Is Frax finance’s DEX bribing operation sustainable?
Have you heard about bribing in DEFI? If yes, you can skip to the 4th paragraph. Bribing is a practice that first became popular by Curve in 2021 through introduction of gauges. Essentially gauges allow some users to incentivize other users for a certain behavior during a certain period of time by giving bribe to whoever commits to that behavior in that period.
Thena is a popular DEX on the bnb chain that is using the concept of gauges to attract more liquidity to its pools. Frax finance is a major DEFI protocol on Thena who owns most prominent FXS, FRAX and frxETH pools. Obviously, this is DEFI we’re talking about. So nobody owns any pools. But what I mean is that Frax finance owns these gauge contracts that incentivize these pools and bribe them.
At each period (each week), Thena, like Sushiswap, offers new emission of its coin to liquidity providers (LP’s) on each pool relative to how important the pool is. The importance of each pool is decided through community voting in the preceding period. Frax finance influences this community decision by assigning bribes(in the form of FXS token) to whoever votes for the importance of its pools. Thereby hoping to attract more LP’s to its pools. Obviously more LP’s equal more liquidity and better price discovery. To get more into the nitty gritty, the way Frax finance actually distributes this bribe is by funding the gauge with the total amount of bribe(FXS) that they decide to pay for a specific pool in a specific period (subsequent period). Voters see the amount of bribes on each pool and usually vote for the pool that has a higher amount of bribes attached to it. A few days later, when the voting is over and the actual period starts, the bribe gets distributed to each voter according to their voting power.
The question we’re trying to answer here is “to what extent the bribing has helped Frax finance’s cause of improving its liquidity on Thena?”
First off let’s look at the profitability of this operation for Frax finance. You might say “well, how can I make profit, if I am only paying bribe?” The answer is that first, Frax finance is a voter itself and is collecting part of its bribes for itself. And second, Frax is not the only entity that is bribing on its pools. Each Frax pool is a pair of 2 tokens, e.g., FRAX-MAI. Therefore the pool is also bribed by the protocol of the other token. So this is also income for Frax as voter. But for this article, we’re only looking at bribing revenue out of only FXS bribes received. Also, we focus only on FRAX pools. A separate analysis is required to track the performance of frxETH pools.
For some reason, it looks like sAMM-DOLA/FRAX and sAMM-FRAX/BUSD pools are the only two that have remained consistently profitable in the past 2 months, you can further expand the period of analysis yourself and see the results here.
Looking at the total bribe for all pools vs total TVL of all pools for the past 2 months, we can see a correlation of 0.73 which is quite good. Ideally, we would like it to be close to 1, which means that there’s a consistent pattern where higher bribe amounts are associated with higher TVL and vice versa.
On the other hand, if we look at the above chart in terms of the % change for each of the 2 time series, the picture is quite different. We only get a mere 0.13 for correlation. Ideally, we would like it to be close to 1, which means that there’s a consistent pattern where higher change in bribe amount is associated to a higher change in the same direction of TVL.
In this case, at least for the past 2 months, there is not a strong evidence that the market buys into the strategy of Frax Finance team when they increase or decrease the bribe amount drastically. It could be that they are not sure what it means in the long run.
It’s worth noting that 2 months, or even 2 years for that matter, does not constitute a period large enough for market to develop a specific pattern. You can use this chart to play with the starting date of the analysis and go as far back as January 2023.
