Composability requires stable accounting. The original 'money lego' metaphor assumes stable interfaces, but volatile assets like ETH or SOL are terrible units of account. A lending protocol's TVL in ETH is meaningless if ETH's USD value halves overnight, breaking risk models for every integrated protocol like Aave or Compound.
algorithmic-stablecoins-failures-and-future
Blog
Why True DeFi Composability Requires a Stable Unit of Account, Not Just a Token
DeFi's promise of a composable money Lego system is broken without a stable numéraire. This analysis dissects the failures of reflexive assets and argues that the future of interoperability depends on robust, non-reflexive stable units.
introduction
THE ACCOUNTING PROBLEM
The Broken Promise of Money Legos
DeFi's composability is crippled by the absence of a stable, universal unit of account, making financial legos impossible to price and combine reliably.
Tokenized dollars are a proxy, not a solution. While USDC and DAI provide stability, they are permissioned and fragmented across chains. A vault's solvency on Arbitrum, backed by USDC bridged via LayerZero, depends on the security of three separate systems, creating hidden correlation risk that undermines true composability.
The result is systemic fragility. The 2022 DeFi summer collapse demonstrated this: cascading liquidations across MakerDAO, Aave, and Curve were amplified because collateral was valued in a collapsing unit (ETH), not a stable one. Protocols were technically composable but financially coupled in the worst way.
Evidence: Oracle dependency as a symptom. The entire DeFi stack relies on centralized price oracles like Chainlink to translate volatile tokens into USD terms. This creates a single point of failure and latency, proving that the base layer lacks the necessary accounting primitive for autonomous lego-building.
key-insights
THE COMPOSABILITY IMPERATIVE
Executive Summary: The Stable Numéraire Thesis
Volatile tokens as the primary unit of account fragment DeFi's liquidity and logic, creating systemic fragility. True composability requires a stable numéraire.
01
The Problem: Volatility Fragments Logic
Smart contracts hardcoded for ETH or SOL cannot natively compose with protocols on other chains or stablecoin-centric systems. This creates protocol silos and oracle dependency for every price feed.\n- Result: DeFi 'money legos' are built on shifting sand, not a stable foundation.\n- Example: A lending pool's liquidation logic breaks if the collateral asset's volatility exceeds its design parameters.
100+
Oracle Feeds
-30%
Efficiency Loss
02
The Solution: Abstracted, Intent-Based Settlement
Protocols like UniswapX and CowSwap demonstrate the power of separating user intent ("get the best price") from execution. A universal stable numéraire extends this to cross-chain state.\n- Mechanism: Users express value in a stable unit; solvers compete to source liquidity across fragmented pools and chains.\n- Benefit: Composability becomes a function of economic intent, not token compatibility.
$1B+
Settled Volume
~500ms
Quote Latency
03
The Enabler: Universal Settlement Layers
Networks like LayerZero and Axelar provide the messaging primitive, but settlement requires a common denominator. A canonical stablecoin (e.g., a properly minted USDC cross-chain) or a synthetic unit like GHO becomes the essential settlement asset.\n- Critical Function: Acts as the neutral reserve asset for cross-protocol debt and collateral.\n- Outcome: Enables native cross-chain money markets and derivatives without constant rebalancing.
50+
Chains Connected
$10B+
Bridged TVL
04
The Outcome: DeFi as a Unified State Machine
With a stable numéraire, the entire multi-chain ecosystem can be modeled as a single, composable financial state machine. Liquidity is global, not siloed.\n- Capability: A position on Aave on Arbitrum can be seamlessly used as collateral to mint a stablecoin on Maker on Base.\n- Ultimate Goal: Eliminates the "chain" as the primary organizational paradigm for DeFi, replacing it with risk and yield as the core primitives.
10x
Composability Surface
-90%
Arb Complexity
thesis-statement
THE UNIT OF ACCOUNT PROBLEM
Reflexive Tokens Are a Dead End for Systemic Stability
DeFi's reliance on its own volatile assets as a pricing standard creates systemic fragility that prevents true composability.
Reflexive token design is a fundamental flaw. Protocols like Aave and Compound use their own governance tokens as collateral, creating a feedback loop of fragility. Price appreciation increases borrowing capacity, which fuels more buying, until the inevitable crash liquidates the entire system.
Composability requires a stable denominator. A volatile unit of account, like ETH or a protocol token, makes risk assessment impossible for integrated systems. Yearn vaults, Gelato automation, and Chainlink oracles cannot price risk when the yardstick itself is elastic.
The solution is exogenous stability. Systems need a non-reflexive numéraire like a properly collateralized stablecoin (e.g., DAI, USDC) or a unit like the Ethereum Gas Token (EGT) proposal. This provides the immutable pricing layer that contracts like Uniswap v3 or perpetual futures on GMX require for reliable integration.
Evidence: The 2022 "DeFi Summer" collapse demonstrated this. When LUNA/UST reflexivity broke, it triggered cascading failures across Anchor, Abracadabra, and leveraged positions on Solana, paralyzing the entire composability stack for weeks.
UNIT OF ACCOUNT PRIMITIVES
The Stability Spectrum: A Comparative Analysis
Comparing the core primitives used as a unit of account in DeFi, analyzing their impact on composability, oracle risk, and systemic stability.
| Feature / Metric | Volatile Native Token (e.g., ETH, SOL) | Over-Collateralized Stablecoin (e.g., DAI, LUSD) | Exogenous Fiat Stablecoin (e.g., USDC, USDT) |
|---|---|---|---|
Primary Stability Mechanism | None - Market Pricing |
| Off-Chain Banking & Reserves |
Oracle Dependency for Valuation | Price Feed for All Assets | Price Feed for Collateral Only | Price Feed for Peg Only (1:1 Assumption) |
Composability Failure Mode | Liquidation Cascades (e.g., 2022) | Collateral Liquidation Spiral | Centralized Blacklist / Depeg (e.g., USDC March 2023) |
Protocol-to-Protocol Settlement Finality | Indeterminate (Value Fluctuates) | Conditional (If Collateral Holds) | Deterministic (If Peg Holds) |
Typical Annualized Volatility |
| <10% (During Normal Operations) | <2% (During Peg Stability) |
DeFi Money Lego Integrity | |||
Censorship Resistance | |||
Required Trust Assumption | None (Code is Law) | Oracle Integrity & Collateral Quality | Issuer Solvency & Regulatory Compliance |
deep-dive
THE UNIT OF ACCOUNT PROBLEM
How Unstable Numéraires Poison the Composable Stack
Composability without a stable unit of account creates systemic fragility, not just user inconvenience.
Composability is not just interoperability. True composability requires predictable, machine-readable financial logic. When a lending protocol like Aave accepts volatile ETH as collateral, its liquidation engine must constantly recalculate safe debt ratios. This creates a fragile dependency where price feed latency or oracle manipulation can cascade.
Unstable pricing breaks automated workflows. A cross-chain money market built with LayerZero or Axelar assumes asset values are consistent. If the numéraire (e.g., ETH) swings 10% during the bridging settlement, the receiving contract's logic executes with stale, incorrect values. This breaks the deterministic promise of smart contracts.
The solution is a stable denominator. Protocols like MakerDAO (with DAI) and Compound (with USDC markets) demonstrate that debt and collateral calculations stabilize when pegged to a unit of account. This allows Uniswap pools and Curve gauges to compose without introducing unpredictable slippage from numéraire volatility.
Evidence: MEV from oracle latency. Flashbots data shows that arbitrage bots extract millions by front-running oracle updates on volatile collateral. This is a direct tax on composability, paid because the stack lacks a universal, stable pricing layer.
protocol-spotlight
THE UNIT OF ACCOUNT PROBLEM
The Next Generation: Building Exogenous Stability
On-chain DeFi is a tower of Jenga blocks built on volatile tokens. True, trustless composability requires a stable foundation.
01
The Problem: Volatility Kills Composable Logic
Smart contracts cannot execute predictable logic when the value of their principal asset swings 20% in a day. This breaks:
- Lending Protocols: Volatile collateral triggers cascading liquidations.
- Perpetual DEXs: Funding rate calculations become unreliable.
- Automated Vaults: Rebalancing strategies fail under extreme price slippage.
20%+
Daily Swing
$100M+
Liquidation Risk
02
The Solution: Exogenous, Non-Collateralized Stability
Stability must be imported from outside the crypto volatility loop. This is the core thesis behind Ethena's USDe and similar designs.
- Delta-Neutral Backing: Synthetic dollar minted against staked ETH and short perpetual futures positions.
- No On-Chain Debt: Avoids the reflexive death spiral of MakerDAO or LUSD under black swan events.
- Yield-Bearing: Captures staked ETH yield + funding rates, creating a native yield curve.
$2B+
TVL (USDe)
>30%
Implied APY
03
The Architecture: Intent-Based Settlement & Cross-Chain Portability
A stable unit of account is useless if it's siloed. It must be the base layer for intent-centric systems like UniswapX and cross-chain messaging like LayerZero.
- Universal Settlement: Stable asset becomes the default quote currency for cross-domain order flow.
- Reduced Fragmentation: Unifies liquidity across Ethereum, Solana, Avalanche without bridge risk.
- Composable Yield: Enables EigenLayer-style restaking of the stable asset's backing collateral.
~500ms
Message Finality
-90%
Slippage vs. Native
04
The Endgame: DeFi's Native Risk-Free Rate
The winner becomes the on-chain risk-free rate benchmark, displacing US Treasury yields as the foundation for all pricing models.
- Pricing Oracle: All derivatives, from GMX perps to Opyn options, are priced in this stable unit.
- Capital Efficiency: Enables 100x+ leverage with predictable margin requirements.
- Regulatory Arbitrage: A crypto-native, yield-bearing dollar operates outside traditional banking rails.
100x
Leverage Capacity
$10T+
Addressable Market
counter-argument
THE UNIT OF ACCOUNT PROBLEM
The Flawed Rebuttal: "Stability Is a Feature, Not a Bug"
Volatile assets as the primary unit of account create systemic risk and break the composability that defines DeFi.
Stability is a prerequisite for functional financial systems. A volatile unit of account forces every protocol to become a de facto FX trader, adding complexity and risk to every transaction.
Composability becomes fragile when contracts must constantly re-price assets. A lending protocol like Aave cannot safely assess collateral value if its unit of account swings 20% daily.
Compare MakerDAO to Frax Finance. Maker's DAI is a stable unit of account, enabling predictable leverage. Frax's FRAX, while stable, is often paired with volatile assets, introducing slippage in every DeFi interaction.
Evidence: The 2022 Terra collapse demonstrated that a non-stable unit of account (UST) creates a systemic failure vector that destroys the entire application layer built upon it.
takeaways
BEYOND PEGS
Architect's Checklist: Demanding Real Stability
Stablecoins that fail as a unit of account create systemic risk and cripple DeFi's core promise of composability.
01
The Problem: Volatile Collateral Breeds Contagion
Algorithmic or crypto-backed stablecoins like TerraUSD (UST) and DAI (pre-2022) are recursive risk vectors. Their stability depends on the very assets they're meant to hedge, creating a feedback loop during market stress.
- Contagion Risk: A 30% ETH drop can trigger mass liquidations in CDP systems, cascading across lending protocols like Aave and Compound.
- Broken Oracles: Price feed latency during volatility makes the unit of account unreliable for settlement, breaking atomic composability.
>99%
UST Depeg
$40B+
TVL at Risk
02
The Solution: Exogenous, Verifiable Reserves
True stability requires assets whose value is derived from and redeemable for real-world, off-chain collateral. This breaks the reflexive risk loop.
- Direct Redemption: Protocols like MakerDAO now hold ~$2B+ in US Treasury bills via BlockTower Capital, providing a non-correlated backstop.
- Transparent Attestations: USDC and USDP provide regular, audited reserve reports. The unit of account is a claim on a verifiable external asset, not a promise.
1:1
Redemption Guarantee
Monthly
Attestations
03
The Problem: Oracle Dependence is a Single Point of Failure
Every on-chain price is an oracle feed. For a unit of account, this introduces latency risk, manipulation risk, and governance risk over the feed itself.
- Flash Loan Attacks: Manipulating the Chainlink price of a stablecoin's collateral can drain an entire lending pool in one transaction.
- Settlement Risk: A 5-second oracle heartbeat means your "stable" unit of account is 5 seconds stale, making high-frequency composability (e.g., Uniswap -> Aave loops) fundamentally unsafe.
~5s
Oracle Latency
$100M+
Attack Vector
04
The Solution: Native Issuance & On-Chain Settlement
The most robust unit of account is one issued and settled natively on its ledger, like EURC on Stellar or USDC on Solana. This eliminates cross-chain bridge risk and oracle dependence for core settlement.
- Atomic Finality: A payment and a trade can be settled in the same ledger state with guaranteed finality.
- Bridge Elimination: Removes Wormhole, LayerZero, and Axelar bridge risk from the stability equation for on-chain activity.
~400ms
Settlement Finality
0 Bridges
Critical Path
05
The Problem: Regulatory Arbitrage is a Ticking Clock
Stablecoins operating in legal gray areas (e.g., Tether's early years) are a binary risk. A regulatory action doesn't depeg the asset—it obliterates it, freezing all composable systems that depend on it.
- Concentration Risk: USDT dominance (~$110B) means a significant portion of DeFi TVL is exposed to a single entity's legal standing.
- Smart Contract Risk: Regulators can blacklist addresses (as seen with USDC), turning a "stable" asset into a frozen, worthless token within a composable stack.
1 Entity
Primary Risk
Blacklist
Kill Switch
06
The Solution: Decentralized, Permissionless Mint/Redeem
Long-term stability requires a decentralized network of licensed, geographically distributed minters and redeemers, as envisioned by MakerDAO's Endgame Plan and Frax Finance v3. No single legal jurisdiction holds a veto.
- Resilient Design: A takedown in one jurisdiction cannot freeze the system; minters in other jurisdictions continue operations.
- Market-Driven Stability: Arbitrageurs enforce the peg by minting or redeeming based on market price, not a centralized entity's discretion.
Global
Minter Network
Arbitrage
Peg Enforcement
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall