Creators Speak | In the New Cycle, How Should Investors Find 'Value Coins'?

'Creators Speak' is a dialogue column launched by Foresight News, where we ask outstanding creators selected each month about hot topics in the market and compile the collected responses into articles, gathering diverse opinions to explore deeper insights.

Written by: Outstanding Content Creators of Foresight News, September 2026

Compiled by: Foresight News

The cryptocurrency market has entered a new cycle, with fluctuating emotions, meme narratives dominating the headlines, and many tokens relying on hype without a solid business foundation. As the tide recedes, the risks of speculative trading are magnified. Beneath the bubble, what truly constitutes a value coin? What metrics should investors use to avoid traps and seize opportunities? The answer may not lie in the loudest places.

This issue of 'Creators Speak' focuses on the 'New Cycle'. We invited Daii, Black Mario, Alice Liu, Head of Research at CoinMarketCap, X-Agent, and Harvey, Co-founder of BigTime, who are on the September 2026 list of outstanding creators from Foresight News, along with special guest Alex Xu, Research Partner at Mint Ventures, to join this discussion.

We posed five questions: 'What changes have occurred in the definition of value coins in the new cycle?', 'How to identify them?', 'Which sectors are promising?', 'How to view the token value capture dilemma?', and 'What are the practical trading strategies?'. Here are the answers we collected.

1. Compared to the previous cycle, what essential changes do you think have occurred in the market's definition of value coins in the new cycle, and which old evaluation criteria are no longer applicable?

Daii: Regarding the so-called 'new cycle', my current judgment is: it may not be the worst time yet, but the turning point should not be far off.

However, for individual investors, guessing the lowest point is not the most important thing. What truly matters is managing tail risks and ensuring that one does not leave the table before the improvement arrives. My new book, 'The Narrow Door of Bitcoin', emphasizes the non-recurrence of investments: the market can go through countless cycles, but each person has only one path to wealth. An irreversible blow can render all subsequent correct judgments meaningless.

In the last cycle, 'value coins' often meant 'having products, financing, TVL (Total Value Locked), and exchange endorsements'. As long as the narrative was big enough and growth was fast enough, the market was willing to temporarily overlook income quality, token dilution, and value distribution.

The biggest change in the new cycle is that the market has begun to distinguish between three things that were often mixed together in the past: whether the project has value, whether the protocol can make money, and whether the token is worth holding. These three are not equivalent.

A product can have a large number of users, but those users may come from subsidies; a protocol can generate income, but that income may be taken by the team, front-end, validators, or LPs; even if the income eventually enters the DAO, it does not mean that ordinary token holders have clear, stable, and unmodifiable claims.

So what I value more now is not 'whether there is income', but four things: who pays the income, why users continue to pay, how the quality of income looks after incentives are deducted, and whether the token can sustainably capture part of that value.

Some old standards have clearly become ineffective, such as looking at TVL, the number of addresses, transaction counts, gas consumption, financing scale, and the number of partners in isolation, or even looking at protocol income alone. A high FDV and low circulation, paired with a beautiful growth curve, does not automatically equate to value.

Of course, Bitcoin is an exception that needs to be discussed separately. It is not a company and does not rely on cash flow valuation; it is more like a digital scarce asset still in the process of monetization. One cannot use the metrics for evaluating DeFi protocols to assess Bitcoin, nor can one assume that a token with 'income' is more valuable than Bitcoin.

Black Mario: In the last cycle or the previous few cycles, I think the market was more about buying into the future imagination and growth expectations of projects, but in this cycle, under the wave of RWA and others, the market is actually paying more attention to the project's ability to create value.

For example, during the bull market of 2020-2021, especially when sectors like DeFi Summer, NFTs, GameFi, and Layer 1 public chains exploded, the market generally referred to several more traditional standards to judge whether a project had value.

For instance, what sector the project belongs to, which well-known VCs are behind it, how much it has raised, whether the technical narrative is advanced enough, how high the TVL is, how many ecological partners there are, and even whether it has listed on top exchanges like Binance and Coinbase.

The entire crypto market was still in a relatively early expansion stage, and everyone was competing for incremental users and funds. At that time, these standards made some sense; many projects did not have mature business models, but the market was willing to pay a premium for their development over the next few years.

This cycle, however, this logic has become outdated.

As the market has developed to the present, it is gradually distinguishing the real from the false. For example, some projects raised tens of millions or even hundreds of millions, but ultimately had very few real users; some public chains claimed to have very high TPS, but the actual economic activity on-chain was very limited; and some DeFi protocols appeared to have tens of billions in TVL, but a significant portion of the funds was actually attracted by token subsidies and airdrop expectations.

Once the incentives stop, funds and users will quickly flee, making it difficult to maintain the so-called ecological prosperity.

So to judge a project's value now, one should first look at whether there is real market demand and whether users are still willing to use it without continuous subsidies.

Another more important change is that the commercial value of the protocol itself and the investment value of the token are gradually being distinguished by the market, which I think is a very important change in the new cycle.

The early logic was that the better a project developed, the higher the token price should naturally be, but in reality, it seems that these two things are not necessarily connected.

For example, a protocol can generate tens of millions of dollars in transaction fee income each year, but that income may all belong to the development company, validators, or liquidity providers, and ordinary token holders cannot gain any actual economic benefits from it.

So you will find that this round of the market has begun to pay more attention to so-called value capture, which is whether the economic value created by the protocol can ultimately be transmitted to the token itself through fee distribution, buybacks, staking demand, or other economic mechanisms.

Hyperliquid is actually a typical example. The market focuses on HYPE, partly because Hyperliquid has a large trading volume and has formed real trading fee income, linking part of the protocol income with the HYPE token economy through a buyback mechanism.

Thus, judging whether a coin is cheap solely based on the token's circulating market value or the project's financing valuation has actually been disproven.

In this round, everyone should have witnessed many high FDV and low circulation VC coins. The project's financing background is good, and the narrative is also nice, but the token price is under huge unlocking selling pressure in the long term. In fact, for ordinary investors, no matter how valuable a project is, if your buying cost has already overdrawn the growth of the next few years, and you still need to continuously absorb the token unlocks from early investors and the team, then it may not be a good investment.

So rather than focusing more on the project's actual income, token circulation structure, future unlock scale, and whether the economic benefits generated by the protocol can cover the continuously released token supply.

In other words, the value the project describes must gradually be validated through real market demand, economic activities, or asset attributes, and cannot remain forever in white papers and financing PPTs.

Alice Liu, Head of Research at CoinMarketCap: What I have observed is basically: In the last cycle, people bought 'stories', while in this cycle, they buy 'accounts'.

In the last cycle, the so-called value coins were generally seen as 'this protocol is very profitable, so its coin will eventually be valuable'. The intermediate step of 'how money flows from the protocol's pocket into the coin' was tacitly skipped. This round, the market has filled in that step, and it has done so quite bluntly: Uniswap turned on the fee switch, Aave was forced by the community to return 100% of brand product income to the DAO, and Hyperliquid simply took almost all fees for buybacks. These three events happened in the same year, not because the project suddenly had a conscience, but because the market voted with their feet for five years, and they finally understood.

Specifically, three things have changed.

The anchor of valuation has changed. Previously, we looked at TVL; now we look at 'holder income'. TVL is rented; once the project incentives stop, users withdraw their investments, and TVL can easily evaporate; income is earned, but we also need to see how much of the earned money actually goes into the coin. The marginal buyers in the new cycle are ETFs, publicly listed companies, and professional market makers. These players focus on buyback yields, FDV (fully diluted valuation) divided by income, and annual dilution rates.

Compliance has shifted from a negative factor to a positive factor. In the last cycle, compliance was a cost; in this cycle, compliance is a ticket to entry. Assets that can be classified as commodities, can enter ETFs, and can be held through compliant channels directly enjoy liquidity premiums; those that cannot enter this pool can only trade among themselves in a small pond.

'Scarcity' has been redefined. Low circulation and high FDV were once packaged as scarcity; now the market sees every unlocked coin as a sell order waiting to be dumped.

As for the old standards that should be discarded, here’s a list: TVL rankings, financing lineups, active address counts, and the number of ecological projects. These four can be looked at, but they should not be the only focus. Because these can be bought with money, and they are not expensive. Another one that should be discarded the most is: 'Having governance rights equals having value.' Over the past five years, several leading DeFi projects' K-lines have verified that voting rights have a market valuation of approximately zero.

X-Agent: My judgment is that a bull market has already started, and this round of the bull market will be led by Wall Street, with US-based funds being stronger. ETFs allow investors to gain exposure to crypto assets through familiar securities accounts, and crypto is starting to enter the traditional financial allocation system more directly.

I believe that the biggest change in this round is that core incremental funds and pricing power will come more from traditional finance. To find value coins, one needs to understand what these funds are allocating, why they are allocating, and which projects can serve their trading, custody, settlement, and risk management needs.

Combining my AI and Web3 product practices, I will also look at whether a project can solve specific problems: whether there are people continuously using and paying for it, whether it can improve efficiency, reduce costs, and how the created value ultimately relates to the token.

Financing background, community size, TVL, and trading volume still have reference significance, but relying solely on these is no longer sufficient. I focus more on whether real demand, sustainable business, token value capture, and buying price are matched. A product has value, a token has value, and whether the token is currently worth buying are three questions that need to be answered separately.

Harvey, Co-founder of BigTime: Let's first talk about cycles. The current four-year cycle has not yet expired. If we cut the cycle by 'Halving Era', we are currently in Era 5: 2024-2028, which is the cycle we are in now. The last cycle was Era 4: 2020-2024.

Definition of value coins:

• Era 4: 2020--2024, Old Definition: Good projects with imaginative potential defined by TVL, user numbers, institutional endorsements, and listing standards.
• Era 5: 2024--2028, New Definition: A business that has "real demand, stable cash flow, and can distribute profits to token holders."

In summary, this round of changes can be expressed as: The "value" of value coins has shifted from "the future in narratives" to "verifiable and distributable present." The market no longer pays for "something that might be valuable in the future," but first asks, "Is it profitable now, and do investors get a share?

Ineffective Evaluation Criteria:

• Worship of TVL

Locked amounts can be inflated through subsidies; once incentives are removed, it returns to zero. It only measures "money attracted," not "sustainable business."

• Pure User Scale / Active Address Count

Airdrop farmers and multi-wallet clusters can generate impressive data, but without paid retention, the scale is illusory.

• Institutional Endorsements

Reliable → Ineffective. This round, institutions are also here to seek liquidity; financing scales and fundamentals have decoupled, and endorsements do not equate to cash flow.

• "Beta Logic of Generalized Monetary Easing"

Ineffective. In Era 4, everything rose with BTC; this round, as BTC's elasticity diminishes, altcoins without independent income no longer automatically gain beta elevation and must prove alpha themselves.

Mint Ventures Research Partner Alex Xu:

1. The earliest value coins:

The so-called value coins or blue-chip coins may refer to those with a longer historical timeline and strong retail investor consensus, including many purely consensus-based cryptocurrencies.

2. Recent cycles of value coins:

Value coins must have actual business operations, preferably with clear PMF, measurable revenue, stable operating status, a certain market cap, and a longer operational duration.

3. Current cycle of value coins:

In this cycle, the standards for value coins will become stricter, essentially resembling those of publicly listed companies and their equity standards. That is, not only must there be real demand, a long-standing business with income, but also positive profits, positive cash flow; and the tokens themselves must provide shareholder returns similar to equity (i.e., buybacks, burns, or dividends), ideally with governance decision-making rights.

Of course, in reality, very few projects in the crypto space meet all these criteria. Even tokens like HYPE, which have buybacks and burns, currently lack governance rights. UNI is the closest to this state.

I believe this overall trend is very correct. Crypto projects should not be treated as exceptions; that is, the notion that "many crypto projects cannot be valued using the same methods as public companies" or that "the valuation multiples given to them are particularly lenient" is incorrect.

The valuation logic for business-oriented crypto projects should be identical to that of all publicly listed companies, differing only in their specific business operations and methods. The intrinsic value of tokens in crypto business projects should also derive from the discounted future free cash flows.

I believe this will become an increasingly dominant consensus over time. Pure narrative-based projects and purely consensus-based projects will see their overall market cap space continue to shrink. This change and trend were already quite evident in the last cycle, where many altcoins did not experience their so-called altcoin season; only altcoins with quality business fundamentals could stand out, a typical example being Hyperliquid.

2. Many projects will artificially beautify revenue and business metrics through multi-wallet trading, bot-driven gas consumption, and statistical adjustments. What key signals can investors rely on to distinguish sustainable real business value from packaged pseudo-value?

Daii: I would first ask a very simple question: Who is actually paying?

If the main payers are associated wallets of the project, arbitrage bots, airdrop farmers, or users cycling operations to obtain more tokens, then this type of income is essentially closer to customer acquisition costs rather than business revenue.

I generally cross-verify from several directions.

First, look at how much of the user base and revenue remains after incentives are withdrawn. Real demand may grow slowly but usually retains; the prosperity generated by subsidies often collapses quickly once the token distribution stops.

Second, look at net revenue, not gross revenue. If a protocol generates $1 million in fees but distributes $3 million in token incentives, it cannot simply be stated as "the protocol's monthly revenue is $1 million." Subsidies, rebates, market-making costs, security expenses, and token dilution must all be accounted for in the same table.

Third, look at the source and path of funds. The number of addresses can be easily fabricated, but the concentration of funding sources, associations between wallets, circular trading paths, unusually synchronized transaction times, and where the funds ultimately flow are usually harder to disguise long-term.

Fourth, look at unit economics. How much does the protocol earn per user interaction? How much must be subsidized to bring this user back? If the larger the business scale, the greater the losses, that is not a network effect but rather converting token holders' money into data.

Fifth, look at pressure conditions. In a bull market, almost all protocols have "demand"; what distinguishes them is who still has users willing to pay voluntarily after the market declines, incentives decrease, and volatility drops.

My simple judgment is: trading volume can be faked, addresses can be split, gas can be burned, and statistical criteria can be changed; but it is difficult to long-term fabricate demand that does not rely on continuous external payment incentives.

Black Mario: One of the biggest characteristics of blockchain is the transparency of on-chain data; all transaction records can be traced. However, the mere existence of on-chain data does not mean that the economic activities reflected by this data are necessarily real market demand.

For example, I could create 1,000 wallet addresses and continuously trade, transfer, or call contracts among them through scripts. From the on-chain browser's perspective, all these transactions are real, gas is indeed consumed, and the protocol generates fee income as a result.

However, these actions may not create any external economic value; their purpose is merely to generate data to make the project appear very active. Therefore, I believe that to assess a project's real business value, one must further investigate the economic logic behind this data.

One of the most important aspects is to look at the source of the project's income and how much cost was incurred to obtain this income.

For instance, a DeFi protocol claims to have generated $10 million in fee income in a month. However, if it issued $15 million in token rewards to attract user trading, provide liquidity, or maintain activity, then economically, this protocol may still be in a phase of continuously subsidizing users.

In fact, we often need to distinguish between Fees and Revenue. Fees are the total transaction fees paid by users, while Revenue is usually the portion that the protocol retains. As for how much the protocol ultimately earns, it needs to further deduct token incentives, operational costs, and other related expenses.

Thus, some projects may have very high trading volumes and fees, but the protocol itself does not retain much profit.

Moreover, some public chains like to use gas consumption to prove their economic activity, but an increase in gas consumption does not necessarily mean that the overall commercial value of the ecosystem is growing in tandem. Who ultimately receives the gas fees, whether they are burned, and what demand lies behind the transactions all need to be analyzed separately.

One thing is that I will pay special attention to whether a project can maintain its user and business data after subsidies are removed.

I believe this is one of the most effective methods to identify real demand. In the crypto industry, many times, project teams can completely buy users, TVL, and even trading volume through economic incentives.

For example, in the early NFT market, LooksRare and X2Y2 exhibited very typical trading mining and volume-faking issues. Some users would repeatedly buy and sell NFTs between wallets they controlled to artificially create large transaction volumes.

These transactions indeed occurred, but a significant portion does not represent normal NFT buying and selling demand. Therefore, for such projects, I am more concerned about whether users are still willing to continue using them when token rewards decrease, airdrop expectations are fulfilled, or mining yields decline. If a protocol had hundreds of thousands of active addresses daily during the subsidy period but quickly drops to only a few thousand after subsidies stop, then I believe the previous data should be significantly discounted.

Of course, there is nothing wrong with early projects using subsidies for customer acquisition; traditional internet companies also adopt similar methods. The key is whether the users acquired through subsidies ultimately convert into real users willing to pay continuously.

Another aspect is the structure of on-chain data.

If we carefully observe on-chain transactions, we will find that many volume-faking behaviors follow patterns. For instance, a large number of wallets originate from the same funding source, transaction times are highly concentrated, transaction amounts and behavior patterns are very similar, and funds circulate repeatedly among a few associated addresses. Alternatively, the number of active addresses for a project may continuously rise, but most addresses only transact once and never use the protocol again; these are all signals worth further investigation.

For ordinary investors, I believe they can also focus on a relatively easy-to-understand metric, which is user retention and payment situation.

For example, if a project has 100,000 active addresses today, how many of those addresses continue to use it a month later? How many are actually paying fees? Is the protocol's income overly reliant on a few large holders?

If a project has a continuously growing base of real paying users, and these users are not retained solely through long-term high subsidies, then its business foundation is usually more solid. Of course, one wallet does not necessarily correspond to one real user, so it is best to consider this data in conjunction with funding sources, interaction behaviors, and retention situations over different periods.

Moreover, we should not fully trust the statistical criteria published by the project team.

For instance, if a project originally reported net income but suddenly starts including all transaction fees in revenue; or merges transaction volumes from different chains and products, or even incorporates business data generated from partnerships into its ecosystem data. In this case, the final data may look much better, but the project's real operating condition has not changed accordingly.

Thus, when I research projects, I typically cross-reference third-party data platforms like DefiLlama, Token Terminal, and Dune, paying particular attention to any changes in statistical criteria, whether they can be verified through on-chain transaction records, and why there are discrepancies between data from different sources.

Especially now, it has become easier; tools like GPT and Grok have data extraction and analysis capabilities, which is very convenient.

Alice Liu, Research Head at CoinMarketCap: All false indicators can basically be filtered out by this fundamental question: Who is paying, and why are they paying? The payers of real business are dispersed and voluntary; the payers of false indicators are either the project team themselves or mercenaries coming for points and airdrops; once incentives stop, the people disperse.

The signals our team repeatedly looks at include several key points.

Are Revenue and Incentives the Same Curve? Look at net income after deducting token incentives. If revenue fluctuates in tandem with incentive switches, it’s not revenue; it’s rent. Over the past year, several trading platforms have been found to have trading volumes closely aligned with the rhythm of reward seasons—when rewards are issued, volumes surge; when rewards stop, volumes drop. Such volumes can only support temporary rankings, not valuations.

Do Fees Keep Up with Trading Volume? Genuine traders pay full fees honestly, while wash traders go to great lengths to evade fees. If trading volume far exceeds liquidity depth but fees do not keep pace, it’s not a good signal. Alternatively, one can check if there is volume without price. Real buying and selling will drive prices.

Don’t Just Look at Total Volume, Look at Addresses. Systematic data research shows that analyzing transaction scale distribution on-chain is not very meaningful; it’s essential to see how much fees are contributed by the top addresses and whether they are circulating among themselves. On-chain activity follows the same logic: after the Fusaka upgrade, new Ethereum addresses surged, but two-thirds of the new addresses had first transactions of less than $1. Researchers need to distinguish between address poisoning and genuine user growth. More than half of the Gas on L2 is consumed by bots, yet their fee contributions are less than 10%. Therefore, one should look at "fee density" rather than just the number of transactions.

Has the metric been subtly altered? If a project suddenly reports data that looks better than third-party sources, quietly modifies statistical criteria, or mixes in unearned income in its revenue definitions, these are all Red Flags. Reliable projects will proactively reconcile with third-party data platforms like CoinMarketCap.

X-Agent: I will separate the study of market funds from project operations. The former answers what funds are being allocated, while the latter answers what value the project is actually creating.

When looking at ETFs, I combine transaction volume, net subscriptions, shares, and position changes to judge, avoiding the direct interpretation of trading activity as new funds. When examining projects, I differentiate between trading volume, user-paid fees, protocol revenue, and token holder returns.

Specifically for business, I will ask: Who is paying, why are they paying, and will they continue to pay after subsidies stop? Is revenue concentrated among a few related clients? How much rewards, rebates, and operational costs were incurred to obtain this revenue?

AI can help organize transactions, identify abnormal patterns, and cross-check criteria, but anomalies can only serve as investigative leads, not direct conclusions of fraud.

Especially in the Agent scenario, bots may also be real customers. The key is what tasks they complete, who bears the costs, and whether they are willing to repurchase. Call volume, wallet numbers, and Gas consumption all need to correspond to verifiable business results.

BigTime Co-founder Harvey: First, a premise judgment: The later the cycle + BTC elasticity diminishes, the more pseudo-value will concentrate—because real projects that can capture cash flow are scarce, and funds have allocation pressures, thus "packaging" becomes the highest cost-performance arbitrage.

This phenomenon is more severe in Era 5 than in Era 4, as the previous round of pseudo-value (pure narrative, no income) has already been debunked once, making fraud more industrialized and covert.

Past methods such as wash trading, bots pulling Gas, and adjusting statistical criteria—essentially all involve faking the two most visible metrics: "revenue" and "activity."

Thus, the underlying logic of discernment is: look for those "same-source signals" where the cost of fraud is extremely high or where fraud leaves indelible traces, rather than just looking at totals and growth rates.

Key Signals We Focus On:

1) Address Homogeneity of Revenue

Real protocol fees come from a large number of independent user behaviors, distributed in a "long tail, random, continuous" manner. Fake revenue is concentrated in a few addresses, over a very short time, within specific amount ranges, and the trading counterparts are highly correlated (circling back to themselves). Examining the concentration of fee-sourcing addresses is a hundred times more useful than looking at total amounts. If 80% of a protocol's revenue comes from 10 addresses, no matter how large the total amount is, it’s fake.

2) Paying Retention, Not Active Scale.

Wallet clusters can fake "active address counts," but they cannot fake whether the same batch of addresses continues to pay after incentives are removed. This is a hurdle that opportunists can never surpass. What to focus on is not "how many people are using it now," but "how many of those who received subsidies three months ago are still around and still paying?" Retention is the mirror of pseudo-value.

3) Matching Gas Consumption with Business Value.

If you say bots pull Gas, this is actually easiest to identify—check whether their transaction content is just spinning (self-rotation, meaningless calls, interacting with themselves on-chain). High Gas but very low business value indicates volume manipulation. Behind the high Gas of a real protocol, there must be corresponding real economic activities (real exchanges, lending, liquidation), and the two must match.

4) Cash Flow Returning to Tokens

This is the ultimate judge of "true value vs pseudo-value." This point is the most critical and counterintuitive: a protocol's revenue can be real, but for you as a token holder, it may still be pseudo-value. If all revenue goes into the team’s treasury, and the token has no capture mechanism (no fee switch, no buyback and burn, no ve sharing), then the business is real, but you don’t get any money. For investors, this is the same as fabricated data; it’s all "visible but not edible." (Note: No ve sharing means that even if the protocol earns fees and revenue, ordinary token holders cannot obtain profit sharing through staking and locking tokens.)

The core is: looking at retention, homogeneity, and revenue direction is much more reliable than looking at total volume and growth rates. Totals can deceive, but structure and direction do not.

Mint Ventures Research Partner Alex Xu: I have always advocated looking at the business side first, then the financial side. The so-called business side means that as project investors, we must first understand:

• What is the real demand corresponding to this crypto project?
• What is the business model it operates based on this real demand? How does it convert demand into revenue and revenue into profit?
• What is the competitive supply in this track? What are the project’s competitive defenses and their sources?
• What is the strength of the upstream and downstream industrial chain in this track?

Only after we have a sufficient and profound qualitative judgment on the business can we discern whether the so-called operational data is real, rather than looking at the so-called on-chain data and directly making rough inferences about the project’s operational situation based on on-chain data. Data is used to validate and expand our qualitative judgments.

For example, the DePIN projects that were quite popular in previous cycles (such as distributed computing projects, distributed storage projects) are, in fact, very uncompetitive compared to mature cloud computing and cloud storage solutions. The scenarios and demands they correspond to have never been genuinely validated, making it difficult from a qualitative perspective to consider their demand as real, let alone form a self-sustaining business model.

Based on this qualitative judgment, when we look at their business data, we can find that their so-called business data actually contains a lot of volume manipulation or data generated due to token incentives, rather than data generated from real demand.

3. From the current standpoint, which tracks do you think have the opportunity to produce new cycle value coins, and which types of tracks will you actively avoid, and why?

Daii: I am more focused on three directions.

The first is still Bitcoin. Strictly speaking, it is not a "value coin" in the usual sense, but if we discuss how ordinary investors can traverse cycles, it remains the most worthy asset to study individually in the entire crypto market. Its value does not come from protocol profits, but from scarcity, security, liquidity, global participation, and the potential for continued monetization.

The second is stablecoins, on-chain payments, and their infrastructure. The advantage of this direction is that demand is relatively easy to verify: Are users really transferring, settling, saving, and flowing across borders without relying on grand narratives to explain? However, just because a track has value does not mean every related token has value. Many values may ultimately be captured by issuers, channels, wallets, or compliance institutions, rather than governance tokens.

The third is financial protocols that have proven real demand, including trading, lending, derivatives, and some asset management infrastructure. Finance is one of the few areas in blockchain that has found relatively clear product-market fit, but I will focus on those protocols that have undergone complete stress testing, transparent risk parameters, and relatively clear token rights.

I will also pay attention to RWA and AI Agents, but with more caution. They both have real demand and significant narrative space; the issue is that real business may exist in off-chain companies, while tokens are merely financing and marketing tools hanging outside. The establishment of a track does not equate to the establishment of a token.

I will actively avoid several types of projects: highly homogeneous public chains and applications that can only rely on subsidies to maintain data; high FDV, low circulation, and tokens with significant future unlocking pressure; "governance commemorative coins" where business revenue is completely disconnected from tokens; and AI, gaming, and social projects that cannot clearly explain why tokens must exist.

I do not believe these tracks necessarily lack market activity. It’s just that market activity and value are two different things. As a personal investor, I am more concerned not about whether it can rise, but about what I will have left when liquidity recedes.

Black Mario: A track having significant development potential does not mean that the tokens within that track necessarily possess investment value. From the current standpoint, I would lean towards those tracks that already have real market demand, can continuously generate income, and have clear value capture mechanisms for their tokens.

I personally see several promising directions.

One is DeFi, especially protocols that have formed mature business models, such as decentralized derivatives trading, lending, DEX, etc.

I believe DeFi has validated some real demand through several cycles. For example, Aave’s on-chain lending and Hyperliquid’s derivatives trading have already formed considerable economic activity. The business models of these projects are relatively easy to understand; users need to pay interest for borrowing and fees for trading, allowing the protocol to earn revenue from real financial activities. Therefore, I will focus more on those protocols that have sustainable income, stable user bases, and can transmit some economic value to the tokens.

The second is stablecoin payments and the financial infrastructure built around stablecoins.

In fact, stablecoins are one of the few products in crypto that have truly achieved large-scale commercial applications. Especially in areas like cross-border payments, commercial settlements, and capital flows, I believe there is still significant development potential as traditional financial institutions and payment companies continue to enter. If more and more economic activities are completed through stablecoins in the future, then the payment, clearing, and settlement infrastructure serving these capital flows will have the opportunity to form sustainable income.

The third is RWA, especially the tokenization of US stocks and the financial infrastructure for bringing real assets on-chain.

We have discussed this direction a lot before. As traditional brokers, exchanges, and asset management institutions gradually enter, crypto is continuously expanding the boundaries of the assets it can accommodate.

If traditional assets such as stocks, bonds, and ETFs gradually move on-chain in the future, the trading, lending, collateral, and settlement demands surrounding these assets will further expand. We can focus on projects that can participate in actual business and capture economic value.

Additionally, I am also quite interested in the infrastructure that combines AI Agents with on-chain finance, such as agent payments, identity, permission management, and automated trading. However, this sector is still in its early stages, and AI products themselves may not necessarily need to issue tokens, so I will be more cautious.

As for the directions to actively avoid, I think there are mainly a few categories.

First, there are a large number of homogeneous public chains and L2s that lack real demand.

In fact, past cycles have already proven that the market does not lack high TPS and low Gas infrastructure; what is truly scarce is users and applications. Many L2s gradually expose issues of insufficient real users and limited revenue after financing, token issuance, and ecological incentives end. Therefore, relying solely on technical parameters and financing backgrounds is no longer sufficient to support the long-term value of a public chain project.

Secondly, projects that heavily rely on token incentives to maintain business scale.

For example, some re-staking, complex yield aggregation, and protocols that attract funds based on points and airdrop expectations. These projects may have a high TVL, but a significant portion of the funds is attracted by subsidies, and once the incentives stop, the funds may quickly dissipate.

Finally, some AI and DePIN projects that lack actual commercial demand, as well as VC tokens with high FDV and low circulation, especially some distributed computing projects. Although the technical narrative is quite good, their actual commercial competitiveness may not be validated.

For example, providing computing power through idle phones, considering costs such as equipment wear and tear, operational stability, and network scheduling, it may not be more advantageous than traditional cloud services. If there are no real paying customers and the network scale is maintained solely by issuing token rewards, then I believe the long-term value is quite limited.

Alice Liu, Head of Research at CoinMarketCap: Block space is a commodity that anyone can create; transaction matching and asset issuance have network effects that cannot be replicated. Money will flow from the former to the latter.

Three promising lines:

The "periphery" of stablecoins, rather than the stablecoins themselves. The issuers of stablecoins take the largest share of profits in the industry, with three-quarters of the daily revenue of protocols going to them at peak times. However, you cannot "hold" shares of USDT and USDC, and with the Genius Act, regulators are increasingly moving towards prohibiting issuers from paying interest to holders. Thus, this profit can only overflow: to distribution channels, settlement chains, and payment fronts. Whoever can become a "distributor" of stablecoins will have an advantage in capturing this profit. Synthetic Yield Stablecoins are a variable in this line, and many issuers have significantly shrunk in scale compared to last year. So I remind everyone that the yield of synthetic dollars is dependent on external factors and cannot be treated as fixed deposits.

On-chain exchange layer. Exchanges are the only repeatedly validated business in the crypto industry, and perpetual contract DEXs have brought this business on-chain, which has also exploded this year. This sector is sliding into a fee war and points war, and projects with real liquidity moats will have better opportunities to compete successfully in 2027.

The "carrying layer" of tokenized assets. Tokenized assets are the strongest category on the CMC narrative list in the past 30 days, with the share of tokenized stocks in RWA tripling in a year. However, it should be noted that the valuable tokens that ultimately emerge from this sector are likely not the "RWA tokens" themselves, but the issuers, settlement layers, and compliance gateways that carry them. The ones making money in a gold rush are the shovel sellers, and this saying holds true on-chain as well.

Three categories to actively avoid.

Narrative tokens without revenue channels. AI agent tokens are the freshest negative example, with the leading market cap shrinking by 90% from its peak. The problem does not lie with AI, but with the lack of a pipeline between the token and the business; no matter how good the business is, it cannot flow into the token.

New public chains with high FDV, low circulation, and unlocking pressure. Infrastructure is already severely oversupplied, with even Robinhood launching its own chain. This indicates that "chains" are becoming cost centers for enterprises, just like renting servers, rather than assets worth buying.

Projects that rely on incentives to buy volume. The numbers supported by points and airdrop expectations will return to zero the day incentives decline, as mentioned in the second question, so I won't elaborate further.

X-Agent: I will look for opportunities in two directions: what U.S. funds need to allocate and what services can accommodate these funds.

First, mainstream crypto assets that can accommodate institutional allocation needs, as well as trading, lending, custody, settlement, and risk management services surrounding them. I believe that providing reliable services to Wall Street will be a significant opportunity this round. When judging projects, we should look at actual customers, sustainable revenue, and whether the tokens can share in business growth.

Secondly, stablecoin payments and asset tokenization-related services. I am interested in whether they can integrate into actual financial processes to improve capital turnover and settlement efficiency. Here, it is also important to distinguish between the returns of underlying assets, the income of service providers, and the rights of governance tokens.

Third, AI Agents, especially research, risk control, execution, and reconciliation services that can enter financial workflows. This is closest to my product practice. My judgment criteria will focus on task completion quality, paid retention, delivery costs, and whether the data and permissions are reliable.

I will avoid projects that mainly rely on subsidies to maintain demand, simply package models to issue tokens, and have unclear long-term income and token rights. The U.S. background itself cannot replace business validation; who can serve, how to serve, and who benefits from the value are the contents that need to be continuously checked.

Harvey, Co-founder of BigTime: The only criterion for judgment: whether a token is a valuable token is unrelated to its sector; it only depends on one thing—by holding it, can I automatically and continuously share in the profits earned by the protocol?

Using this criterion, if we rank all projects by "capture progress," the conclusion will be very clear.

First tier: already sharing profits (true value tokens, allocatable)

The common point of this tier is "mechanism has been launched, money is already being distributed," not "future empowerment."

1）Hyperliquid (HYPE)

• The most thorough model in this round. A perpetual contract DEX on its own L1, with revenue coming purely from fees, no subsidies, no TVL inflation.
• A portion of the fees is used for buybacks and burns, and a portion is directly distributed to HYPE stakers, with no VC private placements or airdrops, distributing money to tokens from day one of design.
• It simultaneously hits "close to money + real cash flow + closed-loop capture" three points, almost the standard answer for "what new value tokens should look like."

2）Uniswap (UNI)

• Activating the fee switch in October 2024, distributing a portion of trading fees to UNI stakers, transforming from "governance token" to "cash flow certificate."

3）Aave (AAVE)

• Launching buy & distribute in 2025, using protocol revenue to buy back AAVE for distribution to stakers, bringing back money that was previously cut off from the front end and not returned to the DAO.

Second tier: currently supplementing capture, completing it will be top value tokens (focus on observation)

1）Chainlink (LINK)

• Demand and revenue are undeniably real (oracles are essential infrastructure for DeFi), but LINK has long been a "consumed commodity"—projects pay LINK as service fees rather than distributing income back to holders, making capture indirect and convoluted.
• This round is transitioning from "indirect use value" to "direct cash flow capture" through Staking (v0.2 has been launched, v0.3 is in progress).

It is a different stage of the same narrative as the first tier: Uniswap/Aave have already completed it, while Chainlink is in the process of completing it.

Whether it can upgrade to a top value token depends on the ratio of "real protocol income" to "inflation subsidies" in the Staking rewards—if subsidies dominate, it remains weak capture; if income dominates, it counts as a true closed loop.

Third tier: good business, but tokens have not yet shared profits (not value tokens, but breeding grounds)

1）RWA(Ondo/ONDO)

• U.S. Treasury interest is distributed to product holders (USDY), but not to ONDO tokens, which remain governance tokens, sharing the same flaw as early Uniswap.

The latter two tiers do not constitute value tokens; they are breeding grounds for potential value tokens, and it only counts when a project lands a "revenue return to token" mechanism. Treating them as value tokens is repeating the old mistake of "protocol earns money, tokens do not share."

Avoidance tier: tokens are dilution / takeover tools, not cash flow certificates

The common point of this tier is—no independent real income, relying on issuance, subsidies, and narratives to maintain.

1. Subsidy-based L1/L2—relying on issuance and subsidies to pull TVL, with fee income far below the cost of issuing tokens; once subsidies stop, funds flow back to the top (Arbitrum, Base, Solana). You are paying for their customer acquisition costs.
2. Pseudo AI—removing the AI label makes it indistinguishable from ordinary protocols, with no independent cash flow, purely narrative-driven.
3. Burn money GameFi—P2E relies on "new players taking over," with no sustainable external income (Axie has been disproven).

In summary, the dividing line is not in the "sector," but in the "capture progress": money is already being distributed (first tier) > currently supplementing (second tier) > not yet distributed (third tier) > will never be distributed (avoidance tier).

The value tokens that can emerge this round are not about "which sector will win," but about which project first directs income back to the token.

Hyperliquid is not an exception; it validates the conclusion itself: new value tokens are projects that "designate money to be distributed to tokens from the outset," rather than "first financing, first governance, and then we'll see."

Alex Xu, Research Partner at Mint Ventures: I have always said that the underlying value logic that blockchain (or crypto) brings to business is a larger and freer market with lower trust costs.

Therefore, when we look at whether the business logic of a certain sector is valid, we must first see:

• Whether the services and products provided by existing enterprises or projects in this sector are significantly lacking in "larger, freer markets" and "trust costs"?
• Whether the solutions provided by crypto projects can offer sufficient incremental value in these two aspects?

In fact, we find that most pseudo-concept crypto projects do not perform well enough in these two aspects, or the incremental value they provide is insufficient. A typical example is DePIN, decentralized social media, and decentralized or on-chain games, etc.

To be honest, besides finance, I have not yet seen other new sectors or products worth looking forward to. Some newer products that have emerged in recent years, such as prediction markets, etc.:

• First, it is also a product with very high financial attributes.
• Secondly, products like prediction markets often have implementations and product forms that are unrelated to blockchain and crypto. There are many successful centralized products that can exist completely independently of crypto and blockchain. So we still need to continue observing.

4. In the early days, Uniswap allocated all trading fees to LPs, and UNI holders could only participate in governance without automatically receiving protocol income. Aave also faced disputes over front-end revenue attribution, with some income not flowing back to the DAO. Have you encountered similar situations where "the protocol makes money, but tokens struggle to distribute dividends"? How do you view the issue of insufficient token value capture?

Daii: I have encountered many similar situations. This is one of the most easily overlooked issues in crypto investing: investors analyze the protocol for a long time, only to end up buying not equity in the protocol, but a certificate with very ambiguous rights.

I break down value capture into four stages: whether the protocol generates income; who ultimately receives the income; whether governance can change this path; and whether token holders have a clear and ongoing right to claim.

If any of these links is broken, the protocol's value may not necessarily be transmitted to the tokens.

Governance rights are certainly not without value. If a token can genuinely control protocol parameters, fee switches, and the treasury, it has some option value. However, this value must be heavily discounted because governance may be controlled by large holders, fee switches may never be activated, regulatory environments may restrict distributions, and teams may alter value flows through front-end products, foundations, or new entities.

Therefore, I would not value based solely on "potential future dividends." Future governance proposals cannot be equated with today's cash flow, and the DAO treasury cannot automatically be considered the assets of token holders.

I prefer value capture paths that have already occurred rather than those still resting on promises. For example, clear buybacks, burns, staking income, or other verifiable mechanisms. At the same time, I continue to question: is this mechanism sustainable, can it be canceled by governance at any time, and is the income at the cost of greater token dilution?

A protocol can be very successful, yet its token may still be a poor investment. Separating these two matters is the starting point for evaluating "value tokens" in the new cycle.

Black Mario: This situation is actually very common in the crypto industry. When I first got into DeFi years ago, I felt this was a significant issue.

Many projects have protocols that indeed have value and can generate real income, but there is no effective connection between the commercial value of the protocol and the investment value of the token.

Uniswap is a very typical example. As one of the most successful DEXs in the entire DeFi industry, Uniswap has long had a huge trading volume, and LPs can earn fees by providing liquidity.

However, the main function of UNI in the early days was governance, and token holders could not directly share the fee income generated by the protocol.

The Uniswap protocol may be a very successful business, but ordinary UNI holders did not feel the dividends of the protocol's development in the early days.

Although in December 2025, Uniswap activated the fee switch and began to buy back and burn UNI through fees, gradually establishing a connection between protocol usage and token economics, I believe this mechanism belongs to indirectly capturing value by reducing supply, and does not mean that UNI holders can directly receive cash dividends.

The situation with Aave is slightly more complex.

Previously, Aave Labs and the DAO had disputes over the attribution of front-end trading fees. The core issue is whether part of the income generated by users entering the platform through the Aave brand and ecosystem should belong to the company responsible for developing the front-end product or be shared by the DAO, which is actually unclear.

Aave Labs believed that the front-end was an independent product developed and operated by them and had the right to receive corresponding income, while some community members argued that these businesses benefited from the brand, liquidity, and infrastructure built by the DAO over the long term.

Later, Aave gradually adjusted the revenue attribution mechanism, redirecting some front-end income back to the DAO.

However, there is actually a second problem: even if income enters the DAO treasury, it does not mean that ordinary AAVE holders can automatically share these earnings. The DAO's funds may be used for development, operations, security reserves, ecological incentives, or may be decided through governance to buy back tokens.

So I think, we must at least distinguish three concepts: how much income the protocol generates, how much income the DAO ultimately receives, and how much economic value ordinary token holders can capture.

These three are actually completely different.

Of course, I do not believe that all protocol income should be directly distributed to token holders.

For example, Uniswap's LPs bear risks such as capital occupation and impermanent loss, so earning fees is reasonable. If the protocol excessively compresses LP earnings to buy back tokens, it may lead to a loss of liquidity and ultimately harm the protocol's competitiveness.

The DAO also needs to retain certain funds for security maintenance and long-term development.

However, I believe that if a project has developed to a relatively mature stage, has sustainable business income, yet fails to establish an effective token value capture mechanism, then its token will struggle to gain long-term fundamental support.

So from an investment perspective, I am now more focused on the project's fee distribution, buybacks, income attribution, and whether the token itself has irreplaceable actual demand.

This is also why we cannot simply view crypto project tokens as stocks.

Stocks typically correspond to relatively clear company equity, while many tokens do not actually represent equity or profit distribution rights. Therefore, the protocol making money and holders making money are inherently two different matters. Projects that can reasonably connect these two through mechanisms in the future are more likely to gain long-term value recognition.

Alice Liu, Head of Research at CoinMarketCap: This is the most common mismatch in the last cycle, which I basically describe as "protocol rich, token poor." Uniswap is a typical example, with tens of billions in fees over five years, but much of it went into LPs' pockets, while UNI holders received a ballot. Aave is similar, with front-end routing swap fees outside the DAO, and the community calls this a "slow-motion coup." This term is very accurate; the value is not taken away all at once, but rather leaks out little by little along the product structure.

My view has three layers.

First, insufficient value capture is never a technical issue; it is a legal and interest issue. Technically, opening a fee switch is a matter of a single vote. The real barrier is: when regulation is unclear, dividends can easily touch the red line of "I am a security"; coupled with the fact that development companies and DAOs inherently find it difficult to align, front-end products, wallets, and aggregators have a natural incentive to keep money on the company’s books rather than in tokens. A significant reason why many fee switches will open in 2026 is not that everyone suddenly figured it out, but because after regulation became clearer, project parties found it difficult to find excuses to delay.

Second, even when the switch is opened, the problem is not solved. This is the most important lesson to remember this year. Uniswap opened the switch and burned 100 million tokens at once, yet the price still hit a new low for the cycle; Pump.fun once took 100% of income for buybacks, yet during that time the token price still fell by two-thirds. Why? Four things overlapped: the buyback scale was too small compared to market value, unlocks and emissions on the other side offset the buybacks, income itself was declining, and buybacks were governance policies, not contractual commitments, which could be cut in the next vote. These uncertainties are difficult to support token prices.

Third, investors often look at these three numbers: ratio, net amount, and reversibility. What percentage of total income do holders receive? Is the buyback yield positive or negative after deducting dilution? Is the mechanism locked in a contract, or can it be changed in the next vote?

Going deeper, the ultimate outcome of this issue is the transformation of tokens from "governance tokens" to "on-chain equity." Aave consolidating all economic rights into AAVE and Hyperliquid designing tokens using the logic of exchange stocks are both moving in this direction. Regulation will determine how far this path can go, but the market has already taken the lead and is pricing according to this standard.

X-Agent: I would layer the income of the team company, protocol income, DAO treasury income, and token holder earnings. The specific rights that investors buy determine how they should be valued.

AI projects especially need to be viewed this way. Users pay for AI services, and the income may go to the operating company; the token only serves governance or other functions. Even if the product grows well, it is essential to continue confirming how this growth affects the token through what mechanism.

The case mentioned in the title has also undergone subsequent changes: since December 2025, Uniswap has had a UNI burn mechanism supported by protocol fees, but holders do not have the right to proportionally receive protocol income (according to Uniswap's official statement). Aave Labs announced in April 2026 that it would redirect related CoW partnership fees and Horizon fees to a DAO-controlled collection address (Aave Labs implemented the update).

These changes indicate that value capture can improve. My focus is on whether the rules have been executed, whether the source of funds can be sustained, who can modify them, and whether the scale of buybacks, burns, or distributions can support the current valuation.

As traditional finance deepens its involvement, I believe token rights need to be articulated more clearly. Only future distribution commitments and mechanisms that have been continuously executed should not receive the same pricing.

Harvey, Co-founder of BigTime: "The protocol makes money, but tokens do not receive dividends" is not a flaw of a few projects, but the default state of token design in this industry.

Dividing by "where the money is intercepted" I have seen four typical forms:

1) All goes to service providers, holders capture nothing.

Uniswap was typical in its early days. All fees went to LPs, and UNI only had governance rights, equivalent to holding equity without dividend rights.

2) Income is intercepted at the front-end/middle layer, not flowing back to the protocol or DAO.

Aave's front-end dispute is typical. Value is intercepted at the business level, and neither the protocol nor the tokens benefit.

3) Tokens are "consumables" rather than "dividend certificates" ------ Chainlink is typical.

Projects pay LINK as service fees, and LINK is spent rather than returned as dividends to holders; capture is indirect and convoluted.

4) Income stays in a treasury controlled by the team, and holders only have governance rights.

Many projects with real income fall into this category. Money is genuinely earned, but it goes into the foundation's treasury, and the team can dispose of it freely; tokens are purely financing tools from back then, and holders receive no cash flow.

I have seen all four forms, and often the same project exhibits them at different stages. Their root problems are interconnected.

The latter part of the question: how to view "insufficient value capture": this is not a bug, but a product of historical stages.

We judge this in three layers:

First, the result of compliance pressure; early "non-capture" was a proactive choice.

At the beginning of the token's birth, it was not positioned as a "security," but as a "financing + governance" tool. Because once cash flow dividends are written into the token, it may be classified as a security by regulators, and project parties deliberately designed tokens to be "without dividend rights" for compliance and fundraising. Therefore, "insufficient capture" was intentionally designed by the creators in the early stages, not an oversight.

Second, there was no capture demand in the bull market; this has been forced out by this bear market.

In the bullish environment of Era 4, cryptocurrency prices can rise based on scarcity and narrative, with holders profiting from price increases while disregarding minor transaction fee dividends. It wasn't until the recent elasticity decay of BTC, the disappearance of beta, and institutional entry that holders began to question whether what they held was truly something that could distribute profits. Capturing demand only emerged after the market cooled down.

Third, the capture mechanism itself has compliance and technical costs.

Fee switches, buybacks, and ve dividends all face legal risks regarding whether they constitute securities, as well as technical challenges in how to automatically and fairly distribute income to stakers on-chain, which is why they have not been widely adopted.

This leads to a more fundamental conclusion: "Insufficient value capture" is not due to laziness of individual projects, but rather a systemic design flaw in the previous round of token economics—tokens were treated as a "financing + governance + speculation" three-in-one tool, without being considered as "cash flow certificates."

What this round (especially projects like Hyperliquid) is doing is addressing this flaw by redefining tokens from a "three-in-one tool" to "dividend certificates." This is the true meaning of "the definition of value tokens undergoing a fundamental change," and it serves as the underlying basis for all my previous judgments (looking at capture progress, income direction, and whether tokens are consumables).

Alex Xu, Research Partner at Mint Ventures: Most tokens and crypto projects lack returns for token holders, or in other words, have insufficient value capture. The primary upstream reason is that the effective demand that the crypto industry can currently support, and the commercial value of this effective demand, is still not large enough. As a result, most projects do not even have revenue, let alone net profits after deducting token incentives and team operating costs, and naturally cannot allocate value to tokens.

Moreover, those projects that do have revenue and net profits also do not provide returns to token holders. This can be viewed in several scenarios. The first scenario is that although the protocol currently has net cash flow, this cash flow has good reinvestment opportunities, such as promising new product opportunities that require R&D and promotion, and existing businesses still have market expansion potential. In other words, after calculation, they believe that investing this capital in products and operations will yield a longer-term ROIC (Return on Invested Capital) compared to directly buying back tokens or distributing dividends. If a project does not conduct token buybacks or dividends under these circumstances, I believe it is acceptable, but in reality, very few projects meet this condition.

Another type of project may not have much reinvestment space left, but still does not provide returns to token holders despite having ample net cash flow, and there is no reasonable explanation for this. There are quite a few such projects, a typical example being Chainlink:

• Chainlink certainly has off-chain business income, but the income data is not disclosed, and this income and operating profit are unrelated to token holders.
• The use case for the LINK token is very weak; currently, it seems to only have staking as a purpose, and the staking APY is also very low.

For projects like this, I believe investors who focus on fundamentals should simply avoid them. You can imagine a publicly listed company whose shares have no voting rights and do not provide shareholder returns; would anyone buy this company's stock? Certainly not.

In cases where the known revenue and cash flow of a project are unrelated to the token, if you still insist on buying, then it is not the project's problem, but rather the investor's problem.

Similar projects also include BNB. The current value capture of BNB is completely unrelated to Binance's net profits. Binance's returns to token holders have been terminated since 2022, and the main value capture and use cases, aside from the main chain's gas, are primarily for participating in some airdrops and new listings on Binance's main site, but this business volume has also rapidly declined in recent years.

5. Based on past experiences in selecting and holding value tokens, what position and profit-taking strategies do you generally adopt after selecting a target, and what practical insights can you share?

Daii: My first principle is not "how much can I earn this time," but rather "if I'm wrong, will I permanently lose the ability to continue participating?"

This reflects the non-reversibility of investment. Even if a trade has a positive expected return, it does not mean one can go all-in without limits. As long as there is a possibility of zeroing out, liquidation, hacking, regulatory changes, or loss of liquidity, the position must comply with survival constraints.

Specifically, I separate my core Bitcoin holdings from high-risk token holdings. For new value tokens, I usually start with a very small recognition or observation position. Only when the product, data, token structure, and value capture continue to be validated do I gradually increase my position. Position size is not a performance of confidence; it is the loss limit when errors occur.

I generally do not use borrowed money and try to avoid leverage. Many people do not fail due to incorrect directional judgment, but rather because they die on the path: the target does indeed rise later, but they have already been liquidated or forced to sell during previous volatility.

In terms of stop-loss, I do not mechanically set a uniform percentage. Cryptocurrency assets are highly volatile, and simple price stop-losses can easily misinterpret normal fluctuations as logical failures. I focus more on three types of stop-losses: first, logical stop-loss, when the core assumptions of the initial purchase are falsified; second, structural stop-loss, such as adverse changes in token unlocking, governance rules, or value distribution; third, position stop-loss, when the losses of a single target begin to threaten the overall investment portfolio, regardless of how optimistic I am, I must reduce my position.

Profit-taking is also not about accurately selling at the highest point. When the price enters a clearly optimistic expectation-dependent range, I will gradually reduce my position, recoup my principal, and convert part of the paper profit into real assets while keeping a small position to deal with right-side tails. It is not shameful to miss out on profits because no one can capture all the gains while avoiding all pullbacks.

My greatest insight in the past has been: good targets cannot compensate for incorrect positions, and correct judgments cannot save an irreversible heavy position. Individual investors' true advantage is not having more capital or faster information, but rather the ability to wait, to make small trial-and-error investments, and to choose not to participate.

The market will have the next opportunity, provided you are still in it.

Black Mario: I believe the importance of position management may be as significant as your ability to select a good project.

Because even if we are correct in our long-term judgment of a project, it does not mean that buying at any time or price is correct, so I clearly distinguish between long-term core positions and opportunity-based positions when allocating assets.

First, the long-term core position.

For assets like BTC and BNB, which I have long been following and have a deeper understanding of their fundamentals and market positions, I tend to hold them long-term and am willing to give them a higher position weight. For these assets, I will not easily change my long-term judgment due to short-term price fluctuations of ten percent or even larger.

However, for an emerging project, even if the fundamentals are good and the narrative is strong, I usually do not assign too high a position at the beginning.

After all, Crypto projects have too many uncertainties, including technological implementation, team execution, token unlocking, market competition, etc.

Therefore, for new targets, I prefer to participate with a small position first, and as the project's business data, revenue, and token economic model gradually get validated, I will consider whether to increase my allocation.

The second is the buying price and valuation.

I think a common mistake many investors make is to confuse good projects with good prices.

A protocol may indeed have real income and an excellent business model, but if the token's valuation has already overdrawn the growth potential for the next few years, then even if the project continues to develop, the token price may not perform well. Therefore, I pay special attention to FDV, circulating market cap, future unlocks, and the actual income of the protocol, trying to avoid buying too much at once during peak market sentiment due to FOMO.

The third is profit-taking and stop-loss.

In fact, I do not like to apply a fixed percentage stop-loss strategy to all tokens.

For mature assets like BTC and newly listed projects with very low circulation ratios, the risk structures are completely different.

For long-term favored targets, I focus more on whether there have been changes in fundamentals rather than short-term price fluctuations.

However, for emerging projects or highly volatile opportunity-based positions, I will control risks more strictly. If the original investment logic of the project is falsified, such as real business falling short of expectations, continuous deterioration of income, or token unlocking pressure exceeding expectations, then even if there are paper losses, I will consider exiting rather than endlessly averaging down.

As for profit-taking, I prefer to cash out in batches. Especially when a project has already seen significant short-term gains and the market valuation clearly exceeds my reasonable expectations for it, I will consider gradually reducing my position to recover part of my principal and profits.

I believe there is no need to be obsessed with buying at the lowest point or selling at the highest point; being able to continuously keep profits in your account is more important than occasionally catching a hundredfold token.

Finally, one of my personal insights is to try not to develop too strong an emotional dependence on the projects I invest in, which is particularly common in the Crypto industry.

Many people research a project for a long time, even join the community and participate in governance, and eventually become loyal supporters of the project. However, investment is investment, and the project is the project.

A project worth long-term attention does not mean its token is worth holding indefinitely, nor does it mean we need to be bullish at any price.

In the Crypto market, while seizing opportunities is important, being able to endure multiple cycles and remain in the market long-term is what I believe to be even more important.

Alice Liu, Head of Research at CoinMarketCap: First, let me clarify that this is not investment advice, but a few rules I summarized after paying my own tuition.

Position size depends on "how much can be verified," not on "how much can rise." Holdings are divided into three layers: core assets with institutional buying, value tokens with calculable cash flow, and speculative positions that purely tell stories. Each serves its purpose.

Profit-taking is based on valuation, not on price increase. If the buyback yield is compressed to be similar to government bonds, or if FDV divided by income is already several times that of similar listed exchanges, it’s time to reduce. Another signal is market behavior: when the CMC Fear & Greed Index and CMC Altcoin Season Index spike, funding rates heat up, and social media buzz far outpaces income, these are practical indicators for profit-taking. Stop-loss should be event-triggered, not percentage-triggered. It’s also important to pay attention to the liquidity of the token; if you clearly see its spot trading volume dropping and market liquidity drying up, these are signals for price stop-loss points.

Holding value tokens is not about buying and forgetting; it’s still necessary to regularly check and verify these. The "value" of value tokens is a verb that needs continuous validation, not a noun that is fixed on the day of purchase. Fortunately, these can be analyzed and monitored on CMC AI, making it easier for everyone to observe the market.

During the holding period, continuously compare new information with the original investment logic. Is the price increase accompanied by improvements in funding and business? Is revenue growth driven by real demand or subsidies? Has the originally promising value capture mechanism been realized? This helps focus attention on changes that truly affect judgment.

In position management, AI can assist in calculating different scenarios, including account risks under rising, fluctuating, and significant retracement conditions, as well as whether multiple positions are exposed to the same market factors. Calculations must be based on real account data and platform rules.

When it comes to crypto-backed leverage, I pay special attention to collateral prices, margin, and the distance to liquidation. The bullish direction, position size, and leverage level need to be judged separately; confidence in a bull market also requires leaving room for mid-course fluctuations.

Stop-loss should consider whether the preset loss boundary and investment logic have been falsified, with the risk boundary set before liquidation. Take profit should look at whether the price has overstretched the business's ability to realize profits, or whether a single position occupies too large a risk exposure, before considering partial liquidation.

AI is responsible for helping me organize, verify, and discover changes, while I decide on positions and execution boundaries. I believe the most useful ability of AI is to allow investment judgments to continuously undergo verification and to timely translate market changes into assessable actions.

Harvey, Co-founder of BigTime: The core of position management is "tiering" and "diversification," not heavy betting.

1. First, determine the total portfolio, then set a limit for individual positions.

Value coins belong to high-volatility assets, so I will first designate a fixed proportion of my total assets for the "crypto position" (for example, 10%-30%, depending on the individual), and I will never use money that affects my living expenses. This rule is more important than any technical strategy.

2. Set a hard upper limit for individual assets, and do not concentrate bets.

Even if I am optimistic about a particular asset, a single position will not exceed 20%-30% of the crypto position. The biggest risk with value coins is not "making the wrong judgment," but rather that you make the right judgment, yet the market makes you lose 70% before it comes back—concentrated bets can force you to exit before dawn.

3. Allocate position weight based on "capture progress."

Combining the previous framework:

• The first tier (money is already allocated, such as Hyperliquid, UNI/AAVE with fee switches activated) can be given to core positions;
• The second tier (currently capturing, such as Chainlink) can be given to observation positions or small positions;
• The third tier (good business but no profit sharing, such as stablecoins/RWA related) should generally not be touched or only traded in segments.

The more certain the capture, the heavier the position can be; the more narrative-driven, the lighter the position.

4. Build positions in batches, primarily on the left side, and add after confirming on the right side.

I almost never buy everything at once. After confirming the logic, I first build a base position (for example, 1/3 of the planned amount), and then add after price retracement or confirmation of a key signal (such as fee switch implementation, Staking launch). I would rather buy at a slightly higher price for certainty than heavily bottom-fish before a significant drop.

Take profit and stop-loss: Discipline is more important than prediction.

1. Taking profit is "partial liquidation," not "waiting for the highest point."

I will preset several target levels, and when they are reached, I will first sell a portion to lock in profits (for example, sell 1/4 for every 50% increase); only what is sold is profit. Many people fail because they think "it can still go up"—only a few can truly capture the entire market movement in value coins, and most just ride a roller coaster.

2. Stop-loss looks at "logical falsification," not "price decline."

• This is the biggest difference between value coins and short-term trading. I do not stop-loss just because "it dropped 20%"; instead, I ask: Is the logic for buying it still valid? If the fee switch is turned off, the capture mechanism is invalidated, the team runs away, or the fundamentals are falsified—then no matter how much it drops, I will exit; if it’s just a market pullback and the logic hasn’t changed, then it’s actually an opportunity to add rather than a stop-loss point.
• Beware of "deteriorating fundamentals but flat prices." This is the sneakiest way for value coins to bleed: indicators are quietly getting worse, but prices don’t drop, making you think everything is fine. In such cases, you should actively exit rather than wait for the price to tell you.

A few insights:

1) Losing less is more important than making more.

In this industry, surviving long enough and not losing all your capital is essential to have the right to wait for the next round. The premise of compound interest is "don’t go to zero."

2) Always keep some bullets.

Retain some cash/stablecoins so that when a black swan event occurs, you have capital to buy at the bottom, rather than being forced to cut losses at the bottom.

3) Desensitize to positions.

Before building a position, write down the logic (why to buy, what signals count as falsification, what price levels to reduce positions), and then strictly follow it without getting emotionally attached to the position. Writing it down can save you when emotions run high.

4) Don’t apply the "value coin" framework to things that shouldn’t be applied.

If a particular asset is essentially narrative/emotion-driven (like Meme or pure AI concepts), it has its own game logic, and using the "long hold, fundamentals" approach of value coins is mismatched—such things should be approached with short-term discipline rather than the patience of value coins.

5) The biggest cost of holding value coins is not volatility, but "holding in the wrong cycle."

In a position like Era 5 with "BTC elastic decay," it’s better to prioritize cash flow certainty rather than aggressive expectations of "doubling again." The cycle and position determine the strategy.

Choosing the right assets determines your upper limit, while position and discipline determine whether you can survive to reach that limit. The former is cognition, the latter is execution; most people who lose money actually fail in execution rather than cognition.

Alex Xu, Research Partner at Mint Ventures: Before we build a long-term investment system, we must first have a thorough understanding of ourselves, which means:

• Are your personality, resources, and endowments suitable for fundamental analysis-based value investing?
• Or are you an energetic, emotionally sensitive speculator regarding market news?
• Or do you have the ability to directly build your unique quantitative trading system?

Regardless of which investment method above, each has its corresponding capability circle, but only after you clarify your investment style can you potentially build your capability circle in a targeted manner.

I belong to the first type, essentially a value investment system.

In this system, it starts with an analysis of the business, estimating future potential profits and free cash flow to build corresponding positions. The higher your confidence and certainty in the conclusions drawn from analyzing this asset, and the greater the difference between your estimated valuation range and its current market value, the larger the position you can take. However, considering uncertainty, each asset should have a position limit, which varies from person to person.

As for stop-loss and take-profit strategies, they are fundamentally based on two points:

• Whether your judgment or preset regarding this asset has changed;
• Whether the asset's prices have risen to a point where its potential return rate is lower than the opportunity cost of other assets.

When either of these situations occurs, you can take stop-loss or take-profit actions.

If there are no clear answers to the above questions, then it means you do not understand, and if you do not understand, do not invest.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.