Key Insights
- Yearn’s soon-to-be-released veToken (vote-escrow) model uses vote-locking and vault gauges to distribute protocol earnings (bought back YFI).
- Vote-locking incentivizes YFI holders to be aligned with the long-term health of Yearn by increasing their governance power and share of YFI ownership based on the size and length of their commitment.
- Yearn’s veToken implementation is unique: Yearn is a yield aggregator (rather than an AMM-DEX), offering users a weighty 10x maximum boost and gauge/vote-lock rewards based on protocol earnings (via bought back YFI) rather than emissions.
- Gauge boosts incentivize Yearn’s largest users to also become long-term YFI holders, while the opportunity cost around the 10x maximum boost is likely to spawn an influential ve-aggregator/bribing ecosystem.
A central storyline in Yearn’s fabled beginnings is the “fair launch” of its native governance token YFI. In mid-2020, lead developer Andre Cronje distributed the entire YFI supply through an open process (liquidity mining), relinquishing control of the project to an engaged and productive community. The apotheosis came about two months later, with YFI trading at over $40,000.
Despite YFI’s gangbuster performance out of the gate and Yearn’s strong fundamentals and enthusiastic community, YFI’s 2021 performance left much to be desired. Not only did the token spend the year sliding against crypto more broadly, from a little over 30 ETH/YFI to roughly 5 ETH/YFI, it also lagged most of the other (relatively) floundering “DeFi 1.0” majors (i.e., UNI, COMP, SNX, etc.).
At the tail end of 2021, to bolster the link between Yearn’s performance and YFI, Yearn tokenholders passed YIP-65: Evolving YFI Tokenomics. The crux of its soon-to-be-released veToken implementation is familiar: to direct governance rights to active, long-term holders. However, the introduction of novel parameters, such as a weighty 10x maximum boost and variable yield (protocol earnings) distributed using bought back YFI, takes us into uncharted territory.
Primer on Yearn
Yearn is the largest DeFi yield aggregator, with about $500 million in assets under management (AUM).
Yearn’s bread and butter is its yVaults, which run dynamic, community-developed yield-earning strategies across DeFi.
While the protocol’s early growth came from individual users, the launch of the Yearn Partnership Program in 2021 made other protocols a sizable segment of its customer base.
YFI Background and Overview
In mid-2020, YFI was released through a week-long liquidity mining program under a “fair launch.” Yearn’s fair launch meant the total initial supply, 30,000 YFI, was distributed to liquidity providers, with no pre-mine for insiders (Yearn developers, early backers, etc.) or the Yearn treasury.
Since then, Yearn has been controlled by YFI holders, who can submit and vote on off-chain proposals, called Yearn Improvement Proposals (YIPs), via Snapshot. In addition to governing the platform and its token, YFI holders have a claim on protocol earnings.
In early-2021, a number of lasting changes were made to Yearn’s treasury and token.
- A new fee structure for yVaults (per YIP-51).
- Yearn’s treasury started to collect a 2% management fee and a 10% performance fee.
- A switch in the way protocol earnings are distributed, from staking rewards to buybacks (YIP-56, BABY: Buyback and Build Yearn).
- A 22% expansion of the initial YFI supply to 36,666 YFI (YIP-57).
- Out of the 6,666 new YFI, a third (2,222 YFI) were given to key contributors through vested retention packages and the other two-thirds (4,444 YFI) were allocated to Yearn’s treasury.
These changes have allowed Yearn to pad its treasury and implement a new way to capture revenue and distribute earnings to YFI holders.
BABY: Buyback and Build Yearn
Under Buyback and Build Yearn (BABY), instead of distributing protocol earnings as staking rewards, Yearn began to buy back YFI and hold it in its treasury. The impetus for this change was to bolster the less than one-year-old protocol’s treasury for future initiatives, while maintaining the link between YFI value and Yearn earnings.
However, a few weeks after BABY (YIP-56) was approved, the yvDAI vault suffered an exploit. It took Yearn about 9 months (until December 2021) to build back its treasury (a $30 million stablecoin buffer) and really start to direct earnings to buybacks. So far, Yearn has directed $22.4 million to buybacks, adding 1,111 YFI (~3% of circulating supply) to its treasury.
While the point is now moot, it is interesting to note that buybacks only really began in size around the same time the community started discussing ways to improve YFI tokenomics. Unfortunately other, major confounding factors, such as market headwinds and the excitement around the new veYFI model, make it difficult to tease out BABY’s impact on YFI price.
Evolving YFI Tokenomics
The tweet heard around the Yearn (community)…
Despite the status of BABY at the time, this tweet sparked off an involved and methodical ideation process for a YFI tokenomics revamp. Roughly two months later, a group of Yearn community members released YIP-65: Evolving YFI Tokenomics. The proposal outlines a sequential, four-step plan that funnels voting power and bought-back YFI to active, long-term YFI holders. The proposal passed in late-December 2021 with over 99% of votes cast in favor.
The proposal outlines a four step plan:
1. Staking vault (xYFI): Distribute bought back YFI to vault stakers.
2. Vote-locked YFI (veYFI): Curve-style vote locking. Voting power and rewards depend on lock duration (up to 4 years).
3. Vault gauges and voting: YFI rewards are distributed through gauges, where yVault depositors stake their yVault tokens (yTokens). Gauges receive YFI to emit based on bi-weekly veYFI holder votes, and YFI lockers earn rewards based on their veYFI weight and deposit balances.
4. “Useful work” features: Expand the scope of veYFI, in terms of protocol control/rewards.
Since Yearn is open-source, the state of development has been public. In effect, this allowed Yearn developers to skip the first two steps of the plan — jumping straight to step three, vault gauges and voting. Since each step builds on the last, the release will still include components of steps one and two. In fact, the initial release will start off with just veYFI, and move to vault gauges and voting shortly thereafter.
How It Works
Zooming out, Yearn’s new tokenomics model is introducing two concepts: vote-locking and vault gauges. Any YFI holder can lock their YFI for veYFI, representing governance power and their claim on protocol earnings. However, with the introduction of vault gauges, veYFI lockers with yVault deposits are the ones really in a position to use the new system in its entirety and maximize rewards.
veYFI Overview
veYFI’s mechanics are quite similar to those of other veToken models: holders can lock their YFI into non-transferable veYFI. The lock can be for a minimum of one week and a maximum of four years. veYFI is weighted linearly with respect to lock duration (four years gives 100% weight, one year gives 25% weight, etc.). While weights decay as the remaining lock duration decreases, locks can always be extended back up to the max lock duration. The one difference, however, is that veYFI holders can exit their lock early, subject to a (variable) penalty. If the remaining lock duration is greater than three years, the penalty is 75% of locked YFI. Otherwise, the penalty (percent of locked YFI) calculated by dividing the remaining lock duration by four years.
In exchange for locking their YFI, veYFI holders get:
- Bought back YFI (from BABY), through boosted yield on yVault deposits and a share of all, unearned gauge rewards
- veYFI lock early exit penalty fees
- Governance rights, including over gauge weights and deciding how staking rewards are distributed across yVaults. In other words, non-locked YFI holders have no governance rights.
At its core, veYFI is incentivizing YFI holders to align with the long-term health of Yearn. Based on their commitment, as signaled by the size and length of their non-transferable lock, veYFI lockers increase their governance power and share of YFI ownership.
How Gauges Work
Building on top of veYFI, at the crossroads of fee distribution and governance, gauge weights are the centerpiece of Yearn’s new tokenomics. On the governance end, every two weeks, veYFI holders vote on how protocol earnings (bought back YFI) should be distributed across gauges (each gauge is associated with a yVault). Vault depositors can then stake their yTokens in the corresponding gauges.
Stakers’ gauge boosts are calculated with the same formula used by Curve. Each gauge has different boost thresholds based on total yVault deposits, staker veYFI locks, and the amount of YFI allocated to the gauge. However, the similarities with Curve’s gauge system end there. In addition to the 10x maximum boost (as opposed to Curve’s 2.5x maximum boost), the distribution method is quite different.
The calculated boost determines the percent of total rewards earned by gauge stakers and distributed across veYFI lockers.
Examples:
Gauge staker Alex, with no veYFI lock, gets a 1x boost. Alex keeps 10% of the gauge rewards he could be earning. The remaining 90% is distributed across all locked veYFI.
Gauge staker Bella, with a veYFI lock, has a 5x boost. Bella keeps 50% of the gauge rewards she could be earning. The remaining 50% is distributed across all locked veYFI.
Gauge staker Chris, with a veYFI lock, has a 10x boost. Chris keeps 100% of the gauge rewards he could be earning. Nothing is distributed to veYFI lockers.
Other Features and Utility
The final and most open-ended component of the upgrade is creating “useful work features on top of veYFI, as explained in step four. While focusing on additional utility may be premature, considering veYFI hasn’t been released and yVaults V3 are still in development, there has been some progress here as well.
YIP-66: Streamlining Contributor Compensation is the first, and so far only, formally approved utility to be built on veYFI. At a high-level, this new system replaces a number of the pre-existing compensation processes with veYFI locking. In particular:
- Retiring YFI vesting contracts and migrating existing vests to veYFI locks
- Giving contributors the option to take part (or all) of their compensation in discounted YFI (determined by their veYFI lock duration)
- Scrapping the Strategist performance fee (10% of yVault profits) and paying out multiples of past net earnings as locked veYFI (multiples determined by duration of lock)
Now all contributors, including strategists, will (by default) be compensated in stablecoins. The treasury will payout the YFI using reserves minted with YIP-57 (as opposed to buybacks). The extra revenue can then be used for more YFI buybacks through BABY.
Other ideas for “useful work” follow a similar vein — namely, using veYFI to compensate community members for adding value to the protocol. Some of the ideas for value-add activities include providing insurance, configuring vault parameters, and setting fees.
DAO Treasuries and ve-Aggregator(s)
The boost on yVault deposits will appeal to both segments of Yearn’s depositor base: protocols and individuals. However, the approach towards acquiring the rewards may be quite different between the two groups.
The largest yVault depositor, DeFi protocol Alchemix, for instance, has already voted to acquire $500,000 worth of YFI for its DAO treasury to be locked in veYFI.
While other treasuries may also have the capacity and willingness to lock the capital required for a high boost, this is unlikely to be the case for many users. Given the 10x maximum boost, however, the opportunity cost of not doing so is quite high. As a result, expect to see a robust ecosystem of aggregators and bribing protocols.
Modeling the Impact
Since the veToken model has been used across DeFi, there are existing points of reference to help understand the associated dynamics. However, some important differences exist. For example, for the most part, so far, the veToken model has been used by decentralized exchanges (DEXs). In particular, DEXs focused on stablecoin swaps (e.g., Curve, Platypus, and Wombat), with aggressive emission-based liquidity mining programs.
Not only is Yearn not a DEX, but the YFI token itself has a capped supply with no preset allocation for incentives. Thus, to really understand the impact of the switch to veYFI we must also understand the system piece-by-piece, from the ground up.
YFI
While the new model will be effectively “unlocking” the ~1,111 bought back YFI, veYFI could become an even stronger YFI sink. We can look at similar lock models to better understand this dynamic.
Every protocol has different variables at play, but for the sake of this exercise we can project an eventual 50% lock for circulating Yearn (18,333 YFI) with an average lock of three years. This would measure out to roughly 13,750 veYFI.
In addition to taking YFI off the market, unlike other veToken models, the base yield (bought back YFI plus early withdrawal penalty fees) will come from the circulating supply rather than emissions. While this does have potential negative implications (discussed below), it could have a powerful impact on flows and float. Of course there will be some sell pressure, as DAO treasuries return earnings (YFI yield) to holders or reach diminishing returns (maximum boosts). However, the core dynamic remains: fixed YFI supply moves from less convicted holders (selling on the market, withdrawing locked YFI early) to long-term lockers.
veYFI Yield
One of the core assumptions underpinning the veYFI model is that yields will be large enough to entice holders and depositors to lock. Since buybacks really kicked into full swing in December 2021, Yearn’s treasury has spent an annualized $21.82 million on 1,193 YFI. Assuming the same buyback rate and 13,750 veYFI, this would amount to an average 8.7% APY in YFI terms.
Of course this is an incredibly simplified view of veYFI yield, as a number of factors come into play. For example, the Yearn treasury will have its accumulated YFI at launch, which it plans to also distribute linearly (over an undetermined timeline), boosting yields. On the other hand, unexpected expenses could hamper earnings (revenue – expenses) which would reduce buybacks and thus depress yields. Bribing protocols and withdrawal penalties are two other sources of variable yield. Moreover, since the fee distribution first goes through yVault gauges, yield will vary considerably based on a user’s lock (boost) and deposits.
These three factors: veYFI yield, protocol earnings, and deposits, are tightly connected in this system.
As opposed to veToken models built off fixed emissions, where rising TVL depresses yields, this system responds with proportional increases in yield. This is a great example of the reflexivity DeFi’s known for. However, a note of caution, reflexivity can cut both ways.
DAO Treasuries
To better understand the calculus involved for protocols using Yearn’s vaults it’s worth looking at a DeFi protocol, such as Alchemix. It’s possible to gain a glimpse into what Alchemix could gain since Yearn will be using the same formula as Curve to calculate vault gauge rewards.
Currently, Alchemix has roughly $43.5 million spread across four yVaults. Those yVaults hold a total of $197.8 million in assets. Assuming the amount of yTokens locked in vault gauges is commensurate with the amount of YFI locked, that would mean $98.9 million (50%) deposited across those four gauges (note: this assumption might be low depending on ve-aggregator gauge deposits). Since Alchemix plans to max lock their 50 YFI, this would give them 50 veYFI out of the assumed 13,750 veYFI. Finally, the previously assumed 1,193 YFI rewards will be distributed uniformly across gauges (weighted by yVault TVL).
Under these assumptions, Alchemix could earn roughly 11.5 YFI for its depositors, or a 23% APY on its 50 veYFI. This calculation also doesn’t include the additional yield provided by the tranche of already bought back YFI. For now, Alchemix plans to max lock YFI rewards, increasing their gauge boosts.
ve-Aggregators and Bribing Protocols
Wherever veToken models have been introduced, aggregator and bribing protocols have followed.
Looking at the relationship between Curve and Convex, Convex has 70% of Curve LP tokens staked in its gauges with no real competitors. However Curve’s maximum boost is 2.5x while Yearn’s gauges will go up to 10x. Given the greater opportunity cost, we can expect even more yVault Token holders to deposit with ve-aggregators instead of directly to Yearn gauges (using their own vote locks).
Yearn will have no whitelisting mechanism, meaning no ve-aggregator will have exclusive access to vote-locking. However, it is unclear whether this will prevent a Curve-Convex (or Balancer-Aura) dynamic of one dominant aggregator, seeing as ve-aggregators have their own flywheel in place.
While the first-mover advantage will be organic, the ve-aggregator that can jump ahead early on (e.g., via incentives) could continue to attract marginal YFI deposits. This is of course assuming all else is equal (brand, user interface/experience, security, tokenomics, etc.). Scratching that assumption and accounting for all the variables at play, there could be an equilibrium with multiple ve-aggregators. At this point, it’s just too early to tell.
The equilibrium found between vote-locking with a ve-aggregator and vote-locking directly with Yearn will depend on aggregators’ ability to accumulate YFI, and thus depositors’ willingness to hold liquid wrapped YFI (e.g., cvxYFI). This equilibrium also influences veYFI’s average time lock since ve-aggregators lock tokens with a max lock in perpetuity.
Bribing protocols could become another major source of yield for vote lockers. The amount paid out will depend on the rewards directed to briber gauges, as bribes will be an arbitrage between said rewards and bribes paid. Hence, bribers will most likely be those with large yVault deposits (i.e., protocols) and those bribed will be vote-lockers without deposits.
Risks
DeFi has shown us time and again how reflexivity cuts both ways. This is especially true when all components of a system are reliant on direction from each other. In Yearn’s case, this means its flywheel can also roll into reverse.
Platypus Finance, an Avalanche-based stablecoin DEX using the ve-model, is one protocol that experienced a version of this feedback loop. However, Yearn’s role as a well established DeFi “baseplate” certainly raises the TVL (and thus earnings) floor, as a whole category of large, independent protocols have built their products around yVaults. The floor’s meta-logic is: even after mercenary depositors are gone, switching costs and risks will still outweigh yields lost, thus keeping the largest depositors, and largest sources of yield, in place. A similar logic exists for YFI price, since YFI isn’t emitted in an inflationary, preset manner, but instead is reduced as TVL (earnings) go down. Therefore, if vote lockers dump their yield, the size of the yield itself will shrink.
The next potential risk is the 10x maximum boost. This will increase reliance on ve-aggregators. Most likely, the dynamic will create a reliance on just one, or at most a handful, of ve-aggregators – since voting power / boost dynamics funnel activity, from all angles to a dominant platform(s). While they can have their share of benefits, such as effectively being YFI blackholes, they also introduce dependencies and risks outside the Yearn community’s control. A second, more general concern is that these ve-aggregators effectively circumvent the long-term alignment dynamic. Voting power begins to think in weeks and months, rather than years.
The principal-agent problem can also be further exacerbated by bribing protocols. This could become particularly problematic for Yearn if the scope of veYFI (under step four) is expanded to include something like vault parameters.
From a broader perspective, the return to fee distribution may also be untimely. While a discussion around fee collection itself is out of scope for this report, it is uncommon for entities (in crypto and otherwise) to distribute cash flows 2.5 years into operations. Preserving capital and investing in growth and development could turn out to have been a better use of earnings, especially in an embryonic, fast-paced industry.
Conclusion
Discussing “the token” may be seen as gauche, but tokens can unquestionably create value for protocols. DeFi 1.0 seems to have gotten the first part down: using tokens to bootstrap value. Unfortunately, as narratives shifted, the reality laid bare that most tokens had virtually no value.
Curve was the exception.
Curve stands out as a diamond in the rough because it uses its token to create value for the entire ecosystem, and, crucially, itself. If creating value is the first part of a sound tokenomics strategy, then maintaining protocol value is certainly the second part.
To Yearn’s credit, it did attempt to capture value. But luck and timing weren’t on its side. However, this may have been for the best as Yearn’s unique veToken implementation seems set to be a win-win, enabling Yearn to increase protocol usage and the value of its native governance token, YFI.
______________________________________________________________________
Let us know what you loved about the report, what may be missing, or share any other feedback by filling out this short form. All responses are subject to our Privacy Policy and Terms of Service.
This report was commissioned by Yearn, a member of Protocol Services. All content was produced independently by the author(s) and any opinion, estimate, or probability does not necessarily reflect the opinions of Messari, Inc. or the organization that requested the report. Paid membership in Protocol Services does not influence editorial decisions or content. Crypto projects may commission independent research through Protocol Services. For more details or to join the program, contact ps@messari.io. Author(s) may hold cryptocurrencies named in this report, and each author is subject to Messari’s Code of Conduct and Insider Trading Policy. Additionally, employees are required to disclose their holdings, which are updated monthly and published here. This report is meant for informational purposes only and should not be relied upon. This report is neither financial nor investment advice. You should conduct your own research, and consult an independent financial, tax, or legal advisor before making any investment decisions. Nothing contained in this report is a recommendation or suggestion, directly or indirectly, to buy, sell, make, or hold any investment, loan, commodity, or security, or to undertake any investment or trading strategy with respect to any investment, loan, commodity, security, or any issuer. This report should not be construed as an offer to sell or the solicitation of an offer to buy any security or commodity. Messari does not guarantee the sequence, accuracy, completeness, or timeliness of any information provided in this report. Please see our Terms of Service for more information.
