Note: this assumes a high level of familiarity with Curve ecosystem tokenomics. For a primer, read this first.

It’s been a while since I caught up on the Curve Wars. When I last wrote about Convex, it was in a power position and I didn’t see much hope for Yearn to compete, at least in terms of veCRV control.

Since then, new Yearn tokenomics have been introduced, and particularly interesting is their apparent focus on capturing more CRV. Note that Yearn’s veCRV share has inched up while Convex’s has cooled off:

The timing makes sense with CVX emissions now relatively insignificant for locking CRV as cvxCRV.

So let’s take a quick look at Yearn’s value proposition for locking up more CRV.

Activated yTokens

While you can find more details in their docs, Yearn allows you to lock CRV and receive yCRV, which you can phase between st-yCRV, lp-yCRV, and vl-yCRV at any time (respecting cooldown restrictions).

vl-yCRV allows you to vote on Curve gauge weights with the respective veCRV that your vl-yCRV represents. You don’t earn any fees and don’t earn bribes. This module is still under development and is supposedly coming “soooon”.

st-yCRV earns you passthrough Curve admin fees in addition to bribes. The underlying veCRV is used to vote in favor of whatever maximizes st-yCRV bribe revenue. This is likely to be the majority of the fees for st-yCRV stakers.

lp-yCRV earns pool fees (yCRV/CRV) in addition to CRV rewards. Its underlying veCRV is used to vote on its own yCRV/CRV gauge.

Yearn will employ its new vault standard for autocompounding of gains.

The Value Proposition for DAOs

While there are individuals that may want to sway the Curve gauge reward weights, DAOs have been historically the biggest contenders and most willing to pay bribes.

So let’s say there is a DAO, and it wants to incentivize its token pairing on Curve by accumulating voting power (veCRV).

This DAO has a choice. It could buy CRV and stake it on Curve and vote directly. This would require a big upfront investment and a 4 year lock for maximum voting power.

Or it could buy CVX tokens. Each staked CVX provides around 5 CRV worth of voting power. Again, this requires an upfront investment.

The DAO could also opt to bribe holders of CVX and st-yCRV (among other CRV derivatives) to vote for more CRV emissions on their Curve pool. Here we avoid the upfront cost of purchasing tokens and may be able to more efficiently bootstrap the Curve pool rewards without emptying the DAO treasury.

So what will this DAO decide to do if it seeks long-term voting power?

If the DAO believes it will be a long-term Curve voter, it may make sense to invest in voting power instead of bribes. If it decides to do so, it probably wont opt to stake directly on Curve given the downsides of the 4 year lock and the need to re-lock or lose voting power.

Instead, it makes more sense to either buy CVX or yCRV (as vl-yCRV).

And if it were to purchase CVX, it would also receive the cvxCRV and FXS rewards on top. However, it’s not so simple since st-yCRV earns the Curve admin fee alongside bribes and vl-yCRV could be seen as having an option on capturing the admin fees if the DAO no longer needed to vote.

So, what does the DAO choose? I would bet that it chooses CVX due to the fact that its primary objective is to buy voting power, not hold an option on Curve admin fees increasing. By choosing CVX it earns a bit of yield while securing its voting power. Remember that it can always sell it if it decides that it no longer needs to direct rewards to its Curve pool.

Conclusion

While there are many reasons to bet on Yearn, including its new vault design and yETH which I believe could multiply its TVL and revenue, I don’t believe that an overwhelming portion of DAOs looking for voting power will opt for vl-yCRV over CVX or other derivatives. This likely means that Yearn’s CRV capture increases somewhat but is unlikely to change the game for CVX which has already locked a significant portion of the circulating CRV supply and who continues to capture more CRV via new locks and the rake on vault profits. Either way, I’m impressed by both of these projects and am looking forward to seeing where things go!

Did I miss something? I often do, so please reach out if you know something I don’t.

vl-yCRV is Not a Threat to Convex