YFI: Frameworks for Fundamental Valuation

Andrew Kang, Benjamin Simon, Daryl Lau

Introduction

It would be an understatement to say that YFI has taken the crypto world by storm since its July launch. Notwithstanding some strongly-worded warnings from its creator (“YFI has 0 value” and “I test in prod”), the price of YFI has exploded, peaking at nearly $44,000 per token and a market cap of roughly $1.3 billion.

YFI’s meteoric rise is due, in no small part, to its mystique. YFI’s “fair launch” and lightning fast emission schedule (all 30k tokens distributed in the first week) have endowed the fledgling token with a great deal of memetic power. Moreover, YFI’s early success has galvanized a uniquely strong and active community.

But memes and community governance only scratch the surface. Beneath it lies Yearn Finance, the veritable behemoth that Andre Cronje has been developing for months, and one he continues to upgrade at breakneck pace. Andre has more than executed on his original vision to build an automated stablecoin yield-optimizer, and he has grand plans to construct an impressive financial edifice on top of these sturdy foundations.

It is clear that the YFI token will play a significant role in the future of the Yearn ecosystem. The goal of this article is to propose and explore some key frameworks for valuing the token.

Total Value Locked (TVL)

One common approach to fundamental valuation in DeFi is to look at a protocol’s TVL relative to its market cap. However, we do not believe that TVL is a reliable metric because protocols vary widely with regard to the earnings they can generate from their “value locked”. Instead of TVL, we focus on YFI as a revenue-generating asset for its holders. As it stands, Yearn’s yVaults charge users a 5% performance fee* (applied to yVault revenue, not TVL) 1 and a 0.5% fee on yVault withdrawals. These fees are distributed to YFI holders who stake in the governance contract.2

TVL, while commonly used for valuation, is thus meaningless for our purposes because it accounts for only part of the earnings equation. Valuing YFI by its TVL would be akin to valuing a company by the size of its labor force as opposed to its earnings. A company like Macy’s may employ many workers and may even generate a great deal of revenue, but its margins pale in comparison to a more profitable company like Netflix, even though the latter has much less “value locked,” as it were, in the production process.

Present Price/Earnings (P/E)

Compared to TVL, a present P/E model fares much better as a valuation metric because it treats YFI as the productive asset that it is. Present P/E provides a rough snapshot of YFI’s current income generation relative to its market cap. Here are our present P/E calculations, assuming a 100% YFI staking rate.3

In order to compare YFI’s P/E to those of other DeFi protocols, we must first distinguish between Price/Earnings and Price/Revenue (also known as Price/Sales). This is an important distinction that is commonly mixed up. YFI’s P/R is about 2.5: the $1 billion market cap divided by the $400 million annualized yield to Yearn users. YFI’s P/E is roughly 11.5, which is calculated by dividing its market cap by the annualized income the market cap divided by the annualized income to YFI holders.

For consistency’s sake, we wanted to compare Price/Earnings ratios across DeFi. To do so, we first determined which protocols generate earnings for token holders. Then, one by one, we calculated the present P/E ratios for these protocols, taking into account each one’s unique tokenomic model. To standardize our comparisons, we used a fully-diluted market cap and annualized earnings from the past 30 days, and we also assumed a 100% staking rate for each protocol, as we did in our present P/E for YFI above.

It’s clear from this chart that, even after the recent price surge and subsequent drawdown, YFI remains the leader of the pack vis-a-vis earnings. However, one perhaps surprising finding worth noting is that much of this income is the result of the high yVaults withdrawal volume in recent weeks. We calculated this volume — which came out to around 57% per week over the past 14 days — by finding the weighted average of the daily withdrawal rates for each of the yVaults.4 Even though the withdrawal fee is applied only to yVaults and the fee is only 0.5%, the withdrawal rate is so high that it generates an annualized income of over $66 million in withdrawal fees alone. We posit that the high withdrawal volume is a result of the constantly changing yield farming landscape leading to users frequently moving their assets. By comparison, the 5% performance fee applied on yVault revenue accounts for less than a quarter of total annualized income to YFI stakers.

Withdrawal volume has already begun to slow down, a trend we expect to continue. Ultimately, we anticipate that withdrawal fee income will account for a smaller portion of the income that the Yearn protocol generates for YFI holders, similar to the portion of income that trading exchanges collect on withdrawal fees as opposed to trading fees.

Performance Fees: A More Sustainable Source of Income

If withdrawal fees likely won’t be a primary source of income for YFI holders, what will take their place? Here we believe that Yearn Finance can take a page out of the playbook of institutional investment firms, especially those that, like Yearn, are non-directional and automated.

Comparing crypto and non-crypto asset management platforms. Credit: Maple Leaf Capital

Hedge Funds typically use a “2/20” fee model: 2% annual management fee and 20% performance fee. The former is applied to all assets under management; the latter is applied only to profits made by the fund. Yearn could adopt a similar structure, perhaps foregoing the annual management fee and instead allocating a fixed portion of performance fees to a development fund. In fact, Yearn has already implemented a strategist fee paid directly to developers whose yVault strategies are approved by YFI holders.

As of now, there is a performance fee on yVault revenue, but this model could potentially be expanded to include a performance fee on yToken5 revenue, as well. We believe a range of 5–10% for performance fees would drive significant value to the token and help fund continued protocol development. We also believe that it is low enough that a forked protocol with lower fees would not have a significant competitive advantage.

A robust performance fee approach has the advantage of aligning incentives properly. As Yearn evolves and its user base grows, it is only natural that withdrawal volume will decline relative to TVL. Such a decline would be bullish for Yearn Finance, since it would mean that the protocol has a growing retention rate. However, a decline in withdrawal volume would also, of course, result in less withdrawal fee revenue for YFI holders. By contrast, performance fees allow Yearn and YFI to accrue value in lockstep by compensating YFI stakeholders who participate in governance with a fraction of the protocol’s growing profits.

Forward P/E and DCF

Regardless of whether performance fees are expanded and applied to both yVaults and yTokens, it is clear that Yearn Finance is still in its early stages. Not only are the withdrawal rates subject to change, but the TVL and APYs of the protocol are constantly in flux. Accordingly, valuation models like present P/E are inherently limited because they do not account for future protocol growth and changes.

To fill in this incomplete picture, we explored two future-oriented models: forward P/E (at end-of-year 2020) and Discounted Cash Flow (DCF). For each model, we came up with three “cases”: a base case, a conservative case, and an aggressive case. The cases differ not only in TVL, APY, and withdrawal volume, but also in their performance and withdrawal fee rates. 

We should emphasize that these scenarios are not meant to encompass the entire range of future possibilities. Permutations of the different cases — for example, high TVL growth but low protocol fees — are, of course, entirely plausible. These three cases are simply meant to provide a ballpark orientation for how to evaluate YFI in light of its future potential. 6

Forward P/E (end of year 2020)

Below are the three cases for our end-of-year forward P/E model.

One note before proceeding: because performance fees are currently only applied to yVaults, we have included two different P/E calculations for each case: one with a performance fee applied only to yVault revenue, and the other with a performance fee applied to both yVault and yToken revenue. The cells that incorporate performance fees on yToken revenue are shaded in light purple.

Base Case:

Our base case for the end of year assumes some growth in the TVL of yTokens and yVaults, but it assumes a slight decline in withdrawal rate for yVaults. We expect the yield farming market frenzy to continue for the next few months, but slight decreases in the APYs of yVaults and yTokens. This base scenario keeps the performance fee on yVaults unchanged at 5% and includes a 5% performance fee on yToken revenue.

Conservative case:

Under this scenario, TVL of yTokens and yVaults are held constant and their respective APYs stagnate. Additionally, this bearish scenario incorporates performance and withdrawal fees that are lower than the current rates, and a weekly withdrawal rate (30%) that is less than half of the 14-day historical average of the withdrawal rate from yVaults. In fact, the lowered withdrawal rate in this projection incorporates the historical withdrawal volume from Yearn’s inception until now, implying that withdrawal rates would have to be significantly lower in order to bring the average down from current levels.

Though this scenario is bearish, it’s worth pointing out that the resulting P/E would still place YFI well within the top tier of DeFi protocols.

Aggressive case:

Here we assume that the TVL and APY of both yTokens and yVaults grow significantly by the end of the year. This scenario also assumes heavy turnover, and it incorporates an increased performance fee (10% as opposed to 5%). 

Certain elements of this bullish projection are more feasible than others. For instance, the recently-introduced wETH yVault demonstrated how much capital can still flow into the Yearn protocol: within 48 hours, the wETH vault saw over 125,000 ETH (~$45 million) pour in and the vault has since reached a peak valuation of well over 250,000 ETH. By contrast, the 80% weekly withdrawal rate projection is aggressive, especially in light of the fact that withdrawal volume has been slowing down in recent weeks.

Discounted Cash Flow (DCF)

The forward P/E models above give end-of-year 2020 estimates for YFI’s income generation. But an even longer-term view is warranted so that we can translate earnings into a potential range of token price.

A DCF approach is instructive here. These projections are meant to estimate the general magnitude of discounted cash flow, not to furnish precise valuations.

For our three DCF cases, the starting values are the same as those in the forward P/E cases. However, our DCF projections incorporate TVL and APY growth/decay over time. Even more noteworthy is the fact that all three DCF scenarios assume that withdrawal volume declines significantly over the course of the next few years. In this way, the DCF model differentiates itself by providing a snapshot of YFI’s future valuation once withdrawal volume has cooled off and performance fees make up the lion’s share of income to YFI holders.

Two last notes before diving into the forecasts. First, we assume a discount rate of 30% and terminal growth of 5%. Second, as with forward P/E, we have included two different value summaries for each case: one with yToken performance fees and one without (and again, the cells that incorporate yToken performance fees are shaded in light purple).

Base Case:

Our base case for DCF produces prices for YFI that are slightly higher the all-time high, with implementation of performance fees on yToken revenue implying meaningful upside to the current valuation.

This scenario projects $5.2 billion in TVL by the end of next year, $15.6 billion the following year, and over $48 billion by the end of 2024. However, given the pace at which DeFi is currently sucking in crypto and non-crypto assets, we believe that this TVL growth is plausible, especially for a protocol with the product-market fit that Yearn already has.

Conservative Case:

Our conservative DCF model assumes much lower growth across the board, and the prices that it produces (13.5k and $15.8k) reflect this sober outlook. In this case, as in the other two cases, because of waning withdrawal volume over time, projected income from performance fees eventually far outstrips projected income from withdrawal fees.

What are the factors that would precipitate a stagnation like this? In our view, the most serious hindrance to Yearn’s future expansion (beyond a black-swan event) is if DeFi yields were to drop dramatically. If this happens—and if Yearn’s other component parts fail to gain traction—the protocol may still see TVL and revenue growth, but perhaps not at the explosive pace we have witnessed until now.

Aggressive Case:

Our bullish DCF case yields prices of $241k and $315k, depending on whether a performance fee is applied to yToken revenue.

A TVL of over $150 billion by the end of 2024 is certainly aggressive — that’s almost 3x the current market cap of ETH! — but given the growth of stablecoins & vaults that we have already witnessed and the fact that we have only implemented a fraction of potential strategies that are planned we do not believe that this scenario is out of the question. We also don’t want to forget that tokenized real world assets are beginning to enter DeFi. Yearn’s total addressable market is orders of magnitude larger than its TVL,  and its yet-to-be-released financial primitives could further catalyze growth and network effects.

Conclusion

Valuing YFI is no small task. The Yearn ecosystem is complex, and YFI’s role within it is still not set in stone. The various models that we have outlined in this article produce a wide valuation range as a reflection of the drastically different outcomes that are possible. They also don’t capture the every potential revenue or cost source such as the evergrowing basket of products built for the yearn ecosystem such as ytrade, yliquidate, yinsure, etc.

The primary purpose of this article was to flesh out objective valuation frameworks, and to determine an order of magnitude of potential valuations for various scenarios. However, we will conclude by laying our own cards on the table.

We believe that Yearn Finance already has a unique value proposition: it allows individual customers to passively benefit from the market’s yield-generating dynamics as if they were actively and professionally “rotating their crops.” If Andre keeps innovating at whirlwind speed and the community remains strong and active, Yearn has the potential to build out a robust DeFi ecosystem with a powerful and highly composable suite of products.


Special thanks to Marc Weinstein, Weeb McGee, Andre Cronje, Milky Klim, Simon Tan, Armand Cao, Alex Wearn and others for their valuable help and feedback.

Disclaimer: Nothing in this article constitutes investment advice.

Notes:

  1. Currently the platform charges a 5% performance fee on subsidized gas, which occurs only when a trade is executed.
  2. The 5% performance fee income began to be distributed to YFI stakers as soon as the treasury reached $500k in reserves.
  3. At the moment, only 21% of YFI is staked in the governance contract.
  4. You can view our withdrawal volume calculations on this sheet.
  5. yTokens are the yield generating individually wrapped stablecoins that underly yCRV (yDAI, yUSDT, yUSDC, etc.).
  6. Since it was impossible for us to encompass all the possible scenarios in this research report, we’ve published a spreadsheet with forward P/E and DCF shells for your benefit. Just make a copy of the spreadsheet and be sure to fill in the green-shaded cells with numerical values.