The P&L of a Decentralized Application

Introduction

The top 50 most valuable protocols all have circulating market caps of at least one billion. Why do these tokens, many of which have unproven business models, trade at such high valuations? One of the core assumptions behind these valuations is that the protocol will someday mature to have product-market fit, justifying its current market value.

For protocols that operate decentralized applications (dApps), assessing profitability is one of the few ways investors can begin to evaluate tokens based on tangible metrics instead of relying on user metrics and topline fee growth. A project with a profitable P&L demonstrates that it is delivering tangible value to its users above its costs, and is moving closer to achieving product market fit. While dApps that continue to rely on unprofitable incentives require a different level of scrutiny from tokenholders.

Most decentralized applications are in the early growth phase, where they must attract users and suppliers to their network through costly token incentives. They may have high user numbers and market caps, but their product market fit is likely distorted by ongoing native token incentives. Users on the network may only be there because they are paid to be, or because the product they use is subsidized through unsustainable supplier payments. While improving unprofitable traction may lead to higher market valuations temporarily, projects cannot operate at loss indefinitely.

Only a rare few dApps have matured enough to retain users and operate profitably. These projects accrue profits to token holders in three ways: token burns, revenue distributions, and DAO accrual. This piece will provide insight into which dApps are currently the most profitable, and the three methods of revenue accrual they use.

Before diving into which types of dApps have reached profitability, we must first understand how P&L looks at the protocol level.

Protocol Expenses

The economics of most dApps are similar to that of a growing start-up. Given the many dApps competing for a growing yet limited pool of crypto users, dApps must utilize incentivizes to remain competitive. Typically these incentives come in the form of the network’s tokens, representing a cost to the network since incentives can expand circulating supply and dilute existing token holders.

Protocol Earnings & Revenue 

Teams establish value accrual mechanisms to offset the cost of incentives to the network. These mechanisms aim to encourage long-term holders through price appreciation or distributions of fees users pay for the protocol’s products.

To understand which protocols are profitable, we must define protocol earnings:

Protocol Earnings = Protocol Revenue – Token Incentives

We mentioned native token incentives distributed to users represent a network’s expenses for customer acquisition. Revenue can be a bit more nuanced based on each network’s tokenomics and fee model. Token Terminal, a widely used data platform in the industry, currently defines revenue under a few buckets:

• Total Revenue: Fees (total amount end users pay to network)
• Supply-side Revenue: Supply-side fees (what network service providers receive from end-user payments)
• Protocol Revenue: Revenue (what the network and token holders receive from end-user payments)

Supply-side revenue can be considered a network expense from a P&L perspective. These fees would have gone directly to the network or its token-holders but are instead used to pay service providers needed to operate the network. Finally, one must subtract the token incentives from the protocol revenue (user fees net of supply-side fees) to calculate protocol earnings (profits).

Fees vs. Earnings

A project may generate substantial fees yet still operate unprofitably from an earnings standpoint. The easiest way to understand the difference between earnings and fees is to look at Token Terminal’s fee leaderboard compared to its earnings leaderboard.

These protocols have verifiably generated these amounts in on-chain fees from their users. Most of the top fee earners are some of the largest public blockchains and dApps, with OpenSea being the only private company in the top 10 where fees go to shareholders.

When ranked by protocol earnings (as defined above), some names drop off, and others enter the top 10. The names that drop off have incentives or supply-side revenue greater than or equal to the fees accruing to the network’s token holders over the last ~ 6 months. Lido and Pancake Swap are unprofitable due mainly to network incentives. Uniswap and Bitcoin operate at break-even from a network perspective, as all fees go to supply-side revenue that accrues to liquidity providers and miners, respectively.

All these top-earning protocols accrue more revenue to token holders than what they pay out in token incentives and supply-side fees. Revenue accrual to token holders is the key component behind protocol profitability.

Uniswap is a great example of a successful product that does not accrue fees to token holders. Its decentralized exchange trails only Ethereum and Tron in total fees but operates at break-even at the protocol level. Despite generating ~$270m in fees over the last six months, none flowed to $UNI token-holders. All fees were paid to supply-side Liquidity providers required to run its DEX. Uniswap is in the early stages of flipping its ‘fee switch’, which would see a portion of revenue from trading fees accrue to Uniswap-related entities.

After the top three most profitable protocols, there is a steep drop in earnings levels, showing how difficult it has been for networks to achieve profitability. The vast majority of protocols operate at a loss for the network and its token holders. Below, we will examine the P&L of several products that have achieved profitability and their respective earnings accrual.

Token/Fee Burns

Similar to how companies use earnings to return value to shareholders through stock buybacks, dApps can implement mechanisms to reduce the circulating supply of their native tokens, thereby potentially increasing the value of the remaining tokens. These mechanisms are often referred to as “token burns.”

Lending

Maker DAO is a lending platform in the top 10 earners with a token burn mechanism. Maker is an over-collateralized lending platform known for its highly successful DAI stablecoin. Users deposit and lock accepted crypto assets into Maker’s smart contracts as collateral in order to mint loans denominated in DAI.

While loans are issued in amounts lower than the deposited value, the appeal of Maker is that they are issued instantaneously without credit checks or proof of income. Additionally, the borrower does not have to worry about market fluctuations expanding their debt, as they must only repay the borrowed DAI plus interest. The borrower may even request more DAI if the value of their collateral rises. In case of downside price action for collateral, the loan is subject to a liquidation ratio, which specifies the USD value the collateral must be above to avoid automatic liquidation for repayment.

Throughout the life of the loan Maker earns fees in the following instances:

• Stability Fee – continuously accruing interest, APY calculated based on DAI generated
• Liquidation Fee – penalty fee charged in case of liquidation on top of the stability fee

The DAI is burned once the principal is returned, and the stability fee paid in $MKR is also burned. This ties $MKRs supply with the amount of demand for DAI loans, with burns constituting revenue to $MKR holders when supply is contracting.

While Maker’s earnings have over tripled what they were 6 months ago, its token $MKR fell 18%. One reason for the underperformance may be that beyond its utility for fees, $MKR also functions as a backstop for recapitalization. Recapitalization happens when the value of its collateral portfolio begins to fall so fast that the protocol can’t recover fair value for its loans from liquidations. To avoid being underwater from the defaulted loans, Maker will mint additional $MKR tokens and sell them for capital needed to close any shortfalls. This represents a dilution risk to existing token holders as the sales expand $MKR supply.

The token also has utility for governance over Maker, giving holders voting rights over strategic decisions for Maker’s operations. Thus the price of $MKR reflects not only fees accruing to the protocol, but also sentiment for proposed changes to its tokenomics and on-going operational risks.

Dividends and Distributions (Real Yield)

Real Yield distributions function similarly to stock dividends, where holders of a protocol’s governance token are entitled to receive a percentage of its revenue. The reason it’s called ‘real’ instead of just yield is that the majority of DeFi yields are emitted from unsustainable farming rewards paid to incentivize network activity. While paying native token yields to jumpstart initial activity on a network has worked in the past, investors have realized many of these incentive programs result in mercenary capital and users who merely use the protocol as long as they are paid. These mercenary users usually sell tokens of protocols that operate undifferentiated products and only have traction due to token emissions, moving on to the next shiniest incentive program after one ends.

While the majority of protocols struggle to maintain traction and become profitable after this initial boost from incentives, differentiated products with sticky users and sustainable economics eventually reach a point where they are able to provide token holders revenue distributions. These distributions are sometimes paid in USD stablecoins or ETH, ‘real’ currencies compared to unproven protocol token emissions. In contrast with the majority of DeFi tokens, whose only utility is a vague form of “governance,” revenue-sharing governance tokens align holders in steering protocol strategy and sharing from the monetary success of the product.

Decentralized Perps Exchanges

Some of the most successful real yield-generating protocols have been decentralized perpetual (perps) exchanges, which our team wrote about in detail in December last year. To recap, perpetual futures products are popular derivative products traders in crypto use to gain leveraged exposure to spot prices. Perps have no expiration dates and are priced according to their underlying spot asset prices. This is achieved through funding rates, which are payments that fluctuate based on market demand for long and short exposure, with one side paying the other based on the deviation of the derivative price and to the spot index. Perps are the preferred product to gain exposure for most crypto traders, with the volume of perpetual futures normally exceeding that of spot for Ethereum and Bitcoin.

Implosions of once dominant centralized perps marketplaces from FTX and growing skepticism of holding assets on centralized exchanges led to a rise in the adoption of decentralized perps exchanges. These operate similarly to non-custodial DEXs, but instead of offering liquidity for different tokens, the marketplace provides exposure to perps derivatives. Similar to other DeFi platforms, perp DEXs must incentivize initial liquidity for their marketplaces through native token incentives. While none of the protocols were profitable at the time the perps deep dive was written, GMX and Gains are now generating enough protocol earnings to offset their token incentives.

To open up decentralized perp positions, traders deposit to the platform’s smart contracts in order to trade on margin. The protocol needs to have the capital to fund these leveraged positions. Funding typically comes from a liquidity pool that incentivizes depositors with yield from the protocol. For GMX, the funding pool (GLP) accepts assets used for swaps and leverage trading , while Gains Network only accepts Maker DAO’s DAI.

Swap and trading fees are paid to both GLP (70%) and GMX holders (30%). GLP is the counterparty to traders on the platform, which means the GLP pool loses value when traders make a profit and rises in value when traders lose. GMX’s revenue comes from trading fees for opening positions and an hourly borrow fee that rises based on utilization of the asset (“Assets Borrowed / Total assets in Pool * 0.01%).

GMX is an excellent example of a protocol that initially gained traction through unprofitable token incentives and reached an inflection point where its product offered enough value to users to stand on its own without excessive incentives. From a weekly earnings perspective, the protocol has been profitable since the second week of the year. GMX’s model for growth will undoubtedly be emulated by future protocols looking to replicate its profitability.

While GMX has cemented itself as one of the top providers for decentralized perps in crypto, the protocol certainly still poses risks as an investment. GMX traders’ cumulative PNL has increased over the last 2 months, which raises concerns about $GLPs sustainability should this trend continue. Additionally, GMX’s fees and growing user base should be taken with a grain of salt as they primarily came from users on Layer 2 Arbitrum. Arbitrum was the beneficiary of widespread airdrop framing, where users flock to the network’s apps with hopes of generating enough activity to be eligible for a forthcoming token airdrop. Now that Arbitrum has airdropped its token, activity may decline since users are no longer airdrop farming. A bet on GMX is a bet that some of this inorganic activity translates to real adoption of Arbitrum long-term.

DAO Accrual

Another form of protocol revenue accrual is earnings flowing exclusively to DAOs controlled by token holders.  Whether this represents actual revenue to token holders is debatable, but we will cover it for completeness. It may be considered revenue because the DAO theoretically controls claims on the protocol treasury, which makes it analogous to a company retaining cash flow to its balance sheet.

The reason it may not be actual revenue accrual is that the token holders only have a vote on how the accrued fees are spent rather than directly benefiting from them. Additionally, control over the DAO’s funds may be centralized to leadership within the development team. While they are supposed to act by what the community decides, there is still a non-zero risk that they use funds in a manner that the community voted against. Regardless, because the funds are allocated for the growth of the community and adoption of the related protocol, while indirect, the benefits are still considered to be accruing to token holders.

Domain Name Services

The only protocol in the top 10 earners with exclusive revenue accrual to its DAO is Ethereum Name Service ($ENS). $ENS is an open-source, decentralized naming protocol similar to the Internet’s DNS (Domain Name System). Instead of registering website domains, users register human-readable addresses corresponding to Ethereum wallet addresses. This helps make wallet addresses more accessible as anyone reading the address reads a single ENS domain such as nick.eth, instead of a string of unreadable keys. Once the address is registered to a corresponding wallet, users can send NFTs or Tokens using the human-readable address. 

The ENS fee model and economics are relatively straightforward. The platform operates on top of Ethereum and does not require validators or miners to secure its network, as transactions are secured by users paying for gas on Ethereum. There is also no need for liquidity incentives or costly emissions to the network. The protocol earns a fee when a user registers, renews or sells an ENS domain name. All fees generated are considered protocol earnings. Thus, this model is the most profitable of the ones discussed. 

While having governance rights over one of the most profitable protocols in crypto sounds great for token holders on the surface, perhaps the lack of direct value flows from token burns or fee distributions have led to $ENS underperforming both $ETH and $BTC since the market bottomed last year. While it’s hard to speculate the rationale behind accruing all value to the DAO, this could be a potential drag on sentiment for the $ENS token.

Conclusion

At their core, dApps can be considered a blockchain’s product, and fees generated indicate demand for that underlying product. But assessing projects based solely on their fee-paying users can be misleading. Some users only use a product because the network incentivizes or subsidizes its usage. In these cases, the protocol is still at a stage where it needs to gain more traction to operate without costly subsidies. Protocols that operate profitably based on their standalone product rather than expensive incentive programs demonstrate strong product-market fit.