Before the 2026 '618' shopping festival even officially began, a company named Qingbei Daoyuan disappeared from JD.com.
This small company, focused on digital solutions for high schools, was founded in 2016 and sells learning tablets. According to its own claims, its team includes dozens of graduates from Peking University and Tsinghua University. Yet this very store suddenly became unsearchable on JD.com on May 26: its shop couldn’t be found, all products were taken down, and clicking on links actively promoted in livestreams returned the message, 'This item is no longer available.'
On May 27, Qingbei Daoyuan issued a statement explaining what had happened: just before the 618 event, another platform had distributed universal consumer vouchers to shoppers. JD.com demanded that the company either refuse voucher usage on other platforms or lower all its product prices on JD.com to match the post-voucher prices elsewhere—thus guaranteeing JD.com’s claim of offering 'the lowest price online.' The company refused. In its statement, it wrote,We cannot agree with JD.com’s practice of shifting the costs of achieving 'lowest online prices' and enhancing 'shopping experience' onto merchant businesses.”。
In the same statement, it added an even more striking remark: 'We fully understand that we do not have equal standing to negotiate with major platforms.' As a company solely focused on learning tablets for high school students, it offered only two products priced at RMB 6,880 and RMB 7,980—neither cheap—and had collectively sold over 10,000 units on JD.com with strong customer reviews, which is already quite remarkable.
It wasn’t until May 28 that the store quietly resumed normal operations, but JD.com has remained silent about exactly what happened during those two days...

On May 26 Beijing time, JOYY released its fastest year-over-year revenue growth in recent years. Starting this quarter, due to changes in its business model, JOYY has adjusted its financial reporting structure and adopted a new way of presenting its performance.Over the past year, beyond its well-known live streaming segment, JOYY’s non-live-streaming businesses have gradually emerged. In its latest earnings report, a new strategic logic—powered by AI integration—has begun to take shape.
Summary
1. In Q1, JOYY reported total revenue of approximately USD 560 million, up 12.4% year-over-year,marking the first time the company disclosed results across three business segments: Social Entertainment, BIGO Ads, and SHOPLINE.. The foundational Social Entertainment segment returned to year-over-year growth, while BIGO Ads and SHOPLINE—representing the company’s second growth curve—both posted strong increases. Together, these three segments form a synergistic strategic growth flywheel, laying the groundwork for a diversified global ecosystem.
2. AI is the connective tissue binding these three businesses together. Serving as the foundational layer of the entire ecosystem, AI enhances content recommendation, advertising efficiency, and merchant operations. Through cross-business synergy, it powers a self-reinforcing flywheel. This collaboration goes beyond conceptual rhetoric; against a backdrop of continuous technological advancement, AI will act as a catalyst, enabling richer value exchange among social data, ad algorithms, and e-commerce.
3. Profitability quality and shareholder returns strengthened in tandem.Both non-GAAP operating profit and EBITDA grew faster than revenue, and as of March 31, JOYY held USD 3.1...

The long-anticipated SpaceX has finally filed its IPO prospectus.
This company, which carries Musk’s vision of “landing on Mars,” is for the first time laying nearly all its details bare before the public. Curious yet skeptical, I immediately downloaded and reviewed the full document as soon as it appeared on the SEC’s website on May 20.
I have to say, Musk really delivered—with freshness turned all the way up.
In over two decades of commercial spaceflight, this is the first time a player of this scale has fully disclosed its financials to the SEC.SpaceX is valued at nearly $400 billion, making it the world’s most valuable private company, and last year it led the entire industry in rocket launches. The outside world already had a general sense of its size, but only through fragmented data points—valuation rumors, Starlink subscriber counts, Musk’s ownership stake—and never from a complete, publicly available financial statement.
Even more novel is xAI. Although U.S. frontier large-model companies have been under intense spotlight over the past two years, none have submitted comprehensive financial disclosures to the SEC—until now. This time, xAI’s personnel and finances are fully integrated into SpaceX’s consolidated financial statements,meaning this prospectus also serves as the first-ever public window into exactly what a cutting-edge large-model company is spending money on—and just how much it’s burning.。
For these reasons, even before reading it, I’d already compiled several specific questions in my mind: How big is Starlink’s satellite internet business right now, ...

A simple line of reasoning is this: in a memory price supercycle, whoever holds memory inventory benefits.Companies like SK Hynix and Samsung have been significantly re-rated by the market precisely because they possess memory production capacity; yet few people associate Aihuishou with the memory theme, even though it has clearly benefited substantially from this supercycle.
In Aihuishou's Q1 2026 earnings report, an interesting financial detail emerged: cash and cash equivalents on the balance sheet declined from RMB 2.187 billion at year-end to RMB 1.72 billion, while inventory rose from RMB 1.07 billion to RMB 1.486 billion. At first glance, this doesn’t seem positive—but when viewed in the context of broader changes across the 3C industry in Q1, a different conclusion emerges.
In Q1 2026, memory prices continued to surge. According to TrendForce’s latest forecast on May 14, high-spec mobile memory prices are expected to jump by 78% to 83% in Q2 alone. As new smartphone prices are pushed higher by rising memory costs, secondhand device prices have followed suit. Aihuishou’s use of its own capital to secure inventory at this stage effectively positions it within the upward pricing cycle, laying the groundwork for further profit expansion.
Summary:
First, the company reported 32.4% year-over-year revenue growth in Q1, a clear acceleration compared to the same period last year, with a more robust underlying structure. Regarding Q2 guidance, the company maintained its usual stance, forecasting year-over-year growth of 25% to 27%—a seemingly conservative outlook. However, over the past few quarters, the company’s guidance has consistently ranged between 25% and 28%, yet actual results have...

Alibaba's Q1 2026 natural year (Q4 2026 fiscal year, hereinafter the same) earnings report was released on May 13. Due to increased investment in AI and technology businesses, the company’s adjusted EBITA fell by 84% year-on-year. However, the market is highly confident in Alibaba's AI commercialization prospects, causing the stock price to rise despite a low opening.
The most fundamental logic is that AI investment has now reached a turning point for commercialization. CEO Wu Yongming placed the most important judgment of this quarter at the beginning:
Alibaba's full-stack AI technology investment has officially moved past the initial nurturing phase and entered a positive cycle of scaled commercial returns.。
If you only read this sentence, it may seem like typical marketing rhetoric from an AI story. After all, these days, whether or not a business is related to AI, companies still mention it. What makes this quarter special for Alibaba is that the entire earnings report backs up this statement with real financial figures. For Alibaba, AI is no longer 'promising' but rather 'rewarding.'
Alibaba Cloud’s external commercial revenue growth accelerated to 40% this quarter (up from 35% last quarter), with AI-related products generating RMB 8.971 billion in Q4 revenue, accounting for 30% of Alibaba Cloud’s external commercial revenue. This marks the eleventh consecutive quarter of triple-digit growth. During the earnings call, management also projectedthat in the next year, AI-related product revenue will account for more than 50% of external commercial revenue, officially becoming the primary driver of Alibaba Cloud’s revenue growth.。
Notably, Alibaba disclosed its AI model and application services ARR (Annual Recurring Revenue) for the first time (...)

This wave of gains in the storage sector is unreasonably strong.
A month ago, SK Hynix's stock price was still at 1.04 million Korean won, but today it surged to a record high of 1.97 million during trading before being pushed back down to 1.835 million by the close; yesterday alone it jumped 11.5%, with cumulative gains of nearly 30% over the past week. Samsung is following closely behind, with slightly smaller increases but a similar upward rhythm. The explosive rise has brought extremely high market enthusiasm, and almost all investment groups are discussing these two companies, with most questions focusing on...whether it’s still worth buying Samsung, which has lagged behind in terms of gains.。
Over the past two years, AI has exploded in popularity, and the side of computational power—companies like NVIDIA, Taiwan Semiconductor, and Broadcom—has seen significant growth, while the storage line has consistently been left behind. However, as large-scale model applications began advancing into the productivity phase in the middle of last year, token consumption grew exponentially, and AI's demand for storage expanded accordingly. The market suddenly realized that the storage-related stocks, which had been neglected for the past two years, have instead become the bottleneck in the expansion of computing power. This consensus expectation has now formed.
For institutional investors, this situation is nothing new; they simply buy. However, for individual investors, at this particular point in time, before deciding to invest in these two companies, especially if considering purchasing Samsung, three key questions need to be addressed:
What do I currently hold in my portfolio? What returns am I expecting from this purchase? Should I adjust my holdings accordingly?
Among these three questions, the second one is the most fundamental. After reviewing analyses from many different institutions and companies regarding this round of the storage cycle, a relatively moderate consensus is that the peak may...

If you were to list the large internet technology companies in Hong Kong's stock market over the past year, you might be surprised by which one ranked first in terms of gains:
Baidu。
Over the past year, Baidu's stock price has risen more than 70%, leading the pack among large Chinese concept tech stocks. If we take into account the performance of the Mag 7 US tech giants, Baidu's stock price increase ranks third, only behind Google’s 163.3% and NVIDIA’s 84.5%, and ahead of fourth-place Apple by approximately 23 percentage points.
True value has always been hidden in non-consensus; and Baidu,is becoming the embodiment of non-consensus within global large internet tech companies.。
Kunlun Chip originated in 2011 and is set to be officially spun off and listed independently by January 2026. In the first quarter of 2026, Baidu Intelligent Cloud once again captured the top spot in both number and amount of domestic market bids. Recently, Wenxin Large Model 5.1 was officially released, achieving the top position on LMArena's domestic search rankings with only about 6% of the pre-training costs of similarly scaled models in the industry. Baidu's DuMate topped the PinchBench evaluation benchmark for intelligent agents, securing three out of the top five spots...
Application-first approach, full-stack self-research, and long-term investment—these are key areas that Robin Li has repeatedly emphasized in public speeches, and they are now being validated one by one by the market. From the financial reports, the company has shifted its disclosure perspective, using an AI-native classification centered on 'Baidu Core AI New Business,' providing a clear and quantifiable path for analyzing the value of AI-driven business operations.
Non-consensus...

After the market closed on April 30, Wuliangye issued 37 announcements at once.
The most eye-catching one was titled "Correction of prior accounting errors." It cut the revenue for the first three quarters of 2025 from 60.9 billion yuan to 30.6 billion yuan; the net profit attributable to shareholders was reduced from 21.5 billion yuan to 6.5 billion yuan, with a 70% downward revision in net profit.
For an ordinary company, this would be enough for the market to digest for a while, but Wuliangye’s financial report “has more tricks.” Turning to the next page of the announcement, the company casually stated: "It does not affect the presentation of the cash flow statement." Comparing this statement with the cash flow statement: cash received from the sale of goods and services amounted to 76.6 billion yuan, net cash flow from operating activities was 28.2 billion yuan, and all items for the three reporting periods showed no changes following the adjustments in net profit and revenue.
Turning to the balance sheet, cash, accounts receivable, contract liabilities, and inventory – comparing last year's original version with the updated version issued on the 30th line by line, there is no difference whatsoever.27 billion yuan in revenue seems to have vanished into thin air, with no changes appearing in the cash flow statement, cash, or contract liabilities.。
What truly adds a surreal touch is the Q1 2026 report disclosed that same evening, showing an 82.57% year-on-year increase in net profit attributable to shareholders. After reading the previous correction announcement and then seeing this figure, it feels a bit disorienting – you just watched the company slash 70% of its profit for the first three quarters of 2025, and the next moment, it tells you in another document,The new year has started with an 80% growth。
Not only...

On April 23, after the market close, Ping An Good Doctor released a voluntary Q1 disclosure report of only two pages. Though brief, it contained sufficient information to set the tone.
As of the end of March 2026, the number of paying corporate clients served by Ping An Good Doctor over the past 12 months exceeded 7,500, representing an 89% year-over-year increase. During the same period, the company’s net profit increased by 138.4% year-over-year, with adjusted net profit rising by 45.8%. The quarterly usage of AI doctors surpassed 5.6 million users. After refining its self-sustaining model in 2025, the pace of Q1 fulfillment exceeded market expectations.
The results for Q1 are solid and provide a foundation for longer-term growth stories, which can be traced back to the financial base established in the 2025 annual report. One of the most critical factors is the establishment of the company’s 'self-sustaining' capabilities.
In terms of specific figures, 'self-sustaining' refers to the company's operating profit turning positive for the first time in 2025, surging from a loss of 170 million yuan in 2024 to a profit of 100 million yuan—a positive swing of 270 million yuan. Meanwhile, net operating cash flow amounted to 451 million yuan, up 354% year-over-year.Not only are there profits on the books, but the business itself is also continuously generating real cash.。
Another notable highlight is the AI segment. The annual report disclosed for the first time that AI contributed 4.5% to the company’s gross profit and pledged long-term tracking. In Q1, the accuracy rate of AI-assisted MDT (multi-disciplinary consultation) platform solutions improved from nearly 90% to approximately 95%. For the first time, AI entered financial statements in a quantifiable manner...

Many investors focused on the AI industry may have noticed that, whether in business models or product positioning, numerous Agent products on the market are converging towards uniformity.
On April 9, 2026, OpenAI raised the price of ChatGPT Pro to $100 per month, precisely aligning with the same pricing tier as Anthropic’s Claude Max, which was launched a year earlier. The three-tier pricing structure of $20/$100/$200 is identical. Five days later, on April 14, Bloomberg reported that Anthropic was receiving investment offers valuing the company at $800 billion, nearly matching OpenAI's valuation.
Does something feel off? In theory, two competing companies in their early growth stages should focus on finding their own differentiated positioning or leverage unique advantages for growth. Yet these two companies increasingly resemble the war between Uber Technologies and Didi, though they haven't yet reached the stage of a price war.
This homogenization at the product level is only the surface; beneath it lies a deeper pricing mechanism worth understanding.。
Starting in 2014, Uber Technologies and Didi burned through tens of billions in subsidies, with their valuations rising from tens of billions to hundreds of billions during the same period. The battles in food delivery and bike-sharing followed the same playbook. But in those scenarios, the valuation was based on the future narrative of 'fight until only one remains, then collect monopoly rents,' premised on eventual dominance. This time, however, the $800 billion valuation for AI isn't just about the future...