Pendle

PT and YT

A large part of my work was looking at what happens after a yield-bearing asset is split into Principal Token (PT) and Yield Token (YT). That separation creates a real market for future yield. PT prices a fixed return to maturity, while YT represents exposure to the yield generated before that date. I spent time looking at how the relationship between underlying yield, implied yield, maturity and market demand changes the economics of both sides of the trade.

Quoted and realized

One question I kept coming back to was whether the yield shown to a user could actually be realized at the size they wanted to trade. Pendle’s AMM concentrates liquidity inside an implied-yield range, which makes yield trading capital efficient, but available depth still matters. A position can look attractive at the quoted APY while a larger entry or exit moves the market enough to produce a very different effective yield.

Exit capacity

That made exit capacity an important part of the work. I looked at liquidity at different trade sizes, the path a position takes through Pendle’s AMM and order book, and the external liquidity required when the desired input or output asset is different from the asset inside the Pendle market. In some cases the constraint was not the yield product itself, but the liquidity one layer further down the route.

Several clocks

sUSDe was a useful example. Pendle can separate the future yield of sUSDe into PT and YT, which creates fixed-yield and long-yield markets on top of Ethena. But there are several different clocks involved: Pendle maturity, the remaining yield represented by YT, the liquidity available for an early exit, and the underlying path from sUSDe back to USDe.

Out of line

I looked at what happens when those layers stop lining up cleanly. A trader may be able to sell a PT or YT position before maturity, but the realized result depends on the implied yield available in the market and the depth at that point. Holding through maturity removes one part of that price uncertainty, but it does not remove the dependencies of the underlying asset.

That distinction became particularly important with sUSDe. Ethena’s native unstaking path includes a cooldown before USDe can be withdrawn. Secondary-market liquidity can provide a faster route, but then the exit depends on available liquidity and price impact. For a small position those differences can be easy to ignore. At larger size they become part of the trade itself.

Routing

I also spent time looking at routing. Pendle can source liquidity through its AMM and limit order book, and trades can be routed from or into other assets. On L2s or with non-native assets, that route can depend on external DEX liquidity as well. So the depth visible in the Pendle market is only one part of the liquidity needed to complete an end-to-end exit.

Maturity

Another part of the work was maturity and redemption. PT converges toward its accounting asset as maturity approaches, while YT loses its remaining time value and expires. This makes time a structural part of the market rather than just another variable. I looked at how that changes pricing, liquidity and exit behavior as a market moves closer to maturity.

Composability

The broader risk question was composability. Pendle positions can sit on top of assets from Ethena and other protocols, while PT itself can then be used as collateral elsewhere. That creates several layers of dependency. A problem with the underlying asset, its exchange rate, an external liquidity venue or another integrated protocol can propagate into what initially looks like a simple fixed-yield position.

What I took from it

For me, the most useful way to think about Pendle became less about advertised APY and more about the full path of the position: where the yield comes from, how the market prices it, what happens as maturity approaches, and how much capital can actually get out without materially changing the economics of the trade.