ve(3,3) DEXES: a profitability analysis

ve(3,3) is a fairly new concept in DeFi.

“(3,3)” is a concept from game theory that can be loosely translated as: “we all win if we collaborate”.

“ve” stands for vote escrow, which refers to locking tokens to gain voting rights.

Putting them together: “ve(3,3)” is a DEX model in which the tokenomics are designed in such a way that it incentivizes players to act in the benefit of the protocol.

The LP rewards are in governance tokens, these can be locked (then become a ve-position). The ve-position gives voting rights, which:

• directs emissions of the governance token
• earns trading fees
• earns bribes (external protocol rewarding you if you vote for their pools)

The ve(3,3) model is theoretically superior to older DEX models. In older DEX models, there is less incentive to lock tokens:

• lockers get smaller share of trading fees
• little to no control over emissions (which determines the APR’s of pools)
• little to no bribes to be earned

In a ve(3,3) DEX, locking tokens is much more attractive because all trading fees go to voters, the system is built to attract bribes and allow control over emissions. Of course, in practice things are a bit more complicated than just tokenomics. For example, older DEXes have a first mover advantage.

I see a lot of poor takes on the ve(3,3) DEX model. For example, it is commonly stated that inflationary tokens will likely go down in price. This is simply wrong (bitcoin has been inflationary its entire lifecycle…).

Such takes are probably the result of post-traumatic stress of Olympus DAO investors, where people learned that high emissions do not always mean that your position increases in value. But the real lesson there was not that “inflation equals bad token”, but rather that you need to understand the actual math of the tokenomics; are you getting diluted or is your relative share and the absolute dollar amount of your position growing and why?

This article will focus on simple models to assess and predict future the profitability and price action of ve(3,3) DEXes.

It will use THENA to use specific numbers, but the principles can be applied to any ve(3,3) DEX.

BASIC EPOCH PROFIT MODEL

Let’s start. with the basics, shall we?

In ve(3,3) DEXes, each week represents an epoch. Each epoch has a known emission schedule and token lockers can vote once.

Thus these epoch represent a logical unit of time to track the profitability over.

Profit = Revenue — Costs

In a ve(3,3) DEX:

• Revenue = Trading fees + Bribes
• Costs = Emissions value

Why are emissions costs?

Because locked token holders (veTHE for THENA) should be considered “the protocol”.

• If emissions are being dumped, it is lowering the dollar value of your token position
• even if the emissions are locked, you’re getting diluted

1. Profit = Revenue - Emissions

Some scenarios:

a) Revenue > Emissions = Profitable
Very good. Lockers receive a yield that exceeds their dilution and/or the price impact from dumping.

b) Revenue = Emissions = Break even
Good. A concern with inflationary DEX tokens is that displayed yields (high APRs) are meaningless because it’s from inflationary tokens. In break even, the revenue keeps up with the inflation and thus the yields you see are actually meaningful.

c) Revenue < Emissions = Loss
Lockers are getting diluted faster than revenue is coming in and/or $THE is dropping rapidly. If this happens for several weeks, it may be the start of a death spiral that can be tricky to escape from.

While this model helps understand the basic dynamics, it’s too simple to be accurate.

REBASES

Imagine you run a business that has both a revenue and cost basis of 1M. You would break even. Probably not great. Probably investing a lot of time and effort and nothing to show for it financially.

But what if all those costs are because you decide to give yourself a 1M salary? All of a sudden your financial situation is a lot more appealing right?

In a ve(3,3) DEX, the lockers are the protocols. Any emissions (costs) sent directly to lockers should not be counted as a cost (as it by definition is not diluting you).

Many ve(3,3) DEXes have a rebase, in which a percentage of emissions is sent directly to the lockers. For THENA, this is 30% of the emissions.

Thus our model should account for rebase:

2. Revenue — (Emissions — Rebases) = Profit

Accounting for rebases does not change the interpretation of the model we discussed earlier; it just lowers the cost basis of the equation by 30% for THENA specifically.

BRIBES CLOSE THE GAP

We need to discuss bribes a bit more.

Bribes are a source of revenue for ve(3,3) DEXes. But protocols don’t bribe out of charity. There’s multiple reasons they bribe, but the main one is “capital efficiency”.

Fancy term. But it basically means it’s just the cheapest way to reach a certain goal.

Protocols need to have sufficient liquidity so a lot of people and/or big whales can buy and sell their tokens without drastically moving the price. For that, they need people to LP with their token. But people won’t do that unless they’re incentivized to do so.

So people LP to farm rewards. Protocols can allow LP positions on their own website and then reward the LP farmers with their token. But this has many drawbacks when compared to having the LP on a ve(3,3) DEX, where:

• their pool reaches new eyes (free marking)
• don’t have to invest in keeping up with the latest automated market maker (AMM) tech (e.g. concentrated liquidity and automatic management developments)
• capital efficacy

The latter comes down to that if a protocol bribes for $1000, they typically get >$1000 worth of emissions. Wait wut? Infinite money hack?

Imagine a scenario where a ve(3,3) DEX has an epoch and emits $100k worth of their token. But they have not generated any trading fees and got no bribes from any protocol.

What would happen if a protocol would now put a $1k bribe?

To simplify, they would get all the votes (this is not true for various reasons, but no need to complicate it). All the votes means that all the emissions go to that pool.

Therefore, instead of rewarding LPs with $1k, their LP would get $100k worth of incentives. That is an absurdly good deal for them. Thus, bribing was much more capital efficient than directly sending their $1k bribe to the LPs.

In practice, the total bribes will be capital efficient if emissions are bigger than revenue (it does depend on exact voting behavior, but for simplicity, let’s assume everyone simply votes in a way to maximize their voting rewards).

Let’s say that epoch emissions are worth $100k and trading fees were $50k. In that case, it is attractive for bribes to make up the difference because their bribes are capital efficient.

As a result, bribes tend to close a negative gap between trading fees and emissions worth.

But we need to critically evaluate what this means in practice.

Does this mean that there will be no bribes if trading fees are greater than emissions? No. During such a situation, the capital efficiency may be negative. But it might still be worth it or even a necessity for the protocol to bribe, as they simply need to incentivize their LPs or their liquidity disappears and their token is likely to dump.

And if bribes tend to close the revenue deficit, does that mean that ve(3,3) are always at breakeven? And if so, why do some seem to have death spiraled?

No, a ve(3,3) DEX is not always breakeven. An example where bribing might not be worth it for protocols, is when the ve(3,3) token price seems to go down rapidly and consistently. It might seem like a capital efficient deal at the time of bribing, but if the emissions token drops 30% the week after, it might be negative return on investment.

But perhaps more importantly is that when a ve(3,3) death spirals, LPs start complaining why the APRs keep going down and being involved with the DEX becomes a negative association instead of positive free marking.

Summarizing this section:
Bribes are likely to close a potential gap between trading fees and emissions because it’s capital efficient for partner protocols. However, this only happens when the ve(3,3) DEX is doing at least decently. When a ve(3,3) DEX is in a death spiral, the protocols will run for cover.

BUYS AND SELLS DETERMINE PRICE

We have discussed that Revenue — (Emissions — Rebases) is a simple calculation to get an idea of whether the protocol is operating at profit or loss. But how is this related to the token price?

Token price is determined by buys vs sells.

The protocol can operate at a loss, but the token price price can still go up. Imagine all the emissions are getting locked (0 sell pressure), and the yield is used to buy the token (>0 buy pressure). In that case, the buy pressure is greater than sell pressure, resulting in an increase in token price.

Of course, the scenario above is not very likely. If the protocol operates at a loss, all the emissions get locked, and voting rewards are used to buy the token, the voting APR would go down really fast (as the token price increase relative to the revenue and relatively small revenue is split amongst more people). A relatively high token price and lower yield makes it attractive to sell the token instead of locking it. Thus, when the protocol operates at a loss, it is likely that the token price will eventually go down due to secondary effects.

Our example shows that buys and sells determine the price. Protocol profitability impacts the yields for the lockers and thereby directs people towards selling or locking (so a slower indirect impact on price).

ESTIMATING SELL PRESSURE

Protocol profitability is an important metric to keep an eye on as a metric of the health of the protocol.

But can we estimate changes in the token price more acutely? Moment to moment is too variable to predict. But we can estimate sell pressure over and epoch based on the locking ratio.

We’ve seen that buys and sells determine token price. But what is the impact of locking tokens?

Currently, ~77% of tokens are being locked on THENA. There are a few other things you can do with tokens such as LP or lend it. But let’s take the most negative scenario and assume the other 23% are all being dumped.

Sell pressure = Emissions x (100%-Locking ratio) = Emissions * 23%

As you can see, the actual sell pressure is a lot lower than the cost of emissions. I’m saying this, because I often see incorrect statements along the lines “they emit 200k dollar worth of tokens per epoch, no way there’s enough buy pressure to compensate for that”. (You can check https://eliteness.network/compare-solidly/ for tons of stats such as emissions value)

The sell pressure needs to be offset by at least an equal amount of buy pressure, or the token price will drop.

The challenge that a DEX faces is that sell pressure can be expected. Mercenary capital just goes to a DEX, farms, dumps the token, and compounds into their LP position. They don’t want to have any price exposure to the DEX token, but just compound their initial position. Often stable assets are used for this (stablecoin pairs or gas token/LSD pair) to avoid impermanent loss.

It doesn’t even matter to them whether the DEX is making a massive profit or not (they probably do not even know how to do the analyses), they just want to play it safe and farm with stable pairs.

ESTIMATING BUY PRESSURE

Where does buy pressure come from?

A combination of:

• Speculators (just have a good feeling the protocols will go up in the near future or next bull)
• DCAers (belief in the long term potential and do consistent investments)
• Buybacks from ALPHA perp yield
• Savvy investors who have identified the protocol is profitable
• Accumulators; lockers who compound their voting rewards in more locked tokens

We could try to estimate buy pressure based on historical data. But let’s try to be as conservative as possible.

So imagine a scenario in which speculators, DCAers, traders on ALPHA, and savvy investors have all thrown in the towel and stopped buying. Even with accumulators, past behavior is no guarantee of future buy pressure.

But there is a group that has essentially opted in for consistent accumulation.

Enter liveTHE

liveTHE stands for liquid veTHE.

As the name suggests, it’s a liquid wrapper for veTHE. The product is by Liquid Driver, which is essentially the same team as THENA. Thus, it’s not surprising they created a product that has synergy with THENA.

It essentially gives you the benefits of locking, without the drawbacks. You use your $THE or veTHE to mint liveTHE (thus for every liveTHE, there is at least one underlying veTHE).

liveTHE:

• Is liquid
• Auto-votes for you (saves time, effort, and stress to vote last minute)
• Auto-compounds the underlying veTHE

So it used the underlying veTHE to vote, swaps those rewards into $THE, mints liveTHE, and passes that newly minted liveTHE on to the liveTHE stakers.

One of the benefits of liveTHE for THENA is that people who would never lock any token, are happy to lock $THE as they get liquid liveTHE in return. Good for the individual, good for THENA.

Because remember what locking does to the sell pressure? An increase in locking, means decrease in sell pressure.

And liveTHE also auto-compounds the underlying veTHE. This means that all liveTHE has essentially opted in as accumulators: every epoch they are using all their voting rewards to buy and lock more $THE.

In other words, liveTHE does everything you want to positively impact the $THE price:

• Increases buy pressure
• Lowers sell pressure (by its locking pressure)

I often see people being afraid that liveTHE is used as an exit strategy and that the holders are not loyal. But so far, the data shows the opposite. I’ve been tracking and reporting liveTHE dominance in my weekly THENA updates, and it’s literally up only.

liveTHE dominance = the amount of veTHE underlying liveTHE divided by all veTHE. While maybe some people have used it as an exit from their veTHE position, the total liveTHE dominance is growing.

Right now liveTHE dominance is about 23%.

On a ve(3,3) DEX, all revenue goes to the lockers. liveTHE holds 23% of all veTHE and auto-compounds the underlying veTHE. This means that 23% of all revenue is used as buy pressure.

change in $THE price is determined by: buy pressure vs sell pressure

• Maximal sell pressure = (emissions-rebase)*(100-locking ratio)
• Minimal buy pressure = revenue*liveTHE dominance

We discussed earlier why revenue is typically at least close to emission cost in a well-functioning ve(3,3) DEX.

THENA has been profitable (again, see my weekly THENA updates on Twitter for numbers), but let’s be conservative again and say they’re typically break even.

In that scenario revenue and emissions minus rebase are even.

And right now the locking ratio and liveTHE dominance are both 23%

So:

• Costs equals revenue
• Maximal sell pressure = (emissions — rebase)*0.23
• Minimum buy pressure = revenue*0.23

Therefore, currently the minimum buy pressure equals maximal sell pressure.

This means that the guaranteed buy pressure from liveTHE alone makes up for all potential selling of emissions.

That is an incredibly comforting stat. This makes it hard for $THE price to go down unless something drastic happens.

Because sell pressure cannot really increase (only if a liveTHE whale exits…in short, this likely will be compensated very quickly because of arb opportunities)

In contrast, actual buy pressure is likely much higher than our minimum estimate. Our minimal estimate assumed not a single person with veTHE will use its voting rewards to buy $THE, not a single person will market buy $THE, and ALPHA does not generate any revenue at all.

Ok, so now you might be thinking that this can’t be right. If this analysis is correct, shouldn’t $THE price been mooning?

Well, the timing of this article is no coincidence.

I’ve been tracking these stats since day 1. And I’ve slowly seen the $THE locking rate crawl up to its current ATH of 77%. Likewise, the liveTHE dominance has been going up to its current ATH of 23%.

This week is the first time that liveTHE buy pressure alone can compensate for all sell pressure (technically it’s a little bit short due to the small performance fee on liveTHE). I’ve been patiently waiting for this moment to drop this analysis.

The liveTHE dominance will likely continue to grow due to its auto-compounding nature. This drastically helps the equation, as it means relatively more buy pressure AND less sell pressure due to increased locking.

Add in other potential sources of buy pressure, and $THE looks bright.

CONCLUSIONS

A simple model to track the health of a ve(3,3) DEX is: Profit = Revenue — (Emissions — Rebases). If this is consistently deeply negative, it’s problematic.

Token price action depends on buys and sells. Maximal sell pressure can be estimated by: emissions*(100-locking ratio). Minimal buy pressure can be estimated by: revenue*liveTOKEN dominance.

If the ve(3,3) DEX is decently functioning, it’s likely that bribes close any potential gap between revenue and costs (emissions — rebase), resulting in a break-even epoch.

THENA has hit the critical liveTHE vs locking ratio this week: the minimal buy pressure from liveTHE should entirely offset the maximal sell pressure from mercenary capital. Any other buy pressure will mean the price moves up. As the liveTHE dominance has been up only, it may soon result in net positive buying pressure even without any other buying pressures.

Similar analyses can and should be done for other ve(3,3) DEXes. But it requires some work to find and calculate the data, especially because there’s often differences in rebases or other differences such oTOKENOMICS.

Please note that the fundamental lesson of this article is not that all ve(3,3) DEXes are “good investments”, that Thena is a “good investment” right now, or that liveTOKEN is hack to turn any ve(3,3) DEX profitable. The fundamental lesson is that profitability analysis and price action models can help evaluate whether projects are potentially promising investments or not. If done properly, such analyses should hold a lot more weight than lines on a chart (“technical analyses”) or parroted narratives (“all DEXes trend to zero”).

IF THIS POST WAS HELPFUL…

Please help spread the word on Twitter here.

Shout out to the following for providing feedback:
@thenascribe
@pantheonfi