What is Pendle (PENDLE)?

Any yield-bearing token is two things at once: a claim on the underlying principal, and a claim on the yield it produces. Traditional finance separates those and trades them separately — that is what a zero-coupon bond is. Pendle brought it on-chain.

Deposit a yield-bearing asset and Pendle mints two tokens. PT is the principal, redeemable one-for-one at maturity, so it trades at a discount today — buy it and you have locked in a fixed yield. YT is the yield stream until maturity, which expires worthless, so buying it is a leveraged bet that yields will be higher than the market expects.

That is a genuine financial primitive, not a rehash of an existing one, and it created DeFi's first real fixed-income market. It also made Pendle the venue where the most sophisticated yield trades were constructed — which is why its numbers fell the way they did.

What the TVL collapse actually was

TVL hit a record $8.3bn in August 2025 and averaged around $1.3bn daily through the first half of 2026. That looks like a protocol failing and it is better read as the protocol working.

Much of that peak was the Ethena carry trade: deposit sUSDe, use Pendle to lever the yield, loop it through Aave. When perpetual funding rates compressed, the trade stopped paying and unwound — taking USDe's own supply from above $14bn to roughly $4bn, and Pendle's deposits with it. Pendle was the venue, not the position. Its TVL is a measure of how much leveraged yield trading exists, and that is a cyclical quantity.

Boros

Boros lets traders go long or short funding rates directly, rather than constructing the exposure through spot and perpetuals. Cumulative volume has passed $14bn.

Strategically this is the more durable business. Funding rates matter in every market condition, including the ones where nobody wants leveraged yield, so Boros gives Pendle a revenue line that is not simply a function of how bullish people feel. The H2 2026 roadmap points the same way: real-world assets, institutional markets and on-chain interest-rate derivatives.

• vePENDLE: lock PENDLE for up to two years for voting weight and a share of protocol fees.
• Directing incentives to specific pools through vote weight, which pool operators bid for.
• A claim on swap fees and a share of YT yield collected by the protocol.
• Governance over listings, maturities and the Boros product.

The vePENDLE model is borrowed from Curve and it works the same way: locking for longer gives more weight, and protocols that want liquidity for their pool bribe lockers to direct it there. It concentrates influence among the committed, which is the intent, and makes the token's yield partly a function of how much others want to rent your vote.

PENDLE has a fixed emission schedule that declines weekly, paying liquidity providers. Locking into vePENDLE removes tokens from circulation for up to two years and entitles the holder to protocol fees plus vote-directed incentives.

The economics are directly tied to activity: swap fees, a share of yield, and the value of vote weight all scale with how much capital is using Pendle. At $8.3bn of TVL those were substantial; at $1.3bn they are proportionally smaller, and the emission schedule did not shrink with them.

Why Boros matters to the token

The yield-tokenisation business is inherently cyclical because leveraged yield demand is. A funding-rate venue is closer to a permanent market — funding rates exist in bull and bear conditions alike and traders want exposure to them in both. $14bn of cumulative Boros volume is the argument that Pendle's revenue can decouple from the cycle that just halved its TVL.

Pendle is not a chain and PENDLE secures nothing, so there is no protocol staking yield. What exists is locking into vePENDLE, which pays a share of protocol fees and voting incentives in exchange for committing the tokens for up to two years with no early exit.

Separately, using Pendle is itself a yield strategy and the two get confused. Buying PT locks in a fixed rate to maturity. Buying YT is a leveraged directional bet on yields. Providing liquidity to a PT/YT pool earns fees with an impermanent-loss profile specific to assets that decay toward maturity — which is unusual and catches people who assume it behaves like a normal pool.

The business is leveraged yield demand

TVL from $8.3bn to around $1.3bn in under a year is the clearest possible demonstration. Pendle earns when people construct sophisticated yield trades, and that happens when yields are high and sentiment is strong. It is a high-beta business on the crypto cycle, and the 2026 figures are what the downside looks like.

Concentration in whatever yield is fashionable

Pendle's deposits clustered heavily in Ethena's sUSDe during the carry-trade era. When that unwound, so did Pendle. The protocol is exposed to the specific yield sources its users favour, and those rotate faster than the protocol can diversify.

Complexity is a real user risk

PT, YT, maturities, implied yields and a pool whose assets decay toward expiry is genuinely difficult, and people lose money misunderstanding it. Buying YT close to maturity, in particular, is a decaying asset that expires worthless — correct behaviour that reads as a bug to anyone who did not read the documentation.

vePENDLE locks are long

Up to two years with no early exit, in an asset whose revenue just fell by most of its peak. The yield compensates for that and it is compensation for something real.

Boros has to prove it is different

$14bn of cumulative volume is a strong start and the thesis — that funding-rate trading persists across cycles — is untested through a full bear market. If Boros turns out to be as cyclical as yield tokenisation, Pendle has one business rather than two.