Under the decline of liquidity, where does the protocol's income come from? Could token buybacks and burns be the answer?

CN
PANews
Follow
14 hours ago

Author: Joel John, Decentralised.co

Translated by: Yangz, Techub News

Money governs everything around us. When people start to discuss fundamentals again, the market is likely in a precarious situation.

This article poses a simple question: should tokens generate revenue? If so, should teams buy back their own tokens? Like most things, this question does not have a clear answer. The path forward needs to be paved with honest dialogue.

Where does protocol revenue come from in a liquidity recession? Could token buybacks and burns be the answer?

Life is just a game called capitalism

This article is inspired by a series of conversations with Ganesh Swami, co-founder of the blockchain data querying and indexing platform Covalent. The discussions covered the seasonality of protocol revenue, the evolving business models, and whether token buybacks are the best use of protocol capital. This also serves as a supplement to my article written last Tuesday about the current stagnation in the cryptocurrency industry here.

Private capital markets, such as venture capital, always oscillate between liquidity excess and liquidity scarcity. When these assets transform into liquid assets and external funds continuously flow in, industry optimism often drives prices up. Think of various new IPOs or token launches; this newly acquired liquidity allows investors to take on more risk, which in turn fosters the birth of a new generation of companies. When asset prices rise, investors shift funds towards early-stage applications, hoping for returns higher than benchmarks like Ethereum and SOL.

This phenomenon is a characteristic of the market, not a problem.

Where does protocol revenue come from in a liquidity recession? Could token buybacks and burns be the answer?

Source: Dan Gray, Chief Researcher at Equidam

The liquidity in the cryptocurrency industry follows a cyclical pattern marked by Bitcoin halving events. Historically, market rebounds typically occur within six months after a halving. In 2024, the inflow of funds into Bitcoin spot ETFs and Michael Saylor's massive purchases (spending a total of 22.1 billion dollars on Bitcoin last year) have created a "reservoir" for Bitcoin. However, the rise in Bitcoin prices has not spurred an overall rebound in smaller altcoins.

Currently, we are in a period of tight capital liquidity, with capital allocators' attention dispersed across thousands of assets, and founders who have been working on token development for years are struggling to find meaning in it all. "Since launching meme assets can bring more economic benefits, why bother building real applications?"

In previous cycles, L2 tokens enjoyed a premium due to perceived potential value, supported by exchange listings and venture capital. However, as more participants flood the market, this perception and its valuation premium are being eroded. The result is a decline in the value of L2 tokens, which in turn limits their ability to subsidize smaller products with grants or token revenue. Additionally, overvaluation has forced founders to confront the age-old question that plagues all economic activities: where does revenue come from?

How Cryptocurrency Project Revenue Works

Where does protocol revenue come from in a liquidity recession? Could token buybacks and burns be the answer?

The above image illustrates the typical operation of revenue in cryptocurrency projects. For most products, Aave and Uniswap are undoubtedly ideal templates. These two projects have maintained stable fee revenues over the years, thanks to their early market entry and the "Lindy effect." Uniswap can even generate revenue by increasing front-end fees, perfectly reflecting consumer preferences. Uniswap is to decentralized exchanges what Google is to search engines.

In contrast, the revenues of projects like Friend.tech and OpenSea are seasonal. For instance, the "NFT summer" lasted for two quarters, while the speculative frenzy in social finance (Social-Fi) lasted only two months. For some products, speculative revenue is understandable, provided that the revenue scale is large enough and aligns with the product's intent. Currently, many meme trading platforms have joined the club of over $100 million in fee revenue. This scale of revenue is typically only achievable for most founders through token sales or acquisitions. For the majority of founders focused on developing infrastructure rather than consumer applications, such levels of success are rare, and the revenue dynamics of infrastructure differ from this.

Between 2018 and 2021, venture capital firms provided substantial funding for developer tools, expecting developers to gain large user bases. However, by 2024, the cryptocurrency ecosystem has undergone two significant shifts:

  1. First, smart contracts have achieved infinite scalability with limited human intervention. Today, Uniswap and OpenSea no longer need to proportionally expand their teams based on trading volume.

  2. Second, advancements in large language models (LLM) and artificial intelligence have reduced the investment demand for cryptocurrency developer tools. Thus, as an asset class, it is at a "liquidation moment."

In Web2, API-based subscription models work effectively due to the vast number of online users. However, Web3 is a smaller niche market, with only a few applications capable of scaling to millions of users. Our advantage lies in the higher average revenue per user. The nature of blockchain allows funds to flow, and ordinary users in the cryptocurrency industry often spend more money at a higher frequency. Therefore, in the next 18 months, most businesses will have to redesign their business models to directly generate revenue from users in the form of transaction fees.

Where does protocol revenue come from in a liquidity recession? Could token buybacks and burns be the answer?

Of course, this is not a new concept. Initially, Stripe charged per API call, while Shopify charged a flat fee for subscriptions, but both platforms later switched to charging a percentage of revenue. For infrastructure providers, the API charging model in Web3 is relatively straightforward. They erode the API market by competing on price, even offering free products until a certain transaction volume is reached, after which they begin negotiating revenue sharing. Of course, this is an ideal hypothetical scenario.

As for what the actual situation will be, Polymarket serves as an example. Currently, UMA protocol tokens are tied to controversial cases and are used to resolve disputes. The more prediction markets there are, the higher the probability of disputes, directly driving demand for UMA tokens. In this trading model, the required margin can be a small percentage, such as 0.10% of the total bet. Assuming a $1 billion bet on the presidential election outcome, UMA could generate $1 million in revenue. In this hypothetical scenario, UMA could use this revenue to buy back and burn its own tokens. This model has its advantages but also faces certain challenges (which we will explore further later).

Besides Polymarket, another example using a similar model is MetaMask. Through the wallet's embedded exchange feature, there has been approximately 36 billion dollars in trading volume, with revenue from the exchange business exceeding $300 million. Additionally, similar models apply to staking providers like Luganode, which can charge fees based on the amount of staked assets.

However, in a market where API call revenues are decreasing, why would developers choose one infrastructure provider over another? If revenue sharing is required, why choose this oracle service over another? The answer lies in network effects. Data providers that support multiple blockchains, offer unparalleled data granularity, and can index new chains faster will become the preferred choice for new products. The same logic applies to trading categories like intent or gasless exchange tools. The more blockchains supported, the lower the cost and faster the speed, making it more likely to attract new products, as marginal efficiency helps retain users.

Token Buybacks and Burns

Linking token value to protocol revenue is not a new idea. In recent weeks, some teams have announced mechanisms for buying back or burning native tokens based on revenue proportions. Notable mentions include Sky, Ronin, Jito, Kaito, and Gearbox.

Token buybacks are akin to stock buybacks in the U.S. stock market, essentially a way to return value to shareholders (token holders) without violating securities laws.

In 2024, the funds allocated for stock buybacks in the U.S. market alone reached approximately $790 billion, compared to just $170 billion in 2000. Before 1982, stock buybacks were considered illegal. Over the past decade, Apple alone has spent over $800 billion repurchasing its own stock. While it remains to be seen whether this trend will continue, we observe a clear differentiation in the market between tokens that have cash flow and are willing to invest in their own value and those that have neither.

Where does protocol revenue come from in a liquidity recession? Could token buybacks and burns be the answer?

Source: Bloomberg

For most early protocols or dApps, using revenue to buy back their own tokens may not be the optimal use of capital. A viable operational approach is to allocate sufficient funds to offset the dilution effect caused by new token issuance, which is precisely what the founder of Kaito recently explained regarding their token buyback method. Kaito is a centralized company that incentivizes its user base with tokens. The company generates centralized cash flow from enterprise clients and uses a portion of that cash flow to execute token buybacks through market makers. The number of tokens repurchased is twice that of newly issued tokens, thereby putting the network into a deflationary state.

In contrast to Kaito, Ronin employs a different method. The chain adjusts fees based on the number of transactions per block. During peak usage, a portion of the network fees flows into the Ronin treasury. This is a way to monopolize asset supply without repurchasing tokens. In both cases, the founders have designed mechanisms to link value with economic activity on the network.

In future articles, we will delve into the impact of these operations on the prices of tokens involved in such activities and their on-chain behavior. But for now, it is evident that as token valuations decline and venture capital inflows into the cryptocurrency industry decrease, more teams will have to compete for the marginal funds flowing into our ecosystem.

Given the core attributes of the blockchain "currency orbit," most teams will turn to a revenue model based on a percentage of transaction volume. When this happens, if the project team has already launched a token, they will be motivated to implement a "buyback and burn" model. Teams that can successfully execute this strategy will emerge as winners in the liquid market, or they may buy back their tokens at extremely high valuations. The outcome can only be known in hindsight.

Of course, there will come a day when all discussions about price, yield, and revenue will become irrelevant. We will continue to pour money into various "dog meme coins" and purchase all sorts of "monkey NFTs." But looking at the current state of the market, most founders concerned about survival have begun to engage in deep discussions around revenue and token burns.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink