Monday, stock a is going up. The above index rose by 1. 78 per cent, the entrepreneurship board by 2. 52 per cent and the double-up phase was higher。
But it's structural, it's wheeled. On the whole, low-level large-scale financial booms, small metals continue to be strong, semiconductor plates that began to rebound last week continue to grow strong, while high-level ais materials resonate。
In the current state of affairs, we have repeatedly discussed whether cattle or cattle, but not all stocks can rise. The days of "showing up and down" are gone. With regard to technology, there is now a two-stage divide。
Why is there division
The mid-july reporting season is approaching and the performance validation window is opening. Hard technologies with real orders, cash flows and technical barriers will be further identified; false technologies that elude concepts, tell stories and fail to perform will be accelerated。
At the same time, the signals at the supervisory level are very clear: the cvm scrutinizes the technology's hot spots, the concept of scrambling, and the fifth set of standards has been expanded to the area of artificial intelligence; and the ministry of industry and communications has published the "intelligence+intelligence" implementation opinion on innovative development, which clearly enhances the development of high-end photo-electric chips and devices。
Policies are being faked and roads paved. Cleaning up concepts while turning green lights on real technology. And the next step, probably, is to deepen the dichotomy between real and fake technology。
And for low-level old deng's rebound, it's more like a super-thump, not too sustainable。
Discussions on the financial crisis have increased in recent times. The united states debt, the volatility of the ai concept unit and the continuation of fighting in the middle east — each single issue is sufficiently insomnia。
But what i want to say is that systemic financial crises are not going to happen at this time。
The following angles give you a cool reference。
When everyone's worried, it's safe
There is a simple pattern in financial markets: a crisis never breaks out when the crowd is alert, and suddenly comes when the song and dance are flat and unguarded。
Prior to the subprime crisis in 2008, few people were truly aware of risks. The “discretion” was the most popular self-consolation at the time. And today? From the media to institutions, from the diaspora to fund managers, everyone talks about bubbles, debt and crises. Such universal vigilance is in itself a safety pad — panic is a risk “vaccination” and not a risk “symptoms”。
Citi's report confirms this: although bear city early warning indicators have reached their highest level since 2008, market sentiment “has not become overly fanatical”. In other words, the alarm went off, but everyone was still awake。
Ii. High united states debt, but different in nature from 2008
See firsthand data: as of april 2026, us public debt had reached $31. 3 trillion, almost equal to the size of the entire united states economy. By the end of the fiscal year, the total debt raised by the ministry of finance was expected to exceed $2 trillion. The total federal debt of $39 trillion is equivalent to $114,000 per american。
The numbers are really scary. But the key is that high debt does not amount to a financial crisis。
The 2008 crisis was not triggered by “high debt”, but by the fragmentation of the debt chain. At that time, sub-prime loans were packaged in layers into cdos (debt mortgage bonds) and held in large numbers by major banks, insurance companies, with billions of dollars in cds (credit default swaps). Once the house price falls, the whole chain breaks down from the bottom, and no one knows who holds bad debts and panics like dominoes。
Today's united states debt is primarily a national debt — one of the most liquid and secure assets on the planet. National debt will not default in bulk as a subprime, nor will it be magnified by layers of leverage into unknown derivative black holes. The national debt is a “money owed to itself” and the subprime is a “money owed to others”, with a completely different risk transfer mechanism。
After three or eight years, the firewalls are already very well repaired
That is the easiest point to ignore。
Following the 2008 financial crisis, the united states introduced the dodd-frank act, which established a strict banking regulatory framework. The core result is stress testing — the federal reserve simulates extreme recession scenarios every year to test the viability of large banks。
Pressure tests in 2026 are progressing, assuming a sharp rise in unemployment to 10 per cent, a 30 per cent drop in housing prices and a 39 per cent drop in commercial property prices. Thirty-two of the largest banks in the united states are going to take the “mock exams”. In 2025, the 22 largest banks were shown to be able to maintain robust capital levels even after hundreds of billions of dollars in losses。
More critically, the fed has frozen pressure capital buffer requirements to 2027 – not deregulation, but a judgement that banks’ capital adequacy rates are strong enough。
Data show that, as of the fourth quarter of 2025, over 99 per cent of united states banks had sufficient capital, with average capital adequacy rates of about 13 per cent for both large and small banks. By contrast, before the 2008 financial crisis, many banks had a much lower capital adequacy rate。
In addition, the floating exchange rate regime provides an important buffer. Prior to 2008, many emerging market countries were burdened with large dollar debt, and the devaluation of their currencies directly increased debt-servicing pressures. The ability to absorb external shocks is greatly enhanced by today's floating exchange rates, which allow individual currencies to adjust autonomously。
It's not the same thing about ai and internet foam
Many people compare today's ai with the year 2000 internet. - from 5132 points to less than 1200 points, $5 trillion of wealth evaporated. But there are fundamental differences。
Look at profitability first. At the peak of the internet bubble in 2000, the total net asset return of the nasdaq 100 index (roe) was -0. 5 per cent and the net interest rate was -10. 8 per cent — that is, these “star companies” as a whole were in deficit. In the first quarter of 2026, the overall roe of mag7 (the seven technology giants) reached 52 per cent, with a net interest rate of 35 per cent。
And look at this ai lead. In the first quarter of the financial year 2027 (as of april 2026), camp weida received $81. 6 billion, an increase of 85 per cent over the previous year; and net profits of $58. 3 billion, an increase of 211 per cent over the previous year. Operating cash flows of more than $50 billion a quarter, free cash flows amounted to $48. 6 billion. This represents an increase of 65 per cent。
This is not a “market dream rate”, it's real silver and silver。
In terms of valuation, the current rate of earnings of the philadelphia semiconductor index is about 71 times, which appears to be quite low, but it was more than 150 times during the internet bubble period. More importantly, ai has been scheduled for 2028, with high performance visibility. Global science and technology giants are expected to spend more than $72. 5 billion on ai capital in 2026, and cloud computing and revenue growth in ai services are supporting these inputs。
Ai is not a concept, but is reshaping industry trends across industries。
V. The ai is divided, foaming but not fatal
Of course, not that the ai plate is completely free of foam. Some of the ai conceptual units without performance support do have valuation bubbles。
When the internet bubble broke down in 2000, almost all of the technology units were hit hard because everyone was not profitable and was telling stories. Today's ai board, the head company has solid profits and cash flows, second-line companies have real technology and customers, and only the most marginal concept shares are swimming naked. This division is in itself a health signal — the market is voting with its feet, not a panic without distinction。
The us-iraq conflict is cooling, not escalating
Geopolitics is another factor that has been overstretched。
The fighting has continued for over 100 days since the united states launched a military strike against iran at the end of february this year. However, after months of saw-sawing and multiple good offices, the united states and iran recently concluded a memorandum of understanding on the truce, which was formally signed on 19 june. The parties will conduct 60 days of negotiations to reach a final agreement。
The reaction of the market is telling: following the news of an agreement between the united states and iraq, brent crude oil prices fell by $84 at a time and global stock markets generally rose. During the conflict, brent's crude oil peaked at about $120 per barrel, while before the outbreak of the conflict it was less than $70 — now oil prices have fallen sharply and the geo-risk premium is retreating。
The reason behind this is that both sides “can't move”。
The united states had debt and inflationary pressures, while iran had suffered huge economic and military losses, and both needed a step below. Although the memorandum of understanding is only a one-and-a-half-page document of principle, the details are left for subsequent negotiations, immediate results such as the ceasefire and the reopening of the strait of hormuz have resulted in “brakes on” the middle east tension that has lasted for months。
As can be seen, geo-risks are cooling, not warming。
Anyway, don't let anxiety dominate your judgment. Return to the opening sentence — when everyone fears the financial crisis, it will not come。
The ai industry is much healthier than the real estate bubble, even if it is a little bubble. Investing in ai is investing in future productivity, which brings benefits, while the real estate bubble is just the next harvester。
For investors, rather than being held hostage by “wolves” anxiety, it is better to look calmly for an ai direction with real performance, rather than a concept of hiatus, in which we can look at our planet。
Risk tips: these are only personal views sharing, not investment proposals, such as references to specific enterprises, but case analyses, stock markets are risky and investments need to be prudent. The latter is self-sufficient。
