Jupiter Perps Lending Initial Parameter Recommendations and Yield Projections

Motivation

The launch of Jupiter Perps Lending represents a pivotal opportunity for Jupiter to extend its ecosystem’s capital efficiency by introducing a lending market purpose-built around the JLP pool. At present, a substantial portion of USDC within JLP remains idle from a capital deployment perspective, despite its utility in facilitating perpetual trading. By enabling the borrowing of this unused USDC, Jupiter Perps Lending can unlock a powerful new yield stream for JLP holders while maintaining risk controls appropriate for a volatile, multi-asset-backed collateral base.

This proposal seeks to define a conservative yet effective framework for the initial rollout of Jupiter Perps Lending, starting with JLP deposits and USDC borrows. The goal is to generate immediate, tangible yield improvements for JLP holders while preserving the protocol’s long-term stability and solvency. Our recommendations are grounded in a data-driven approach that takes into account the historical behavior of USDC in Jupiter perps markets, the risk characteristics of JLP, and comparisons to industry-standard lending parameters across leading DeFi protocols.

Launch Parameters

The table below outlines our recommended risk parameters for the initial deployment of Jupiter’s stable lending protocol, focusing on JLP deposits and USDC borrow functionality. These recommendations are informed by a synthesis of historical USDC utilization in Jupiter’s perpetual markets, the volatility profile of the JLP asset, and benchmarking against established risk frameworks used by leading lending protocols. In the subsequent analysis, we examine historical USDC borrow APRs and utilization rates across top-tier lending platforms to estimate the incremental yield that can be driven to the Jupiter protocol and the JLP token as a result of this initiative.

Parameter Recommended Value Recommendation Rationale
borrows_limit_in_bps 5000 bps (50%) USDC’s average utilization in Jupiter perps over the past 180 days was 9%, with a maximum of 21%. A borrowing limit of 5000 (50%) provides for a significant liquidity buffer relative to USDC peak utilization rates.
liquidation_margin 8600 bps (86%) Set slightly above the typical LTV ratios for WBTC, WETH, and WSOL on leading DeFi lending protocols, reflecting the added risk diversification inherent in the JLP asset.
maintainance_margin_bps 8000 bps (80%) The maintenance_margin_bps is set at a prudent buffer above the liquidation_margin to ensure a smoother user experience and to mitigate the risk of immediate liquidations during periods of heightened market volatility.
protocol_fee_bps 2500 bps (25%) Set equal to Jupiter’s current perpetuals fee take rate.
liquidation_fee_bps 600 bps (6%) Provides a sufficient buffer to ensure profitable native liquidations by the protocol, even in the absence of a liquidation bonus. In most cases, there is more than enough liquidity to facilitate low-fee liquidations. However, under high-utilization conditions, liquidators will not act at the recommended fee level. Moving forward, we recommend a risk oracle to dynamically incentivize liquidations.
min_rate_bps 0 bps (0%) Incentivizes borrow growth along the path to target_utilization_rate.
target_rate_bps 850 bps (8.5%) A competitive borrow rate compared to USDC borrowing rates on Solana.
max_rate_bps 1500 bps (15%) A relatively shallow slope allows for high direct borrowing demand even as perp traders’ borrowing increases.
target_utilization_rate 6000 bps (60%) Allowing for 50% utilization by direct USDC borrowers + 10% borrowed by perp traders. Of course, higher utilization by perp traders may take place, possibly triggering a decrease in direct USDC borrowing.

The below volatility and utilization data informed our liquidation_margin and borrows_limit_in_bps parameter recommendations, respectively.

Annualized Volatility of Hourly Returns (last 365 days of data)

The annualized standard deviation of hourly returns for JLP (virtual) is meaningfully lower than that of BTC, ETH, and SOL.

Asset Volatility
JLP 38.13%
BTC 43.38%
ETH 59.03%
SOL 75.95%

Similarly, at the P99 and P99.9 level, JLP’s absolute 1H virtual price changes are less extreme than each of the constituent volatile assets in its basket.

Absolute Price Changes (Hourly)

Percentile JLP BTC ETH SOL
P50 0.16% 0.19% 0.25% 0.38%
P90 0.57% 0.66% 0.88% 1.18%
P95 0.80% 0.95% 1.25% 1.58%
P99 1.49% 1.77% 2.24% 2.82%

As such, we anchored JLP’s core collateralization requirements — maintainance_margin_bps and liquidation_margin — above those of WBTC, WETH, and WSOL on leading DeFi lending protocols.

Utilization Metrics

USDC’s peak short utilization was 21% over the last 6 months. As such, even at borrows_limit_in_bps 5000 (50%), perps short utilization could spike by more than 100% above its trailing 6mo high before encroaching on USDC that may be lending-utilized. A target_utilization_rate of 60% for USDC is only provided to allow for high utilization at all times, while capitalizing on high interest rates during interest rate spikes.

Asset WBTC WETH SOL USDC USDT
Average Utilization 39% 7% 21% 9% 0%
Min Utilization 19% 5% 15% 2% 0%
Max Utilization 95% 52% 100% 21% 12%

Yield Projections

Over the last 12 months, Aave v3’s USDC Ethereum Core market — the largest USDC lending market in DeFi — has sustained an average borrow APR of 6.73%, though this rate has trended down to a relatively steady state of ~5% today.

However, on Solana, we traditionally see a higher borrowing yield on stablecoins, as can be seen on Kamino’s USDC market:

New JLP Yield Projections Table

The table below shows the expected yield from direct borrowing of USDC only, not including yield from perp traders borrowing USDC under the assumption of full utilization of direct USDC borrowing at all times, i.e., 50% of the available USDC, considering the current supply of 494M USDC and the current borrowing interest rates.

Average Utilization Rate Average APR Expected yield (3mo) Expected yield (6mo) Expected yield (1yr)
60% 8.50% 2.125% / $5.2M 4.25% / %10.5M 8.5% / $21M
70% 9.80% 2.45% / $6M 4.9% / $12.1M 9.8% / $24.2M
80% 11.10% 2.775% / $6.8M 5.55% / $13.7M 11.1% / $27.4M
90% 12.40% 3.1% / $7.7M 6.2% / $15.3M 12.4% / $30.6M
100% 13.70% 3.425% / $8.5M 6.85% / $16.9M 13.7% / $33.8M

Summary

The proposed Jupiter Perps Lending launch parameters are designed to unlock significant yield potential for JLP holders by safely deploying idle USDC from the JLP pool. Our recommendations are grounded in a comprehensive analysis of historical utilization patterns, risk-adjusted volatility metrics, and prevailing practices across leading DeFi lending markets. By adopting a conservative initial configuration, anchored in both quantitative rigor and market precedent, Jupiter is well-positioned to drive sustainable value accrual to the protocol and its stakeholders.

As utilization scales and market conditions evolve, we anticipate opportunities to further optimize these parameters, particularly via the introduction of dynamic risk management mechanisms such as a responsive risk oracle. We invite feedback from the Jupiter community and governance participants to refine these parameters ahead of deployment, and look forward to a collaborative path toward growing Jupiter Perps Lending into a core yield engine for the Jupiter ecosystem.

Disclaimer

Chaos Labs has not been compensated by any third party for publishing this recommendation.

Copyright

Copyright and related rights waived via CC0

5 Likes

In total support of this proposal and the initial setup of things. Just wanted to mention a few things specific to Jupiter Lend (aka Fluid) and how few things need to be factored differently.

Insights

I think renaming borrows_limit_in_bps to allocation_limit_in_bps might be better and more precise.

Jupiter Lend allows to set a few extra parameters, it has 3 parameters:

  • Loan to Value (LTV): The ratio/percent to which user can borrow, if set to 80% then it means user can borrow $80 for every $100.
  • Liquidation Threshold (LT): The ratio/percent at which user’s position becomes available for liquidation, if set to 86% then above 86% user’s position can be liquidated.
  • Liquidation Max Limit (LML): The ratio/percent at which user’s position becomes available for 100% liquidation, if set 92% then above 92% the position will be fully wiped off, kind of like giving 8% liquidation penalty bonus.

Between LT to LML, liquidation only happens to LT, for example, if position becomes 88% then liquidation can at max move the position to LT, in this case 86% so not as much liquidation is require to liquidate making liquidation much more optimized.

The goal is to keep Liquidation Penalty super low between LT & LML to minimize the liquidation losses for end users like in this case 1-2% Liquidation Penalty should be enough. Above LML, position will get fully liquidated to make sure protocol remains safe of bad debt. This link should help understand the benefit of Jupiter_Lend / Fluid’s liquidation.

  • borrows_limit_in_bps: To start with 50% sounds good, however should easily be able to expand it to 70% or 80% as Jupiter Lend grows, USDC can be instantly withdrawn from Jupiter Lend to be made available for Jupiter Perps.
  • liquidation_fee_bps aka Liquidation Penalty: 2% to be conservative on launch. Later upgrade it to 1% when liquidation mechanism gets widely integrated across Solana ecosystem.

Lastly, I think we should also have an analysis on SOL market and how it’s utilization reacts in different market conditions. SOL is the biggest market on Jupiter Perp with liquidity of $728M. Allocating SOL to Jupiter Lend to be used in JUPSOL-SOL looping strategy can be very beneficial and also add up nicely on fees & revenue.

2 Likes

@chaoslabs
Honestly i like idea about create lending protocol
Also its might could meet my proposal about CAT FIRE NFT (Financially Independent Retired Early)

:sparkles: Upgrade: Introduce Tokenized Gold as a Lending Asset or Collateral

:white_check_mark: Option 1: Gold as a Yield-Bearing Collateral

Allow users to stake tokenized gold (PAXG, XAUT) into the lending system.

Use it to back stablecoin loans (USDC).

Gold-backed loans = lower volatility risk than DeFi tokens.

:white_check_mark: Option 2: “Gold-Boosted Borrowing” Incentive Layer

Users who stake JUP or JLP + gold get lower borrow rates

Adds incentive to lock value and reduces protocol default risk


:chart_increasing: Benefits

Benefit Description

:shield: Risk Reduction Gold is less volatile than crypto collateral
:money_bag: Real Yield Tokenized gold can earn yield in real-world markets
:locked: Capital Efficiency USDC-backed by gold = lower loan risk
:globe_with_meridians: Diversification Makes Jupiter a leader in RWA-integrated DeFi