远山洞见

In the crypto industry, money has never cared about your aesthetics.

There’s more USDT on TRON than on Ethereum. You might need to read that again.

TRON: $89 billion. Ethereum: around $65 billion. 47% of the world’s USDT runs on CT—almost nobody is seriously talking about it on-chain.

Honestly, the first time I saw these numbers, I was also stunned.

On CT, TRON is basically a joke. Whenever Justin Sun is mentioned, people instinctively start debating and teasing. Well, except for Sun Xue.

But take the emotions out and look at the numbers only: Q2 stablecoin market share was 28.7%, with 3.5 million daily active users. Protocol fees were $89 million—second in the whole industry, only behind Hyperliquid.

In the breakdown data: for small USDT transfers under $1,000, TRON’s share was 43% in Q1, rising to 52% in Q2—over half.

But there’s been one problem with those $89 billion: it’s semi-closed.

TRON’s cross-chain channels with other networks aren’t very usable. The money is there, but it doesn’t move.

deBridge recently connected its MCP server to TRON. Developers and AI agents can programmatically call TRON’s cross-chain liquidity to do routing and execution.

A lightweight interface that connects to the world’s largest stablecoin pool. After launch, TRON contributed about 40% of deBridge’s monthly transaction volume—an underestimated chain, with usage higher than most “proper” chains.

In a bear market, I’ll keep backing deBridge.

What is truly safe in the crypto world?

First, the industry said CEXs aren’t safe—so you should self-custody.
Then Coldcard was stolen for $100 million. Did the hackers touch any device at all?
CEX isn’t safe, and self-custody also isn’t safe. So what is safe, exactly?

In the past 30 days, three exchanges announced they were shutting down. A project valued at $3 billion still has less than $500,000 in assets left. The “most secure” hardware wallet was stolen for over $100 million.

A supposedly “most secure” storage method in the crypto world: a firmware vulnerability hidden for 5 years. In 72 hours, more than $100 million was drained. The hackers never touched a single Coldcard device. They didn’t even need to.

In March 2021, a firmware update moved the random-number source used to generate private keys from the hardware chip to software. This means that over the past five years, all wallets generated with this batch of firmware had far less “randomness” in their private keys than you thought. The attacker computed the private keys offline, then swept 4,585 addresses in four waves—and the fourth wave hasn’t stopped yet.

Coinkite has stopped shipping and destroyed all inventory machines.

Over the last month, AscendEX shut down. BitMEX announced it would close on September 23. BitMart followed, saying it would shut down in January next year. Three CEXs in one month. Q2 perpetual contract trading volume fell overall by 10%. The BMX token briefly hit zero.

At a time like this, you remember your holdings in deBridge. The more you look at its architecture, the more it feels like its design philosophy is the exact opposite of the “security incidents” above.

It does cross-chain without locking any funds—no liquidity pool. When you want to bridge, you send an intent. A solver competes to execute; settlement is completed while the assets are in motion—never stopping, turning it into a target.

Without a liquidity pool, there’s nothing that can be stolen. It’s not “putting money somewhere safer.” It’s that there isn’t even “that place” to begin with.

Compare a set of data and you’ll know what real demand looks like: Movement raised $141 million, but its on-chain revenue is about zero. deBridge raised a total of $10.5 million; the protocol’s cumulative revenue is over $13 million—and 100% of that revenue goes into buying back DBR.

Coldcard destroyed inventory, BitMEX is counting down to close, and Movement is calculating in court who owes whom. Last month, deBridge went live on TRON and captured 40% of cross-chain traffic from the TRON USDT ecosystem—86 billion USDT.

Honestly, there are fewer and fewer things in this industry that can survive this round.

I wrote for you, and I wrote for ordinary people.

Buffett once said something that’s been quoted to exhaustion, but very few people actually listen to it.

He said that after he died, 90% of the money he left to his wife would go into an S&P 500 index fund, and 10% would go into short-term government bonds.

The world’s best stock pickers choose an index for their own families.

Are we better at picking stocks than he is?

I personally also hold some individual stocks, but the proportion is quite small. I shared this with everyone before: overall, I focus mainly on indices. And I always keep cash. The allocation among these three—individual-stock offense, index as the core holding, and cash as the ticket to enter the market—varies from person to person. But if any one of them takes up 100%, something will go wrong sooner or later.

Especially for people who have just entered the U.S. stock market. We haven’t even experienced a full cycle of ups and downs there. And then we rush in to buy individual stocks right away—this is no different from someone who can’t swim jumping straight into deep water.

Start with an index first. Go through a full round of volatility, then feel what level of psychological tolerance you truly have. Only then consider whether to use a portion of your money to touch individual stocks.

This month, many people lost money, and many people made money. But whether you gained or lost, while that memory is still fresh, you should ask yourself a few questions.

Is my money “spare” money? If it isn’t, withdraw the portion that isn’t spare money first—regardless of whether you’re currently up or down. If you use money you might need to repay as investment money, you’ll never be able to hold on.

Do I have leverage? If you do, take it off today. Leopold’s 20 billion fund was right on direction—and still died by leverage. Do you think you can withstand more than he could?

Do I have cash? If you’re fully invested, then in the next major downturn you’ll only have two choices: either tough it out uncomfortably, or cut your losses. With cash, you have a third option: add to your position.

What exactly am I buying? If you can’t even clearly explain what this company does, what money it makes, and why it’s worth that price, then what you’re buying isn’t an investment—it’s a bet.

Every market crash is a free medical checkup.

Rather than cursing the market, when your memory is clearest, adjust your positions properly.

The next market crash will definitely come. The only difference is whether you’re prepared.

Although I don’t know which one to believe, there’s always someone lying.

Storage is down 30%, optical modules are down 60%, and orders are booked for delivery in 2027.

Samsung, SK hynix, and Micron—this year’s total DRAM and HBM production capacity for all three has already been fully allocated. NAND is also basically pre-sold out. It’s for 2027. Next year’s supply has already been locked in this August.

As SK Group’s chairman himself said, AI semiconductor demand in 2027 may rise another 60% to 100% compared with this year. But the combined supply those three can provide is only enough for about 60% to 70% of downstream demand.

Demand exceeds supply, orders are locked, and prices are still rising. Micron’s HBM4 unit price is expected to jump from $2/GB to $4–$5/GB by 2027—doubling.

Then take a look at the secondary market. Micron fell from its 52-week high of $1,255 to $830—down more than 30%. AAOI is even worse: it was $233 in May and is now below $110—cut in half and then another 30% off.

On the industry side, people are疯狂抢产能 (rushing to secure capacity) and signing 3- to 5-year long-term orders like crazy. The secondary market is crushed.

These two scenes exist at the same time. Someone must be wrong.

Either everyone on the industry side is crazy, spending tens of billions to抢 a product that’s about to collapse. Or everyone in the secondary market is crazy, dumping an asset that will be in short supply until 2027.

I lean toward the latter being crazy. Because what the industry signs is real money-long-term contracts; if they default, they have to pay penalties. What the secondary market sells is sentiment—no need to pay default penalties for panic.

Now look at AAOI. The white-haired stock “god” Serenity called it out in early June, saying it could be 4 to 5x in 12 to 24 months. Two months later, it first dropped 60%.

Today, another variable has arrived.

The FCC is preparing to prohibit the import of new Chinese optical module models, with implementation expected within the year. One company, Jieshi?（中际旭创）, has about 27% of the global data center optical module market. In June, it was already listed on the Pentagon’s military-related entities list.

But, to be honest, if they truly impose a total ban, North American manufacturers won’t be able to fill that gap for years. Chinese optical modules make up too large a share of global capacity. If you ban them today, your own U.S. data center buildout pace will be dragged down tomorrow. So it’s likely to be phased restrictions, not a full stop immediately.

Even with phased restrictions, it’s a long-term positive for a domestic substitution play like AAOI. Because every batch of China-made products that gets restricted will send a batch of orders toward it.

Back to the very first question.

Storage is down 30%, optical modules are down 60%, and orders are booked for 2027. So who is lying?

In a market full of air coins,
"being able to prove you’re making money" is itself the strongest moat.

deBridge’s June monthly report is out.
Cross-chain transaction volume is $863 million, protocol revenue is $684,000,

The reserve fund has accumulated 605 million DBR tokens, accounting for 6.05% of the total supply.

Most DeFi projects rely on token emissions to make their numbers look good—using high APR subsidies to attract capital. When the subsidies stop, the money runs off, and the token price crashes along with it.

deBridge takes the opposite approach. Using real protocol revenue, it repurchases its own tokens on the open market and locks them into the reserve fund. It’s not printing tokens and paying subsidies—it’s earning money and buying back.

This cycle has been running for a long time.

A protocol that can form a positive feedback loop of "real transactions → real revenue → buy back its own tokens" is scarce on-chain. When the reserve fund grows to a size of 6% of total supply, that’s very strong.

On the U.S. stock side, the AI theme just completed the strongest quarter since 2020. The Nasdaq rose 20% in Q2, and market sentiment is already very hot. In crypto, BTC is still around $60,000, and altcoin valuations and sentiment are both at low levels.

If you’re positioning for altcoins, you should definitely pay attention to DBR, because the first batch of capital to be found will be projects that have real revenue, buybacks, and fundamentals that can essentially verify themselves.

TVB teacher’s insights! In the new cycle, layout shanzhai coins—must see $DBR #deBridge

Users are loyal to opportunities, not chains.

After Phantom integrated with Hyperliquid, it brought $45 billion in trading volume.

Users of Solana wallets moved to Hyperliquid to trade perpetual contracts. Two completely different chains are in between, but users don’t care.

They see a trading opportunity on Hyperliquid—and their money goes there.

In the past, whenever people talked about DeFi, it always turned into an argument over which chain would win. "Are you a Solana user, or an ETH user, or an Arbitrum user?" It feels like team sports.

Now, it’s clear that the argument itself is wrong. Nobody cares which chain they use. What everyone cares about is where they can make money today.

What users want isn’t "to trade on some chain." It’s "to complete trades as fast as possible."

There’s a very real bottleneck in this process. Your money is on Solana. You see a trading opportunity on Hyperliquid—how do you get there?

The old workflow looks like this: first, find a bridge to move SOL to the target chain, wait for confirmation. Then you discover you don’t have enough gas tokens. So you spend more on gas, swap into the required assets, and only then place the order. By the time you’ve done all that, the market may already have moved on.

And traditional bridges have a bigger problem too: they require locking a large pool of funds in the contract. In the past few years, cross-chain bridges have been repeatedly hacked—the root cause is that those locked funds in the pool are just too tempting.

deBridge’s architecture is completely different from these bridges. It has zero TVL—no funds are locked anywhere inside the protocol.

When you send a cross-chain transaction, in essence you’re broadcasting an intent: "I want to swap USDC on Solana into assets on Hyperliquid."

Then a group of professional solvers compete with each other to execute the trade. Whoever can give you the best price takes the order.

Getting $45 billion in volume isn’t only because Hyperliquid’s product is good or because Phantom users are many—it’s also because deBridge sits in the middle.

If the route isn’t viable, users can’t get to the opportunity they see—or by the time you’ve found a workaround, the opportunity is already gone.

deBridge is doing exactly that—connecting everything. It doesn’t compete with any chains, and it doesn’t compete with any exchanges. It ensures that funds on all chains reach any place where there’s an opportunity as fast as possible.

The chain isn’t the destination. Opportunity is. Whoever can deliver capital to the opportunity first gets the next round of traffic distribution rights.

$DBR

1/ When doing swaps on-chain, most folks just fire up an aggregator, plug in the amount, and hit confirm. They never bother to compare prices.

2/ In traditional finance, there's a concept called best execution, where brokers are legally bound to help clients find the best market price.

On-chain, that doesn't exist; whatever aggregator you use is the price you get, and you have no idea what the spread is.

For the same transaction, the price differences between different aggregators can be much larger than most people think. I bet few would go through the hassle of manually comparing prices; it’s just too much work.

3/ After watching @Sherrypeter's video, I realized someone is actually tackling this issue seriously.

There's a program that checks multiple aggregators' prices simultaneously, compares them, and automatically selects the best route to execute. It’s not just linking to one DEX; it treats several aggregators as suppliers to find the cheapest one for you to transact.

4/ Yep, I'm holding $DBR.

In the last three months, the trading volume for same-chain swaps on deBridge hit $750 million, accounting for 40.38% of the total trading volume of $1.8575 billion. The number of same-chain transactions exceeded 450,000, making up 53% of the total 840,000 transactions.

5/ I've been holding DBR for a while now. The logic when I bought was straightforward: the team has strong delivery capabilities, and the data background is solid.

Looking back now, it’s clear they haven’t just stuck to the "cross-chain bridge" label. The bridge is the starting point, but it’s not the limit.

Feeling bullish~!

Real builders, real numbers.

CEX's perpetual contract monthly trading volume dropped from $7.11 trillion in 2025 to $4.69 trillion in 2026.

At the same time, Hyperliquid, an on-chain DEX, hit a monthly trading volume of $190 billion.

Last week, deBridge released some data showing that over $1.7 billion flowed into Hyperliquid through cross-chain bridges in the past year. When I saw that number, I thought, isn't that just what I was doing a few days ago?

I remember around the 15th, everyone was speculating on alt fun on Hyperliquid, the on-chain version of pump fun, but with perpetual contract leverage.

A lot of folks in the community wanted to get in but didn’t know how to move their funds over. I tried a few options and found #deBridge was the smoothest with the least slippage, so I recommended it in the group.

When you look at one person's experience in the context of $1.7 billion in funds flowing in, it suddenly becomes more than just 'user-friendly.' It's the siphoning effect of the entire chain at work.

Hyperliquid now has a TVL over $5.1 billion, up from $4.2 billion in early May. HYPE has surged more than 156% from its low at the beginning of the year.

Just yesterday, it hit an all-time high of $64. Last week, institutions poured in $72 million, with a16z's wallet buying nearly $10 million of HYPE and staking it right away.

What matters to me isn’t the price, but a comparison. CEX's average monthly perpetual contract volume is shrinking, while on-chain DEX's average is rising. In 2025, DEX averaged $531.6 billion, and in the first four months of 2026, it jumped to $611.5 billion.

The direction is no longer debatable.

To be honest, I have two typical friends. One has been telling me to dollar-cost average into $HYPE since it was $20 last year, claiming it's a major on-chain trend. I didn't listen back then.

The other bought in directly below $50 a couple of days ago. Two types of people: one sees the trend early, and the other acts fast. I'm neither.

Currently, I don’t hold any HYPE. But I’m looking for new alpha opportunities as funds continue to migrate to this chain. That's how I noticed alt fun.

In this process, there's a type of project quietly raking in profits – cross-chain bridges. Whether you're chasing memes or making serious trades, money has to cross the bridge.

deBridge is one of them; their protocol revenue is real and transparent, and the foundation uses this money to continuously buy back DBR.

Not all infrastructure can withstand the test of time. Personally, I hold $DBR.

The history of cross-chain bridges getting hacked can basically be summed up in one sentence: if there's money in the pool, hackers are gonna show up.

Ronin got hit for 625 million, Wormhole was breached for 320 million, and Nomad lost 190 million. Each incident tells the same story: the funds locked in the bridge turned into a giant honeypot.

Alex's thread is about security, but what he really wants to highlight isn't 'we're doing security well'; it's a more fundamental issue: you shouldn't be fortifying the honeypot, you should make the honeypot disappear.

The 0-TVL structure of deBridge embodies this idea. The protocol locks no funds; users only express intent, and market makers cover the execution with their own capital.

No shared pool, no TVL, and when hackers come, they find the vault is empty. This flips the security logic of most projects on its head.

The industry still revolves around who has done more audits, who monitors faster, and who offers higher bounties, but all of this assumes 'the pool is definitely going to exist' and just slaps on patches.

Alex is spot on; audits and monitoring are downstream measures, and architecture is upstream. Of course, 0-TVL isn't without its costs. You need enough solvers willing to front the capital, and the market maker network has to be deep enough. This isn't a model every team can pull off.

But in terms of direction, removing 'things that need protection' from the protocol could be the biggest paradigm shift in DeFi security.

AI making attacks cheaper only solidifies this judgment.

Where's the next 10x opportunity in AI data centers?

This is a post in the English-speaking area with 570K views, discussing a hot topic.

It's not about GPUs, not light modules, not storage chips. It's about the chip that powers them.

NVIDIA is pushing something that most investors in the Chinese market haven't noticed: the 800V DC architecture.

Current data centers rely on traditional AC power. Electricity comes from the grid and must pass through several layers of conversion before reaching transformers, UPS, rectifiers, and distribution units, losing some energy at each step.

In a large data center, the losses from light-to-power conversion account for 10%-15% of total electricity consumption.

NVIDIA's solution is to switch the entire data center power distribution system from AC to 800V DC.

Electricity comes from the grid and gets converted directly to 800V DC, delivered straight to the server racks, eliminating all those layers of conversion.

The current results are:
-- 5% improvement in end-to-end power efficiency,
-- 70% reduction in wiring,
-- 70% decrease in maintenance costs,
-- single rack power capacity increased from the current 100-150 kW to 600 kW, even up to 1 MW.

Vertiv, one of the largest data center power distribution manufacturers globally, confirmed in their latest earnings call that 800V is "a 2027 thing," with products launching in the second half of this year.

Schneider's outlook is even bolder: a large-scale rollout starting in 2028.

What does this architecture revolution need? It requires power semiconductors that can withstand 800V high voltage—silicon carbide (SiC) and gallium nitride (GaN).

Traditional silicon chips can't handle efficiency and heat dissipation at this voltage; only SiC and GaN can play.

Everyone's eyeing GPUs. Few are watching the chip that powers them.

Next, let's chat about specific targets with high consensus in the English-speaking area.

Is the US stock market going to crash this year?

To be honest, I don’t know, and nobody really does. But I can lay out a few key indicators for you to judge yourself.

Let’s start with the bullish side.

Q1 earnings season just wrapped up, and the overall earnings growth for the S&P 500 constituents is 28.6%, the best quarter in the past three years.

Apple's Greater China revenue up by 28%, Microsoft’s AI business is doubling annualized, Palantir's revenue up by 85%.

The Fed hasn’t cut rates yet, but inflation has dropped from 9% to the 3% range. The market expects at least one rate cut in the second half. Once the rate cut cycle starts, it's a direct win for the stock market.

AI is a variable we haven’t seen in past cycles. This isn't like the 2000 bubble where “internet companies had no revenue but were valued in the billions.”

NVIDIA made over $40 billion in a single quarter, and Microsoft’s AI business is on track for $37 billion annualized. AI commercialization is already generating real cash flow.

Now, let’s talk about the risks.

The S&P currently has a price-to-earnings ratio of about 22x, which isn’t cheap. Some AI stocks in the Nasdaq are already at 50-100x PE. If earnings growth slows down, these valuations could get compressed quickly.

Geopolitical factors are the biggest uncertainty. Tariffs, Taiwan Strait tensions, the Middle East—any black swan event could trigger a short-term nosedive.

Plus, market congestion is increasing. For example, last week the storage chip sector dropped 3%-6% in a single day, and the Philadelphia Semiconductor Index fell 3% in one day, indicating that certain sectors are getting crowded and even a slight shake could cause a sell-off.

So, what’s my conclusion?

What’s the probability of a 10%-15% correction in the US stock market by 2026? Quite high, you could almost say it’s a certainty. Over the past 46 years, the S&P 500 has experienced an average maximum drawdown of 14% each year. Every year. A correction isn’t a surprise, it’s the norm.

What’s the probability of a crash of 20% or more in the US stock market by 2026? Not zero, but not high either. Historically, a crash usually requires two conditions to be met simultaneously: worsening earnings + tightening liquidity.

Right now, earnings are accelerating, and while liquidity hasn’t loosened, it hasn’t tightened further either—so both conditions aren’t met.

The real question isn’t here. The real question is: what are you waiting for?

Is the US stock market going to crash in 2026? What are the odds of a crash?

I received a DM last night, and this question has been asked at least ten times recently, phrased differently but meaning the same: Is the US stock market too high right now, and will I be holding the bag if I jump in?

I get the anxiety. The S&P has surged over 30% from last year's lows, the Nasdaq is up 40%, and Nvidia's market cap is bigger than the entire German stock market. Every number seems to be screaming "bubble."

But before answering "will it crash," let’s do a little hindsight analysis.

In the past 50 years, the US stock market has experienced 7 significant drops of over 20%.

The 1973 oil crisis, the 2000 dot-com bubble,
2008 financial crisis, 2020 pandemic, and the 2022 rate hike cycle.

Every time the market dipped, everyone thought it was the end of the world.

However, if you bought into the S&P 500 at any point in the last 50 years and held for 10 years, your returns were positive.

At any point in time. Including buying at the peak just before the 2007 financial crisis.

Stay tuned for more.

What payments need is certainty, not probability.

deBridge's Exact Output launched last year isn't really about tech issues; it's about trust.

Knowing exactly how much you send and how much you get—that's the prerequisite for cross-chain payments to truly go live.

On one hand, we’re shouting about replacing SWIFT, and on the other, we can't even ensure precise transactions. This industry has been a bit wild before.

Sit back and enjoy $DBR.

Last night I caught CZ's AMA, and he said he only looks for one type of founder during a bear market.

It's the ones who keep launching protocols and building while the market is tanking, "I'll be there for two years too."

Anyone can see the valuations in a bull market, but it’s hard to tell who will survive when the tide goes out. CZ's criteria are pretty straightforward: those who keep grinding after a dip are likely the ones who genuinely want to make a difference and have a sense of mission.

Today I happened to read a long piece from the deBridge CEO.

He talked about how in 2021, when they decided to support Solana, $SOL dropped from 260 to 10. The sentiment was gone, liquidity dried up, and they were in a state that felt like "building against the wind" for two years,

but they still managed to compress cross-chain transfers from 15 minutes down to just a few seconds. After launching, they only saw 10-20 new users daily, and they kept at it for several months. Now that figure is 100 times what it was back then.

The CEO's exact words: "Focusing solely on the current reactions will make you miss out on what's forming beneath the surface."

As of today, deBridge has processed a cumulative $19.1 billion in cross-chain transactions, with 5.83 million transactions and DAO revenue of $23.41 million. They’re at the top of their niche.

CZ's standard perfectly aligns with deBridge.

A team like this deserves a big respect. Wishing them all the best.

I have never bought LABUBU and have no feelings about this kind of product. There are many people around me who like it, but I have always not really understood.

Previously, when I saw those videos of people queuing to grab blind boxes, the word that came to my mind was "bubble"; a harsher word would be "money laundering" and "Ponzi". It's that kind of situation where you see something popular without logic, and your first reaction is "there must be a problem here."

This time, Duan Yongping's card flip made me honestly think about this matter. It's not about "should I buy Pop Mart"; it’s about whether my previous judgment of "not understanding means it’s a bubble" is itself a form of laziness.

Not understanding is just not understanding,
not understanding does not mean it has no value,
nor does it mean it’s a scam.

A person who has been in value investing for twenty years spent eight months repeatedly looking at it, and finally said, "I want to take back what I said." At least he is telling me one thing: when some things are not understood, don’t rush to make a determination.

The insight here is not in Pop Mart itself, but in how we treat the three words "not understanding."

So what does this have to do with ordinary people?

It’s not about having you copy Duan Yongping’s homework to buy Pop Mart; the 13F holdings that come out are all lagging, and ordinary people copying homework are basically just taking over. The truly useful things are the other two.

First, distinguish whether you are looking at "acceleration" or "total amount."

The assets in your hands, whether they are A-shares, U.S. stocks, or BTC, what do you look at when you open the app every day?

Is it how much it has risen or fallen this week (acceleration), or what this company or asset will look like five years from now (total amount)? This question will filter out 80% of anxiety.

Second, admitting when you are wrong is harder than being right, and you have to do it when the price is falling.

Most people can change their minds when prices are rising; that’s called chasing up. Very few can change their minds when it has dropped by half; that’s what true reassessment is called.

Next time you face something you previously judged incorrectly, ask yourself a question—if it is at a high point now, would I admit I was wrong? If you wouldn’t, then you haven’t really thought it through yet.

【After the Cut, What Did Duan Yongping Actually See】

Duan Yongping has rejected Pop Mart three times.
This week, he publicly retracted all of his previous statements.

What makes a person who has been engaged in value investing for twenty years willing to admit they were wrong?

In the past year and a half, he has spoken to Pop Mart four times.

The first time he said "I don't understand"
The second time he said "I don't understand 10 years later"
The third time he said "What if in two years no one wants it?"
The fourth time, he said, "I want to retract my previous statements."

The time was this week. Pop Mart's stock price dropped from a peak of 340 HKD to 149 HKD, halving its value, with a market value evaporating by 185.8 billion HKD.

Why is this a big deal?

Because "not understanding, not touching" is an almost unbreakable rule in Duan Yongping's investment bible. He has held onto Apple for twenty years without selling, and has held Moutai for over a decade, rejecting things he doesn't understand far more often than he has made purchases.

This time he publicly said "I retract", which in the value investment circle is akin to a monk admitting to breaking a precept.

I went to check Pop Mart's 2025 financial report. Revenue was 37.1 billion, up 185% year-on-year. Net profit was 13 billion, up 293% year-on-year.

Overseas revenue accounted for 43.8%, with America growing by 748% and Europe by 506%. This is not just "acceptable", these are epic numbers. But on the day the financial report was released, the stock price fell by 22%.

The market saw "2026 guidance is only 20% growth", meaning the acceleration is decreasing.

Duan Yongping saw something else, he wrote on Xueqiu: "The speed of economics is actually the acceleration of physics, and investment buys the total amount of the future."

You focus on the decreasing acceleration, while I focus on the speed itself which is already quite fast, multiplied by a sufficiently long time, the total mileage of the future remains astonishing.

To be continued, next post.

The deBridge monthly report is out, and several key signals are worth highlighting.

1/ Let's look at revenue first. In March, the protocol generated $481,000 in revenue, with a cross-chain transaction volume of 533 million and 217,000 transactions running on 23 chains.

In this current market environment, protocols that can generate this level of cash flow can be counted on one hand.

2/ Next, let's see where the money has gone. The foundation is using protocol revenue to continuously buy back DBR in the open market, with a total of 486 million tokens accumulated, accounting for 4.86% of the total supply.

Real cash has been bought back from the market. The money earned by the protocol is flowing back into the token itself, and this path is becoming more solid.

3/ Lastly, let's look at the direction. The most interesting part of this monthly report is the deBridge MCP, a cross-chain trading protocol aimed at AI agents. AI workstations like Claude and Cursor can directly call it, allowing agents to autonomously complete complex on-chain operations.

Revenue supports the fundamentals, buybacks lock in the chip structure, and the product points to the narrative of the next cycle.

No wonder I have high hopes for $DBR!