There's a version of this article I drafted three weeks ago that talked about Pendle's $5.2B TVL like it was a stable, growing protocol metric. I had to throw it out.
The actual numbers tell a much more dramatic story. Pendle peaked at ~$13.1B in TVL during the 2025 yield-tokenization mania. By August 2025 it was sitting at $8.9B with Ethena USDe accounting for over 70% of total TVL ($6.1B in PT-USDe and related positions alone). Then USDe yields compressed and the leveraged loops unwound. By Q1 2026 Pendle TVL had collapsed to $1.96B — a multi-year low and roughly 85% off peak.
That's not a Pendle protocol failure. The smart contracts work. The PT/YT mechanics still produce fixed yield. The protocol revenue is still flowing. What happened was a textbook concentration risk event: when you build a $9B TVL protocol where one asset is 70% of the deposits, that asset's yield trajectory becomes the protocol's trajectory. USDe's funding rate compression took down Pendle's headline numbers because the leveraged USDe loop on Aave that drove PT-USDe demand became less profitable.
Meanwhile, Pendle's actual most interesting product right now is Boros — a funding rate trading platform launched August 2025 that did $2.9B in monthly volume by January 2026 and now accounts for 58% of on-chain yield trading. The Boros story is what to focus on. The PT/YT story has matured and is now structurally smaller. Below is what actually happened with Pendle in 2025-2026, why the concentration risk is the lesson worth remembering, and how my own Pendle positioning evolved through it.
The TVL Trajectory Nobody Wants To Frame Honestly
| Period | Pendle TVL | Ethena USDe share | Comment |
|---|---|---|---|
| Peak (early 2025) | ~$13.1B | ~70% | USDe leverage loops at maximum |
| August 2025 | ~$8.9B | ~70% ($6.1B in USDe-related) | Already declining from peak |
| Late 2025 | ~$3-5B | declining share | USDe unwind accelerating |
| Q1 2026 | ~$1.96B | ~30-40% | Multi-year low |
The drop from $13.1B to $1.96B is 85% off peak. That's a brutal contraction by any measure. A normal DeFi protocol losing 85% of TVL would be a death spiral story.
Pendle isn't dying. The contraction is largely about one specific strategy unwinding rather than the underlying protocol mechanics breaking. The PT/YT architecture still works. Fixed-yield positioning still has structural demand. New product launches (Boros) are growing strongly. The protocol is fine; the concentrated TVL number was always inflated by the USDe leverage loop.
What Was The USDe Leverage Loop?
The structure that drove Pendle to $13.1B at peak:
- Buy Ethena sUSDe (yields 8-12% APY from funding rate spread)
- Convert sUSDe to PT-sUSDe on Pendle (lock in fixed ~8.8% APY)
- Deposit PT-sUSDe as collateral on Aave V3 (Aave accepted it as collateral)
- Borrow USDC against PT-sUSDe at ~4-6% APY
- Use borrowed USDC to buy more sUSDe
- Repeat (recursive)
At maximum leverage, this produced effective yields of 12-20% on the original capital with manageable IL exposure (because PT was locked-in fixed yield, not floating).
When Aave raised PT-USDe collateral cap by $600M in September 2025, it was fully utilized within an hour. That's how aggressive the leverage demand was. The strategy worked when funding rates were high and Aave's USDC borrow rates were low. The spread captured was real.
What killed it: USDe funding rate compression in late 2025. As ETH funding rates dropped from sustained 15-20% annualized to 6-9% range, sUSDe yield compressed to ~6-8%. That made the leverage loop unprofitable at high leverage. Capital unwound. PT-USDe positions on Aave got reduced. Pendle TVL collapsed proportionally.
The lesson: protocol TVL can be 70% concentrated in one strategy and that strategy can compress. When it does, headline TVL numbers fall hard even though the underlying protocol mechanics still work. Don't anchor protocol valuations to peak TVL when peak TVL came from one specific loop.
What's Actually Happening In Pendle Now
Q1 2026 Pendle is a different protocol than 2025 peak Pendle. Three structural shifts:
TVL is more diversified. USDe share is ~30-40% (down from 70%). LRT positioning (eETH, ezETH, rsETH) is ~25-30%. Stablecoin yield positioning beyond USDe (sUSDS, sDAI, USDM) is ~20-25%. Other yield-bearing assets are ~10-15%. The remaining $1.96B TVL is structurally healthier than the $13.1B peak because the concentration risk has been wrung out.
Boros is the growth story. Boros launched August 2025 as a funding rate trading platform. By January 2026 it had monthly volume of $2.9B and cumulative volume crossing $9.8B. Boros became 58% of on-chain yield trading. This is materially bigger than legacy Pendle PT/YT trading. The product is genuinely working.
Pendle's role for institutions has shifted. Pre-collapse, Pendle was where retail and institutional capital chased aggressive USDe leverage yields. Post-collapse, Pendle is where institutions hedge funding rate exposure (via Boros) and where users seeking fixed-yield positioning find PT products. The use case is more mature even if the headline TVL is smaller.
Boros Is The Story Now
Boros is structurally interesting because it solves a real problem nobody else solves cleanly: trading funding rates as their own asset class.
If you're running an Ethena USDe-style basis trade, you're long spot ETH and short ETH perpetual to capture funding rate. Your P&L depends on funding rates staying positive. If funding rates compress, your strategy underperforms. You can't easily hedge funding rate exposure on traditional CEX or DEX venues.
Boros lets you:
- Take a position on funding rate direction (long if you expect rates to rise, short if you expect them to fall)
- Hedge existing funding rate exposure (lock in current rates to immunize against compression)
- Trade funding rate volatility specifically rather than ETH price
For Ethena (which has $5.8B in USDe needing funding rate stability), Boros is critical infrastructure. For other protocols and traders running funding rate-dependent strategies, Boros is the only real venue for that hedging.
Boros monthly volume Q1 2026:
- January: ~$2.9B
- Cumulative since August 2025 launch: ~$9.8B
That's growth. The product is working and traction is real. PENDLE token capture from Boros activity is meaningful even if legacy PT/YT trading is smaller than 2025 peak.
What I Learned From The Pendle Peak
I was running ~$80-150K notional in Pendle PT positions through 2025 peak. Concentrated in weETH PT and sUSDe PT. At the time it felt safe — fixed yield, locked in rates, predictable maturity.
Mistake: I had ~$60K in sUSDe PT positions when funding rate compression started in late 2025. Pre-compression my PT-USDe yield was ~10% locked. As USDe yield compressed below ~8%, the PT secondary market for sUSDe started pricing the maturity at a discount. I held to maturity (got the locked rate) but if I'd needed to exit early I'd have taken a loss.
Bigger mistake: I had no concentration awareness at the protocol level. I knew my own Pendle position was concentrated in sUSDe. I didn't think about the fact that Pendle as a whole was 70% concentrated in sUSDe-related positions. When that concentration unwound, every PT market on Pendle saw secondary trading impact even if the specific PT wasn't sUSDe.
The lesson I'm still working through: protocol-level concentration risk is real and rarely visible to retail users. You can do everything right at your own position level and still face protocol-level dynamics that compress your liquidity or pricing. Position sizing has to account for what the protocol overall is concentrated in.
My Updated Pendle Allocation
Post-USDe-unwind, my Pendle positioning:
- ~$30-50K in weETH PT positions (LRT fixed yield, structurally cleaner exposure)
- ~$10-20K in sUSDS PT (Sky savings rate fixed)
- Reduced sUSDe PT exposure to ~$15-25K (still meaningful but no longer overweight)
- ~$5-10K Boros positioning (testing the funding rate trading product)
Total Pendle exposure ~$70-105K — down from $80-150K pre-unwind. Diversified across multiple yield-bearing assets. Lower concentration in any single strategy.
I held vePENDLE through the TVL collapse. PENDLE token traded $4.20-5.80 across Q1 2026 (down from 2024 peaks). vePENDLE yield is 18-28% APY which is competitive even at compressed protocol size. The economics are: vePENDLE captures a share of Pendle protocol fees, fees scale with trading volume (PT/YT + Boros), Boros volume growth offsets PT/YT contraction.
Where Pendle Is Vulnerable
Three risks to monitor:
Boros is now the protocol's growth driver. If Boros adoption stalls or competing funding rate trading platforms emerge, Pendle's growth story compresses fast. The PT/YT side is mature; future growth has to come from Boros.
USDe ecosystem health. Ethena's $5.8B USDe supply still has substantial Pendle integration. If Ethena faces another funding rate compression cycle or operational issue, Pendle's USDe-anchored TVL takes a hit even at current reduced concentration.
Yield environment dependency. Pendle PT economics depend on underlying yields existing. If T-bill yields, ETH staking yields, and stablecoin yields all compress simultaneously (which would happen in a Fed cutting cycle), Pendle's PT product becomes less interesting because there's less yield to lock in.
Competing yield-tokenization platforms. Term Finance, Spectra, and emerging alternatives compete for Pendle's market share. None has matched Pendle's network effects yet but the threat is real if Pendle stumbles operationally.
Where Pendle Is Defensible
The structural reasons Pendle survives even with TVL compression:
Boros is genuinely novel. Funding rate trading as its own asset class doesn't have direct competition. As long as Boros works, Pendle has a unique product.
PT positioning has institutional demand. Family offices, fund vehicles, and DeFi-active institutions all use PT positioning for fixed-yield exposure. That demand is structural and unlikely to disappear.
vePENDLE economics align long-term holders. PENDLE locked as vePENDLE captures protocol revenue. Long-term holders are economically aligned with protocol health, which provides governance stability.
Multi-asset coverage. Pendle integrates across 50+ yield-bearing assets. Concentration risk in any single asset is bounded by diversification across the protocol's footprint.
Decision Framework
If you want fixed yield on yield-bearing assets: Pendle PT positions, but watch the protocol-level concentration. Don't take a position larger than ~10% of the specific PT market liquidity.
If you want to trade funding rates: Boros is currently the only structural option. Worth using if you have specific funding rate exposure or directional view.
If you hold PENDLE/vePENDLE: The compressed TVL + Boros growth story means the token economics are roughly balanced. Hold if you believe in continued yield-tokenization adoption; rotate if you think Boros stalls.
If you're running USDe-style leveraged strategies: Be aware of protocol-level concentration. The 2025 USDe leverage trade worked until it didn't. Position size for unwind scenarios.
What I Watch For
Boros monthly volume trajectory. $2.9B/month in January 2026. If that grows to $5-8B/month by end-2026, Pendle's recovery is on track. If it stalls below $3B/month, the protocol growth story is in trouble.
USDe TVL share on Pendle. If USDe share creeps back above 50%, concentration risk is rebuilding. If it stays under 35%, the protocol is healthier structurally.
Aave PT-USDe collateral utilization. The September 2025 cap raise was a leading indicator of leverage demand. Watching this cap utilization tells you what the leveraged USDe loop is doing.
Pendle V3 development. Pendle has been building V3 mechanics. Major architecture shifts could either revive growth or introduce new risks.
Sources
The TVL trajectory data ($13.1B peak, $8.9B August 2025, $1.96B Q1 2026) is from DeFi Llama and Pendle's published dashboards. The Ethena USDe concentration figures and Boros volume data are from third-party analysis and Pendle's own disclosures.
Pendle TVL on DeFi Llama — current and historical TVL data
Pendle Settles $69.8B in Yield (DL News) — coverage of Pendle's cumulative yield settlement scale
Boros yield trading analysis (ChainCatcher) — third-party analysis of Boros traction
Pendle: One Venue, All of Fixed Income (FalconX) — institutional perspective on Pendle's positioning
DeFi Yield Engines analysis (CoinDesk) — broader yield engine context including Pendle/Ethena/Aave
Caveats
The TVL trajectory and Ethena USDe concentration figures are from publicly disclosed Pendle data plus third-party analysis through April 2026. The "85% off peak" framing depends on which peak you anchor to ($13.1B all-time high vs $8.9B August 2025); the directional pattern is robust either way. The Boros volume figures are from Pendle's published metrics and have grown rapidly so any specific number ages quickly. The concentration risk lesson is my read — others might frame the unwind differently. None of this is investment advice — Pendle PT positioning involves protocol-level concentration risk and yield-environment risk you should size around.