And just like that, MegaETH’s @aave deployment enters an unsubsidized, sustainable era, only a calendar quarter after TGE. We see today a completely borrower-funded lending APY on par with Ethereum Aave. Rewards distributed by Merkl had been tapering for some time, as the deployment ended up with excess liquidity compared to demand, and with its major collateral, which could not support Aave’s targeted 90% utilization. The tapering of those rewards finally hit the elastic point in some large LPers’ demand to lend USDM, resulting in a more rational size. Simultaneously, the introduction of stcUSD from @CapApp as collateral meant there was finally a yield bearing collateral able to sustain borrowing at the traditional Aave utilization target (borrow ~4%), which boosted rates as well. There are a couple lessons I think we should take away from this. 1) In a post-Kelp world, it’s a long process for Aave to onboard new collateral assets. I personally think they need to find ways to streamline this, because much of Morpho’s success has been the ceding of vast swaths of the lending market to them voluntarily. This is good and bad for them - they have kept their nose clean about onboarding *financially weak* assets, unlike the independent curators. But it also leads them to existential risk-level concentration for rail risks, as we saw with Kelp. An Aave with 50 collaterals that builds in an expectation of some losses as part of the business is stronger than an Aave with 10 collaterals and needs to seek external financing in my opinion. 2) This is a low-yield environment, and even many moderate-risk assets simply can’t support borrowing even below the risk-free rate. (s)USDe is an excellent example. You have what is a multi-strategy, actively managed credit fund, and it can only pay a few bps premium over a 4-week tbill? Even if you are willing to sit with that risk-reward on the belief the team will bring you better days in the future, it’s just not an asset you can borrow against at any reasonable rate. Even assets like syrupUSDC/T and stcUSD only get you to a modest rate in lending markets. 3) On rewards: MegaETH Aave rewards worked fairly rationally, but not perfectly so. Initially begun in a world where Aave could/would onboard multiple collaterals and e-modes, it was rational for a new deployment to err on the side of oversupply of stablecoin inventory, since no supply means no lending. (s)USDe also had higher yield 3 months ago, and a softening of the returns from the workhorse collateral on the deployment made the slow speed of post-Kelp asset listing even more painful. 4) Collateral uniqueness. For any market not named Ethereum, Aave really needs more differentiation. stcUSD is only listed in MegaETH Aave, so there is no other venue. But when you look at the most recent deployment, on Monad, you only see MM USD as a novel asset, which is not yield bearing. You can see the ossification of Aave risk tolerance in real time, as Monad launched with only familiar assets otherwise. That those familiar assets listed even in the face of literally zero liquidity is an indicator that Aave risk tolerance is very low, and makes the future of non-Ethereum deployments a question mark. If those deployments only offer leverage against assets all competitors take, and any given deployment is unlikely to have an asset different from the mainnet Aave, what is the competitive advantage? Add in that the typical interest rate curve on stablecoins only gets lenders to the risk-free rate at 90% utilization, and there is no room for a risk premium, except in the form of rewards. And all rewards have to be planned with their sunset in mind. But mostly? Low rate environments are just really challenging for everyone until DeFi discovers a way to lend to someone for purposes other than leveraged crypto exposure (whether asset price or asset yield)