[SIP2.8] Adding Fluid to ETH Higher Risk Strategy

Summary

This SIP proposes integrating Fluid Lite ETH into the ETH Mainnet (Higher Risk) vault in the Lazy Summer Protocol.

Extensive discussion and technical diligence have confirmed that Fluid Lite ETH’s redemption mechanics, fees and other technical considerations meet Lazy Summer’s standards, see previous thread [RFC] Onboard Two New ETH Vaults: Including Lazy Summer's first higher risk strategy

Adding Lite ETH will diversify ETH yield strategies, tap into Fluid’s unique liquid staking model, and expand Lazy Summer’s yield-generating sources beyond standard DeFi lending protocols.

Motivation

  • Expand yield sources – Provides exposure to Fluid’s staked ETH-backed product, adding a complementary ETH yield stream alongside existing ARKs.
  • Attract new user types – Engages stETH and Fluid Lite ETH holders, deepening penetration into the ETH staking ecosystem.
  • Increase TVL – Adds a ETH ARK, supporting Lazy Summer’s ETH vault growth objectives.

Specification

Parameter Value
Vault ETH Mainnet – Higher Risk
Network Ethereum Mainnet
New ARK FLUID Lite ETH
Contract 0x9600A48ed0f931d0c422D574e3275a90D8b22745
Risk Level Higher Risk
$SUMR Rewards In line with ETH Higher Risk SUMR Rewards

@jensei @Recognized_Delegates @chrisb @halaprix @0xtucks

4 Likes

Hereby voicing support for this addition.

2 Likes

Fluid Lite ETH brings another diversification angle to the ETH Higher Risk vault. This helps Lazy Summer offer broader exposure across the ETH yield landscape.

Let’s move this forward after any final review from @BlockAnalitica.

2 Likes

Big fan of adding this too. It’s exactly the sort of thing that the Higher Risk strategies are for IMO.

3 Likes

Instadapp Lite Vaults v2 - Risk Assessment by BA Labs

Assessment Summary

  • BA Labs supports onboarding the Instadapp Lite v2 ETH vault as an ARK to the Higher risk ETH Fleet on Mainnet
  • Reference parameters for Instadapp Lite v2 ETH vault will be provided in a separate post below

Overview

Lite vaults (v1) were launched in March 2022 by the Instadapp team who have a proven track record considering their presence in Ethereum DeFi since 2018 by building Instadapp Dashboard (now Instadapp Pro) for managing MakerDAO CDPs initially, and later on aggregate numerous DeFi lending and DEX protocols.

The idea behind the Lite vaults v1 was to offer a 1-click yield solution while depositors receive the proof-of-deposit token, aka iTokens (e.g. iETH). The approach of Lite vaults is pool-based LP-ing, meaning the capital deposited is used to create a single (leveraged) position (e.g. stETH/ETH) in the underlying lending protocol (e.g. Aave v2), while the management of the capital was essentially done by the Instadapp team.

stETH depeg in 2022

While Lite v1 vaults reached ~$200M across four vaults (4 different deposit tokens), the ETH Lite v1 vault was put through the test when stETH lost its peg in the aftermath of the Luna meltdown, around 2 months after the Lite v1 launch. Being one of the biggest stETH leveraged positions in the ecosystem, Instadapp team was advised by LidoDAO to reduce the size of Lite positions due to the increased risk from stETH depegging and market illiquidity. Since this would essentially increase the chances of cascading effects and liquidating many other leveraged positions widening the peg furthermore, the Instadapp team and investors took actions of topping up collateral to the Lite vaults (~$35m) to protect the vaults from a stETH price depeg.

To further protect the Lite vaults, the Lite vaults were unwinded by 25,000 ETH, which resulted in ~500 ETH loss due to the price variance between stETH and ETH at that time.

Instadapp team made sure that Lite depositors’ losses were fully covered by paying back from their Treasury.

Lite Vaults v2

In Feb 2023 the v2 of Lite Vaults was released, introducing a single ETH vault. It leverages the new ERc-4626 standard, opening the door for the composability of iETH tokens across DeFi protocols, although, as of now there were no known (to BA Labs) lending venues offering iETH as collateral.
Lite v2 also brought the option to distribute stETH/ETH positions across multiple underlying protocols (v1 utilized Aave v2 only) as an act of risk diversification and targeting highest net APY. At the time of writing, supported protocols are: Aave v3, SparkLend, Compound v3, Morpho Blue, and Fluid.

Risk considerations of pool-based leveraged vaults in v2

Counterparty risk

  • Mgmt fee - The ETH vault charges 20% on profits, and a 0.05% exit fee. All fees earned by Instadapp Lite go to the InstadappDAO.

Lite v2 vaults are upgradable by the DAO governance which can adjust fees, add/remove underlying strategies/protocols, assign (permissioned) roles (deposit, withdraw, leverage, and refinance (reallocate between protocols)) to “Rebalancer” accounts, etc.

It’s important to note that team multi-sig can be used to reduce ratios for a particular protocol and can ‘pause’ the vault by either pausing withdrawals or rebalancing. These functions do not change the ownership of the assets and do not prevent users from trading or transferring the underlying iToken.

Withdrawal risk

Lite vaults essentially have limited withdrawals, although v2 introduced 50% of capital to be withdrawn instantly without incurring any loss, thanks to the multiple underlying lending (yield) protocols.

Depeg/Unwinding risk

Here we note that with the launch of Lite v2, the stETH depeg risk has been significantly reduced with the combination of Shanghai upgrade in April 2023 (enabling staked ETH withdrawals), and the Instadapp’s stETH redemption protocol introduction.

The stETH Redemption Protocol significantly reduces the costs associated with deleveraging stETH/ETH (Lite v2) positions. By taking short-term loans against ETH in Fluid, protocols/users can unwind their leveraged positions at a fraction of the cost compared to traditional selling on the market.

The stETH Redemption Protocol serves as a specialized mechanism for efficiently deleveraging stETH/ETH positions by utilizing Lido’s receipt NFT as collateral during the redemption process. Once the ETH is withdrawn from Lido, the position is deleveraged using the newly released ETH to pay back the ETH debt (deleveraging the entire position) and sending the remaining ETH back to the user. This eliminates the need to maintain a 25% reserve for Lido redemptions, allowing for those 25% of the vault’s assets to be actively engaged in its strategic operations, potentially enhancing returns for users by 30-50%.

Rebalancing (gas) costs - also became negligible according to the e.g. 6m or 1y gas price data (thanks to the Dencun upgrade). This reduction in gas costs makes rebalancing operations much more affordable than in the past. The “Rebalancer” role, which is an on-chain account responsible for executing rebalancing transactions, is crucial for the vault’s health. While the exact incentives for “Rebalancers” are not explicitly documented, we still need to consider whether the operations are sustainable. Gathering historical data, we found that the total gas cost in the last two months was around 0.78 ETH, paid by the rebalancer 0x10F37Ceb965B477bA09d23FF725E0a0f1cdb83a5. The total value generated by the vault during the same two months can be estimated at around 596 ETH, taking into account the current APY of 5.56%, which means gas costs were approximately 0.13% of the yield generated in the same period. Thus, the overall APY of the vault would need to be slightly adjusted down by approximately 0.0073 percentage points, e.g. from 5.56% to 5.5527%, if the gas costs were passed on to users, which shows that it is completely sustainable at the current TVL.

4 Likes

Very good addition to the high risk strategies!

2 Likes

That seems to be a typo, though… This is going into the Higher Risk strategy

2 Likes

edited now! thanks for pointing it out :slight_smile:

The proposed Fluid Lite Ark for the Higher Risk Fleet Mainnet ETH has been evaluated. Since this product is based on looping positions across multiple lending protocols, we’ve introduced the following adjustments to our risk model:

  1. Liquidity risk and collateral volatility weights have been increased, due to the leverage involved. At the time of writing, the strategy shows a leverage of approximately 7x. This amplifies the impact of any sudden price changes, high volatility, or liquidity shortfalls during unwinding.

  2. Due to the 5% withdrawal fee, we propose adjusting the maxInflow formula based on the fleet’s cooldown period. The goal is to protect the fleet from losing up to 95% of its cap solely to fees. The table below summarizes the recommended maxInflowFleet (i.e., the inflow cap computed using this logic) for different cooldown settings and assumed reaction times to fix a malfunctioning keeper:

maxInflowFleet (% of fleet cap) cooldown period Reaction time
2.7 10 min 15 days
5.8 10 min 7 days
16.2 60 min 15 days

The final maxInflow will then be computed as: min(maxInflowFleet, maxOutflow). Note that currently, all Arks across all fleets have maxInflowFleet set to 20%.

To optimize fleet operations, we will start by increasing the cooldown period to 1 hour and assuming a 15-day reaction time, resulting in a maxInflowFleet value of 16.2% of the fleet cap.

  1. We apply an additional penalty to the available liquidity in Fluid Lite, scaled to the strategy’s observed leverage, to reflect the increased risk exposure, particularly if unwinding at fair value becomes problematic.

  2. If any protocol used by the Fluid Lite vault results in a negative net APY, we will fully subtract its allocated TVL from the vault’s available liquidity, so that the risk model automatically reduces both maxArkCap and max %TVL.

Finally, we have considered the instant-withdrawable liquidity from the Fluid Lite vault to add extra constraints to the maxCap of the ark. Following the previous considerations, we propose the following initial parameters:

Ark Symbol/Vault maxCap Max. %TVL maxInflow maxOutflow
fluidLite WETH 16,400 35.0% 2,430 5,300

The parameters above represent the final caps after three weeks of progressively increasing them. This progressive increase is due to new implementations at the smart contract level from the summer.fi side, and will serve as a slow roll up of the full caps after verification that the ark is working as intended. Thus, the maxCap and Max %TVL will be planned to be executed as (it may differ from actual execution if any issue arises):

We will monitor the Ark and adjust its parameters as market conditions evolve, including revisiting the fleet cooldown based on the discussion outlined here.

3 Likes