lending in these contexts, where the credit need is tied to productive enterprise activity rather than speculative positions.

The same logic extends to South Asia. In countries like Bangladesh, India, and Pakistan, the non-margin credit model maps closely to informal sector lending and microfinance structures that already exist. The eUTxO model's deterministic transaction behavior also carries practical value in low-bandwidth environments, where predictability reduces the risk of failed or ambiguous submissions. Borrowers in these markets are more likely to need credit for inventory, harvest cycles, or small-scale trade than for leveraged speculation, making default-based liquidation a more appropriate risk structure than price-triggered margin calls.

Stablecoins already represent 43% of Sub-Saharan Africa's digital asset transaction volume, the highest share of any region, according to 2024 data, suggesting an existing user base oriented toward utility rather than speculation.

The Cardano Africa Tech Summit, held in Nairobi this past February following a developer program across 12 African cities, drew more than 500 participants and signals a deliberate shift from hackathon-stage activity toward production-stage ventures. The Cardano Foundation's 2026 roadmap includes more than $30 million allocated toward liquidity, real-world asset tokenization, and Hydra scaling.

What Comes Next

The IOG blog post does not announce a specific product or launch date. It is a design philosophy document intended to steer protocol development on Cardano. The institutional DeFi lending market offers a concrete benchmark for how much room for improvement exists: despite more than $78 billion in global DeFi lending TVL as of 2025, Aave Arc, the permissioned institutional lending product from Aave, held just $50,000. Whether the DeFi Kernel framework can close that gap depends on whether builders on Cardano adopt its constraints as design principles, and whether the "deliberate slowness" argument survives contact with developers accustomed to high-throughput chains. That debate is already active in Cardano governance forums and in cross-chain developer communities, where the trade-off between base-layer throughput and long-term data integrity remains an open design question rather than a settled one.