Waller Said He Could Wait. The Market Didn't.

$S&P 500(.SPX)$ closed 1.06 per cent higher at 7,747.71 and $Dow Jones(.DJI)$ added more than 600 points, the best day for both in a month, and the reason was a sentence from Christopher Waller, a Federal Reserve governor. He said on Thursday that he is willing to support holding rates steady in September as long as inflation keeps moving towards 2 per cent. Bets on a September hike fell from above 60 per cent to a little over half, and the ten-year Treasury yield came back to around 4.75 per cent, a day after touching its highest in more than two years.

Waller's reason for holding was to give disinflation a chance: "We can wait one meeting." In the same remarks he left himself an exit: "if inflation comes in hot, I would consider a rate hike". He also said policy is only slightly restricting demand, and that it would not take much of a pickup in inflation to move him towards tightening. He tied his September vote to the August inflation figures due on 11 September, and to nothing else. That does not sit easily with what Kevin Warsh, the Fed chairman, said last week.

The first things to move were the ones tied most closely to rates. Bitcoin rose more than 5 per cent and went back above US$80,000. $Strategy(MSTR)$ closed 17.56 per cent higher at US$144.82 on 45.2 million shares, close to twice its usual volume; $Coinbase Global, Inc.(COIN)$ rose 10.14 per cent; and $Circle Internet Corp.(CRCL)$ rose 16.46 per cent, having picked up a New York trust charter and a federal bank charter, with its own Arc mainnet set to go live on 16 September and Visa, Mastercard and BlackRock among the founding validators.

There was also something specific to Strategy: on 31 August it spent US$370 million on 4,603 bitcoin, ending roughly ten weeks without a purchase. In the same week Bernstein cut its target by US$100 and pushed its US$150,000 bitcoin call out to 2027, and a rule MSCI still has out for consultation could see it dropped from MSCI's indexes, a proposal Strategy has publicly opposed. It is up 47.22 per cent over a month and still down 56.15 per cent over a year.

The biggest gainer had nothing to do with rates. Snowflake closed 16.55 per cent higher at US$356.47 after touching a 52-week high of US$384.56 during the session. Its second quarter of fiscal 2027, which ended on 31 July, came out after Wednesday's close: revenue of US$1.55 billion against the US$1.48 billion analysts expected, and adjusted earnings of 62 cents a share against 45 cents. Product revenue was US$1.49 billion, up 37 per cent, which Brian Robins, the chief financial officer, called a third straight quarter of acceleration.

Full-year product revenue guidance went from US$5.84 billion to US$6.07 billion, and the full-year adjusted operating margin outlook from 13.5 per cent to 14.5 per cent. The cost sits in the same release: the product gross margin outlook was cut to 74 per cent because AI workloads are a bigger part of the mix. Management said about half the acceleration came from AI. Six firms raised their targets that day, Goldman Sachs from US$300 to US$436 and J.P. Morgan from US$285 to US$426.

$Tesla Motors(TSLA)$ closed 5.42 per cent higher at US$376.37, having been up around 7 per cent during the session. Its Cybercab began carrying passengers in Austin with no steering wheel and no pedals, by invitation, from a fleet of 45. Reuters reported that US safety regulators opened an evaluation the same day.

$Palantir Technologies Inc.(PLTR)$ rose 7.71 per cent to US$182.53 after expanding its alliance with PwC to build an AI-native platform for mergers and acquisitions and to modernise ERP systems. The argument for the other side arrived the same day: Michael Burry included the company in what he called a US$100 billion warning, saying what holds it up is the fear of missing out on AI. It trades at 156 times earnings, or 80 times expected earnings.

The rally left one group behind. Micron closed 0.22 per cent higher at US$958.16, SanDisk 0.10 per cent higher at US$1,554.99 and SK hynix 0.79 per cent lower — on a day the S&P rose 1.06 per cent.

What held them there was a market-share report. Second-quarter figures for this year from Counterpoint, a research firm, put CXMT's share of DRAM revenue at 10 per cent, up from 4 per cent a year earlier and the first time a Chinese maker has reached double digits; Samsung, SK hynix and Micron went from 94 per cent between them to 87 per cent. In NAND, YMTC went from 9 per cent to 14 per cent and passed $SanDisk Corp.(SNDK)$, which slipped from 12 per cent to 11 per cent.

None of those seven points came off Micron. $SK hynix(SKHY)$ is the one giving ground: its DRAM share fell from 39 per cent a year earlier to 25 per cent, which Counterpoint puts down to falling high-bandwidth memory prices and long-term supply deals signed early. The same report has Micron's DRAM share rising from 22 per cent to 24 per cent and its NAND share from 13 per cent to 15 per cent. Of the three stocks held down together, the ones actually losing share are SK hynix and SanDisk.

The share was taken during a boom, not a slump. Counterpoint has the NAND market growing 70 per cent from the previous quarter in that same period, with prices up 55 per cent. Micron's high-bandwidth memory output is committed under take-or-pay contracts through December 2026, and the Chinese makers are competing in commodity memory for now.

On expected earnings, SanDisk trades at 5.90 times and Micron at 6.17 — one of them losing share, one taking it, and the prices do not tell them apart.

$Broadcom(AVGO)$ closed 2.74 per cent lower at US$357.16 on 60.2 million shares, 2.35 times its usual volume, leaving it about 27.9 per cent below its 52-week high. Wednesday's after-hours fall carried straight into Thursday's regular session. The sell side did not agree on what to do with it: Macquarie raised its target to US$490, Rosenblatt resumed coverage at Buy with US$600, and Bank of America cut its target while raising its earnings estimates. The average target stands at US$538.33.

The number Hock Tan, Broadcom's chief executive, has put on AI revenue for fiscal 2028 is US$230 billion. Building to it needs memory to keep up, and one piece of coverage this week called that Broadcom's memory problem. What that share table counts is who supplies the memory.

The rest of the large caps all closed higher without a reason of their own: Microsoft up 2.68 per cent, Meta 3.01 per cent, Nvidia 1.80 per cent, Intel 1.80 per cent and Apple 1.00 per cent. Alphabet rose 1.59 per cent on a day when Waymo closed its first debt raise, US$3 billion from Pimco and Blackstone, and a judge turned down the request to break up Alphabet's advertising business, ordering remedies instead.

The August payrolls report comes out today. The market is looking for a gain of 56,000 with unemployment at 4.1 per cent and 45,000 of it from the private sector; July came in at 23,000 jobs lost and June was revised down to a gain of 20,000. Average hourly earnings are expected up 0.3 per cent on the month and 3.0 per cent on the year, which would be the weakest annual pace since May 2021.

Waller said his September vote turns on the inflation figures due on 11 September, not on the jobs figures out today. The market did not wait; it repriced on his words. Memory went the other way: a share table held three stocks in place, and both of Micron's rows in that table point up. Some information gets paid for the day it lands, as Waller's remarks did on Thursday. Those two rows have to wait until Micron reports on 30 September.

The above is personal analysis, not investment advice.

💬 【Talking Point】

$Snowflake(SNOW)$ raised its revenue guidance and cut its gross margin outlook in the same release, because AI workloads cost more to run. The stock closed 16.55 per cent higher anyway. Would you put money in after a day like that?

💰 【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!

# 45 Cybercabs on the Road — Enough to Justify a $1.49 Trillion Autonomy Narrative?

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

Comment3

  • Top
  • Latest
  • 苏36
    ·18:18
    TOP
    I wouldn’t rush into Snowflake after a 16.5% single-day surge. The fundamentals are clearly improving: product revenue grew 37%, AI contributed roughly half of the recent acceleration, and full-year product revenue guidance was raised to $6.07 billion.

    However, there’s an important catch: management lowered its product gross-margin outlook to 74% because AI workloads are more expensive to run. That means revenue growth is accelerating, but profitability is not moving in the same direction.

    My view: SNOW is a strong long-term AI/cloud story, but I’d rather buy the pullback than chase the breakout. If growth keeps accelerating and margins stabilize, today’s valuation could eventually look reasonable. If margins keep deteriorating, the market may quickly reconsider the premium.

    @Marktomarket [贱笑]

    Reply
    Report
  • glitzy
    ·18:52
    I care more about the revenue guide than the margin cut here. AI workloads are expensive now, but the market is basically saying growth matters more for Snowflake
    Reply
    Report
  • Jerry Lam
    ·18:38
    我会选择 先不追,等回踩再看。SNOW这次财报确实很强,产品收入重新加速、全年指引上调,而且管理层明确说AI贡献了相当一部分增量,这些都说明增长逻辑在改善。

    但我会特别盯住一个细节:收入指引上调的同时,产品毛利率预期下调到74%。这意味着AI工作负载虽然能带来更多收入,但也更“烧算力”。如果未来AI收入增长很快,却长期压低毛利率,那么市场最终还是会追问:这些新增收入到底能不能变成更多利润和自由现金流?

    所以财报后一天涨16%以上,我觉得更多是在交易 “增长重新加速”,而不是已经证明“AI利润率模型完全跑通”。在这种情况下,追高的赔率不算特别舒服。

    我的策略会是:长期看好SNOW的AI+数据平台逻辑,短期等情绪降温。 如果后面股价回踩,但产品收入、NRR和经营利润率继续改善,我会更愿意加。

    一句话:AI把Snowflake的增长重新点燃了,但真正决定估值能不能站住的,是这把火最后能烧出多少利润。

    Reply
    Report