0 – As per popularly requested, here’s my preliminary thoughts on the $LEND / $AAVE 2.0 tokeneconomics restructuring, with the #’s based on @The3D_ ‘s proposal at governance.aave.com/t/initial-disc…
1 – In the new token model, $AAVE effectively becomes the “equity tranche” of a co-operative bank, enjoying net interest margin of balance sheet & any add’l fees + taking 1st loss upon default, whereby depositors / capital-providers get to enjoy part of the $AAVE equity issuance.
2 – In this framework, assuming 50% of $AAVE is staked (6.5 mm), then this “bank’s initial book value tops ~200 mm USD, with today ~500 mm as deposits on liability side of the balance sheet + ~100 mm on the asset side as “yield-generating”: aavewatch.now.sh
3 - …Notably of the 500 mm deposited today, ~150 mm is net $LEND anyways, so upon the shift it’d really be a 150-200 mm equity tranche with 350 mm of deposits + 100 mm of earning assets. ”Ultra-safe” compared to typical banks where it can be levered 10-20:1 asset: equity.
4 - A bit more of banking 101 – typical net interest margin for US banks (asset yield – liability cost) = ~3% with typical return on asset of 1%. Levered 10-20x on equity you get 10-20% return on equity. For $AAVE today it seems the NIM spread on borrowed = 2%...
5 - …whereby 2% * 100 paid 90% to equity tranche of staked 150-200 mm = ~1% return on equity on fees alone. This is obviously quite low – and the way to juice it (without $AAVE inflation) would be via (a) more people borrowing and (b) more asset-lite services that generates fees
6 - …so if say the depositor base tops 5 Bn (10x from here) and borrowing utilization goes to 50% @ 2.5 Bn (25x from here). Even if NIM margin compresses to 1-1.5% (similar to compound), this would mean a 10-20x asset:equity ratio and 30/200 = ~15% ROE for $AAVE stakers.
7 – back to the v2.0 design, as per the proposal, AAVE’s inflation in year-1 would be ~4%, compared to that of ~33% of Compound. The 2 trade at similar floated market cap today, while COMP’s daily “inflationary mining” reward to ecosystem would be ~9x higher. 
8 - …however, $COMP owners today don’t get any inflation emission, while staked $AAVE owners would get ~50% of the inflation. OTOH, compound borrowers & lenders split the inflation 50/50, and it would look like $AAVE lenders would get to enjoy all of the remaining inflation.
9 - …so on a like-for-like basis, assuming any user put in $10k into token itself, lending, or borrowing, $COMP farmers today would get a ~4% inflation if lending, 7% if borrowing; while $AAVE lender would only get ~1.6% on lending (60% lower than $COMP lenders)
10 – One could certainly imagine the current proposal of $AAVE not being enticing enough for share-gain against $COMP, it’s hard beating 33% annualized inflation! But there’s one confounding factor – likelihood of $AAVE being dumped upon emission is MUCH lower than $COMP.
11 – Compared to $COMP that accrues no ecosystem benefit at the meantime, $AAVE staked token acting as the equity tranche has a return profile of [90% fee + 50% inflation + x% BAL farming – tail-risk of ecosystem liquidation].
12 – … with fee = ~1% and inflation = ~3.5-4% (assuming 50% staked). The very interesting thing is that the 80/20 $AAVE / $ETH pool on BAL (likely topping 200-250 mm TVL) would likely make it the largest there, thereby soaking up a significant amount of $BAL emission.
13 – Assuming this $AAVE / ETH pool is ~40% of TVL and $AAVE staker itself makes 75% of the $BAL emission, this stream of returns could contribute another ~12.5% return to $AAVE stakers, bringing the overall $AAVE staking return to 15-20% (assuming no blowups).
14 – In other words, the return for staking $AAVE at low asset:equity today is drastically better than holding $COMP, and such incentives could mean all $AAVE will be hoarded and staked (very similar to $SNX) instead of dumped like the case of $COMP.
15 – The ETH pool within the $AAVE/ETH Balance pool could be a good farming opportunity for $YFI yVault, considering part of $AAVE fees +$BAL inflation can go towards it. Overall, this is a very virtuous cycle that accrues value to $AAVE holder while minimizing dump pressure.
16 – I would have hoped fore more aggressive ponzinomics with a higher inflation rate of $AAVE (larger total supply) whereby not only $AAVE staker gets bigger yield, but borrowing / lending reward on per $ deployed is higher than $COMP as well…
17 - …but I concede that such a model of aggressive market share gain when pie is still small doesn’t seem healthy for longevity. What I could foresee is $AAVE continuously grinding up in value thanks to value-accrual S/D mechanics and more innovation on products front.
18 – stepping back to the bank analogy, how banks juice their returns are 1 of 3 ways – either they compress cost via scale (JPM), find plebs to expand ROA (low deposit rate + large cross-sell, WFC), or go big on asset-lite products (various type of referral & service fees)
19 – In $AAVE’s case, it’d be a combo of more asset-lite services (flashloan + P2P lending utilizing $AAVE platform a great start), designing more products that utilize $AAVE borrowing (ex: BTC or 32 ETH mortgage taking out a fixed loan, repackage into CLO, etc)…
20 – …or expanding touchpoints like Chase in NYC into all types of digital-native financing scenarios (work w/ @argent, @Fold, and other B2C apps). A lot of wood to chop still for @StaniKulechov, @hiFramework, @zhusu , and @paraficapital , but exciting roadmap ahead.
21 - All the best to the team exploring the new frontier. Fair disclosure – my math on $BAL could be wrong given complexity and there’s a ton of risk in smart contracts that the VC must help the team on. Feedback very welcome if there’s any mistakes!
