This is going to be a rant about revenue. When I write about PE (price/earnings), all the work goes into calculating revenue. I’d like to discuss some of the choice points I have made deciding what counts as revenue in my PE calculations using examples from the ecosystem.
- Does all protocol revenue count or only revenue that benefits token holders?
When evaluating a token I’m evaluating the benefits I personally accrue for owning that token, not the benefits other participants in the ecosystem accrue from the platform. For example, when I look at tokenterminal for Aave right now they show a P/S ratio of 11.96. When I last did a calculation I came up with something closer to 2000. So, did I miss 99% of the revenue in some glaring oversight? Not exactly. Tokenterminal is calculating all the lender revenue in their P/S value. But that doesn’t benefit me as a token holder. I can decide to use a platform such as Aave or Alchemix without deciding to be a token holder. Thus my opinion about a platform as a user can be very different than my opinion about the financial opportunity of its token.
- Different classes of stakeholders?
Let’s take something like KNC. There are different classes of stakeholders. Stakers earn staking revenue (in ETH) that other KNC holders such as AMM LP’s don’t earn. When calculating PE, I’m calculating PE specifically for the class of user I would fall into. Taking KNC staking as an example, all holders benefit from the 6.2% burn but only stakers get the 62.7% of the protocol fees. If you plan to be a staker you need to calculate the staker revenue, which requires calculating what percent of the KNC is staked, projecting trading fees, etc.
As a different example, RPL has inflation, but not everyone receives that inflation equally. They basically have an elite council that receives extra inflation. If you aren’t part of this elite group that issuance is a debasement cost to you personally. I’m not saying this is bad incentive design, not all participants in an equal system provide equal benefit and in a well designed system participants are rewarded according to the benefits they provide. I’m only saying that you need to be aware of the different classes of stakeholders and be aware of how you are going to be treated by that ecosystem as part of that class.
- If a governance token manages a treasury, does treasury revenue count, or only the dividends?
I’d like to think a token should be worth at minimum the value of the treasury it manages. This point came up when the DGX DAO decided to dissolve itself when its token was worth like 1/3 of its treasury. It’s also a more nuanced point than you might think. What does a treasury consist of? Is it just funds that could be taken by whomever holds the admin key? Even if AAVE stakeholders could vote to dissolve AAVE and take all the deposited funds (I’m not sure they physically can) I wouldn’t call those funds a treasury. Even in a more benign case, if treasury funds are earmarked for something other than me the token holder, should I count that in revenue? Back to point 1, maybe not. So, if Maker has fees earmarked to a surplus buffer, should we count that as revenue in PE? How about if Alchemix has revenue earmarked for a transmuter to defend the alUSD peg? In the former case, they have it structured as a temporary measure. In the latter case they don’t but I can certainly imagine a future where the transmuter black hole creates a reserve so large that ALCX voters vote for a cut of it. When I calculate the MKR PE I am including the revenue that goes to the surplus buffer because of a notion I call time horizon. In the latter case in my ALCX post I do show the PE if the transmuter revenue was considered for PE but I think it’s more fair of a calculation if it isn’t included. In other cases like YFI where all treasury overflow eventually gets used for YFI buybacks, I include most revenue as a slush pool. It’s a nuanced point.
- What time horizon do we estimate revenue on?
When calculating something like trading volume for Uniswap the volume looks exponential. So, do you do an exponential projection of volume or a linear projection of volume? Even in the former case, how far back do you go when calculating average volume? The further back you go the lower the average volume will look, but too short a horizon is equally dishonest.
This concept of time horizon comes up in other ways. If there are short term costs such as something like BREAD that is paying off a hack harvest once had do you include that as a cost in the PE? How about something like the surplus buffer in MKR which should fill and then be done? How about issuance which is known but decaying over time? Something like FARM issues 4% less of itself every week. Do you project the debasement cost out 30 days, 90 days, a year, or an average of the next several years? If deciding to hold you might model it based on the time horizon of your hold. If deciding to buy, I’d personally wait until shorter time horizons indicate the system is in profit right now. In the case of farm above I did a calculation projecting out over 30 days and then a year to make the case that I could get excited about the coin but not until debasement pressure had died down a good bit first. It was something to watch. And it’s a good thing I watched instead of bought because it’s down from $410 the day of that post to $55 today.
There are a lot of these little points that come up across the ecosystem and any calculation you find of PE is going to have some opinionated answers. The best I can do for you all is being transparent about my calculation, which is something I don’t see from tokenterminal or theblock.