This proposal builds on Chutoro’s proposal and community conversations that have ensued in Silo’s discord channel. In this proposal I advocate for the veTokenmics and explain why this model is better suited for a nascent lending protocol with potentially hundreds of money markets.
The virtuous cycle of borrowing
The previous proposal did a good job explaining the virtuous cycle of borrowing – borrowing drives lending. This hypothesis has been tested and can be used to build a case for a model that incentivizes borrowers especially in the early days of the protocol when markets should attract borrowers to drive demand for liquidity.
To incentivize borrowers in a competitive market, Silo can offer a “discount” on the borrowing rate in the form of SILO rewards. I think the majority of us have agreed on this point.
Let’s now get to the most important aspect of the token model:
- Who gets the discount?
- What are the requirements for receiving the discount?
Token model centered around power users
The 80/20 rule applies in Defi more than anywhere else - Take any Defi protocol today, activities are clustered around a small number of users (I use the term “user” in a generic sense - retail users and otherwise).
Because borrowing drives lending, the system is better off attracting users that borrow large amounts frequently (and deposit large collaterals therefore) - let’s call them Power Users.
Across the borrowing spectrum you have a user with small borrowing activity on one end and a power user who is constantly borrowing for leverage, farming, shorting, etc on the other end.
Power users bring their large collaterals with them and fund lending rates, therefore Silo should design its token model in a way that helps the protocol attract and retain them. Offering a discount on borrowing rate in the form of rewards likely attracts power users.
If the discount is offered on borrowing in the right context, it is a great incentive to build a basis for large borrowing activities. With large depositing and lending, TVL grows. Though TVL is not a success metric, it’s a great social signal that any nascent protocol needs to signal that the marketplace is “safe” to use.
Who gets what?
Here is where things get hairy. What about token holders? Is it fair that the discount goes only to power users?
You can look at any token model from the perspective of two players:
1- The borrower: A user seeking to borrow at the lowest rate possible
2- The token holder: A user seeking to maximize the token value
The overlap between these two players creates a distinct subgroup: A power user who is willing to commit to buying and locking the token for a certain period of time in exchange for an immediate incentive - that’s basically the essence of veTokenomics.
In our token model, users would lock tokens to receive an immediate discount on the borrowing rate. The discount is a tiered system and a function of the lock time. However, unlike other token models, the discount is granted immediately. Because the incentive is immediate, power users are incentivized to collateralize and borrow.
Back to the question - What are the requirements for receiving the discount?
- Lock SILO in the system (everyone has to do it)
- Choose a lock duration congruent with your projected borrowing business that you will bring with you. Duration can range from a week to years.
- Receive the discount and use it immediately
Similar to Curve’s model, Silo can offer a boosted discount up to 2.5x depending on the lock duration the user chooses. Lock durations can range from a week to 4 years.
From the protocol’s perspective, the veTokenomics offers a fair distribution of incentive where discounts are given to those who will likely borrow now. These double user actions (locking Silo and borrowing) drives value to the entire ecosystem including lenders and token holders.
VeTokenomics vs. Time-based tokenomics
The two models incentivize two important user behaviors:
- Retention: Both models reward users who stay longest.
- Borrowing: Both models encourage users to borrow though I argue late in the post that veTokenomics performs better on this aspect.
In the following paragraph I make comparisons between the two models across different aspects.
Speed & defensibility
The velocity of the borrowing cycle is likely faster when veTokenomics is applied because large-scale borrowing can happen straight away since discounts are offered immediately. You can make a strong argument that time-based tokenomics would lead to increased borrowing activities, however that is likely to happen over time as the discount grows (multiplier points grow in volume over time). This is a vital aspect to Silo because borrowing at high velocity and volume is much needed now.
Another point is defensibility. As a borrower, you are likely to use Silo over a competitor even if the competitor offers a similar or even better borrowing rate. That’s likely to happen if you think the value of your rewards will grow over time.
Rewarding the right user
veTokenomics incentivizes both retention & large borrowing volumes if you are willing to exchange locking your tokens for a discount. However, this is not a permanent state since the boost decays. If you don’t renew your max lock - because you are not interested in the borrowing discount - your value to the protocol decreases and so does your discount. Discount goes to those who need it the most in a given period of time and are willing to lock SILO for the max boost.
While time-based token models reward retention, they don’t differentiate between power users and others in the system. If two users locked SILO on the same day, one of them with expected borrowing of $10M and the other $10K, the system rewards both with points growing at the same rate for both users.
Entering the system
In veTokeonmics, anyone can lock for the max duration today and receive the boost (discount). In time-based tokenomics, this is not the case. If Will has locked SILO before Alan, there is not a way for Alan to catch up with Will’s voting power and boost even though Will happens to be a power user.
PS: From the user’s perspective, exiting the system is obviously harder when veTokenmics are applied but I argue that making the system fairer to enter is far more important than fairer to exit.
Everyone benefits from veTokenomics
If you borrow little or don’t borrow at all, you will find it hard to stomach veTokenomics. However, if you believe that incentives are better delivered to those who will drive the growth of the protocol, your perspective changes.
While token holders don’t receive direct rewards, they still benefit indirectly. Since borrowers will likely buy SILO and lock it to receive discounts, the supply shortage will cause the token to appreciate. This drives indirect incentives to token holders to hold the token and participate in governing the protocol.
Convex: a catalyst of price appreciation
Silo needs to find ways to appreciate the price of its tokens in order to maintain competitive borrowing discounts across its growing markets. The Silo treasury is large, but when Silo grows into having hundreds of markets, emissions will be spread thin driving down the value of the discount.
In the beginning, the majority of “power users” will likely wait for the dust to settle down before they lock the token for longer durations. This effectively means rewards will likely not constitute a meaningful discount on the borrowing rate unless the token value is well-established. Until more users lock for the max duration driving the token price up, you need a catalyst for the price to appreciate.
The catalyst can be meta governance dApps like Convex. Meta governance projects could scoop up SILO from the market and drive up its value. Projects that apply veTokenomics are more attractive to meta governance projects for many reasons beyond the purpose of this post.
I think the Silo community should not fear working with meta governance projects because they offers great utility to borrowers, token holders, and Silo alike:
1- Users get the max discount on the borrowing rate and can still exit the system
2- It drive the project’s token price higher
Silo veTokeomics - recommendations
- Follow Curve’s token model (veSILO/boost) and emission program with shorter duration (ideally 3-5 years). Governance can extend the program if needed.
- Consider forking Curve’s bribe system - the system can co-exists with other dApps.
- Plan to emit 35% of the treasury - keep 10% as reserves
- While veSILO holders should vote where emissions go, it is not a bad idea to guarantee voted gauges a base amount of rewards. This mechanism dampens the negative impact of bribes skewing emissions to Silos that benefit influential users. The protocol benefits from borrowing taking place across all its markets, not just among the top 10%.
- Implement mechanisms to break the locks to migrate users to a new token model preserving current voting power and boost. This action should only happen through governance and with an ample notice period like 6-12 months. I don’t expect the DAO will ever resort to breaking the locks, but it is good to code it.
I am happy to hear your suggestions.