The Real "Blood and Tears" History of Crypto VCs: 300 Projects Invested, Nearly 30% Went to Zero or Rug-Pulled

Recently, well-known venture capital firm a16z announced that it is renaming its crypto investment arm a16z Crypto to a16z Frontier, expanding its investment scope from Crypto to frontier technology fields such as AI and robotics. This change also reflects how some crypto VCs are re-examining their investment direction.

This episode of the Wu Blockchain podcast invited Dashan, co-founder of Waterdrip Capital, and Xiao Yang, a practitioner at a crypto VC, to discuss the real state of survival in the crypto primary market. The two guests come from investment institutions of very different sizes, yet face similar problems: a large number of projects struggle to keep operating, investors have difficulty exiting, traditional Crypto investment opportunities are shrinking, and capital and attention are gradually shifting toward emerging fields such as AI.

Since its founding in 2017, Waterdrip Capital has invested in close to 300 projects and currently manages several hundred million US dollars in assets. Dashan disclosed that fewer than one-third of its portfolio projects are still operating well, and that about 20%–30% of projects involve rug pulls, going to zero, default, or damage to investors' rights. In 2026, of the dozen-plus projects Waterdrip Capital invested in, only three or four are in the Crypto field, with the rest mainly concentrated in AI and chips. Dashan believes Crypto no longer needs to be the protagonist of the digital economy, but should become infrastructure for areas such as AI Agent payments, data privacy and operation verification.

By comparison, Xiao Yang's fund is about $20 million in size and has invested in roughly 100 projects in total, including more than 70 in 2024 alone, but it has stopped making new investments since the first quarter of 2025. Xiao Yang said about two-thirds of the portfolio projects are still operating, but more than half are difficult to maintain effective communication with investors. The fund's book returns are close to breakeven, and actual exits still face difficulties. Both guests agreed that future investing should pay more attention to founder integrity, real user demand and sustainable business models, rather than short-term market hype.

Guest remarks do not represent Wu Blockchain's views and do not constitute any investment advice. Please strictly abide by local laws and regulations.

Part One: Waterdrip Capital's Dashan — From Crypto Investing to AI

Starting with 5,000 ETH, nearly 300 projects invested to date

Maodi: Could you introduce Waterdrip Capital's history, fund size and investment style?

Dashan: Waterdrip Capital was founded in 2017 and was one of the earlier batch of Token Funds in China. At the beginning, a few co-founders pooled about 5,000 ETH and mainly invested using ETH and BTC; only after 2019 did we gradually start using USDT.

In 2021 we established an LPF compliant structure in Hong Kong, and in 2023 we set up a compliant fund in cooperation with traditional financial institutions. At present, counting both our own capital and funds we manage jointly, the overall scale reaches several hundred million US dollars, covering primary market, secondary market, quant and RWA, among other directions; the primary market fund is around several tens of millions of US dollars.

To date we have invested in close to 300 projects. The team is about 15 people, kept at under 20 in total.

The name Waterdrip Capital comes from Liu Cixin's The Three-Body Problem. Several of our co-founders are fans of The Three-Body Problem, and at the time we thought blockchain might, like the "water drop" in the novel, become a technology that upends the traditional financial system, so we named it Water Drip.

Maodi: How has your investment direction changed over the years?

Dashan: During the 2017 ICO boom, we mainly invested in public chains and various "blockchain+" projects, including ecosystems such as Cosmos and Polkadot. In 2018 and 2019 we began to focus on the combination of blockchain and traditional industries, and afterwards moved into areas such as DeFi and Layer 2.

The Bitcoin ecosystem is also one of our more representative investment directions. We took part in Bitcoin mining early on, and later set up a dedicated Bitcoin ecosystem fund, Satoshi Lab, investing in projects such as the Lightning Network, RGB and Bitcoin Layer 2.

We also took part when inscriptions emerged, but we thought they were closer to Meme Coins, so we did not invest on a large scale. Although interest in the Bitcoin ecosystem has declined somewhat, technologies such as the Lightning Network and RGB are still developing, and we will keep watching them.

Overall, we increasingly prefer to invest in projects with real applications, rather than projects with very high valuations and nothing but a technical concept.

Crypto VC has gone through a major shakeout — so why is Waterdrip Capital still expanding?

Maodi: Over the past two years many crypto VCs have had a hard time. Why is Waterdrip Capital expanding instead?

Dashan: In 2018 there were more than a hundred crypto investment institutions in China, but many of the funds that were active that year have now disappeared. Waterdrip Capital has been able to hold on partly because early investments in Bitcoin and mining accumulated a certain amount of assets, and partly because the team is small and operating costs are relatively low.

We have launched four funds in the past; some returned ten or twenty times, and some did not even return the principal, but we did not carry out large-scale layoffs because of that.

We have indeed added people over the past two years, but mainly for AI, chips, US equities and listing-related business, rather than expanding the Crypto investment team. Apart from Hong Kong, we now also have colleagues in the United States and Japan.

I have always wanted to keep the team under 20 people. AI can now help complete a lot of research and supporting work, so a large staff is not necessary. At the same time, I prefer flat management and want to communicate directly with every colleague every week; a team that is too large actually reduces efficiency.

After nearly 300 investments, founder integrity matters more than technology

Maodi: After so many investments, what do you consider the most important criterion for choosing a project?

Dashan: If I could choose again, I would first look at whether the founder is honest, and only then at sector trends and investment timing.

We have stepped into many pits in the past. Some project founders had untrue résumés, some exaggerated their fundraising materials, or failed to deliver on promises. Sometimes we knew perfectly well the team was bragging, but when we saw well-known institutions participating, we still invested with the mindset of following the crowd to make money, and in the end we often suffered heavy losses.

What is more troublesome is that even if a project successfully issues a token and the token price rises, investors may still not make money. Some teams do not unlock tokens according to the agreement, some conceal their financial situation, and some even sell tokens privately, leaving investors unable to exit normally.

So a project succeeding and investors earning a return are two different things. Whether a founder is trustworthy matters more than technical ability and fundraising background.

Maodi: How should investment timing be judged?

Dashan: Many crypto sub-sectors have a life cycle of only a few months to a year or two. In the past, a large number of technical concepts ultimately failed to create real demand, but directions such as DeFi, DEX and asset tokenization have already proven to have long-term value.

For example, zero-knowledge proofs are valuable in themselves, but that does not mean every technology needs its own separate public chain and token. Many technologies can perfectly well be integrated into existing ecosystems as infrastructure.

We have also made money in the past in sectors that were ultimately disproven, because we entered early and exited in time. But I think a truly successful investment should be one where investors earn a return and the project really does solve a problem and exists for the long term.

Investing too late makes you the one holding the bag, while investing too early may mean the market never matures in time. Finding the right timing is very important.

Rug pulls and defaults: nearly 30% of investments ran into problems

Maodi: Of the nearly 300 portfolio projects, how many are still operating normally?

Dashan: We did a tally in the first half of this year. Fewer than one-third — about 60 to 70 — are still operating well, continuing development, with tokens trading normally and funds sufficient to support long-term development.

If we also count those that successfully issued tokens and performed normally at the time but later stopped being active because of operating difficulties, then overall more than half should have performed normally.

But about 20%–30% of projects involve rug pulls, going to zero, default, or damage to investors' rights — a very high proportion.

Many of these problems are related to founder integrity. Some teams did not manage the funds according to the agreement after raising money, some were thinking about hyping and cashing out from the very beginning, and some simply stopped communicating once the money ran out.

Maodi: When a project breaches its agreement, how do you usually defend your rights?

Dashan: If the amount is not large, we may just have to give up, because cross-border litigation is very costly. But if a large amount is involved, we will also take legal action.

We once fought a lawsuit in the United States, spending more than $900,000 in legal fees and eventually recovering more than $1.3 million. Although the profit after costs was limited, it at least showed that we will pursue accountability.

There was also a project that used Waterdrip Capital's name without authorization to promote its fundraising and attract retail investors to buy tokens. In such cases we even took the initiative to report to the police, clarifying that we had not participated in the investment.

These experiences have made us attach more importance to due diligence, including verifying founders' backgrounds, financial conditions and credit records.

Sharply reducing Crypto investment in 2026, shifting to AI and chips

Maodi: How many projects did Waterdrip Capital invest in during 2026? Why reduce Crypto investment?

Dashan: This year we invested in a dozen-plus projects, of which only about three or four are purely Crypto-related; the rest are mainly AI and chips.

Even these few Crypto projects mainly revolve around practical applications such as AI Agents, Agent payments and tokenized stocks.

We have reduced our investment in the Crypto primary market, but our overall investment amount has actually increased — it is just that more money has gone into AI and chips.

We started researching these new directions in 2024. Blockchain has been developing for more than a decade, its infrastructure is already relatively mature, and there have been no particularly major underlying technological breakthroughs in the past two or three years. As a VC, we need to look for new technological opportunities.

Crypto is no longer the protagonist; in the future it should serve AI

Maodi: Do you think crypto VCs shifting to AI is an industry trend?

Dashan: I think what matters more is that the positioning of Crypto has changed.

In the past we thought blockchain would completely change traditional finance and even replace some institutions on Wall Street. But looking at it now, although Bitcoin, DeFi and others have made great progress, blockchain has not completely transformed the traditional financial system.

I now think Crypto is better suited to being the infrastructure of the digital economy rather than the protagonist. The protagonist of the future may be AI, especially the AI Agent economy.

Blockchain should think about how to serve AI, including payments between Agents, data security, privacy protection and operation records.

I am no longer bullish on projects that circulate only within Crypto itself, such as developing a new public chain just to raise funds, or attracting users simply by issuing tokens and hyping them.

In the future, two categories deserve more attention: RWA serving traditional finance, and infrastructure serving AI.

Opportunities in AI Agent payments, data privacy and zero-knowledge proofs

Maodi: Which directions specifically are worth watching?

Dashan: First, data privacy and AI security. Many governments, hospitals, schools and enterprises want to use AI, but dare not hand sensitive data directly to large models.

Blockchain, encryption technology and zero-knowledge proofs can help protect data and record key operations, making AI systems more transparent and auditable. In the future these technologies may be combined with specialized hardware and computing power; we have already invested in some related projects.

Another important direction is AI Agent payments. For ordinary users, traditional payment tools such as credit cards and Alipay are already convenient enough, and they do not necessarily need Crypto.

But Agents are different. They do not have ID cards and bank accounts in the traditional sense, and it is also hard for them to complete KYC like humans do. In the future, Agents may purchase large amounts of data, computing power and other services from each other, requiring automated payment and settlement, and Crypto may be a more natural solution.

To some extent, much of the blockchain infrastructure we have built in the past may not have been prepared for humans, but for future AI Agents.

Maodi: When might AI Agents really drive Crypto's development?

Dashan: I think we first need to wait for AI Agents to become widespread. Agents can already book tickets, write research reports, send emails and even execute on-chain transactions, but not many people use them at high frequency yet.

Only when Agents enter work and daily life on a large scale will demand for payments, identity verification, data security and operation auditing really emerge.

This process may take one to three years, or it may be faster. The next development opportunity for Crypto may appear after AI Agents become widely adopted.

Part Two: Xiao Yang, a Small Crypto VC — Projects Losing Contact, Difficult Exits and a Halt in Investing

A $20 million fund that invested in more than 70 projects in 2024

Maodi: Could you introduce your fund's size, team and investment strategy?

Xiao Yang: We are a small VC with about $20 million in assets, mainly investing in early-stage projects in Crypto and frontier technology, including infrastructure, DeFi, AI, data security and privacy.

We focus on the Pre-Seed and Seed stages; our capital comes from individual LPs, family offices and other VCs, and the team is about 10 people.

The fund roughly began investing in 2023, and I joined in 2024. In 2024 alone we invested in more than 70 companies, and have invested in about 100 projects in total.

But in 2025 we invested in fewer than 10 projects, basically all concentrated in the first quarter. After that we stopped all new investment — not only in Crypto, but also in AI and other fields.

Why stop investing? About one-third of projects can no longer operate normally

Maodi: Why did you suddenly stop investing after 2025?

Xiao Yang: Mainly because the existing portfolio already had about 100 projects, which requires a great deal of effort to manage. At the same time, we have become increasingly clear-eyed about the risks of early-stage investing, and the fund's principals have gradually shifted their energy to other matters.

Starting in 2025, some portfolio projects stopped operating or scaled back their business one after another. Even if the sector itself has potential, that does not mean every team can build the product or secure the next round of financing.

So we decided to pause new investment and concentrate our energy on existing projects, helping them find cooperation opportunities and connect resources.

Maodi: Of the roughly 100 projects at present, how many are still operating?

Xiao Yang: Based on actual tracking, about two-thirds are still operating. But some of them are merely scraping by and may not last much longer.

More seriously, more than half of the projects are hard to maintain effective communication with investors. Some teams are still updating their X accounts but rarely reply to messages anymore, so we can only judge from their official websites and social media whether they are still operating.

The number of projects that genuinely keep reporting progress to investors is far smaller than the number that appear to still be operating.

When projects go dark for a long time, small VCs have little recourse

Maodi: If a project team does not reply to messages, what do you do?

Xiao Yang: We try to reach them by email, instant messaging and other contacts, but some teams will not reply no matter what.

Because our individual investment amounts are relatively small, we may not be seen as very important in the project's eyes compared with large funds.

Legal measures are also difficult. Even if the other side replies and tells you the project is still under development, it is hard to verify actual progress. When contact is completely lost, it is even harder to gather evidence and take action.

So I feel small VCs are in a very weak position in post-investment management.

Maodi: How do you evaluate projects when investing early?

Xiao Yang: We place relatively great emphasis on the founder and team, and many projects come from long-term partnerships or referrals from acquaintances.

What matters a lot in early-stage investing is betting on people: We hope the founder is capable of adjusting the business even if the original direction fails — that is, pivoting.

But we also assess whether the technology solves a real problem and whether there is a sustainable revenue model. Advanced technology does not mean a project will necessarily succeed.

The fund is roughly at book breakeven; real exit returns may be worse

Maodi: After investing in about 100 projects, is the fund as a whole profitable or losing money?

Xiao Yang: On paper we still have a small profit, but not by much — basically close to breakeven.

If we take into account how the projects are actually operating and how much can really be exited in the future, I think there may be a loss.

The main problem is that many tokens have not yet unlocked, or the projects have not yet reached the exit stage. Book valuation does not equal the amount that can actually be cashed out.

Early-stage VCs usually rely on a small number of successful projects to cover losses on other investments. We are relatively lucky in that our portfolio includes more than 20 pure AI projects, some of which are performing well. If these projects can exit smoothly in the future, the fund as a whole may still achieve a positive return.

Maodi: Which projects have the best returns? Have any gone completely to zero?

Xiao Yang: The projects currently showing better book returns include Caplabs, Prime Intellect and Excel Labs, many of which come from the AI field.

On the loss side, we once invested in an AI graphics analysis project related to the fashion industry, putting in close to $300,000, and its valuation ultimately went to zero, basically with no return.

This shows that AI is not a sure-thing direction either; early-stage investing always carries high uncertainty.

No investment plans for the next year or two; crypto VCs are broadly slowing down

Maodi: Do you plan to resume investing in the future? How are other crypto VCs doing?

Xiao Yang: As far as I know, we probably will not resume new investment this year or next year. We still have dozens of projects that need continued tracking, and we will put our energy into the existing portfolio.

Although we have stopped investing, the team has not laid off staff because of it.

From what I have seen, many crypto VCs are investing less frequently, and some have not invested in new projects for a long time. Some funds have made personnel adjustments, while other institutions have kept their original teams.

I have also heard that some funds are facing dissolution because of heavy losses, but these reports have not been verified and cannot represent the whole industry.

Maodi: Which investment directions do you favor in the future?

Xiao Yang: We pay more attention to real user demand, technology implementation and sustainable revenue, rather than short-term hype.

At present we are relatively focused on AI generation, data technology, intelligent financial infrastructure, and film and television, among other areas. But following these directions does not mean we will immediately resume investing.

A star project that raised tens of millions of dollars eventually shut down

Maodi: You have worked at this VC for more than two years — what has been your biggest impression?

Xiao Yang: After many projects finish fundraising, they are not very proactive in communicating with investors. Some do not update their progress on their own initiative, and some do not reply even when investors ask.

It is not that we only want to supervise projects; we also want to help them connect with resources and partners. But some teams are unwilling to build long-term cooperative relationships with investors.

For small VCs this situation is especially obvious. When a project raises funds smoothly and attracts a lot of market attention, our investment amount is small and it is hard for us to gain enough say.

Maodi: Is there a particularly typical case?

Xiao Yang: We once invested in a well-known project that raised tens of millions of dollars and drew a lot of market attention, but it later stopped operating.

At the time we had a meeting with the team to ask about some basic business progress and operating conditions, and their attitude was very poor. We raised several questions in a row, and they answered directly that it was "not convenient to disclose."

Although our investment amount was not large, as investors we should at least have the right to understand the basic situation of the project. Looking back now, I am not surprised that this project ultimately failed.

Maodi: Is this kind of disrespect toward investors common?

Xiao Yang: Actually it is not common. Most projects, even if they are slow to reply, at least maintain basic respect.

But for small VCs this situation is hard to handle. Especially when a project has raised a large amount and draws a lot of market attention, it is very difficult for us to require them to report regularly.

I would also very much like to know whether large VCs face the same problems, or whether it is because our investment amounts are too small that we are more likely to be ignored by project teams.

Join the official Coincamps community:

X: https://x.com/coincamps

Telegram: https://t.me/coin_camps