The few that survive are the core of the entire game.

Written by: Raoul Pal, Founder of Real Vision

Compiled by: Chopper, Foresight News

In your portfolio, there must be a public chain that you bought in the last cycle because you thought it would become the next Ethereum. Better technology, combined with the founder's persuasive words in podcasts… You weren't foolish to buy in; many smart people did the same.

But looking back now, that chain is essentially a ghost town. Only a few validation nodes are still running, with hundreds of wallets transferring small tokens back and forth, and the token prices are long below the purchase cost of the believers.

According to CoinGecko, as of the end of last year, over 13 million tokens had stopped trading, with the vast majority having vanished by 2025. Many of the public chains they rely on have also met the same fate.

In the past thirteen years, I have seen similar viewpoints on social media during every bear market. Cryptocurrency is a scam; it has always been a casino, and I told you so long ago.

But this mass extinction is not a scandal; it is merely basic probability.

I wrote this in December 2020 when I felt it was obvious, but others did not share that view. Tokens are merely products of real-time venture capital valuations. If you hold early seed-stage projects, and the prices are being repriced every second, you will see the same result: about 80% will ultimately be worthless, 19% will create some value, and only 1% will earn the majority of returns. The crypto market is simply broadcasting this process live for everyone to see, and such real-time price fluctuations can easily provoke panic.

At that time, I also examined myself with this standard. If Real Vision had evaluated its valuation in real-time every day since its founding, I suspect its value would have nearly dropped to zero six times. The value of a project comes from its adoption rate, which ultimately comes from revenue, and the vast majority of projects will never make it that far. Most projects do not survive. This is the reality, whether you like it or not; this is capitalism.

So the real question has never been, "Will most public chains perish?" They are destined to. The real consideration is: which ones can survive, and how can we identify them before the prices provide the answers?

We have seen the same script

Twenty-one years of writing experience has taught me to look for precedents, and the precedent to keep in mind here is the burst of the internet bubble.

In November 2018, Bitcoin was in a deep bear market, and I wrote that this scene was incredibly similar to the internet bubble bursting in 2000 and 2001: grand ideas, but flawed business models. In the end, a phoenix will rise from the ashes. I was certain that the crypto market would repeat this process, but I could not judge which projects were phoenixes and which were just ashes; I honestly stated this at the time.

Admitting the inability to predict is the right answer because in 2000, no one could pick the final winners. The infrastructure was overbuilt, and the Nasdaq index plummeted by 78% from peak within two and a half years, with the vast majority of assets losing all value. However, a few companies that rose from the ashes grew into the largest enterprises in the world, relying on the fiber infrastructure built by those bankrupt firms. In hindsight, the crash was just an interlude on the way up; being in it at the time made it feel like the entire industry's philosophy was coming to an end.

All significant platform markets since then have ended up this way. In the cloud computing field, only about five leading players remain; the operating system and mobile phone sectors are similar, with the vast majority of value concentrated in the top four. Public chains and these platform markets are fundamentally the same.

A public chain is an economy, not just a token

Since it belongs to a platform market, what is its product? The answer is simple: block space.

The winners will use either technology or narrative (or both) to make their block space more attractive, drawing more users and funds to circulate on-chain. Everything else is just marketing packaging.

This is why generic public chains cannot be viewed as companies, let alone simply as a token. Ethereum, Solana, Sui, and a few other public chains are independent economies with their own GDP, money supply, and monetary policies. They should be assessed like an economy—by looking at how much economic activity is generated on-chain.

Four paths to survival

The competitive landscape is much simpler than people think; projects only have four paths to survival:

First, become an independent economy. Ethereum, Solana, Sui, and a few other public chains have sufficient throughput, composability, and developer density, making them capable of supporting a global token economy.

Second, pursue a vertical specialization route. Hyperliquid is the most typical example; it is an integrated exchange that just happens to use a public chain as its execution layer. Specialized exchanges can often turn a profit, but their revenue ceiling is constrained by their niche, so they should be valued like an exchange with a solid moat, rather than an open economy without limits.

Third, operate as a layer two network. Build a business on top of a layer one public chain, renting its underlying security guarantees. Some layer two networks can become quality businesses, but they are essentially franchises, much like stores operating on Shopify. Even if a particular L2 perishes, its on-chain status can be migrated elsewhere, which is why they cannot retain value like layer one public chains: L1 has data permanence guarantees, while L2 just rents.

The fourth category is somewhat special: BNB. The market prices it as an independent economy, but it is not. It is not decentralized; its true underpinning is Binance's balance sheet. It's a decent business, but essentially, it is a centralized platform token dressed as an L1.

Why most public chains struggle to survive

The vast majority of remaining public chains—the many projects that launched in the last two cycles—do not belong to the aforementioned four categories. They are generic public chains trying to compete with those independent economies, but they lack the conditions for victory; they are not specialized enough to own a niche market; and their level of centralization is insufficient to rely on the balance sheets of large platforms. They are merely "a public chain." A public chain without a moat will ultimately have its value reduced to zero.

This is the structural root. On a practical level, what truly kills them is the token issuance mechanism.

Since the summer of 2024, I have been studying low circulation, high fully diluted valuation (FDV) tokens, and this pattern has not changed: the market dislikes these kinds of assets. In the initial phase of token listing, the circulating supply is typically small, often plummeting by 70% or more, as the market anticipates future unlocking sell pressure. In a bull market, with real demand behind it, low circulation tokens can soar. But when a bear market arrives, unless unlocking has nearly completed or demand is truly sustainable, these assets will significantly underperform the market. And public chains without a moat never have real demand. They will only gain short-term hype, but turning that hype into long-term adoption is difficult; most of the hype evaporates in an instant.

Thus, the issuance structure merely accelerates the process of these homogenous public chains reaching zero; even without the issuance mechanism issue, the market would ultimately arrive at the same conclusion.

The race towards zero cost

Most valuation frameworks in the industry will fail here.

The core dimension of competition among public chains is operational efficiency. In other words, a public chain that wins collaborative competition will inevitably continue to drive the marginal cost of each transaction toward zero. If you apply a cash flow discount model, the conclusion is that leading public chains are valued lower than challengers because they earn the least transaction fees per transaction. This model gives a completely inverted answer.

To assess the value of public chains, one must think in reverse. Consider if Ethereum were shut down, all layer two networks became incapacitated, most stablecoin markets vanished, and all DeFi, NFTs, and tokenized assets were impacted, the likelihood of future financial systems relying on this infrastructure would cease. This is the value of Ethereum. No one assesses TCP/IP protocols by tallying the earnings per data packet. To evaluate it, one looks at the entire internet that operates on it.

This competition will continue to accelerate; a new class of users is about to enter. I previously explained that AI entities cannot use traditional banking systems; they must rely on these kinds of on-chain channels for settlement. AI beings have no inherent preferences and will always automatically choose the cheapest, fastest, and most efficient network; if a lower fee option becomes available, they will immediately leave the more expensive public chains. Solana and Sui have pushed down the cost of stablecoin transfers to nearly zero, which is precisely what Wright's law is demonstrating right before us.

Some businesses require the underlying security of Ethereum, while others only pursue extreme speeds. Thus, different public chains accommodate different scenarios; the market landscape has the top chains holding the vast majority of the market share, rather than one chain dominating all. But the lower tier public chains do not qualify to share the pie.

Everyone has misunderstood the measurement metrics

I can explain why the industry's measurement standards are inherently flawed; I have fallen into this pit myself.

Earlier this year, I planned to build a valuation framework for layer one public chains, and like all analysts, I started with fully diluted valuations and the metrics everyone tracks: Total Value Locked (TVL), fee income, DEX trading volume, and stablecoin market cap. The data correlations appeared to be high, and I thought I had found the right approach.

But the reality was different; the whole logic was circular reasoning. The TVL valuation unit is itself the token to be valued; DEX trading volume fluctuates with speculative trends, and fee income fluctuates with transaction volume. It is equivalent to using data derived from token prices to explain token prices. We were only observing the surface, not the essence.

So I eliminated price interference and converted to ratios unaffected by cryptocurrency prices. I selected the peak in January 2025, covering the data of 12 public chains, simply asking one question: Is there any metric that can predict which public chains will have the smallest declines in the next 13 months?

Fee yield had absolutely no predictive power. In a bull market, fee income can only reflect speculative enthusiasm, not measure real value creation. TVL/FDV (the crypto version of the price-to-book ratio) is equally ineffective.

The only metric associated with asset returns is the capital turnover ratio: the speed at which capital circulates in on-chain exchanges. However, the correlation is inverse. In January 2025, Solana's capital turnover rate reached 200 times its locked value, while Ethereum was only at 7 times. The faster the asset turnover on a public chain, the worse its performance over the next 13 months; chains that retain capital deposits have stronger resistance to price declines.

Now my understanding is that turnover rate represents entropy increase, while capital stickiness represents order. When capital comes in and settles down, it is used for lending, exchanging, and re-staking, continually establishing economic connections, attracting more capital into the market. Quick in-and-out capital merely treats the chain as a casino, and after the speculation wanes, it leaves no residue. Therefore, the high trading volumes that everyone clamored for at the peak of the bull market are actually risk warnings; chains that appear calm, with capital long-term settled, are instead capable of withstanding declines.

Characteristics of survivors

Thus, I have summarized three key observation dimensions, all unrelated to price:

First, during a sharp drop in token prices, can the economic density per user hold steady? From the peak in December 2024 to February 2026, the token price of one public chain I tracked fell by 83%, TVL dropped by 78%, and the number of active addresses decreased by three-quarters. Yet the deployment funds corresponding to each active user hardly changed: around $2,100 at the peak, and remained around $2,100 at the trough. Speculative visitors and bots exited, but the remaining users continued to engage in the same scale of economic activity. Among the public chains I tracked, only three layer ones maintained this density indicator during the decline: Ethereum, Solana, and Sui. All other public chains saw their economic density shrink simultaneously with price drops; one well-capitalized mainstream layer one saw a direct shrinkage of 83% in user economic density, indicating that retained users had largely halted all on-chain activities.

As a side note: I do indeed work for the Sui Foundation. But it was only after completing the data analysis that I realized Sui was one of these three projects; the data does not lie.

Second, during the token price decline, can the supply of stablecoins continue to grow? The value of on-chain dollar stablecoins does not diminish as token prices plummet. If dollars continue to flow in and settle during a bear market, it signifies that external capital has made a long-term commitment. All three of these public chains showed an increase in stablecoin balances during the downward cycle; that public chain that saw its price drop by 83% still saw its stablecoin size increase by about half.

Third, continuous flow of developers is also my highest-weighted metric. Developers are the most scarce capital in public chains. Users can withdraw all their funds in minutes; developers invest years into learning specific languages and architectures, with high switching costs. The ongoing migration of developers toward a particular public chain during a bear market indicates that the most knowledgeable professionals in the industry have made a high-certainty choice at a significant cost. Price charts will never reflect this.

Growth in wallet numbers, increase in on-chain transaction values, developers entering first, and all these metrics remaining steady during market pullbacks—these are the characteristics of projects that can weather the cycles. Projects that meet this condition are worth in-depth study; if they do not, all you hold is a homogenous, commodity-like token.

The S-shaped growth curve of such projects may experience a 90% price drop midway. But a few of the surviving projects will eventually achieve exponential growth, continuously setting new highs. After the life-and-death filtering of thousands of projects, the network effect kicks in, and the entire ecosystem takes off.

Projects in the long tail sector will continue to increase, with a massive number of tokens issued in every cycle. Many people always daydream of finding a hundredfold return among the vast long-tail projects while neglecting the main opportunities right in front of them. Focus on adoption, not on token prices.

The vast majority of projects will ultimately perish. The few that survive are the core of the entire game.