Pendle is a yield-tokenization protocol that splits any yield-bearing token (sUSDe, sUSDS, weETH, rsETH) into two tradable assets: a Principal Token (PT) that redeems 1:1 for the underlying at maturity, and a Yield Token (YT) that captures every drop of variable yield until expiry. As of September 15, 2026, Pendle holds roughly $1.24B in TVL per DeFiLlama, down from about $2.08B at the end of Q1 2026, with sUSDe and Ethena-linked pools accounting for the bulk of activity. The split turns a single staking position into a fixed-rate instrument, a leveraged yield bet, or a hedge, depending on which side a trader holds.
​

This guide explains how PT and YT work, why expiry dates matter, and how to use Pendle for fixed-yield strategies, leveraged yield exposure, and rate hedging. TVL numbers are dated to the September 15, 2026 pull and sourced from DeFiLlama and Docs Pendle.
​

Pendle is a permissionless yield-trading protocol on Ethereum, Arbitrum, BNB Chain, Mantle, and Base that wraps yield-bearing tokens into a Standardized Yield (SY) format, then mints two tradable derivatives: a Principal Token (PT) and a Yield Token (YT). PT plus YT always reconstitute the original yield-bearing asset until expiry.
​

The protocol launched in mid-2021 with vanilla LP yield, then pivoted in 2023 to liquid-staking and stablecoin yield. The 2024 Ethena partnership (sUSDe pools) and 2025 RWA expansion (USDY, USYC) pushed Pendle to a peak of about $13.4B in TVL in September 2025, per DeFiLlama, making it the largest yield derivatives venue in DeFi. Every pool has a fixed expiry date (typically 3-12 months out), after which PTs redeem 1:1 for the underlying and YTs become worthless.
​

Pendle's contracts are non-custodial. Underlying assets sit in audited vaults; PT and YT are ERC-20s that trade on Pendle's custom AMM. The protocol has been audited by Ackee, Chainsecurity, Spearbit, and WatchPug, with the full list in the public repo.
​

The PENDLE governance token uses a vote-escrow model (vePENDLE) where holders lock tokens for up to two years to direct emissions, vote on new pool listings, and capture protocol fees. That vote-escrow contract is now winding down: per the sPENDLE docs, sPENDLE replaces it, 80% of Pendle V2 Yield and Swap fees fund PENDLE buybacks, and the repurchased PENDLE is distributed to active sPENDLE holders every two weeks rather than weekly to lockers.
​

Pendle wraps a yield-bearing token (the SY layer), then routes it through a tokenization contract that mints equal amounts of PT and YT against the wrapped balance. PT holders surrender all variable yield in exchange for a guaranteed redemption at expiry. YT holders receive every unit of variable yield the underlying accrues, but the token itself amortizes to zero.
​

Mechanically, depositing 1 sUSDe into a Pendle pool with a March 2026 expiry mints 1 PT-sUSDe-MAR2026 plus 1 YT-sUSDe-MAR2026. The PT trades at a discount to 1 sUSDe, say 0.94, and that discount IS the fixed yield. Hold the PT until expiry, redeem for 1 sUSDe, and the 6.4% gap is the locked APY. The YT, meanwhile, streams whatever sUSDe yield Ethena pays out across the holding period (variable, typically 8-25% APY per Ethena's dashboard).
​

Trades happen on Pendle's custom AMM, which uses a yield-aware curve that prices PT and YT against the underlying based on time-to-expiry and implied APY. The curve is described in the Pendle V2 whitepaper. LPs earn swap fees plus Pendle token incentives.
​

Standardized Yield (SY) is the wrapper layer that normalizes different yield-bearing tokens into a single interface the tokenization contract can split. SY-sUSDe, SY-weETH, SY-rsETH, and SY-USDe are separate wrapper contracts, each handling the deposit, withdrawal, and reward-claim logic specific to its underlying. The SY layer is what lets Pendle support a new yield asset by deploying one wrapper rather than reworking the AMM. The full SY spec is in the standardized yield docs.
​

PT and YT are two halves of the same yield position. PT captures the principal: buy it at a discount, redeem at par. YT captures the yield stream: pay upfront, collect variable APY until the token expires worthless. The trader's view on whether realized yield will exceed implied yield determines which side to hold.
​

Consider a sUSDe pool with March 2026 expiry, where PT trades at 0.94 and YT trades at 0.06 (PT + YT must equal 1 sUSDe). A trader buying $10,000 of PT receives ~10,638 PT-sUSDe; at expiry they redeem for 10,638 sUSDe, locking ~6.4% APY assuming a 6-month tenor (the PT and YT split follows Pendle's minting mechanics). A trader buying $10,000 of YT receives ~166,667 YT-sUSDe units, each of which streams whatever yield sUSDe produces between purchase and expiry. If Ethena pays 15% APY for that window, the YT buyer earns ~$12,500 on $10,000; if Ethena pays 3%, they lose money.
​

The YT side is implicitly leveraged. With YT at 0.06, a $10,000 outlay controls yield on ~$166K of underlying, roughly 16x exposure to the yield rate, not the price. This is why Pendle's YT pools attract traders who want to bet that variable yield will spike (points farming, funding rate surges, liquid restaking incentive seasons).
​

To lock in fixed yield, buy PT directly on Pendle and hold to expiry. The implied APY shown in the UI is the fixed rate the buyer secures; no further action required. At maturity, the PT auto-converts to the underlying yield-bearing token (e.g., sUSDe), which can then be unstaked or redeployed.
​

The fixed-yield workflow is the largest use case by volume. The fixed rate a PT buyer locks is whatever the market prices on the day, and it typically sits above the prevailing passive stablecoin lending rate. Read the live implied APY for each expiry from Pendle's market dashboard and compare it against the same day's supply rate on Aave rather than against a number printed here. The premium reflects Ethena's funding-rate-driven yield and the basis between expected and implied rates.
​

Risks: PT is exposed to the underlying's depeg risk. A PT-sUSDe holder takes Ethena's smart contract, custody (Copper, Ceffu), and basis-trade risks for the full tenor. If sUSDe depegs at expiry, the PT redeems to a depegged sUSDe, not par USD. See the Ethena risk disclosures for the underlying mechanics.
​

To take leveraged exposure to a yield rate, buy YT. The capital outlay is the upfront cost of the YT (typically 3-15% of the notional underlying), and the payoff is every unit of yield the underlying accrues between purchase and expiry. If realized yield exceeds the implied yield priced into the YT, the position is profitable.
​

YT trading is the dominant strategy for points farming. When Ethena, EigenLayer, or a liquid restaking protocol announces a points campaign, YT holders typically receive points multiplied by their YT notional, not their dollar outlay, a 10-30x amplification on the same dollar of capital.
​

The trade-off is asymmetric loss. If realized yield underperforms implied yield, YT decays to zero faster than the yield stream replenishes it. A YT bought at 0.10 implying 25% APY over six months loses money if realized yield comes in below 25%, since the YT only ever pays out the yield the underlying actually produces (see Pendle's minting mechanics). YT is best understood as a yield call option with no upside cap and a 100% downside cap.
​

Pendle is the dominant yield derivatives venue but not the only one. Spectra (formerly APWine) offers a similar PT/YT split on Ethereum and Base with deeper RWA coverage. Sense Finance shipped a comparable design in 2022 but never reached significant TVL. Element Finance, an earlier yield-stripping protocol, wound down in 2024. The table below summarizes the active venues.
​

TVL figures pulled from DeFiLlama on September 15, 2026; the Term Finance figure combines its lending and vault deployments. Pendle's lead comes from sUSDe and weETH pools that no other venue matches in depth. Term Finance is included for completeness; it solves the same fixed-yield need without splitting yield tokens, using auction-based lending instead.
​

Every Pendle pool has a fixed expiry date (typically 3, 6, 9, or 12 months from launch), after which PTs redeem 1:1 for the underlying and YTs become permanently worthless. The expiry mechanic is what makes the protocol work. Without a fixed maturity, there is no "principal" to separate from "yield".
​

Expiry has three practical consequences. First, PT discounts narrow as expiry approaches, which means the locked APY for a fresh PT buyer drops over the pool's life. Second, YT pricing decays toward zero on a curve set by the AMM. Third, traders rolling fixed-yield positions must repurchase PT in a new-expiry pool, paying the prevailing implied APY rather than the original lock. Pendle publishes the full expiry schedule and historical APY data at Markets (app.pendle.finance).
​

Pools at or near expiry can become illiquid. The AMM curve flattens as time-to-expiry approaches zero, and arbitrage tightens spreads, but exit liquidity for large positions can thin in the final weeks. Plan exits ahead of expiry or accept redemption at maturity.
​

Pendle does not generate yield; it tokenizes yield from external sources. PT discounts and YT prices reflect market expectations of what underlying yield-bearing tokens will produce between trade date and expiry. The real yield engine sits in the underlying protocol: Ethena's basis trade, EigenLayer's restaking rewards, Sky's collateral, or Aave's borrow demand.
​

For sUSDe pools, the yield source is Ethena's perpetual futures basis trade, whose current rate is published on Ethena's internal yield dashboard. When funding rates are positive and high, sUSDe APY spikes and YT becomes profitable; when funding compresses, YT decays. For LRT pools (weETH, rsETH), yield comes from Ethereum staking plus EigenLayer restaking incentives plus partner-protocol points. RWA pools (USDY, sUSDC variants) draw yield from short-term Treasury bills and tokenized money market funds.
​

This sourcing matters for risk decomposition. A PT-sUSDe buyer is not exposed to Pendle's solvency; the contract simply holds wrapped sUSDe until expiry. The PT buyer IS exposed to Ethena's solvency, basis-trade collapse, and custody-partner failures. Pendle's own smart contract risk is layered on top, audited but non-zero. The Pendle V2 public repo hosts the contracts and audit history.
​

Pendle pools live on five chains, and the deepest sUSDe and weETH pools rotate between Ethereum (highest TVL) and Arbitrum (lowest gas). Moving USDC or sUSDe between chains to chase the best implied APY is the standard friction point. Eco Routes settles cross-chain stablecoin transfers across 15 supported chains, including Ethereum, Arbitrum, Base, and Optimism, so a treasury rebalancing from Arbitrum PT pools to a fresh Ethereum expiry takes one transaction rather than a bridge plus a swap. See Eco for the chain coverage and live route execution.
​

For institutional treasuries and DAO operators, Pendle's combination of fixed-yield PT and yield-leveraged YT replaces the off-chain fixed-income desk for stablecoin-denominated capital. Treasuries that hold USDC reserves can buy PT-sUSDe to lock the prevailing fixed APY for six months, then roll into the next expiry, producing predictable quarterly returns without manual rate hunting. The downside is concentration: a PT-sUSDe ladder is a leveraged bet on Ethena's continued solvency and basis-trade profitability for the full ladder duration. Diversifying across PT-sUSDS (Sky-collateralized) and PT-aUSDC (Aave-collateralized) lowers protocol risk at the cost of blended APY. Operators running this strategy typically split capital 50/30/20 across high-yield, mid-yield, and money-market PT, rebalancing quarterly as expiries roll.
​

Sources and methodology. TVL figures pulled from DeFiLlama on September 15, 2026. APYs are not quoted here because they reprice daily; read them live on App Pendle. Protocol mechanics verified against Docs Pendle and the Pendle V2 whitepaper. Comparison venue TVLs cross-checked on DeFiLlama.
​