Frax is essentially a CRV Farm that rewards TVL with CRV emissions.
Frax pools currently control 35% of CRV emissions, we estimate that it can potentially expand to 50% influence as a likely upper bound.
Given Frax's level of incumbency on Curve, we don't expect Frax's TVL to grow secularly.
We expect Frax to rotate towards newer pegged assets and capture new narratives. It's a cyclical, narrative-driven bet on future pegged assets.
Otherwise, markets don't seem to react to Frax's internal updates (FraxSwap, FraxLend, FraxChain, Tokenomics, etc.) We also deem these developments immaterial compared to its main Curve Farm business.
（For the sake of clarity, in this essay, we use "FRAX" to refer to the stablecoin token, "Frax" the conglomerate in the abstract, and "FXS" the Frax Share token.）
Motivation: Why do people hold FRAX and frxETH?
We start our investigation with an actor-based analysis:
What drives Frax's TVL?
What's the incentive behind holding FRAX and frxETH?
A quick tour of Defi Llama yields a few insights:
Curve and Convex Concentration: More than 80% of all outstanding FRAX and frxETH are used in farming among Curve, Convex, and Frax.
Little organic yield: Most FRAX and frxETH farms have <0.1% organic yield. In most cases, the rates are made competitive only with CRV/CVX reward APY.
If we look at both the FRAX and frxETH pools, the utilization ratios are a meager 2% and 0.1%. In other pools, the utilization ratio (24h volume/Pool TVL) is generally at 30-50%. Essentially, frxETH and Frax LPs are paid in CRV to park money on Curve.
With this evidence, we can conclude with Frax's Curve flywheel:
User converts USD and ETH into FRAX and frxETH
Park FRAX and frxETH on Curve
Frax uses its voting power to boost Curve's reward to competitive levels
Frax charges some of the yield* into FXS Treasury
FXS Treasury acquires or borrows more CRV and CVX voting power
User and TVL stay for as long as Frax can steer enough CRV emission APY
*: Frax charges some of the yields in mostly indirect ways, for example
Use USDC deposits to buy real-world assets. (proposal)
Stake their USDC in friendly Defi farms.
Charge part of sfrxETH's surplus yield, which is essentially paid for by frxETH holders who choose to farm CRV instead of ETH PoS yield.
An interesting observation from the flywheel:
Token Agnostic: The flywheel is agnostic about which pegged token it boosts. Frax can and does rebalance its CRV gauge votes among its family of pegged tokens. This also means when Frax wants to launch the next token (RWA-flavored FRAX?), it can shift its current gauge votes quite easily. Since 2023, FRAX's market cap has been dropping while frxETH is climbing, but FXS price and attention have been up a lot.
Quantifying the Frax-Convex-Curve Farm
This section further quantifies the status quo and potential of FRAX's Curve farm.
What's Frax's current influence on Curve/Convex?
How much room is left for Frax to farm more CRV emissions?
Implication on Frax's TVL upside.
Current CRV Influence Level: ~35%
Curve‘s gauge vote distribution is relatively stable on a week-on-week basis. Thus we use the vault distribution for the week of 8/17/2023 for our analysis. We describe our methodology below. You may access Curve's source data here and our model here.
We take the biweekly CRV vote gauge data from dao.curve.fi after the vote closes. This makes sure we calculate all votes including those cast with Convex.
We select pools with FRAX, FXS, FPIS, frxETH, FRAXBP, and their wrapper tokens.
We then calculate to sum of the vote share of all selected pools.
The calculation yields a CRV influence sum of 35.1% for all Frax family tokens.
Frax only controls about 6% of all vlCVX, but is somehow able to command 50% of vlCVX voting power and another ~10% of total veCRV votes. We have not yet found a definitive answer as to where the influence leverage comes from. We speculate it is through either or both of
Collaboration with the CVX team and whales
Using vote-rental platforms like Votium to borrow votes
Let's sum it all up in one visualization:
sfrxETH：4.78% base + 2.26% Fraxlend = 7.04%
stETH: 4% base + 4.44% Curve = 8.44%
From the simple comparison, we can say Frax's yield is in line with its bigger competitors. This also implies there is not much space for Frax to further dilute its yield. If Frax wants to increase TVL, it has to farm more CRV as compensation.
With around 80M USD of assets on the balance sheet, we want to do a rough calculation as to how much additional CRV or CVX Frax team may acquire to further boost their CRV farm.
For the sake of simplicity, we assume
(obviously wrongly) The Frax team uses all of their treasury to buy CRV/CVX and causes no FXS slippage in the process.
We assume no CRV/CVX price impact as they conduct buy operations.
We assume they are gonna increase the total vlCVX/veCRV outstanding as they lock new CVX/CRV into vl/ve.
We used August 2023 prices.
The resulting rough calculation shows that Frax can theoretically increase its CRV influence by another 10-18%, resulting in a 45-53% influence upper bound.
We believe this 50% influence is also reasonably an upper bound for 3 reasons:
Curve neutrality: There might be social and community-political repercussions if other Curve ecosystem players believe Frax has unfairly "captured" Curve at the cost and risk of other holders.
Increasing Cost: Frax will bid up CVX/CRV prices if they conduct a large-scale buy operation.
Decreasing Marginal Utility: As Frax is already the biggest player in the CRV/CVX ecosystem, locking more CRV/CVX also dilutes whatever CRV/CVX they already use.
What drives TVL？
(Trigger warning: a bit mathematical. You can jump to the final conclusion in red.)
We view TVL as the most important metric of the health of Frax. Since we have investigated how the predominant driver of Frax TVL is yield farming, we try to derive the mathematical relationship between several key factors in the ecosystem. In this derivation, we make a few assumptions:
We only count CRV emissions as the source of yield. This leaves out some minor yields denominated in CVX or FXS.
We start with expressing Frax's across-the-board additional APY as a relationship between CRV farming yield and Frax TVL. If we plug in today's numbers, we can see the 1.2% resulting additional APY is in line with observation (in reality a bit higher since not all FRAX/frxETH are parked in CRV farms).
We then assume that due to free market liquidity competition in Defi, Frax's additional APY is linear with Defi beta yield.
With some arithmetic manipulation, we have our final formula for Frax TVL.
In human language, Frax TVL is linearly correlated with
CRV Farm Emission Volume
Frax's share of CRV influence
And is negatively correlated with
Defi's beta yield (which competes with Frax for liquidity).
If we want to further conceptually simplify this formula:
We assume CRV price moves together with Defi's beta yield in a bull market (so they cancel each other out)
We assume CRV Farm Emission remain somewhat constant
Then we get: Frax's TVL is linearly related to Frax's CRV influence. It's quite intuitive indeed: the more you can farm, the more TVL you can sustain with the same competitive rate.
Here's the entire process:
In its current form, Frax has about another 50% TVL expansion potential if it can further increase its CRV influence.
Frax is linearly exposed to the CRV token's own upside and downside. This might be especially relevant as CRV token price has been volatile recently.
CRV emission will decrease again in August 2024 from ~163M to ~140M/yr, putting slightly more pressure on Frax.
In totality, we don't really count on Frax to realize another multi-time TVL increase. Instead, we view Frax as a pegged-asset launching platform that can freely shift its focus to whatever is trending in the market. We view it as a cyclical, hype-driven Defi bet (with some CRV exposure).
Research Lead, Maverick Crypto
Researcher, Maverick Crypto
DMs are open.