After three months of interuption due to hectic times, let’s pause here to look back over all this chaos and its impact over birb markets.
A lot happened over the past few months. The start of the summer looking well, Stake DAO locked ~25m CRV in a few weeks, birb output was going up, good yield were coming from everywhere, and then, a bunch of hacker decided that this had to stop. The Curve hack (coming from a Vyper vulnerability) started a spiral of fear, uncertainty and doubt around Curve and its founder’s credit quality. Weeks of dramas, OTC, and scandals started. Is this the beginning of the famous death spiral eagerly expected by Curve’s professional fudders? I guess we’ll see…
When the dust settles down, what’s left? CRV lost c.50%, Curve’s TVL lost 30% ($1bn), steady volumes and fees in line with historical levels (better efficiency of the TVL?), crvUSD supply still above 100m, and an exciting pipeline being rolled out steadily as if nothing ever happened.
The broader market also suffered a massive haircut. BAL lost 30% notably. Combined with CRV drop and CRV inflation reduction (daily CRV issuances reduced by 16% in August), the total birb market is under significant pressure, having lost approximately 40% in total output in the past 3 months, but the bigger losers are not the ones you would think…
Curve
Surprisingly, the Curve birb market held very well compared to CRV price action. This is actually a very good sign of user (and team) confidence and trust in the product, despite everything that happened. The total output went from $3.3m in round 45 (June) to $2.1m in round 52 (first week of voting on September 7, 2023). It lost 36%, which corresponds to 25% if we adjust for the inflation reduction that happened in August. As we see, birbs outperformed CRV price action (c.-50%).
Did this lead to a better return for veCRV holders? No, yield stayed consistent (around 23% APR from vote incentives) as in this period of “uncertainty”, smart money took the opportunity to accumulate more CRV and lock it massively (which is yet another bullish factor). Total veCRV supply rose from 579m to 613m between during Q3.
From a voter perspective, the voting APR has been pretty similar between vlCVX and veCRV, but this is probably going to change in the future since this proposal just passed. We could expect an increase in yield to vlCVX since there will now be a point in staking CVX. The overall revenue for Convex doesn’t change, but there should be a yield premium when locking compared to staking, which should skew the APR in favor of vlCVX.
From a protocol perspective, it’s still generally more profitable to address vote incentives to veCRV holders than to vlCVX holders. It has been the case 80% of the time since June 2022 when we started tracking the data. The discount however tends to reduce, which highlights a more efficient market.
Consistently, birb ROI have also decreased historically, but due to CRV’s free fall of the past few weeks, we are reaching dangerously low values. The past two rounds have been the first one in Curve’s history where birb ROI via vlCVX entered in negative territory.
This unprecedented amount of unprofitable vote bounties is over represented on vlCVX, but also represented in veCRV birbs. the question that remains: how will it evolve going forward, and will bribers start to incentivise liquidity directly again?
Balancer
The veBAL vote market is undergoing massive changes, which creates some bumps on the birb road.
First, Hidden Hand shipped its v2, which tried to achieve more flexibility and more protection for bribers against gamers, by replicating some of the features developped by Votemarket (long term programs, blacklist), but it created a bit of confusion which led to a serious reduction in bribes deposited on the platform.
Furthermore, Aura facing the structural (and announced) decrease in their token price (-75% in the last six months), even though having adjusted the Convex tokenomics to bring more value to AURA (25% performance fee on LPs, only 50% of the supply dedicated to Balancer), decided that it was time to ajust those tokenomics to bring even more value to AURA. AIP 42 which passed recently, mainly says that 1) inflation to LPs is reduced by 60%, but 2) a new inflation stream of 180,000 AURA per cycle directed to pools proportionately to their vlAURA votes. This is approximately neutral in inflation terms (-33,000 token issued per week at the beginning, decreasing progressively as inflation to LP reduces). However, what it really changes, is the value it brings to vlAURA. It increases the efficiency of vlAURA vote incentives by a flat amount (that adjusts between cycles depending on Aura’s veBAL votes) of $0.026/vote.week, which will therefore lead to more vote incentives directed towards vlAURA, so more yield for locking AURA and more value to the token, so price 1000x right? This is probably a very good move from Aura, and I think it will give them a decent advantage, but there is one thing which I think they didn’t take into account, and which could backfire. This move is effectively decreasing veBAL’s value (vote incentives directed to veBAL are being less competitive), while not increasing auraBAL’s value in any way. So the total yield going to locked BAL/ETH LP is going down, and just like AURA should benefit if yield goes up, BAL should suffer if yield goes down. But wait, isn’t AURA backed by its locked BAL/ETH LP? Yes, so a reduced BAL price will mean a negative impact on AURA. So in financial terms, if market is efficient (which it tends to be in the birb ecosystem), the appreciation of AURA should be partially offset by BAL decrease. Who will suffer from that? Liquidity providers who will see their BAL denominated yield go down, and BAL and veBAL holders.
“But wait, why would AURA pay for Balancer’s yield?”, one could say. True, but let’s not forget it goes both ways: the biggest vote incentiviser is actually Balancer themselves, and they currently direct 40% of their vote incentives towards vlAURA, even if they are less efficient than veBAL ones! For last round, despite AIP 42, Birb ROI was still higher via veBAL (98%) than via vlAURA (61%). This is actually free money given by Balancer to subsidise AURA. How long until Balancer whales figure out that the deal is not fair anymore and would decide to allocate all incentives to veBAL? Probably not long: veBAL supply has already started to shrink as Balancer whales stop relocking, pissed off by this change in the rules. I guess the future will tell.
After this rather long update, let’s dive into the figures. Historical birb output decreased significantly over the past six months: it lost 66%, which is explained by the 50% drop of BAL and the 75% drop of AURA. Nothing very special to see here. However, we can say that despite the bumps on the road this summer, the market stayed very stable, which is a good sign to be sent to veBAL lockers.
As expected, it is now skewed very much in favor of vlAURA, with 63% of birbs deposited for 39% of veBAL supply.
This over representation of AURA leads to a higher price per veBAL vote via AURA compared to via direct veBAL birbs. This has been a general trend though, mostly due to Balancer’s vote incentives, and to the higher consistency of vlAURA birbs compared to veBAL ones. On voting round ending on Sep-7, one veBAL vote for two weeks costed $0.0557 via direct incentives, compared to $0.1006 through Aura (nearly twice as expensive!)
As a consequence, the ROI of veBAL incentives was significantly higher than the ROI of vlAURA incentives. This is still true with the new AURA incentive distribution method: it brings the ROI of vlAURA birbs from 10% to 61%, which is a massive boost, but they are still on average lower than the 99% ROI of veBAL birbs.
In terms of incentives efficiency it is a bit all over the place, except for Aura, which benefits from Hidden Hand optimisation giving strong confidence to bribers that their program is going to be filled. It’s also important to take into account that some whales (such as Humpy for example) vote for incentivised pools, but benefit also from the farmed rewards, which skews the analysis.
Frax
Frax’s birb market has been following its path nicely and quietely over the summer. There was a peak in total incentives in August coming mainly from unclaimed IQ incentives.
In terms of votes, Temple (red) and Spiral (dark purple) have been the main winners of this late summer.
Finally, the ROI has been fairly stable, always above 150% through the winter. It is currently sitting around 200%.
