Aave - Initial Treasury Strategy

Authors from Llama: @MatthewGraham, @Dydymoon and @AcceleratedCapital.


Llama would like to present the Aave community an opportunity to deploy the Reserve Factor (RF) to earn additional yield by acquiring BAL rewards on Balancer V2. The balance of the RF and Ecosystem Reserve (ER) are shown below:

Wallet USD Notes
Ecosystem Reserve $821M 100% AAVE
Reserve Factor V1 $0.42M $224K aDAI, $176K aUSDC, $340K ETH
Reserve Factor V2 $13.88M $7.44M aUSDC, $2.4M aUSDT, $2.3M aDAI, $0.15M awETH
Reserve Factor Polygon $5.06M $1.2M amUSDC, $1.2M amDAI, $0.6M amUSDT, $0.5M amWETH, $0.76M amWBTC

This strategy has two goals:

  • Earn additional yield with the aTokens from the treasury
  • Accumulate governance influence with protocols that interact with Aave

Llama proposes depositing a portion of the aTokens held in the RF into Balancer V2 pools to earn BAL rewards. The aTokens are wrapped, creating a waToken, and the yield generated from holding the aToken accumulates within the wrapper. Aave is still earning yield from the aToken as if it was held within the RF wallet, but will also accrue trading fees and BAL rewards by providing liquidity on Balancer V2.

An alternative strategy is to use AAVE from the ER as collateral in order to borrow stables to fund the yield generating strategy. The cost of capital is reduced by the borrowing incentives on V2, the fee income that flows to the RF, and the interest received by the aTokens. Provided the loan is sufficiently overcollateralized and managed, the risk to the Aave treasury is largely derived from Balancer V2 Smart Contract risk. This risk seems very low, at the time of writing Balancer V2 has more than $1.6b TVL (mainnet and polygon combined) and offers the largest, unclaimed bug bounty (1000 ETH) in the industry. The investment generates a positive ROI due to the low cost of capital, Balancer trading fees, and BAL incentivises.

Further to the above, there exists the potential for the Balancer V2 liquidity provider (LP) token to be integrated into Aave’s AMM platform. This creates a further use case and presents an opportunity for Aave to provide liquidity to the underlying pools.

Utilising Reserve Factor to Earn Yield

The general strategy is to deploy DAI, USDC, USDT equivalents held within the RF into Balancer V2 pools to earn yield from Aave V2, trading fees and BAL rewards. ETH and BTC equivalent assets are to be deposited into Balancer V2 to earn trading fee income and BAL rewards. The BAL rewards are to be accumulated over time and held as a strategic asset. Other governance tokes, like LDO are to be sold for ETH and reinvested into the ETH strategy.

Stable Coins (DAI, USDC, USDT)

Llama has been discussing with Balancer a yield strategy that involves depositing Aave’s aDAI, aUSD, and aUSDT holdings into the next iteration of the metastable pool on V2. The staBAL3 pool deposits around 90% of the stables back into Aave V2 making them productive whilst also receiving additional BAL incentives.

In addition to the yield from Aave accruing for liquidity providers, Balancer is expected to offer incentives. The staBAL3 metastable pool will boost productivity as the yield from the 10% of stables not deposited in Aave is more than offset by the entire stable balance earning BAL rewards plus trading fee income. The stables are now earning yield from Aave and BAL rewards, this is expected to roughly double the overall yield generated by these assets.

For example, the Balancer V2 DAI / USDC / USDT pool is $163M in size and generates 12.21% APR, comprising 0.03% in swap fees and 12.18% in BAL rewards. This pool still doesn’t lend its underlying stable coins to AAVE, which will be the case with the new version of the staBAL3 pool.

We expect the pool to be launched in the coming months, potentially early Q4, 2021. When the full details are released, we will be able to refine the strategy further and present an ARC to the community. At this point in time we are considering depositing around half of Aave V2’s DAI, USDC and USDT equivalent into the pool which is a bit over $6M. The final amount will be defined in the ARC and will depend on feedback received from this post.

The overview for deploying the stable coins is summarised below:

  • Deposit half of Aave V2 RF’s aDAI, aUSDC, and aUSDT holdings on Balancer V2 to earn BAL rewards
  • Expected APR is difficult to predict, but is anticipated to be greater than 10% in addition to the aToken yield
  • BAL rewards are to be held as a strategic asset and not sold

Acquiring the BAL tokens whilst only forgoing a fraction of the yield otherwise earned is a near-zero cost way of acquiring a Balancer holding. Given the relationship between Balancer and Aave, the BAL holding becomes a strategic asset that further aligns the two communities through ownership.

Ethereum (ETH) and Bitcoin (BTC)

For the ETH and BTC allocation, we’ve identified two potential strategies to generate additional yield.

The first strategy is to take advantage of the recently launched metastable pools from Balancer. The first pool Balancer has launched is wstETH / wETH. The stETH yield accumulates within the wstETH token. Currently, the pool is yielding 18.73% APR, comprising 0.46% swap fee, 3.07% stETH staking rewards, and 15.20% from liquidity mining (2.10% BAL and 13.1% LDO).

The alternative strategy is to deposit ETH into the Curve stETH / ETH pool and earn CRV and LDO rewards. The yield is currently around 6.7%, comprising 6.36% LDO and 0.27%-0.68% CRV.

When comparing these strategies, the rewards on Balancer are greater than Curve, with both Lido and Balancer offering relatively higher rewards than available on Curve. Due to this, Llama’s preference is to earn BAL rewards and concentrate on acquiring Balancer protocol governance influence.

ETH is typically deposited into Aave as collateral and the utilisation rate is less than 5% on V2, so the the yield generated from holding aETH is low relative to other productive use cases.

  • Aave currently holds the equivalent of 391 ETH across the RFs and have a current market value of $1.3M (0.11% of the total treasury; comprising 156 amwETH, 44 awETH and 101 ETH)
  • Deposit 90% of the ETH into the wstETH / wETH pool on Balancer V2
  • Retain 10% for gas costs
  • BAL rewards are to be acquired as a strategic asset and not sold

Similar to the ETH strategy, Balancer offers 7.15% APR on the stable pool wBTC / renBTC / sBTC of which 6.98% comes from BAL incentives. By comparison, Aave offers less than 1% on BTC assets.

  • Aave holds the equivalent of 18.18 aBTC across the RFs with a current market value of $850K (0.1% of the total treasury; comprising 16.25maBTC, 1.93 aBTC)
  • Deposit all of the BTC-equivalent holdings into the Balancer pool to earn BAL rewards
  • BAL rewards are to be acquired as a strategic asset and not sold


This forum post introduces the concept of deploying the Reserve Factor to earn additional yield through BAL rewards by providing liquidity on Balancer V2. There will be a Snapshot vote to gauge community sentiment towards this proposal. If the sentiment and feedback in the comments below is favourable, then this strategy will be developed further and presented to the community at a later date when the details around the pools are known.


I really like the proposal and just wanted to say that the Balancer community will really appreciate this being implemented as well!


I really like the proposal. Although BAL is for sure a defi blue chip and an amazing asset, the community in the past has already showed interest to accumulate more CRV through various allocations of the treasury (there is also a snapshot vote created by @dydmoon here Snapshot) which i fully agree. I would personally consider shifting the strategy towards a more consistent accrual of CRV (50% BAL - 50% CRV)?

Side note, there are another 800K in the V1 token burner address that can be moved to the V2 tresury Aave: Token Burner Address | 0xe3d9988f676457123c5fd01297605efdd0cba1ae


I really like what you are doing. Please keep up the progress.


Authors: @elisafly, @eek637


We enjoyed reading both this post and the Asset Management Guidelines Llama has proposed. We largely agree with the approach to diversifying risk and thinking carefully about the treasury’s long term goals. However, we believe that given the rapidly-changing nature of the space, the treasury would be better served by more active management. We’ll touch on that separately in the Asset Management Guidelines comments section.

Here, we’ll outline some specific strategies we believe would be beneficial to the diversification of the treasury, focusing mainly on the stable coins from the Reserve Factor, which we believe should be implemented first. We also touch on the large AAVE balance in the ecosystem fund.

Stable Coins (DAI, USDC, USDT)

As a reminder, the Reserve Factor represents funds collected as a share of the protocol’s interest and is used to pay protocol contributors. In other words, it will be distributed in regular and predictable amounts at a regular and predictable cadence, and grow linearly with the protocol’s utilization. Its balance has no impact on the safety of the protocol, and staking rewards are paid out of the Ecosystem Reserve (which is 50-75 times larger than the Reserve Factor). These factors combine to make it an appealing first step in decentralizing the Aave treasury’s asset management program.

We like the idea of allocating 50% of the reserves to the Balancer V2 DAI / USDC / USDT pool (later migrating to staBAL3 pool when it becomes available towards the end of the year). The returns promise to be attractive, and we agree that building a strategic stake in BAL is a worthwhile endeavor.

We would note, though, that even then - diversifying across strategies would be prudent. There’s another immediately-available opportunity to run a similar strategy in another protocol where building a strategic stake of tokens also makes sense. Indeed, Curve offers a pool with a similar APR that could perhaps be boosted further over time by increasing AAVE’s CRV holdings. Whilst the Curve governance forum did not agree to a token swap in recent forum discussions there is no reason why we could not just DCA these in the market and leverage some influence over the gauge rewards as and when that becomes interesting.

With that in mind, we propose that in addition to the OP’s Balancer proposal**,** that we allocate the remaining 50% of the Reserve Factor’s stable coins to the Curve AAVE pool and running a yield farming strategy on that position. The yield on this pool may be additionally be enhanced through accumulating a position in CRV.


Borrowing stable coins against the AAVE position and deploying them in to various yield-bearing strategies is a good first step. However, given the sheer size of the position, we do not believe it should be the only strategy the treasury employs. We have spent time researching the utilization of Uniswap v3’s liquidity pools to manage a highly concentrated position. By concentrating liquidity provision effectively on different parts of the price curve, in addition to earning LP fees, we can strategically lighten up on AAVE after big rallies whilst buying more on big sell offs. We are happy to present and model this further if of interest.

Ethereum (ETH) and Bitcoin (BTC)

We agree that leveraging Balancer for the ETH/BTC strategy makes sense given the superior yields it earns. It is also worth keeping an eye on the Curve tricrypto2 pool which is yielding 11% (base) to 25% (max boost) in rewards, especially if it is decided to purchase CRV as per above suggestion.

Who We Are

Our team is Avantgarde Finance. We have been building in the DeFi space for five years and share more than 40 years of cumulative experience in TradFi. More recently, we have broadened our focus to providing Treasury Management Services. This can be anything from developing custom integrations and reporting tools required by DAOs to managing the assets on behalf of DAOs on-chain and implementing strategies.


Hey Emilio, unfortunately the Curve Team denied the proposal about a treasury swap, and buying more than 2.5M$ worth of CRV was not a good option, so we looked at other possibilities and BAL is one of the most interesting.

However, the treasury is not only growing on mainnet, but also on Polygon and Avalanche, which could open new possibilities and strategies if we consider having different strategies on each chain where Aave is deployed.

Because the most important is to follow the community choice on this point, @llama published a snapshot vote that present the benefits and the drawbacks of both options: Focus on mainnet or work on deploy strategies on different networks too (the snapshot vote will be live at 4PM CET today):