Disclaimer: This is not financial advice. Anything stated in this article is for informational purposes only and should not be relied upon as a basis for investment decisions. Triton may maintain positions in any of the assets or projects discussed on this website.
TL;DR
Introduction
Throughout this series, we have frequently mentioned that the often-voiced narrative that digital assets have no intrinsic value is an absurdly outdated trope given how far development has come. In part VII we laid out a high-level framework against which one can view most digital assets and in this post we will take a closer look at crypto-native projects to reveal just how much ‘underlying’ intrinsic value there is in some of these projects.
As a warning, this will likely be a serious narrative violation from what one typically hears on CNBC or Bloomberg; one well captured by the quote at the top and voiced by a host on Bloomberg’s Surveillance podcast (53:10). Thankfully, the guest on that show, David Bailin (former CIO at Citi) has evolved his view on digital assets over the years:
“You know, I have to say that, um, six years ago I was a wild skeptic about the digital world and about cryptocurrencies in general and I don’t like even calling them that but there are a handful of companies that are online today that I think are going to be game changing. They are literally gonna be the next generation of banks or the next generation of lending and so I think that we’re ignorant if we don’t think we should give our clients exposure to that”
Specifically, we will focus on two token types – native network tokens and DeFi tokens – to provide a better sense of how an investor should think about each type of token in terms of value capture and then, more importantly, highlight just how much economic value some of these protocols are actually creating. And not to bury the lede, there is some immense value creation at work.
Network Value Generation
Likely this class of tokens is what most are familiar with. Bitcoin, Ethereum and Solana are the most prominent examples of these, and each has their own native token that is used to transact on the network – BTC, ETH and SOL, respectively. We won’t spend more time defining what they are here, but interested readers can review our earlier post exploring these in more detail.
As with any network business, one should begin framing a valuation based on Metcalfe’s Law, coined by Robert Metcalfe in 1980 to explain how the value of a telecommunication network grows exponentially relative to the number of nodes on the network (e.g. V ≈ θN2). At their cores, blockchains are network effect businesses and this provides a strong place to start. While the size of the network grows, from a financial perspective, there is another factor far more important: the ability to sustainably capture rent from that network, θ. To do this, blockchains by and large charge a usage fee to transact on the network. The implementation and model of these differs across most but generally consist of a base transaction fee and some form of priority fee. Others, like Ethereum, also have explicit fee markets for other chains to use when posting data, in essence expanding Ethereum’s settlement capabilities to those chains.
Most basically, one can simply think of these networks as selling blockspace. Their customers (users) pay the network in order to have their transactions included in the network’s blocks, thus allowing their desired position to be updated in the new state of the network. And because there is a limited number of blocks any one network can provide, certain networks can command premiums over others. This can show up as higher fees paid at a base level, but because development is still so early and these networks continue to optimize on throughput, one more often sees this premium appear in the ‘priority’ fees users pay the networks above and beyond the base fee to transact for ‘top-of-block’ inclusion or to have certain transaction bundles added.
Generally, the base fees to transact on most networks have been increasingly compressed to just fractions of a cent. The chart on the left below shows that of the most popular chains, Tron consistently has the highest median fee to transact, typically around $3.00. This is a massive outlier, and the next most expensive chain is Ethereum, nearly two orders of magnitude cheaper at just $0.16 per transaction. From there, BNB has the next highest, another order of magnitude cheaper, at just $0.014 per transaction. If we remove all three of those highest-cost chains, we get the chart on the right. Here you can see that essentially every chain now has median transaction fees of just $0.001 or less. For the average transaction, these are now essentially free for all intents and purposes.
However, the cost of a transaction varies based on its complexity, with simple sends being the lowest baseline cost and more complex interactions like multi-hop token swaps or contract creation being more computationally heavy, thus requiring more expensive transactions (though still remaining in the essentially-free range for most chains). The economic value that chains generate is most reductively a product of the number of transactions and the cost to transact, modulated by the complexity of those transactions and the willingness for a user to pay extra for their transaction to be included quicker. Below shows the growth in transactions for Ethereum and the biggest EVM L2s, an ecosystem facilitating over 24 million transactions per day now. Base (Coinbase’s own L2) is by far the largest driver of this, accounting for nearly 9M transactions on its own. So as long as transactions count growth outpaces fee compression, these networks will continue to see increased economic value generation.
One metric that attempts to define this rent capture is called ‘Real Economic Value’ (REV) by measuring the value accrual generated by user activity. It consists of the transaction fees paid by all users (individuals, applications, etc.) as well as any out-of-protocol fees paid for transaction inclusion, typically in the form of validator tips or priority fees that users pay for expedited inclusion. These fees are paid to the validators that help run the network, and a substantial portion of which is then passed through to entities that stake the native token of these networks with those validators (90%+). In some cases, such as with Base, there is a single sequencer that orders transactions rather than a decentralized set of validators. With Base, that single sequencer earns all fee revenue that users pay. And it can be lucrative: Coinbase regularly earns over $200,000 in profit per day from its Base sequencer and has seen its monthly earnings spike to almost $17M in periods of high market activity. Coinbase’s peak 1-day profit from Base was $1.4M, straight to the bottom line (well, 99% net margins, but close enough). You can easily track this in real time on analytics sites such as growthepie.xyz and as a result, Coinbase may be the only company in the S&P 500 with real time public visibility like this into its financial and product performance. With data so easily available and transparent, periodic 10Qs start to feel antiquated.
However profitable Base is, Solana undoubtedly leads REV creation by far, generating roughly $3M per day right now. This is down considerably from its January 20 peak, when the network generated $50M in REV on the day President Trump launched his own memecoin. That is $50M in real yield that flowed through to SOL stakers and validators, in a single day. A keen reader will note the many spikes in the chart below. This highlights the volatility in REV generation on the networks, which can reach eye popping numbers at times of acute heightened demand. Importantly though it also points to the REV generation potential of these networks at a future state when activity reaches these levels in a sustainable way. Blockchains currently produce blockspace in amounts that far outpace demand, and as such, it remains relatively cheap to access most of the time. On top of REV, most networks also pay out a low-single-digit inflation issuance as well, primarily as a way to incentivize validator activity as a bootstrapping mechanism. Combined, yields can frequently reach 10%-15% for stakers on Solana. As these networks mature, one should expect REV generation to account for an increasingly higher portion of that yield as net new issuance reduces over time while activity increases.
Bitcoin, likely counter to what most would assume given its market value and prominence, generates relatively little REV. It is far more unique in its position as a buy-and-hold store of value asset and as such, transaction activity is far lower on the network than that of Ethereum or Solana (e.g. you spend cash, not gold). Median transaction fees on Bitcoin are ~$0.50 while the mean skews mildly higher to $1-2 with median transaction values of ~$450 and mean transaction values of $200,000. But importantly, though overall network transaction volumes are high ($40-60 billion per day), the number of transactions is relatively low (300K per day vs. 60M non-vote transactions on Solana, for example) and thus lower REV generation, to the tune of $500k-$1M per day. For Bitcoin, however, that is okay given its proof-of-work security model via new BTC issuance to miners (at least for the foreseeable future). Proof of stake networks such as Ethereum, Solana and essentially every other major network, are much more reliant on the long-term REV generation to incentive validator participation and staking and as a result, maximizing network throughput is the name of the game.
Fat App Thesis
Network value creation is just one way to look at digital assets. Though the native tokens of these networks represent the largest market caps, by far, applications are beginning to generate a far greater share of digital asset revenues. For example, whereas Solana generated $122M in REV over the past 30 days, apps on Solana generated an additional $227M. In aggregate, that amounts to nearly $350M in value generation on Solana alone over the past few weeks. Ethereum ($50M) and Hyperliquid ($50M) represent the next two largest chains for app revenue generation. We’ll take a closer look at a few apps in particular to better understand the economics at work here.
What’s the, uh, quarterly cash flow of a Hyper Liquid?
Well for starters, looking for quarterly reports is a relic of the past for crypto businesses because we can track these cash flows in real time. Hyperliquid, the top perpetual futures exchange, consistently earns a few million per day in revenues ($8-15M per week), putting it on pace to earn just shy of $600M this year. Pretty impressive for a less-than-2-year-old company and one that was entirely bootstrapped with zero private funding or outside capital.
But here’s the real kicker: because Hyperliquid has such a lean team and operating structure, it passes through 97% of revenues to HYPE token holders by buying back tokens off the open market. To date, Hyperliquid has purchased nearly 23M HYPE tokens valued at $596M at current prices. To put it plainly, a $600M-per-year business with a 97% net margin that passes through 100% of earnings as buybacks has no equivalent in traditional financial constructs. Full stop.
But to answer the question of “a Hyper Liquid’s quarterly cash flow”: $130M in cash to token holders in 1Q 2025. And we already know that there has been $75.7M in cash to token holders so far in 2Q, on pace for another ~$130M or so.
Beyond perps, lending is one of the core financial primitives that has been built as decentralized protocols on these chains. The leverage and liquidity that they provide are, along with decentralized exchanges, the lifeblood of most ecosystems. And as a result, there are numerous highly successful protocols that are printing cash.
Aave is by far the largest and has been slowly building up the tagline ‘Just use Aave’ due to its prominence. Currently, there is over $40 billion in assets deposited in Aave smart contracts across all chains. Aave has earned over $210M in revenue since going live and based on the last 30 days has a current run rate of $52M. Its annual OpEx is $18M, giving $34M in profits and a net profit margin of 65% for token holders.
To better put in perspective just how large Aave is, if it were a US commercial bank, it would nearly be in the top 50 in total assets, slightly ahead of Deutsche Bank and closing in on Raymond James and Barclays.
Kamino is another lending protocol, similar to Aave but on the Solana network. It is not nearly the same size (just $2.5 billion in assets on its platform) but is currently on pace to earn $27M in revenue this year. As for who is using Kamino? Well, Apollo Global’s tokenized credit fund asset ACRED is now available on the platform (yes, that $500B AUM Apollo Global).
Maple takes a more ‘traditional’ approach to lending and while it has similar DeFi aspects to Aave and Kamino, it also has a core credit team that provides overcollateralized lending to institutions. Maple is smaller than Aave or Kamino but is growing rapidly. It now has nearly $1.7 billion on its platform (with almost-vertical growth in 2025) and from that, looks to generate nearly $15M this year based on current asset levels. Given the growth and product expansion, one could see this breaking $20-$25M very easily (it was trading at $100M market cap just a few weeks ago but is now up to $400M). Like Hyperliquid, Aave and Kamino, all value created goes to Maple’s SYRUP token holders. (Yes, Maple’s token ticker is SYRUP, something your Canadian author very much appreciates)
There are Caveats
Now, there are major differences in the nature and quality of revenue between onchain protocols. Given the rapidity of capital flows and nature of these protocols (e.g. self-custody with no vendor lock-in), it is far harder to treat all revenue as equal.
It's okay, you can call digital assets intrinsically valuable in 2025
If there is one takeaway from all of this, it is that anybody viewing digital assets in this day and age needs to immediately discard the increasingly luddite perspective that these tokens have no purpose, intrinsic value or underlying cash flows. We have highlighted just a few in this article (collectively generating hundreds of millions in quarterly cash to token holders), and there are many, many others that fit the bill in a similar manner. It is true that Bitcoin has little intrinsic value in the traditional cash-flowing sense, much like gold, but it frankly does not need to, much like gold. Capturing cash flows is not the purpose of BTC the asset, and as such, it is not valued in that way. But for other tokens, one can and should look at the underlying metrics to value them. And the industry has reached the point where those underlying metrics are real and there are numerous investable businesses with tokens that directly capture that value and with operating efficiency that make even the best run ‘traditional’ companies blush.
If you, the reader, still hold the view that all of crypto is essentially just BTC and worthless memecoins, it is a good time to reconsider your priors and evaluate the space with a fresh set of eyes. It is still a nascent industry with many, many warts, but it is also important to understand that there are real, valuable, explosive crypto-native businesses out there, with more and more hitting the market every single day.
Digital assets in 2025 generate real, scalable cash flows—via tokens, DeFi apps, and network usage—proving clear intrinsic value beyond outdated narratives.
How crypto’s modular architecture unlocks powerful financial primitives like perps, yield-bearing stablecoins, and tokenized real-world assets—reshaping the future of global finance.
Exploring quantum computing's potential threat to crypto.