How DAI works
DAI is the original decentralized stablecoin. You lock collateral (ETH, stETH, USDC, whatever the protocol allows) and borrow DAI against it. Simple. Unlike USDC (backed by actual bank vaults), DAI has no reserve fund. It maintains its $1 peg through interest rates and arbitrage.
The protocol charges a stability fee on borrowed DAI. When DAI dips below $1, the fee drops to encourage borrowing and push prices up. When DAI climbs above $1, the fee rises to discourage new issuance. It's algorithmic peg management. The Peg Stability Module sweetens the deal by swapping DAI for USDC at 1:1 rates. That creates an arbitrage floor that stops runaway depegs.
Liquidation keeps the protocol solvent. If your collateral falls below the 150% threshold, liquidators buy your debt at a discount and seize your collateral. That removes bad debt before it stacks up beyond the protocol's loss-absorption capacity. Liquidators earn a 13% discount for their trouble.
Single collateral to multi-collateral
MakerDAO started requiring ETH only (SCD era). That was restrictive. Multi-Collateral DAI (MCD) opened the doors. Stablecoins get 90-95% LTV because they're safe. ETH sits around 50-80% depending on volatility. This diversification let DAI reach $5+ billion in circulation.
The Risk Core Unit evaluates new collateral candidates with rigor. Technical security? Market quality? Oracle reliability? Counterparty risk? Each asset goes through quantitative stress testing. The protocol won't accept garbage. This discipline and diversification are why DAI outlasted dozens of dead stablecoins.
Real-world assets reshape the game
MakerDAO bet big on RWAs—bonds, invoices, real estate tokenized and brought on-chain. An institution with a bond portfolio tokenizes it, posts it as collateral, gets DAI liquidity without selling the underlying. That bridges the gap between traditional finance and crypto. Centrifuge pioneered this by creating blockchain-compatible tokens representing real asset claims.
This was philosophically contentious. Purists wanted pure crypto collateral. Pragmatists saw institutional adoption requiring traditional finance integration. The compromise: RWA accepted but conservative parameters and caps. No single RWA asset dominates.
The peg and its fragility
The PSM is the main defense. 1:1 swap into USDC. When DAI trades cheap, you buy it and swap it back at par. Arbitrage converges price to parity. The Dai Savings Rate (DSR) provides demand-side support—pay interest on DAI holdings and people accumulate instead of spending. This reduces supply and pushes price up.
But governance faces a bind. Maintain peg (expensive), stay profitable (higher fees), avoid token dilution, build reserves. These sometimes conflict. Black Thursday 2020 broke this model temporarily. Liquidation cascades spiked DAI to $1.30. The PSM couldn't absorb the demand shock. Governance responded slow. Activist voting forced emergency parameters. That exposed governance's speed limits during crises.
DAI's systemic importance now creates feedback loops. It's the DeFi settlement currency. Widespread adoption means a depeg could break multiple protocols. That makes healthy DAI critical to DeFi as a whole.
SubDAOs and governance scaling
Governance can't make real-time parameter tweaks. You need specialized SubDAOs with domain authority. The MetaMorphosis SubDAO handles protocol upgrades. Others manage risk, treasury, etc. They propose changes without full governance approval for routine decisions. Major changes go to the full MKR voting.
This trades decentralization for operational speed. It works only if SubDAOs stay accountable and don't capture their domains. The architecture remains under development. Balancing autonomous efficiency against community control is the puzzle.
MKR token and voting
MKR holders vote on everything. Total supply is 1 million tokens. Early supporters, investors, and team members own most of it. MKR accrues value through buyback-and-burn—protocol revenue purchases MKR from markets and burns it permanently. Shrinking supply plus stable revenue creates value. Delegation reduces voting friction. Token holders delegate to governance delegates who coordinate votes on their behalf.
This created a robust delegate ecosystem with various philosophies. But concentration in early hands remains a governance weakness. The community discussed expanding distribution through incentives and airdrops to broaden participation.
Sky rebrand and the Endgame Plan
MakerDAO announced a major redesign. The rebrand to "Sky" signals evolution from a simple stablecoin protocol toward institutional financial infrastructure. The Endgame vision simplifies governance structures, phases out complexity, and emphasizes institutional features. Stability mechanisms get strengthened for corporate confidence.
The new vision includes tokenized treasury assets where governance participants stake wealth on outcomes. This reduces voting token concentration. Multiple assets could participate in governance instead of just MKR. It acknowledges that pure token voting tends toward concentration and proposes fixes. Backward compatibility with DAI holders ensures existing integrations survive.
The rebrand is risky. Existing governance expectations require careful management. But the vision reflects real learning about DeFi governance challenges.
Decentralized authority and its limits
Early MakerDAO governance put all authority in MKR holders but lacked mechanisms ensuring responsiveness. The protocol evolved toward specialized committees managing risk, collateral, treasury. But true decentralization remains aspirational. Core decisions concentrate in active governance participants. Broader community engagement hits participation friction.
Multisig governance distributes power across multiple signers with thresholds (usually 3-of-7 requiring three signatures). No single compromise kills the protocol. But it requires coordination overhead. Sky's roadmap emphasizes better mechanisms—delegation improvements, token redistribution, specialized governance structures. None have been proven yet.
Liquidations and keeper economics
Liquidation isn't automatic. Keepers—specialized agents and market makers—monitor vaults, execute liquidations, and capture bonuses. The keeper ecosystem is now professional. They run monitoring infrastructure at scale. The liquidation bonus varies by collateral (7-13% discount) reflecting liquidity and execution difficulty. Illiquid collateral needs higher bonuses. Stablecoins accept lower discounts.
Keeper profitability depends on liquidation frequency and bonus size. This creates natural incentives: protocols with sophisticated keeper infrastructure attract better collateral by proving liquidations execute reliably. Weak keeper ecosystems accumulate bad debt.
DAI's worst moments
March 2020 was brutal. Flash crashes triggered liquidation cascades. DAI spiked to $1.30. The PSM couldn't handle the demand shock. Governance moved too slow. Activist voting forced emergency responses. The protocol survived but exposed governance bottlenecks. Subsequent updates expanded PSM capacity and improved emergency mechanisms. But structural limits remain—if market dislocations spike DAI demand beyond the PSM's capacity, the peg breaks.
DAI's systemic role means its depeg could cascade failures across DeFi. This is an inherent risk to decentralized stablecoins without traditional reserve backing.
Treasury and sustainability
MakerDAO built substantial reserves from stability fees and other revenue. The treasury funds Core Units (specialized organizations managing functions) without governance token dilution. Long-term sustainability requires treasury growth keeping pace with operational costs. Shrinking treasury forces uncomfortable choices—dilute the token or cut development spending.
Treasury diversification into RWA or uncorrelated assets provides stabilization but adds complexity and risk. Governance balances treasury building against operational efficiency constantly.
Competition and the path forward
Aave's GHO competes on lending efficiency. Curve dominates stablecoin liquidity. Lido's liquid staking derivatives offer alternatives. Each protocol has different tradeoffs. GHO prioritizes lending integration. Curve focuses on stablecoin swapping. MakerDAO distinguishes itself through institutional adoption and RWA integration.
Rather than compete on simplicity, MakerDAO positions as institutional-grade infrastructure with sophisticated risk management. The Sky rebrand reflects that. The roadmap includes RWA expansion, corporate governance features, and mechanism refinement. MakerDAO's evolution shows DeFi's maturation toward institutional infrastructure requiring real governance and risk management instead of pure speculation.
Recent developments: Governance discussions continue around SubDAO structure refinement and the Endgame Plan implementation timeline.
FAQ
What's the difference between DAI and USDC?
USDC is custodial—Coinbase holds dollars and issues tokens 1:1. DAI is algorithmic—collateral on-chain backs issuance. DAI is non-custodial but more complex and subject to peg risk during crises.
Why do liquidations matter?
They remove bad debt before it kills the protocol. Without liquidations, underwater positions would accumulate losses that governance can't absorb. Keepers executing liquidations keeps the system healthy.
How does the PSM prevent DAI depegging?
It creates an arbitrage floor. When DAI trades below parity, you buy cheap DAI, swap it for USDC at par through the PSM, and profit. That arbitrage activity pushes DAI back to $1.
Can DAI break its peg permanently?
Yes, if market dislocations spike demand beyond the PSM's capacity. It happened briefly in March 2020. The protocol has improved defenses since then, but structural limits exist.
What's the MKR token actually worth?
Governance rights and protocol revenue distribution. The buyback-and-burn mechanism creates token scarcity as revenue purchases MKR from markets. Stable revenue plus shrinking supply drives long-term value.