Pendle Finance, the yield-trading protocol that splits yield-bearing assets into principal tokens (PT) and yield tokens (YT), crossed $8 billion in total value locked this week. The growth has accelerated through Q2 from roughly $5 billion at the start of the year, and the composition of the new flows has shifted in a meaningful way.
A year ago, the bulk of Pendle's TVL was in ETH-staking-derived markets — primarily Lido's stETH and the various restaking tokens. Today, tokenized treasury markets — Ondo's USDY, Mountain's USDM, BlackRock's BUIDL, and a growing list of similar assets — account for the largest single share, and the marginal growth flow has tilted further in that direction.
The Mechanic, In One Paragraph
Pendle takes a yield-bearing asset (say, stETH at 3.5% staking yield) and splits each unit into two tradable components. The principal token (PT) entitles its holder to one unit of the underlying asset at maturity. The yield token (YT) entitles its holder to all of the staking yield generated by that unit between now and maturity. The two components trade independently, which lets sophisticated users do things like lock in fixed yield (by buying PT at a discount), lever up on yield expectations (by buying YT), or hedge yield-rate risk (by selling YT against a held position). The mechanism is more than three years old and has proved robust through multiple market cycles.
Why the Mix Is Shifting
The shift toward tokenized-treasury yield has three drivers.
First, the absolute level of treasury yield. Tokenized treasuries currently yield somewhere in the 4.4% to 4.8% range depending on the issuer and underlying maturity. ETH staking yield has compressed below that level — currently in the 2.8% to 3.2% range as the validator set has grown and as MEV revenue has normalized lower. For yield-seeking depositors, the absolute level matters, and for several quarters the tokenized-treasury side has offered better risk-adjusted economics.
Second, the composition of who is depositing. The depositor base for Pendle's PT markets has tilted toward institutional and treasury allocators — DAO treasuries, fund vehicles, and individual depositors using Pendle to construct fixed-yield positions on tokenized treasuries. That cohort has both a strong preference for treasury-backed yield over crypto-native yield and the operational sophistication to use Pendle's split-token structure efficiently.
Third, the maturation of the tokenized-treasury issuers themselves. A year ago, the universe of tokenized-treasury assets that could plausibly be plugged into Pendle's framework was small, and the issuers' redemption mechanics weren't proven enough for protocols to take collateral exposure. That has changed. The major issuers now have well-tested redemption infrastructure and clear regulatory positioning, and Pendle has built integrations with the most liquid of them.
The Risk Profile Has Also Shifted
The Pendle of 2024 was mostly a market for crypto-native yield — staking, restaking, LP positions. Risk concentration was within crypto, where the underlying-asset volatility was the dominant variable. The Pendle of 2026 is increasingly a market for off-chain yield, where the underlying-asset volatility is low but the dependencies extend off-chain in ways that aren't fully transparent.
The clearest example is the tokenized-treasury redemption risk. Pendle's PT markets work cleanly only if the underlying tokenized treasury can be redeemed at par at maturity. If a tokenized-treasury issuer were to face a redemption disruption — a brief operational failure, a custodial issue, a regulatory action — the principal token's claim on the underlying could be impaired in ways that the protocol's liquidation logic isn't well-suited to handle. The largest issuers have strong operational track records, but the failure modes are different from crypto-native ones, and the protocol's risk management has had to adapt.
There is also a maturity-date concentration risk that has grown with TVL. Pendle markets cluster around standardized maturity dates — 90, 180, 365 days from issue — and the volumes at each cluster point have grown to levels where rolling positions across maturity boundaries can move the market materially. The protocol's mechanisms for managing this concentration are functional but haven't been tested at the scale they're now operating.
The Competitive Picture
Pendle remains the dominant yield-trading protocol by a wide margin. Several competitors have launched similar PT/YT split mechanisms — Element Finance was an early one, with various successors emerging since — and none has gained meaningful share against Pendle's lead. The network effects of liquidity concentration in Pendle's markets are real, and the marginal yield-trading flow continues to consolidate there.
The more interesting competitive dynamic is between Pendle's split-token structure and the simpler yield-bearing wrappers that don't decompose principal and yield. Most institutional depositors who want to earn tokenized-treasury yield don't actually need the PT/YT split — they want to hold the underlying tokenized treasury and earn its native yield. Pendle's value-add for that depositor is the fixed-yield lock that buying PT at discount provides, which is meaningful but not essential.
If a competing protocol — or a tokenized-treasury issuer directly — offered a comparably liquid fixed-rate product without the operational overhead of Pendle's split-token framework, the institutional flow could rotate. So far that competing product hasn't appeared at scale. Whether it will is the central competitive question for Pendle's next year.
For now, the $8 billion milestone is real, the composition shift toward off-chain yield is the more interesting story, and the protocol has positioned itself at the center of how DeFi interfaces with tokenized treasury yield. That is a meaningful strategic position even if the underlying mechanic is several years old.
