Lukegao: by freem.ai...

2026-01-24T05:38:57Z

by freem.ai...

New page

{{DISPLAYTITLE:Economic Risks}}

''This conversation was exported from ChatGPT''<ref>{{cite web|title=Economic Risks|url=https://chatgpt.com/share/6973721c-28b0-8007-a991-09211709d274|publisher=ChatGPT|access-date=2026-01-23}}</ref>

== gpt-5-2-thinking ==
=== User: What’s the economic risks of pendle PT looping on morpho or Euler ===
What’s the economic risks of pendle PT looping on morpho or Euler

Also let’s do optimal LTV settings calculation

=== Assistant: Below is how I’d think about economic not just smart-contract risk when you loop Pendle PT as collateral and borrow on … ===
Below is how I’d think about economic (not just smart-contract) risk when you loop Pendle PT as collateral and borrow on Morpho (Blue) or Euler, and then how to set an “optimal” LTV (really: optimal vs liquidation risk + execution risk + funding risk).

==== 1) Economic risks of Pendle PT looping (Morpho / Euler) ====

===== A) You’re short funding + long a “zero-coupon bond” =====

A Pendle PT is like a zero-coupon bond: you buy it at a discount and it converges toward redeemable value at maturity. docs.pendle.finance<ref>{{cite web|title=docs.pendle.finance|url=https://docs.pendle.finance/ProtocolMechanics/YieldTokenization/PT/|publisher=docs.pendle.finance|access-date=2026-01-23}}</ref>
Looping = levering that convergence trade, but you’re also short the borrow rate (variable funding).

Failure mode: borrow APR spikes (utilization shock) while PT implied yield compresses → carry turns negative, you de-lever into bad liquidity.

===== B) Liquidation is driven by oracle + LLTV/LT, not “fundamentals” =====

Even if PT “should” converge, your position lives/dies by the money market’s collateral valuation rule.
* Morpho Blue: liquidation eligibility is when LTV > LLTV, and health factor is computed directly from collateral value, debt, and LLTV. Morpho Docs<ref>{{cite web|title=Morpho Docs|url=https://docs.morpho.org/learn/concepts/liquidation/|publisher=Morpho Docs|access-date=2026-01-23}}</ref>
* Euler: liquidation is triggered when risk-adjusted collateral value < debt, and can happen due to price moves or debt growth. Euler Docs<ref>{{cite web|title=Euler Docs|url=https://docs.euler.finance/concepts/risk/liquidations/|publisher=Euler Docs|access-date=2026-01-23}}</ref>

Key economic risk: oracle design (spot vs smoothed vs PT-specific risk oracle). If the oracle can be pulled by secondary-market dislocations, you can be liquidated even though “maturity value” is fine.

===== C) “Liquidity gap” risk (the silent killer) =====

Liquidators must sell seized PT somewhere (or hedge it). If the PT market is thin / out of range / stressed, liquidation execution can fail or clear at huge discounts → bad debt risk for the market and forced liquidation losses for you. This is explicitly called out in PT risk discussions for money markets. Aave<ref>{{cite web|title=Aave|url=https://governance.aave.com/t/llamarisk-insights-aaves-pt-token-exposure-risk-outlook/22312|publisher=governance.aave.com|access-date=2026-01-23}}</ref>

Looping amplifies this because everyone is crowded on the same unwind path.

===== D) Maturity + roll risk (cliff events) =====

As maturity approaches, PT volatility typically declines and convergence strengthens, but your loop introduces a “roll decision”:
* close before maturity (market liquidity)
* hold through maturity (redemption mechanics, settlement timing, post-maturity oracle handling, etc.)

If many players roll at once, PT liquidity + borrow liquidity can both gap.

===== E) Underlying asset + wrapper risk =====

PT inherits risk from:
# Pendle contracts (tokenization + redemption flow)
# the underlying yield-bearing token / protocol (LST/LRT, stable yield token, etc.) A failure in either can reprice PT sharply and instantly. docs.pendle.finance<ref>{{cite web|title=docs.pendle.finance|url=https://docs.pendle.finance/pendle-v2/Developers/Oracles/PTAsCollateral|publisher=docs.pendle.finance|access-date=2026-01-23}}</ref>

===== F) Reflexivity + “loop convexity” =====

Looping makes your liquidation distance nonlinear:
* small adverse move → higher LTV
* higher LTV → less buffer
* any further move → liquidation cascade

This is why “2–3x is fine, beyond that gets hairy” is a common heuristic in PT looping guides. Medium<ref>{{cite web|title=Medium|url=https://medium.com/contango-xyz/how-to-enter-a-pt-loop-like-a-pro-2f279e72ac3f|publisher=medium.com|access-date=2026-01-23}}</ref>

==== 2) “Optimal LTV” calculation (practical framework) ====

There isn’t one universal optimum because it depends on:
* oracle type + update frequency
* liquidation bonus / close factor / auction mechanics
* PT liquidity + slippage under stress
* your rebalance frequency (do you monitor hourly, daily, weekly?)
* borrow rate volatility (util shock probability)

So the right way is: choose an acceptable liquidation probability over a horizon and back out a safe LTV.

===== Step 1 — Define variables (per market) =====
* LLTV (Morpho) or LT (Euler equivalent trigger)
* h = risk horizon (e.g., 1 day if you monitor actively; 7 days if you want “sleep well”)
* σ = volatility of collateral value in loan terms (PT priced in borrowed asset)
* r_b = borrow APR (use a stressed value, not today’s)
* LB = liquidation bonus / penalty (how much collateral you lose when liquidated)
* s = stress slippage haircut for unwinds (PT liquidity + DEX depth)
* z = quantile (e.g., 2.33 for 99%, 3.09 for 99.9%)

===== Step 2 — Compute a conservative “survival buffer” =====

Two things push LTV up over horizon h:
# collateral value can drop
# debt grows via interest

A simple (usable) bound:
* Collateral shock factor:

q=exp⁡(−z⋅σh)q = \exp(-z \cdot \sigma \sqrt{h})q=exp(−z⋅σh)
* Debt growth factor:

g=exp⁡(rb⋅h)g = \exp(r_b \cdot h)g=exp(rb⋅h)
* Execution haircut for liquidation mechanics + slippage (rule of thumb):

η=1−(LB+s)\eta = 1 - (LB + s)η=1−(LB+s)
(If LB=7% and you assume 5% slippage in stress, η≈0.88\eta \approx 0.88η≈0.88)

===== Step 3 — Safe initial LTV =====

For Morpho Blue, you want to stay below LLTV even after shocks:

LTV0,max⁡≈LLTV⋅qg⋅η\textbf{LTV}_{0,\max} \approx LLTV \cdot \frac{q}{g} \cdot \etaLTV0,max≈LLTV⋅gq⋅η
This gives you an LTV that has roughly “(1 − tail probability)” chance of staying solvent over horizon h, under your σ/r assumptions, and acknowledges that liquidation is lossy.

==== 3) Translating LTV into “how many loops” (quick) ====

If you loop by borrowing and buying more PT repeatedly, leverage roughly scales like:

Leverage≈11−LTV\text{Leverage} \approx \frac{1}{1-\text{LTV}}Leverage≈1−LTV1
So:
* LTV 0.70 → ~3.33x gross exposure
* LTV 0.80 → 5x
* LTV 0.85 → 6.67x This is why tiny LTV changes massively change liquidation fragility.

==== 4) What I need from you to compute your optimal LTV numbers (no fluff) ====

Send:
# Which PT (e.g., PT-sUSDe-YYYYMMDD) and chain
# Borrow asset (USDC? DAI? ETH?)
# Market + parameters (Morpho: LLTV + oracle type; Euler: LT / risk params if shown)
# Your desired monitoring horizon: 1d / 3d / 7d
# Your stress assumptions (or I’ll apply defaults): - σ (daily vol of PT/borrow) - stressed borrow APR (e.g., 2× current or p95)

If you don’t want to pull vol data, I can still give you a parameterized LTV curve (LTV as a function of σ and horizon) so you can plug numbers in instantly.

===== Default “founder-safe” settings (if you want an immediate starting point) =====

Without your specifics, a conservative institutional heuristic is:
* target liquidation probability ≤ 0.1% over 7d
* assume stressed borrow APR = 2× current
* assume stress slippage 5–10% on PT
* then set LTV 10–20 points below LLTV/LT for looping (bigger buffer if oracle is spot-like)

This aligns with how PT-as-collateral risk is discussed: liquidation execution + oracle behavior dominate outcomes, not the “it converges at maturity” story. Aave<ref>{{cite web|title=Aave|url=https://governance.aave.com/t/llamarisk-insights-aaves-pt-token-exposure-risk-outlook/22312|publisher=governance.aave.com|access-date=2026-01-23}}</ref>

If you reply with the exact PT + market, I’ll compute:
* recommended LTV band (aggressive/base/conservative)
* max loops / target leverage
* break-even borrow APR vs PT implied yield
* liquidation distance under 1d/7d shocks

== gpt-5-2 ==
=== User: Thanks ===
Thanks

=== Assistant: Anytime 🙂 ===
Anytime 🙂
If/when you want, we can turn this into a one-pager risk memo (Morpho vs Euler side-by-side, LTV bands, stress table) or run exact numbers for a specific PT. Just say the word.

== References ==
<references />

[[Category:ChatGPT Shares]]
[[Category:openai]]

Openai/6973721c-28b0-8007-a991-09211709d274 - Revision history

Lukegao: by freem.ai...