Building a Successful DEX Pt 1: Bribing Analysis

The Pearl thesis is a ve(3,3) flywheel that runs in reverse with bribes driving the ecosystem forward. As a protocol predicating its success on the economic power of bribes within a Solidly-style ecosystem, we spent a great deal of time researching bribes and their correlation to a positive flywheel.

Our investigation looked to answer two primary questions:

1. Do bribes increase token value?
2. Do bribes help build and maintain TVL?

Intuitively the answers to this are, of course, yes. But what does the evidence say?

Below we’re sharing our exploration into two mature, vote-escrow ecosystems and the data we’ve collected to help substantiate our thesis.

VELODROME

We’ll start the exploration with a look into Velodrome, the most mature and largest Solidly fork.

Summary findings:

• VELO returns can be strongly associated with bribes, but there have been periods where this has gone both ways. However, the macro moves are generally consistent with our view that bribes move token price.
• Total liquidity does increase as bribes increase. This is true moreso for stables, less for volatile assets, but the relationship is fairly clear regardless.

In the first set of charts, we’ll look specifically at the price of VELO. When bribes are high, is that associated with positive returns? From what we’re seeing, there is a positive relationship between VELO returns and bribes, however this is mostly evident in the earlier epochs. As time has progressed, the relationship is now less clear, with high bribes being associated with both positive and negative returns, though neutral to positive outcomes are more likely than not.

Changes in bribes actually appear to be more predictive. But even that varies depending on the timeframe. On the left is the first half of the data, which does not seem to exhibit the desired behavior. Returns on VELO do not move positively as bribes increase. On the right we can see the second half of the data above, which does start to show a correlation between bribe values and token price. This is consistent with a market that’s starting to become more efficient.

Stepping back and looking at the longer-term trend, we can clearly see this relationship over time. From a purely statistical perspective, there may be some periods of inconsistencies, negative correlation and lag, but the general trend is clear and obvious.

So we’ve established a fairly clear pattern supporting the notion that high bribes lead to positive returns for the protocol governance token.

What happens when we look at the connection between bribes and TVL? Once again, a clear pattern starts to emerge showing increasing bribes increases TVL for our surveyed pools.

The correlation is most pronounced when looking at stablecoin pools. Below are three Velodrome stable pools plotted out. We clearly see that as bribes increase, liquidity for the pools increase. Bribes are driving up TVL.

We note that this effect is less pronounced for tokens that are considered “risky” vs stables. This also explains some of the friction wUSDR faces in the chart above compared to more established or traditionally designed stablecoins. At the moment, USDR charts more similarly to volatile tokens/pools. Bribes are clearly a positive influence on liquidity for volatile/risk assets, just to a lesser extent than they are with stablecoins. As markets become more efficient in their evaluation of USDR with the product becoming lindy, it should behave more like a standard stable. As such, we believe there’s a lot of upside Pearl TVL as per this analysis.

CURVE

Curve is the original bribing economy in crypto. While it doesn’t fit the same exact mechanics as Velodrome, Pearl or other Solidly forks, it’s still informative on the relationship between bribes, token returns and TVL.

Summary Finding:

• The CRV market is larger, more efficient and demonstrates the desired relationship more clearly than Velodrome
• Although there’s a strong bribe to price association, this cannot protect against larger macro factors, like the fact that CRV is structurally inflationary and crypto moves in cycles

There was a tremendous amount of anticipation and demand for CRV at launch and what we see are bribe increases trailing CRV returns. To some extent, CRV prices may have been preempting the anticipated bribes, especially during the froth of a bull run. The chart also shows a period of CRV decline while bribes were increasing. However, there’s still a statistically significant relationship here. And over time, we see a clear correlation between bribe values and CRV returns, as both begin to trail off.

Lastly, we winsorized the data, just to be sure that outliers were not biasing the result. The trend held and we’re 99.88% confident the correlation is there:

CONCLUSION

What we see above is a clear correlation between ve(3,3) token value and bribes. Our thesis is substantiated by the data, though how it will play out in the market is yet to be seen.

Each new “flywheel” DEX attempts to solve the problems of its predecessors, however are new protocols just dressing the same ineffective tokenomics in more complex costume? How can Pearl truly avoid the ve(3,3) death spiral?

While our bribe data is generally positive, it’s an open question as to whether you can count on saving the token price with bribes alone. The key is to ensure that bribe values provide a clear and consistent backstop to token value, ensuring the market never reaches a point where the token isn’t valued. In our following articles on protocol analysis and the Pearl Flywheel, we’ll dig into the mechanics behind Pearl and how we plan to build a ve(3,3) DEX that lasts.