DeFi's Revenue Meta: Fundamentals Matter for Valuations | Felix Machart - Greenfield Capital

Hi guys. Good morning everyone. Hope uh hope everyone's doing great. Um I'm talking today about uh DeFi's revenue meta that fundamentals matter for valuations. Uh we have been looking at data from 2021 to 2025 to see if that's the case and how the market has been evolving.

Just a quick spoiler. Um, so since the beginning of 2024, fundamentals have added uh 8.7 percentage points and explained variance versus just a simple Bitcoin and Epher based model for uh six month market cap growth of DeFi uh projects and um protocol fees and TVL growth have been the most important drivers with protocol revenue being at the at the third place. So very brief words about uh myself and I'm a partner at Greenfield Capital, a European crypto fund investing in uh DeFi infrastructure and consumer. Um here a few of our investments on the on the bottom.

Um and yeah the reason why we have been doing this uh this research is that one of our core D5 thesis next to many other interesting properties is that uh cryptographic availability of real-time financials should should lead to um more efficient markets um due to um lower information asymmetries leading to more efficient efficient uh asset pricing and better risk management. And so yeah, we wanted to basically see if if the market has has been pricing tokens accordingly actually. So we have been looking into basically whether um what what explains DeFi valuations is it uh largely Bitcoin and Ephera prices is it protocol specific KPIs or narratives on social and how has that changed over time horizon and market regimes. here a bit about the methodology we applied. So we have taken a bunch of DeFi protocol 77.

Um we have taken uh Bitcoin and Ether prices as um proxies for for market beta within the crypto market. also capturing um macroflows um and uh and then we've taken a bunch of different onchain KPIs as protocol specific fundamentals for example uh protocol fees protocol revenues TVL users uh transactions and and taking the growth rates of those and then we built a bunch of um random forest machine learning models um the first one being only consist or only taking bitcoin and eer price growth uh the market only model, then a model only taking into account the protocol specific KPIs, a simple last valuation benchmark, um a model with market beta and um Twitter followers for measuring social narrative and then um a combined model of market beta and protocol specific fundamentals uh to basically control for just Bitcoin and Ether price being um a heavy driver of um KPIs like TVL growth and also DEX trading volumes to kind of um kind of separate out what the impact is of the protocol um idiosyncratic uh KPIs. And then we have basically trained the models over 12 months um uh rolling 12 months. And then we have tested them um on the subsequent 12 months basically seeing um what is the explained variance of the the prediction of the model and the actual data and to see basically how um how the market has been evolving and uh based on the uh feature importances and also the the um performance. And we've taken different time horizons 1 month, 3 month and 6 month uh growth rates to see how long does it take for pro uh the fundamentals to be reflected.

And um yeah, one thing to highlight is we didn't try to predict the future. We rather wanted to see to what extent um market participants have priced in fundamentals over time. So we looked at um the same uh periods of the fundamental KPI growth and uh market cap growth. So here one of the key findings uh being that over a six-month time period fundamentals actually significantly I would explain um the simple market beta model on this chart you can see basically on the x-axis the different evaluation periods the 12 months each towards the end it's getting shorter up to a minimum of three months to incorporate the latest data as well and then each uh bar shows the the additional explained variance in terms of pear squared um of like the fundamentals models over over the uh simple market beta model and when we can see actually um yeah there is a significant uplift um and um it has to some extent increased over time and and that that pattern actually has held as well over different market regimes like bull bear and flat markets here uh one month time period we can see that um there is very little uplift or even negative at times. So at the short time frame market beta dominates at 3 months we have a medium uplift 2 percentage points on average.

Um and then here uh 6 months again we can see basically that uh it takes several months for fundamentals to really play out to make a big difference in terms of um the additional variance being explained um by by fundamentals over just simple market beta. Um so that being the key uh second key finding um takes actually yeah quite a significant time for real-time financials being priced in. So uh despite despite real-time financials, there are significant time lag and so question is um does that mean defa markets are very inefficient um due to maybe slow information dispersion and decision-m time lags and I guess to some extent that's the case but I think it it can also um basically indicate that um that there's rational behavior actually that investors market participants actually uh wait to what extent and uh a certain um KPI momentum is sustainable and and uh uh continued versus just being a blip in the pan and and a short um a short spike and and so investors probably are waiting for persistence um to to really uh price uh the fundamental uh KPIs into valuations. here the most important drivers. um the we have basically shown the the pre204 um KPIs uh feature importance on in the background and the foreground it's from 2024 onwards uh with TVL always having being historically the the most important still the most important but it has um ceased in importance uh in in the more more recent periods um where you can see like in the foreground the full colored bar being the the later periods protocol fees and protocol revenue news having gained um significantly in importance.

So now monetization matters much more than before um uh over for example like more leading indicators like uh user growth, daily transaction growth also to I guess that are also more easy to be manipulated with incentives um since yeah it's rather cheap to just like spin up an account and and perform some transactions um with like the kind of hard financials being uh more difficult to to be pushed with incentives and and to be faked and yeah a clear clear signal basically that the market has been evolving towards pricing in uh real monetization over time. So we believe actually that um indicates a maturing DeFi market. Um that revenues matter still take several months for for them to be priced in. What are implications for builders? Um sustainable fundamentals do pay off.

So it does make sense to really focus on uh sustainable fee fee revenue protocol re protocol fees fee growth versus just mercenary activity. uh and right now topline growth P revenue has been more important than uh than the revenue metric that's basically accounted for to be like really uh clearly attributable to token holders and but I think over time we will see an increasing importance actually in the latest data we do see an increasing importance of of protocol revenues over protocol fees um and and since we there is actually the significant time lag and sometimes obviously qualitative factors is also matter uh to a great extent. it it does make sense to um to help investors to interpret the data to to also basically be integrated in the various uh data providers uh publish KPI dashboards yourself because sometimes it's not entirely straightforward to parse the data directly on chain run earnings calls regularly to explain the latest trends um strategic positioning give some context to basically increase the confidence in the onchain onchain KPIs that that are visible to potentially have Basically investors price it in uh more quickly, maybe attain higher multiples or get provide some um some some forward guidance as well. um uh which can can be beneficial and uh for investors it means there is uh increased under underwriting clarity uh since there's a much clearer relationship between um monetization and and valuations these days and and that should actually open up uh the gates for further institutional capital coming into the space legitimizing the space which should benefit valuations the overall space and and just default projects generally and Um at the same time the timelines indicate that there's significant alpha potential um to be early to a trend to basically spot uh breakouts and in in revenue growth for example on chain before before others can can see them and uh and interpret them correctly in terms of like to to see whether a certain um certain tra certain traction is going to be sum sustainable or not or is it just a incentive driven fad and and so Yeah, I think like still you could say uh quantitative signals obviously increasingly matter but at the same time you need to combine it with just qualitative assessment, strategic positioning and basically because it's mostly about the future uh and um to kind of build a holistic picture here. If you scan the QR code you can find uh the full report and uh my contact detail.

feel free to reach out, follow me Twitter and um happy to answer any questions. I think there's not super much time left. Um but yeah, thanks a lot. So, so what we've looked at here is is all basically fundamentals that are reflected on chain based on Arteimus data and and so those basically if you look at uh the revenue metric that has been that we've used here that's directly attributable to token holders like uh revenue that acrews to a treasury that goes into uh uh revenue share or buybacks and so any labs revenue um is separate to that and I mean I guess going forward it's going to be I think yeah very important to kind of really try to align incentives and have like I mean I guess we see that with AVA more more recently basically like I guess you have you have to really I think work on on better governance models of how to to make sure that um whatever you invest into future growth is really benefiting um token holders after all and I think yeah generally like we're looking also a bit into buybacks. I think they don't really make so much sense.

They make sense because there's not so much trust that actually earnings will really benefit token holders, but in the end that that should be the case. And I guess it's that's a bit uh dynamic. So I think time is over. Thanks a lot.