If I swap USDC to USDS directly through PSM vs. Curve, what's the difference?
A very practical question — which method to choose depends on several factors. PSM advantages: near-zero slippage (fixed 1:1); extremely low fees (current sky.money PSM fees are 0%); instant execution (one Ethereum transaction); no need for LP pool depth. PSM limitations: only swaps USDC ↔ USDS (can't directly swap USDT → USDS); if you need large-scale swaps (exceeding PSM's available USDC balance), may need batch execution; Gas fees on Ethereum mainnet can be non-trivial (but near-zero on Solana or L2). Curve advantages: can swap more stablecoin combinations (USDC/USDT/USDS triangle); deep liquidity, large swaps (>$100,000) may be more efficient than PSM batching; LPs can simultaneously earn fee income. Curve limitations: has slippage (though 3pool is usually very small, <0.01%); has fees (~0.02–0.04%), slightly higher cost for small swaps. Practical recommendation: amounts <$50,000, only need USDC ↔ USDS → use PSM (simpler, lower cost); need USDT → USDS path → use Curve (PSM doesn't support); large amounts (>$500,000) → consider combining PSM and Curve, absorbing with PSM first, then handling remainder through Curve.
Does PSM make USDS 'over-dependent on USDC'? How does Sky Protocol address this?
One of PSM design's most fundamental criticisms — and the problem Sky Protocol (formerly MakerDAO) worked hardest to solve from 2022–2025. The core problem: PSM design causes USDS (formerly DAI) reserves to increasingly consist of USDC. At peak, USDC in MakerDAO's PSM comprised over 40% of DAI reserves. This makes USDS stability substantially dependent on Circle (USDC issuer) credit — Circle problems mean USDS problems. 2022 Tornado Cash event lesson: in August 2022, OFAC sanctioned Tornado Cash, and Circle froze Tornado Cash-related address USDC. MakerDAO community immediately realized: if Circle were required to freeze 'all DeFi protocol USDC' (an extreme hypothetical, but not impossible), PSM's USDC could also be frozen, collapsing DAI's 1:1 redemption mechanism. This directly triggered the 'de-USDC-ification' community debate. Sky Protocol's response strategy (Endgame): introducing RWA as PSM supplements — like accepting Ondo's USDY (Treasury-backed, unrelated to Circle) as partial PSM collateral; TPPM deploying some USDC to Coinbase for yield (but making Coinbase the new counterparty); long-term goal to reduce USDC's share in USDS reserves below 20%, increasing non-USDC diversified reserves. Current situation (2026): Sky Protocol has partially achieved decentralization, but USDC remains the primary PSM reserve asset. If you're concerned about USDC risk and prefer not to hold USDS, this concern has merit — USDS and USDC fates haven't fully decoupled.
Besides USDS/DAI, which other stablecoins use PSM or similar mechanisms?
PSM's design has been widely borrowed across the industry, with different implementations per project. FRAX (partially collateralized stablecoin): FRAX has an 'AMO (Algorithmic Market Operations)' controller similar to PSM — can directly mint/burn FRAX when needed to maintain peg. FRAX's AMO is more complex than MakerDAO's PSM because FRAX is partially algorithmically collateralized; AMO must dynamically manage collateral ratios. Liquity Stability Pool (LUSD): Liquity has no standard PSM — instead uses a 'Stability Pool' mechanism letting LUSD holders exchange LUSD for $1 of ETH collateral at liquidation value, indirectly maintaining the peg (but mechanism differs from PSM). Curve's crvUSD: has a 'Peg Keeper' mechanism letting the protocol directly buy and sell crvUSD in Curve's liquidity pools when crvUSD deviates from $1 — more AMM-dependent version rather than pure mint/burn. PayPal PYUSD: has 1:1 fiat redemption (through PayPal itself) — users can exchange $1 USD for 1 PYUSD or vice versa anytime. This is the most traditional 'PSM equivalent mechanism,' just not executed on-chain. Common principle: all these mechanisms share the same logic — 'establish a predictable arbitrage opportunity so market forces maintain the peg' — whether through direct on-chain mint/burn (PSM), protocol-controlled AMM operations (AMO/Peg Keeper), or issuer fiat exchange guarantees (PYUSD).
How are PSM fees determined? Does a free PSM affect protocol financial sustainability?
PSM fee design: PSM fees have two directions: 'tin' (in) — fee rate for depositing USDC into PSM to receive USDS; 'tout' (out) — fee rate for swapping USDS back to USDC. Both rates are decided by Sky Protocol's MKR/SKY holders through governance voting, settable from 0% to theoretically any value. Fee strategic significance: governance can use fees to manage PSM fund flows — raising 'in' rate increases cost of more USDC entering PSM (suitable when PSM approaches Debt Ceiling); raising 'out' rate increases cost of redeeming USDS for USDC (suitable when needing to preserve PSM USDC liquidity); setting both to 0% maximizes peg maintenance (any deviation is immediately arbitrageable) but the protocol earns no fees. Free PSM financial impact: PSM fee income contributes minimally to Sky Protocol's overall finances — because the protocol's primary income comes from: CDP collateralized lending stability fees (borrowing rates); SSR rate management (income > interest expense differential); RWA investment returns. A free PSM sacrifices a few basis points of fee income but gains stronger peg and better market confidence. Currently (2026) sky.money's PSM maintains near-0% rates — governance assessment concluded this best serves USDS ecosystem long-term health.
If you've ever tried buying USDS on a DEX, you may have noticed that regardless of how much you buy (within reasonable amounts), USDS's price is almost always exactly $1 — rarely deviating more than 0.1% even during market stress. In contrast, trying to swap FRAX or some niche stablecoins outside Curve might show much larger slippage and deviations.
Behind USDS's 'near-perfect peg' is a key infrastructure component called the PSM (Peg Stability Module). PSM enables 1:1 direct exchange between USDS and other fiat-backed stablecoins (USDC, USDT), without relying on AMM supply-demand pricing, establishing a powerful arbitrage mechanism that keeps USDS from straying far from $1.
PSM was introduced by MakerDAO (now Sky Protocol) in 2020, originally designed to solve a long-standing DAI problem: DAI had 'premiums' (>$1) when market demand was high and 'discounts' (<$1) when demand was low, while traditional interest rate adjustment mechanisms (DSR and stability fees) reacted too slowly to handle sudden demand shocks. PSM's design is simple: Sky Protocol builds a smart contract letting anyone: exchange $1 USDC for 1 USDS (minting direction); exchange 1 USDS for $1 USDC (redemption direction). Exchange is nearly instant (one Ethereum transaction), with minimal fees (usually 0–0.1%). How PSM maintains the peg (arbitrage mechanism): suppose USDS falls to $0.99 on the market (below $1). Arbitrageurs immediately have a risk-free opportunity: buy 1 USDS on the market for $0.99 → use PSM to swap 1 USDS for $1 USDC → profit $0.01 (1%). This arbitrage action simultaneously: consumes USDS on the market (buying pressure pushes USDS price up); burns USDS through PSM (reducing supply). Both effects push USDS price back to $1. Reverse similarly: if USDS rises to $1.01, arbitrageurs use $1 USDC to mint 1 USDS through PSM, then sell it on the market for $1.01, arbitraging $0.01. This increases USDS supply and pushes price back down. PSM works because it establishes a 'hard arbitrage boundary': any time USDS deviates more than PSM's fee (0–0.1%) from $1, there's profitable arbitrage, and arbitrageurs' self-interest automatically restores price. This is far faster than DSR adjustment or governance voting mechanisms.
Many people confuse PSM with Curve's stablecoin liquidity pools (like 3pool). Both enable stablecoin swaps, but the mechanisms are completely different. Curve 3pool (AMM model): liquidity providers (LPs) deposit USDC, USDT, and USDS/DAI into the pool. Token swaps are priced by AMM formulas — supply-demand determines actual exchange ratios. If large amounts of USDS are sold, USDS accumulates in the pool and price begins deviating from $1. Pool depth (TVL) determines how much shock can be absorbed without large slippage. PSM (hard 1:1 mechanism): not a liquidity pool, but a 'mint/burn' smart contract. Deposit USDC → Sky Protocol directly mints equivalent USDS at the protocol level; deposit USDS → Sky Protocol directly burns USDS and returns USDC. No AMM formula, no slippage (theoretically), no 'liquidity depth' concept — as long as PSM has sufficient USDC reserves, redemption is 1:1. PSM and AMM's complementary relationship: PSM provides the 'hard peg boundary'; Curve provides 'deep liquidity for large-scale swaps.' PSM's existence makes Curve's USDS pool more stable — arbitrageurs know USDS can swap 1:1 for USDC through PSM, so Curve pool USDS is unlikely to severely depeg (arbitrageurs preemptively correct deviations).
PSM isn't omnipotent — it has several key limitations. Limitation 1: PSM's USDC reserve cap (Debt Ceiling). Sky Protocol governance sets a 'Debt Ceiling' for each PSM — the maximum USDC deposits accepted. Once reached, new USDC-for-USDS operations can't execute (minting direction paused). If market demand for USDS surges and hits the PSM cap, USDS may briefly trade at a premium (>$1). Governance can vote to raise the cap, but this takes time. Limitation 2: USDC reserves exhausted in redemption direction. If many users simultaneously redeem USDS for USDC through PSM, PSM's USDC may be exhausted. Redemptions then pause; exit only through other means (Curve AMM, selling USDS on CEX). During the 2023 SVB crisis, many USDS/DAI holders tried to redeem through PSM for USDC, but USDC itself was depegging — PSM protection effectiveness was limited. Limitation 3: PSM concentration risk. PSM causes USDS reserves to have a large USDC proportion. Per Sky Protocol data, PSM's USDC once comprised 20–40% of USDS reserves. This means USDS's stability partially depends on USDC's stability — if USDC has problems, USDS may be collaterally affected (the brief 2023 USDC depeg caused brief DAI/USDS deviation from $1). Limitation 4: PSM's opportunity cost. USDC sitting in PSM just sits there — not generating yield. Sky Protocol later introduced RWA and TPPM (Third-Party Protocol Modules) mechanisms to deploy some PSM USDC into Coinbase for interest — but this introduces new centralization and counterparty risks.
PSM's design continuously evolved throughout Sky Protocol's (formerly MakerDAO's) history. 2020 original PSM: initially supported USDC and USDT as PSM input assets, with 0.04% (in) and 0% (out) fees. This rapidly stabilized DAI's peg but made DAI's reserves increasingly dependent on centralized stablecoins (USDC). 2022–2023 decentralization pressure: as USDC regulatory risk grew (especially after the 2022 Tornado Cash sanctions led MakerDAO to briefly consider swapping PSM USDC for ETH), the MakerDAO community recognized PSM's over-dependence on USDC. This directly drove part of the 'Endgame' plan — diversifying PSM assets from pure USDC toward RWA. 2024–2025 USDS era: Sky Protocol upgraded DAI to USDS, largely preserving the PSM mechanism but introducing 'TPPM (Third-Party Protocol Module)' letting some PSM USDC be deployed to institutions like Coinbase to earn yield — partially resolving PSM's opportunity cost problem. Currently USDS's PSM primarily accepts USDC and some RWA assets (like Ondo's USDY) as collateral.
PSM has several direct practical implications for you as a stablecoin user. You can directly use PSM. If you have USDC but want USDS (to deposit into SSR for 8.5% yield), you can directly swap through PSM on sky.money's interface (1:1, near-zero fees) without paying Curve swap fees. Reverse also holds: holding USDS but needing USDC to off-ramp, PSM provides the most direct exit channel (no need to wait for Curve AMM execution). USDS's PSM makes it more stable than most stablecoins under pressure. During market panic, arbitrageurs using PSM's 1:1 exchange quickly correct deviations, keeping USDS from deviating more than 0.1–0.2% from $1. By contrast, stablecoins without PSM mechanisms (like apxUSD) can deviate 20%+ under market stress. But understand PSM's limitations: under extreme stress (like simultaneous USDC depegging), PSM cannot protect USDS from impact; PSM's USDC concentration partly links USDS and USDC fates. Understanding this connection before holding large USDS positions lets you better assess your actual risk exposure.