From the era of "gambling" to "investment," the institutional and AI-driven crypto market pricing logic is being rebuilt.

Written by: Castle Labs

Translated by: Glendon, Techub News

2026 has not started well for cryptocurrencies. Most asset prices are falling; Bitcoin reached its all-time high six months ago and has since been steadily declining. Recently, there has been no positive news, ETF funds continue to flow out, the market has lost interest in cryptocurrencies, companies are going bankrupt, and venture capital firms are no longer actively investing. It seems the once abundant source of opportunity in the crypto space is drying up.

Although these conditions are true and the current situation lacks optimism, we are heading toward a major shift: tokens unrelated to protocol revenue will significantly depreciate, and projects without income support will struggle to survive. The “harvesting” behavior in the crypto world is shifting from “gambling” to “investment.”

The event accelerating this change was the liquidation incident last October, followed by a series of macro events, such as gold outperforming Bitcoin significantly, prompting us to question: do crypto assets still have investment value? Do they still hold the upward potential that initially attracted people? This article will focus on explaining this transition and its impact on crypto assets and underlying investment models.

From “Gambling” to “Investment”

Cryptocurrencies have gone through several stages: the initial exploration phase, where it was just a tech toy for geeks with no one seeing its practical use; the ICO boom characterized by extreme speculation; a period of regulatory absence marked by major events like Luna’s collapse and FTX’s explosion; and the current phase of gradual institutional entry.

For a long time, the crypto industry has been dominated by a “value extraction first” model, creating a norm of gambling rather than investing. Products like pump.fun that allow users to quickly create meme coins illustrate that the market has long been in a gambling bubble, with new users pouring in just hoping to get rich overnight. This “value extraction” model can be divided into three types:

Low investment, low output (Memecoins)

High investment, high output (scam projects and slow “exit” projects)

Low investment, high output (celebrity coins)

Among these, the simplest method is Meme Coins, which remain effective today and may continue to be so in the future, though their growth rate might slow. Meme Coins are easy to issue, with no need to explain their purpose or reason for issuance, because the core profit motive is only one: exit before others do. Participants in Meme Coin trading are usually aware of this, and losses in some cases are just natural market outcomes. Another category involves projects that overpromise, hype heavily, and then slowly “exit” after a period of hype. There are also a few exceptions, such as projects that maximize value extraction with minimal investment, like celebrity coins.

Taking last year’s token generation event (TGE) as an example, most projects were ultimately classified as poor investments because they resulted in significant losses for token holders. Reasons for price declines include poor tokenomics, overvaluation at issuance (most cases), and market or project sentiment downturns.

For a long time, crypto projects focused on building top-tier technology but neglected achieving product-market fit (PMF), leading to these technologies being unused. But since 2026, the situation seems to be changing. As institutions gradually go on-chain, the “value extraction first” model is gradually receding. Institutions want to leverage the infrastructure built over many years in crypto, but with an important premise: they are not interested in the tokens issued during our technical development. They appreciate the code and underlying architecture, will use them, but this does not necessarily have a positive impact on most existing assets.

Previously, the New York Stock Exchange (NYSE) announced it would adopt blockchain infrastructure to support 24/7 trading. Robinhood has begun testing a layer-2 network based on Arbitrum technology for tokenized stocks and ETFs, allowing users to hold “stocks” in self-custody wallets. BlackRock’s BUIDL and Franklin D. Roosevelt’s Benji are currently excellent on-chain real-world asset (RWA) products. These innovations enable instant settlement, solving long-standing issues caused by traditional finance (TradFi) trading time limitations.

In the RWA field, it is expected that in the next few years, the scale will reach trillions of dollars. Tokenization of private credit, public stocks, and short-term US Treasuries is rapidly growing on-chain; users can now leverage trade commodities and stocks on platforms like Hyperliquid and Ostium, and these figures continue to rise.

Everyone is going on-chain because these are the foundational infrastructures capable of elevating finance to the next level. The dream of widespread decentralized finance (DeFi) is becoming reality—institutions and retail users are using the same infrastructure we rely on today, resulting in increased transparency, faster settlement, no delays, and stronger control over funds.

Applications that have laid a solid foundation will continue to thrive in this new era. Existing leaders in lending, such as Morpho, Aave, and a few others, will maintain dominance because they have withstood real-market downturns, performed well across various market conditions, and continue to innovate. Moreover, protocols like Hyperliquid are becoming some of the deepest sources of on-chain liquidity, supporting leveraged trading of public stocks and commodities. As institutional scale expands, they need venues capable of handling their trading volumes.

Oracle networks, cross-chain interoperability protocols, L2/L1 scaling solutions, and token standards are the key determinants of success. Clearly, when institutions fully bet on on-chain infrastructure, there are no guarantees of assets delivering the best returns; but projects with a solid track record will not be eliminated—they will be adopted by both institutions and retail investors.

Revenue Is King

Coingecko lists over 17,000 tokens. DeFillama tracks about 5,700 protocols; if we only count protocols that generated over $100k in revenue in the past 30 days, there are only about 200 protocols or products, accounting for just 3.5%. The truly investable pools in crypto are far fewer than most imagine—most tokens are simply not investable.

If I analyze these data more realistically, I would consider “holder income”—the income returned to token holders in any form. Surprisingly, in the past 30 days, only about 50 protocols had holder income exceeding $100k, accounting for less than 1% of all protocols on DeFillama.

Raising these benchmarks makes sense; perhaps they could be increased to $100k per month, as most tokens have trading market caps in the hundreds of millions or even billions of dollars.

Delving into the root causes of low holder income, it stems from the long-standing issue of aligned interests in crypto and the token structure itself. Most projects involve two entities: Labs and DAO/token holders. Labs are the “team” in the token economy—they are the initial developers, selling part of the company’s equity and issuing tokens early to raise funds for growth. However, tokens are not legal ownership certificates of the business, nor do they grant holders any actual rights to profits—unlike equity. Investors holding tokens only have rights through their equity stake. But in aligning product and token interests, token holders are often entirely dependent on the project team’s will.

However, over the past year, this situation has begun to change. People are increasingly reducing bets on speculative projects and focusing more on how much revenue protocols actually generate. This shift will push crypto toward the heights that the “extract first, develop later” model has failed to reach in recent years.

Below, we will discuss key metrics every crypto investor should consider when analyzing tokens. We examined the top tokens with the highest revenue in the past 30 days, including Hyperliquid (HYPE), Pumpdotfun (PUMP), Tron (TRON), Sky (SKY), Jupiter (JUP), Aave (AAVE), and Aerodrome (AERO).

Market-to-Sales Ratio (P/S)

The P/S ratio is calculated by dividing a protocol’s market cap by its annualized revenue. It reflects how much the market is willing to pay for each dollar of revenue. The premium indicated by this ratio reveals how investors value the protocol’s future capabilities and growth potential.

We compared some of the most profitable protocols and their tokens based on annualized revenue and P/S ratios. The data are derived by multiplying the past 30 days’ revenue by 12 to estimate annual revenue. The results are shown below.

We set the overvaluation threshold at 20, referencing the P/S ratio of top US listed companies. Most protocols have P/S ratios at or below this threshold, with Tron being an exception, trading at a much higher ratio than others. Another threshold we considered is based on revenue levels, taking the average annualized revenue of the discussed protocols, approximately $250 million. Only Pumpdotfun, Hyperliquid, and Tron exceed this threshold, collectively accounting for about 80% of the listed protocols’ revenue.

Token Holder Revenue

Next, we discuss the key factor of token holder revenue. This mainly depends on the total revenue of the protocol and how much of it is genuinely returned to holders through buybacks, token burns, staking rewards, etc. This metric is now highly regarded, even more than total protocol revenue, because it is the core mechanism for token value accumulation.

We again categorize based on the past 30 days’ holder revenue and multiply by 12 to estimate annual figures. Overall, most protocols perform fairly well in treating holders, with nearly all revenue used to create value for tokens.

This is just one side of the picture, indicating that buybacks are ongoing. Maintaining the current pace could add millions of dollars to token value. To better understand the actual impact of value accumulation, we also compare these tokens’ relative performance since the liquidation event in October, assessing the real effect of token value-building activities.

In the chart, some anomalies are visible, such as TRON, HYPE, and especially the relatively rising SKY. Among these, TRON’s fluctuations are modest, with a sideways trend; HYPE diverged sharply from others in late January.

This shows that buybacks alone are insufficient to sustain token value growth. Other factors are equally critical, including overall market retracement, token release schedules and lock-up periods, narrative shifts in the sector, and overall market sentiment. These will be further discussed in subsequent sections. Before that, we compare the performance of the two highest-revenue protocols: Pumpdotfun and Hyperliquid. The chart clearly shows that during periods when both actively executed buybacks, HYPE performed better (annualized holder revenue of about $660 million versus PUMP’s $380 million). This is mainly due to better overall protocol sentiment and market expectations priced into future supply shocks and token unlocks.

Tokenomics and Supply Overhang

In crypto, tokenomics design aims to help projects raise funds from investors, incentivize users, sometimes conduct community fundraising, and allocate token supplies to teams. Currently, there are no absolute rules for tokenomics design; projects decide their structure independently. This step is crucial because it not only determines the short-term supply pressure but also influences how value is accumulated into tokens, whether there are mechanisms to offset selling pressure, and whether interests between tokens and holders are aligned.

The next chart shows the unlocking speeds of several fixed-supply tokens. Although most tokens will eventually fully unlock, the pace varies significantly: PUMP unlocks fastest, HYPE slowest. Generally, slower unlock schedules are preferred because they reduce sudden supply shocks and market dumping risks. For tokens like AAVE and SKY, most of the supply has already been unlocked; for JUP, the long-term unlocking plan is governed by DAO governance, representing a non-deterministic release.

It’s also important to note that unlocked tokens can be further subdivided into investor unlocks, team unlocks, and community unlocks. The community portion can be used for staking rewards, incentives, or airdrops. Such analysis must be done on a per-token basis and plays a vital role in understanding seller dynamics.

The Lindy Effect

“The longer something survives, the longer it is likely to survive.”

This is the essence of the Lindy Effect, which applies to nearly all enterprises, including on-chain projects, where innovation is a key variable—those unable to sustain innovation will eventually be eliminated.

Last year, crypto protocols generated about $16 billion in revenue, highly concentrated among top projects: the top 10 protocols contributed 80% of net revenue, with the top three accounting for 64%, and Tether alone contributed 44%. Notably, not all high-revenue protocols issued tokens. For example, Circle, with revenue second only to Tether, is listed on the NYSE (ticker CRCL), and Tether itself has no token. Among the top ten revenue-generating protocols, only three issued tokens, indicating that whether to issue tokens should be a cautious decision based on protocol design, not an automatic choice.

Referring to the Lindy Effect, in most crypto sectors, the top two protocols usually dominate the market share. This phenomenon is especially evident in stablecoins: Tether (USDT) and Circle (USDC) together hold 84% of the market, with others like Sky (USDS) and Ethena (USDe) following. Other sectors show similar patterns—e.g., in lending, the top two by TVL, Aave and Morpho, hold 64%. Prediction markets, yield protocols, liquidity staking, and re-staking also display similar dominance patterns.

The Lindy Effect’s importance is also highlighted by the frequent security incidents at the protocol level. This year alone, over $130 million has been lost from smart contracts, with long-term losses reaching hundreds of millions. Over time, users become increasingly wary of new protocols because they cannot predict when attacks might occur. Therefore, the longevity of a contract and the protocol’s survival become critical trust indicators—systems that withstand the test of time without failure are a form of trust endorsement. Even in rare cases of anomalies (such as recent false reports from Aave’s CAPO oracle), as long as the protocol’s treasury can cover liabilities, users can still be compensated. Moreover, the longer a system survives, the more resilient it proves during market downturns. Leading protocols that continue to operate in bear markets send a strong signal of “tested and reliable” systems.

Conversely, continuous innovation is equally vital. Market leaders keep iterating their products. For example, Morpho attracts institutional capital through its vault architecture, allowing highly customized vaults for specific needs; Aave will implement similar features in v4 via Spokes. Additionally, Aave’s Horizon instance now supports institutions to borrow against tokenized RWAs on-chain. The next wave of crypto growth will be driven by institutions and agentic finance—protocols prepared for both will have the greatest growth potential.

Crypto Doomsday Theory

In Citrini’s article “The Global Smart Crisis of 2028,” it is stated: the main way to repeatedly save users money (especially as agents begin autonomous trading) is to eliminate fees. In machine-to-machine business scenarios, 2-3% card interchange fees become obvious targets.

Agents are seeking faster, cheaper alternatives to card networks. Most ultimately choose stablecoins on Solana or Ethereum Layer-2s because settlement is near-instant, and transaction costs are just a few cents.

This marks a new chapter, shifting focus from institutional adoption of crypto to agentic finance and broader blockchain adoption. This trend has already begun, with many protocols integrating AI agents to optimize user flows and eliminate long-standing user experience bottlenecks in crypto products. All these efforts are part of a new category emerging at the end of 2024: the combination of decentralized finance and artificial intelligence (DeFAI). While it initially followed the old “extract first, develop later” narrative, it also highlights AI’s enormous potential to enhance crypto experiences.

Imagine it’s June 2028, most crypto trading is done by AI agents without human intervention. Agents will automatically seek the best yields based on user risk preferences. For non-crypto sectors, blockchain’s low cost, high efficiency, and verifiability make it the preferred platform for executing most transactions. Over time, blockchain space becomes cheaper, transaction costs plummet. Crypto is no longer complicated—you just give an AI agent an instruction and some funds, and it will help you maximize returns. Cryptocurrency and blockchain finally go mainstream and are widely used.

To improve overall capital efficiency, agents will withdraw funds from low-yield protocols or underutilized liquidity pools and concentrate them into platforms offering the best returns. Most public chains and protocols, due to lack of usage, will be phased out. The value of tokens you invested in will have fallen to the lowest point since your initial investment, and you’ll regret not exiting in 2026. Only a few tokens will rise, mainly those that continuously generate income and accumulate value through revenue. Other tokens’ values will shift into these truly performing, practically useful protocols. Although the total market cap of crypto has increased since March 2026, most tokens have not benefited from institutional adoption and agentic finance development. The crypto dream has finally been realized—it is widely used by the public, but the token part of the vision has not unfolded as many expected.

Now, whether you believe this scenario will come true or not, protocols with positive cash flow will be able to sustain long-term growth, and their tokens will flourish.

Conclusion

For years, crypto protocols have focused on technical issues but have never truly addressed product-market fit, which is the biggest long-term risk investors have ignored—yet the market will eventually price this in. Today, most tokens have been declining for years; their all-time highs are distant memories. The situation is clearer than ever: change is imminent. The rise of some tokens in 2026 reflects the importance of revenue data and token-centric value distribution mechanisms—investors are shifting from “gambling” to “investment.”

Bad actors in crypto have long profited from the “extract first, develop later” narrative, while most participants exit with losses, creating a toxic liquidity cycle. This reality is even more apparent with institutional entry—they are less interested in holding our assets and more focused on the infrastructure we have built and proven over time.

As institutional and AI-driven crypto infrastructure further develops, this trend will intensify: more investors will seek verifiable hard metrics to convince themselves to buy a token or equity.