Yield Tokenization Explained: Pendle PT and YT
PT and YT both come from splitting one yield-bearing token — the difference is who keeps the rate risk. PT buyers lock 4.92% on sUSDe to 28 January; YT buyers bet on the floating rate, and Pendle's TVL still fell 93%.

Introduction
Pendle lost half of its peak deposits in seven days in September 2025, while the supply of Ethena's USDe stablecoin, its largest asset, barely moved (DefiLlama, 2026). The explanation is a maturity date, and it sits at the centre of yield tokenization: every Pendle market ends on a fixed day, and capital leaves unless it is rolled over. Pendle's total value locked (TVL) stood at $0.98B on 16 November 2026, 93% below its September 2025 peak, a far steeper fall than decentralized finance (DeFi) as a whole ↗. This guide measures the mechanism on live markets. It prices a Principal Token (PT) that locks 4.92% on Ethena's sUSDe until 28 January 2027, shows why a Yield Token (YT) buys 105 times the yield exposure yet barely clears Pendle's fee at an unchanged rate, explains Boros, and traces the unwind and the risks that remain.
Key Takeaways
- A PT bought at 0.9905 USDe on 16 November 2026 redeems at 1.0000 on 28 January 2027, a fixed 4.92% annualised.
- $1,000 of YT-sUSDe controls the yield on $105,211 of principal, yet returns only +1.5% if the underlying rate stays at 5.10%.
- Pendle's TVL fell 49% in the seven days after its 19 September 2025 peak because 33 pools matured, while USDe supply stayed flat.
- Ethena assets fell from 67.6% of Pendle's TVL to 5.1%, and Morpho now lists $260M borrowed against PT collateral.
- Boros trades funding rates through Yield Units and held $6.6M of TVL against $0.98B for Pendle V2.
How Does Yield Tokenization Split One Token Into PT and YT?
Yield tokenization splits a yield-bearing token into a claim on its principal and a claim on its future yield, so each can be bought and sold separately. Pendle is the dominant venue: its total value locked (TVL) was $0.98B on 16 November 2026, against $12.4M for Spectra, another yield-tokenization protocol (DefiLlama, 2026).
One Wrapper for Every Yield Source
A Standardized Yield (SY) token wraps a yield-bearing asset in one common interface, so Pendle handles Kelp's rsETH, Aave's aUSDC and Ethena's sUSDe the same way. In most cases 1 SY represents one unit of the wrapped asset. SY has no maturity date and receives every unit of yield and every airdrop point the asset earns (Pendle docs, 2026). The sUSDe token, the staked version of Ethena's synthetic dollar USDe, shows why the wrapper matters: one SY-sUSDe was worth $1.26 on 16 November 2026 because sUSDe accrues yield against USDe (Pendle API, 2026). Pendle therefore tracks value in an accounting asset, USDe in this case, rather than in the wrapper. A separate guide to yield-bearing stablecoins such as sUSDe ↗ explains where that yield comes from.
PT Plus YT Equals the Underlying
Depositing SY into Pendle's yield contract mints equal amounts of a Principal Token (PT) and a Yield Token (YT) that share one expiry date. PT redeems 1:1 for the accounting asset at maturity. YT streams every unit of the asset's yield until then and is worth nothing afterwards (Pendle docs, 2026). The two prices always add up to the accounting asset: on 16 November 2026 PT-sUSDe-28JAN2027, on the Monad blockchain, traded at $0.9900 and its YT at $0.0095, a total of $0.9994 against $0.9996 for USDe (Pendle API, 2026). Before expiry both tokens are needed to recombine into SY; after expiry PT alone redeems.
Buying the principal half alone converts a floating yield into a fixed one.
How Does a Principal Token Turn a Discount Into Fixed Yield?
A principal token is a zero-coupon bond on a yield-bearing asset: bought below face value, redeemed at face value on a fixed date. On 16 November 2026 PT-sUSDe-28JAN2027, the deepest sUSDe market with more than a month to run, on the Monad blockchain, cost 0.9905 USDe and redeems for 1.0000 on 28 January 2027, a fixed 4.92% annualised (Pendle API, 2026).
A Dated Example: PT-sUSDe
An illustrative $10,000 of USDe buys 10,096 PT at the 0.9505% discount shown in Pendle's market data at 09:00 UTC on 16 November 2026. On 28 January 2027 those tokens redeem for 10,096 USDe, a gain of $96, or 0.960% over 72.6 days, which annualises to 4.92% (Pendle API, 2026; author's calculation). The result excludes gas and slippage and assumes USDe holds $1. The sUSDe underneath paid 5.10% on the same day, so the buyer gave up 0.18 percentage points to swap a floating rate for a fixed one. The fixed rate applies to a PT held to maturity: a PT sold on 15 January realises the pool's implied rate on that day, which differs from 4.92% in either direction, and a PT held to 28 January realises the full discount. Other on-chain yield sources, from tokenized Treasuries to lending and staking ↗, set the benchmark that a fixed rate competes against.
Four Live Markets Compared
The same formula prices every market, and the discount scales with time left. The 409-day wstETH market carries a 2.82% discount for a 2.59% annual yield, while the 73-day sUSDe market carries 0.95% for 4.92% (Pendle API, 2026). The table lists each market's implied annual percentage yield (APY), the rate its PT price locks in, beside the underlying APY the asset pays today. Staked ether quotes 2.59% against 4.9% to 5.9% for the stablecoin markets, a difference in what each asset pays. The wstETH underlying rate is Pendle's own reading of 2.37%, in line with the 2.38% that Lido's API showed for stETH that day. Pool liquidity ranges from $0.98M to $15.6M across the four, so a $10,000 position is small against each pool.
Data current as of November 2026.
The fixed rate has a mirror image: the buyer of the other half takes the variable one.
Why Is a Yield Token a Leveraged Bet That Decays to Zero?
A yield token pays all the yield of the underlying asset until maturity and is worth nothing afterwards. On 16 November 2026 YT-sUSDe-28JAN2027 cost 0.9505% of the principal it earns on, so $1,000 buys the yield on about $105,211 of sUSDe (Pendle API, 2026; author's calculation).
105 Times the Exposure
The leverage comes from price, not borrowing. A YT costs a fraction of the asset, and Pendle's docs describe the result as leveraged exposure to the asset's yield, where a small change in the underlying rate moves the YT's return by a large percentage (Pendle docs, 2026). The $1,000 position holds 105,211 YT, a 105-fold exposure. The buyer cannot be liquidated, and the loss is capped at the price paid. The same ratio applies to airdrop points: 1 YT earns the points of one unit of the accounting asset, which is why YTs are a popular points-farming tool ↗ (Pendle docs, 2026).
Break-Even, the Fee and the Decay
YT value falls toward zero as maturity approaches; at a constant implied yield a YT with 15 days left is worth about half of one with 30 days left (Pendle docs, 2026). Pendle also takes 5% of all yield that YT accrues (Pendle docs, 2026). With 72.6 days left, YT-sUSDe breaks even at an underlying rate of 4.78% before the fee and 5.03% after it (author's calculation). If sUSDe keeps paying 5.10%, the $1,000 returns $1,015, a 1.5% gain. sUSDS is worse: its YT is priced for 5.93% while the asset pays 3.80%, so an unchanged rate returns $629, a 37.1% loss.
Data current as of November 2026.
PT and YT are priced off the same pool, so one implied rate sets the price of both.
What Do Implied APY and Underlying APY Reveal About a Market?
Implied APY is the annual rate that a PT's price locks in; underlying APY is what the asset pays today. The gap between them is what traders are paid or charged to bet that the rate will change. On sUSDS the market implied 5.93% against 3.80% paid on 16 November 2026 (Pendle API, 2026).
The AMM Quotes a Yield
Pendle's automated market maker (AMM) trades PT against SY in one pool, and trades YT through flash swaps in the same pool. Liquidity concentrates inside a preset implied-APY range, and fees scale with time to maturity, so a trade a year before expiry costs more than the same trade a month before it (Pendle docs, 2026). A quote is therefore a rate, not a price. The range has an edge: when implied yield leaves it, liquidity is too thin to push the rate further (Pendle docs, 2026). The sUSDe pool on Monad held $15.57M of liquidity and handled $2,265 of volume in 24 hours on 16 November 2026 (Pendle API, 2026).
Maturity: Convergence, Redemption and Exit
PT price converges to 1 at expiry, and the AMM curve is shifted over time to push PT toward its redemption value (Pendle docs, 2026). A PT can be sold at the pool's current implied rate before then, so the exit price depends on pool depth. The wstETH market held $3.76M of liquidity and traded $10,493 in 24 hours on 16 November 2026 (Pendle API, 2026). After expiry PT redeems without YT, and yield on matured but unredeemed PT goes to Pendle's treasury (Pendle docs, 2026). The Ethereum sUSDe market that expires on 26 November, ten days after this snapshot, shows the endgame: a 0.14% discount remains for a PT worth 1.0000 on that date.
Pendle also applied the same split to a rate that is not an asset yield at all.
What Is Boros and How Does It Trade Funding Rates?
Boros applies the Pendle idea to perpetual-futures funding rates, the periodic payment between traders who are long and short. Pendle's team opened it to the public on Arbitrum on 6 August 2025 (Pendle Team, 2025). Boros held $6.6M of TVL on 16 November 2026, against $0.98B for Pendle V2 (DefiLlama, 2026).
Yield Units on Funding Rates
Boros trades Yield Units (YUs). One YU-ETH represents the yield from 1 ETH of notional value until maturity, and Pendle describes it as akin to a YT (Pendle Team, 2025). A trader who expects funding to rise goes long: pays a fixed rate and receives the floating funding rate. A trader who expects it to fall goes short: receives fixed and pays floating (Pendle, 2026). Pendle's docs name Ethena's delta-neutral strategy, which earns funding on short futures, as the use case, because a short YU turns variable funding income into a fixed one (Pendle docs, 2026). The first markets covered Binance's BTCUSDT and ETHUSDT funding rates.
Boros After Year One
The launch was capped on purpose: open interest of $10M of notional per market and 1.2x leverage, with higher limits promised as the system was validated (Pendle Team, 2025). Pendle declared Boros fully launched as v1.0 on 18 September 2025 and stated that it would have no token of its own. PENDLE stays the value-accrual token (Pendle Team, 2025). DefiLlama files Boros under interest-rate derivatives, as a child of the Pendle parent. The companion guide to Pendle's RWA markets, token economics and Boros revenue ↗ covers those topics, so this guide leaves them there.
Boros is small, and the larger Pendle markets show what a fast exit of capital looks like.
Why Did Pendle's TVL Fall 93% From Its September 2025 Peak?
Pendle's TVL fell from $13.39B on 19 September 2025 to $0.98B on 16 November 2026, a 93% drop, and the first half of it was over within seven days (DefiLlama, 2026). TVL had fallen 49% by 26 September 2025. The cause was a maturity date, not a depeg or a liquidation wave, and Ethena's USDe supply shows it.
The Maturity Cliff
Pendle TVL went from $12.24B on 24 September 2025 to $6.84B on 26 September. USDe supply over the same two days moved from $14.42B to $14.35B (DefiLlama, 2026). The date was 25 September: 33 Pendle pools expired that day, and Ethena's Season 4 incentives had ended on 24 September (FalconX, 2026). A matured PT redeems and leaves the protocol unless its holder rolls into a new market, so the peak had been inflated by capital positioned for the maturity. The tokens did not lose value; the deposits left.
Looping PT on Aave
Aave first listed Ethena PTs as collateral in April 2025, and Pendle's TVL rose with USDe supply through leveraged looping (FalconX, 2026). A loop buys PT, deposits it as collateral, borrows a stablecoin and buys more PT, one case of the yield stacking that the guide to composability in DeFi ↗ describes. The June 2025 Aave proposal for PT-sUSDE-25SEP2025 recommended a USDe efficiency mode with a loan-to-value of 89.1% and a liquidation threshold of 91.1% (Aave governance, 2025). Those figures allow a loop of up to 9.2x, and a 2.2% fall in PT value then reaches liquidation (author's calculation). On 10 October 2025 USDe traded at $0.65 on Binance but near $0.99 in Ethereum pools, and Ethena's governance analysis recorded about $47,000 of USDe and sUSDe liquidations on Aave (Ethena Governance, 2025). The loops survived the shock and left slowly.
What Remains After the Unwind
USDe supply fell from $14.66B on 10 October 2025 to $8.10B on 15 November and $4.87B on 16 November 2026 (DefiLlama, 2026). USDe and sUSDe made up 67.6% of Pendle's TVL on 19 September 2025 and 5.1% on 16 November 2026, which accounts for $9.00B of the $12.41B decline, or 72.5% (DefiLlama, 2026). Everything else fell too: from $4.34B to $0.93B. TVL bottomed at $0.93B on 19 June 2026 and has recovered 6% (DefiLlama, 2026). Aave accepts PT as collateral only inside an efficiency mode (E-Mode): the normal-mode loan-to-value is zero, E-Mode allows 89% to 94%, and $59M of PT was supplied on 16 November 2026, made up of $41.4M on Monad, $17.6M on X Layer and $0.1M on Ethereum and Plasma combined (Aave API, 2026). Expired Ethena PTs on Ethereum are frozen at a loan-to-value of zero. Morpho lists 66 markets with PT collateral: $319M of collateral and $260M borrowed, 4.9% of Morpho's borrowing (Morpho API, 2026). Together about $378M of PT backs loans on the two protocols.
Data current as of November 2026.
A smaller market still carries every risk the larger one did.
What Are the Real Risks of Fixed Yield on Pendle?
A PT fixes the rate but not the credit. The buyer holds a claim on the underlying asset, valued by an oracle and sold through a pool, and each of those three links can fail while the arithmetic stays correct. Pendle's developer docs state that in a sustained drawdown even a PT's redemption can fall below what was expected, because the SY backing has shrunk (Pendle docs, 2026).
Underlying and Collateral Risk
A depeg of the accounting asset passes straight to the PT. The 10 October 2025 episode shows how venue-specific that can be: Binance used its own order book as the price source for USDe, and the token printed $0.65 there while Ethereum pools held near $0.99 (Ethena Governance, 2025). Aave priced USDe against USDT and avoided mass liquidations (LlamaRisk, 2025). Off-chain managers add a separate failure: an external fund manager lost about $93M of Stream Finance's assets on 4 November 2025, and its xUSD token fell 77% in 24 hours (Decrypt, 2025). A PT on any such asset inherits the loss, because the principal it redeems is the damaged token. A companion article on vault curators ↗ covers that collapse.
Liquidity, Decay and Contract Risk
YT value reaches zero at expiry by design, so a YT buyer who is right about direction can still lose if the rate rises too slowly. Exit risk is measured in the pools: the wstETH market traded $10,493 in 24 hours on $3.76M of liquidity (Pendle API, 2026). The AMM's yield range adds a limit, since a rate outside it meets thin liquidity. Smart-contract risk applies to the SY wrapper, the yield contract and the AMM at once, and a borrower who loops PT on a lending market ↗ adds that market's contract and oracle risk on top.
Data current as of November 2026.
Conclusion
Yield tokenization turns one asset into two bets, a fixed rate and a variable one, priced against each other by a single implied rate. The arithmetic is sound: PT and YT always add up to the accounting asset, and every market resolves on a known date. The measured markets show the split working as designed, and the measured history shows what that design does to capital when the dates arrive.
The good case and the bad case look identical until a date or an underlying asset changes. Pendle's September 2025 peak was inflated by capital positioned for maturities and by loops stacked on an incentive programme, and the decline that followed returned the protocol to organic demand rather than exposing a contract failure. Fixed yield protects a holder against a falling rate. It gives no protection against a falling underlying asset or a thin exit.
Before buying PT, check what the accounting asset redeems for and who prices it, how much of the quoted annual percentage yield (APY) is the market's expectation rather than today's rate, and how deep the pool is against the size of the position. Before buying YT, compute the break-even underlying rate after the 5% fee. An implied rate above what the asset pays today means the price already assumes a rise.
Quick Stats
- $0.98B — Pendle TVL on 16 November 2026, 93% below the 19 September 2025 peak
- 4.92% — fixed annualised yield on PT-sUSDe-28JAN2027 on 16 November 2026
- 105x — yield exposure per dollar of YT-sUSDe with 72.6 days to maturity
- $4.87B — USDe supply on 16 November 2026, down 67% from the $14.82B peak
- $260M — borrowed against PT collateral on Morpho's listed markets, 4.9% of its borrowing
- $234.6M — PENDLE market capitalisation on 16 November 2026, rank 155 on CoinPaprika
Data current as of November 2026.
FAQ
?What happens to a PT that is not redeemed after maturity?
The PT stays redeemable, but the underlying asset behind it keeps accruing yield and points that Pendle redirects to its treasury fee wallet. Pendle designed that rule to push holders to redeem promptly and roll into a new market (Pendle docs, 2026). A holder who waits forgoes the yield accruing after expiry but keeps the principal.
?Who can open a new Pendle market for a yield-bearing token?
Any yield-bearing token can be wrapped into Standardized Yield (SY), which gives it the common interface that Pendle's yield contract needs, and the standard is permissionless by design. In practice pools are created by Pendle's team: its docs state that currently only the Pendle team deploys pools (Pendle docs, 2026). A token therefore needs a listed market before PT and YT trade.
?How is a Pendle PT different from fixed-rate lending on Morpho?
A PT is a discounted claim bought from a pool, and its yield is fixed by the purchase price when held to maturity. Morpho Midnight is fixed-rate, fixed-term lending, a loan to a borrower rather than a claim on an asset's yield, and it carried $199.8M of TVL on 16 November 2026 against $0.98B for Pendle V2 (DefiLlama, 2026).
?Are Boros Yield Units the same as Pendle YT?
Pendle describes a Yield Unit (YU) as akin to a YT: both pay the yield on a notional amount until maturity (Pendle Team, 2025). The difference is the source and the structure. A YT is minted from a yield-bearing token, while a YU trades on margin against a perpetual-futures funding rate, with leverage capped at 1.2x at launch.
?What does Pendle charge, and who receives the fees?
Pendle takes 5% of all yield that YT accrues, plus a swap fee that scales with time to maturity. Liquidity providers receive 20% of swap fees. The rest of the swap fees and all YT fees split 80% to PENDLE buybacks ↗, 10% to the treasury and 10% to operations (Pendle docs, 2026).
?Does a YT holder receive the underlying token's airdrop points?
Yes. One YT earns the same points as one unit of the underlying asset while costing a fraction of it, which makes YTs a popular points-farming tool. PT holders give up those points along with the variable yield (Pendle docs, 2026). The YT's price still decays to zero at expiry, so the points have to outweigh that decay.
?Is Pendle the only protocol that tokenizes yield?
No. Spectra splits interest-bearing tokens into the same PT and YT pair, and its docs state that YT heads toward zero at expiry (Spectra docs, 2026). Spectra V2 held $12.4M of TVL on 16 November 2026 against $0.98B for Pendle V2 (DefiLlama, 2026).
References / Sources
Market Data & APIs
Live figures pulled on 16 November 2026.
- DefiLlama: Pendle TVL, USDe Supply, Protocol Listings (api.llama.fi, Nov 2026)
- Pendle: Market Data API, PT and YT Prices (api-v2.pendle.finance, Nov 2026)
- Morpho: Listed Markets with PT Collateral (api.morpho.org, Nov 2026)
- Aave: PT Reserves on Ethereum, Plasma, Monad and X Layer (api.v3.aave.com, Nov 2026)
- CoinPaprika: PENDLE Ticker (coinpaprika.com, Nov 2026)
Protocol Documentation
Primary mechanics, fee and launch material.
- Pendle: SY, PT, YT, AMM and Fees (docs.pendle.finance, Oct 2026)
- Pendle Team: Boros Launch and v1.0 Posts (medium.com, Sep 2025)
- Spectra: Principal & Yield Token (docs.spectra.finance, Oct 2026)
Research & Governance
Independent analysis and governance records.
- FalconX: Pendle, One Venue for Fixed Income (falconx.io, Jan 2026)
- Aave Governance: PT-sUSDe September 2025 Proposal (governance.aave.com, Jun 2025)
- Ethena Governance: USDe Analysis, 10-11 October (gov.ethenafoundation.com, Oct 2025)
- LlamaRisk: When Pricing Breaks, a USDe Case (llamarisk.com, Oct 2025)
- Decrypt: Stream Finance xUSD Collapse (decrypt.co, Nov 2025)
Original CoinPaprika guide.