The Chip Sellers Were All Sold. The Buyers Closed Higher

Hello. On Monday the chip chain was sold from end to end.

$Micron Technology(MU)$ closed down 5.83 per cent at US$910.43, $SanDisk Corp.(SNDK)$ fell 6.45 per cent, $SK hynix(SKHY)$ 4.92 per cent, $Advanced Micro Devices(AMD)$ 3.49 per cent, $Marvell Technology(MRVL)$ 3.27 per cent and $Intel(INTC)$ 3.12 per cent, while the optical name $Applied Optoelectronics(AAOI)$ dropped 13.77 per cent. $Direxion Daily Semiconductors Bull 3x Shares(SOXL)$fell 7.83 per cent.

$NVIDIA(NVDA)$ itself fell 2.91 per cent, and it reports after Wednesday's close.

The two that buy the chips closed higher: $Meta Platforms, Inc.(META)$ up 1.66 per cent and Alphabet 0.83 per cent.

The indices barely moved — $S&P 500(.SPX)$ fell 0.28 per cent and $SPDR S&P 500 ETF Trust(SPY)$ 0.29 per cent, while $Invesco QQQ(QQQ)$ fell 1.00 per cent. That gap says this was not a broad flight to safety but chip weightings dragging the index down.

The trigger came from memory's own supply side. A Chinese memory rival listed, and Barron's put Micron's fall down to competitive concerns; there was also a report about CXMT and Apple, though one analyst said the pre-market reaction to it was an overreaction.

Samsung fell 9 per cent that day and took the memory chain with it. $Taiwan Semiconductor Manufacturing(TSM)$ fell 2.3 per cent, with coverage describing confidence in AI chips as having suddenly cracked.

The macro pushed the same way: coverage attributed the day's technology losses to remarks from Trump and Bessent, and $NVIDIA(NVDA)$ was being sold ahead of its results.

But nothing arrived on the demand side. No figure that day said memory was no longer short or that compute demand had slowed.

What changed was positioning. Reports say big hedge funds have started cutting $Micron Technology(MU)$ after a 255 per cent run, under headlines about smart money leaving memory stocks; another said Wall Street is rotating out of this year's biggest winners.

So what was sold on Monday was not the fundamentals but the position. Results and the central bank conference fall in the same week, and whatever has run hardest gets taken down first.

$Alibaba(BABA)$ offered a sharper contrast. Sunday's HK$80 billion discounted placement diluted existing holders, and on Monday the US line slipped 0.73 per cent while the Hong Kong line rose 0.71 per cent; it is at HK$113.30 intraday today.

Then the insiders stepped in. Chairman Joseph Tsai bought 720,000 shares for about US$10.29 million and chief executive Wu Yongming bought 350,000 for about US$4.98 million — roughly US$15.3 million between them, on the first day after the placement.

Michael Burry had sold before it. His phrasing is unkind and accurate: issuing shares is now its new paradigm.

The same day Alibaba released a new AI video model — raising US$10 billion with one hand and showing you where it goes with the other.

On one side, executives buying with their own money. On the other, a short seller saying the company now lives by issuing stock. Both are true. The disagreement is only about how long that money takes to become cloud revenue.

$PDD Holdings Inc(PDD)$ reported before Monday's open, another case of beating and falling anyway. Earnings per share beat by about 5.2 per cent, but revenue missed by about 2.1 per cent, second-quarter revenue growth slowed to 8 per cent, and profit fell under heavy reinvestment. The shares rose intraday on the beat and still closed down 1.48 per cent.

$Tesla Motors(TSLA)$ fell 3.83 per cent. Half of the pressure was sentiment: two notes landed side by side, one arguing Xpeng's second quarter made it the better choice, the other saying SpaceX is Tesla's only hope.

If the market really treats an unlisted company as a carmaker's only hope, that says the pricing of the car business is already cautious. SpaceX itself fell 1.44 per cent on Monday.

$Strategy(MSTR)$ also closed higher, up 2.83 per cent. Bitcoin has risen about 24 per cent in this move, taking its coin position from a loss into profit, and it announced another US$2 billion raise to keep buying.

That structure is an amplifier while the coin price rises and a double hit when it falls, because the financing side and the price come at you together.

The week's fuses are packed close. $NVIDIA(NVDA)$ reports after Wednesday's close; Jackson Hole runs from 27 to 29 August, with the market framing it as Warsh and Bessent meeting head-on, where the monetary stance and the debt-issuance approach have to be settled; July PCE lands in the same stretch; and Marvell reports after Thursday's close.

What was sold on Monday was whatever had run hardest. What was left standing was the group that does not have to build its own chips and still gets the AI benefit. The difference is not whose business is better. It is whose price has not yet been lifted to the point where it has to deliver first.

The above is personal analysis, not investment advice.

💬 【Talking Point】

Alibaba's chairman and chief executive bought about US$15.3 million of shares the day after a discounted placement diluted holders, while Michael Burry had already sold, saying that issuing shares is now the company's new paradigm. Whose read would you take?

💰 【Bounty】

Drop your view in the comments and there are coins in it for you! 🎁

🔔 Better shared than saved — tag a friend and split the coins!

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Report

Comment4

  • Top
  • Latest
  • 苏36
    ·08-25
    TOP
    I’d lean slightly toward Alibaba’s insiders—but with a big caveat.

    Joe Tsai and CEO Eddie Wu buying roughly US$15.3 million of shares immediately after Alibaba’s discounted US$10.2 billion placement is a meaningful signal. It shows management is willing to put personal capital behind the same AI strategy they are asking shareholders to finance.

    But Michael Burry’s argument should not be dismissed. Issuing 710 million new shares creates real dilution, while Alibaba’s latest quarter already showed a 75% profit decline as AI investment surged.

    So I see this as a “prove it” moment. Management has bought time and computing power; now Alibaba must turn AI spending into cloud growth, higher margins and eventually stronger free cash flow.

    For me, insider buying is bullish—but execution is the real catalyst. If AI revenue accelerates, this placement could look brilliant. If returns disappoint, Burry’s warning will age very well.

    Reply
    Report
  • Jerry Lam
    ·08-27 17:07
    如果只看这次分歧,我会 更重视阿里管理层的买入信号,但不会因此忽略Burry的担忧。

    蔡崇信和吴泳铭在配售后马上用自己的钱增持,至少说明他们认为当前价格并不贵,而且愿意继续押注AI和云。但Burry讲的核心问题也很关键:如果未来还需要不断发股融资,股东稀释就会从一次性事件变成结构性问题。

    所以我不会简单站“内部人一定对”或者“空头一定对”。真正决定结果的,是这100亿美元融资之后,阿里能不能把钱变成 云收入、AI客户、利润率和自由现金流。如果AI收入增长跑得比股本稀释更快,这次融资就是扩张;如果ROIC继续下降,那Burry的担忧就会越来越有道理。

    一句话:管理层买入证明他们有信心,Burry卖出提醒我们要看资本回报;最终不是看谁名气大,而是看AI投入能不能跑赢稀释。

    Reply
    Report
  • Burry on BABA, but on chips this still looks like pre-earnings risk cleanup, not a broken AI cycle. NVDA does this into prints all the time lol
    Reply
    Report
  • For now, core positions are simply being squeezed, triggering spillover selling.
    Reply
    Report