To The Moon

Today, global memory stocks finally staged a meaningful rebound. It felt like weeks of pent-up pressure were finally released.

South Korea’s KOSPI rose roughly 3.6%, while Japan’s Nikkei gained about 3.3%. Samsung Electronics, SK Hynix, and Kioxia $KIOXIA HLDGS CORP(KXIAY)$ all rallied. Japan was closed yesterday, and Kioxia came back today with a 17% surge—nearly brought me to tears.[Cool][Smart][Sly]

In U.S. premarket trading, SNDK, MU, WDC, and STX were also all higher. Asia-Pacific market performance

There was also a fundamental catalyst today: South Korea’s exports during the first 20 days of July rose 52.3% year over year, with semiconductor exports surging 180.6%. At the very least, this confirms that memory shipments have not suddenly weakened. Reuters

But my view remains unchanged:

Today was a rebound, not yet a reversal.

What determines whether this rebound can continue will not be another memory manufacturer’s earnings report. It will be Alphabet’s earnings tomorrow after the close. $Alphabet(GOOG)$

Why is Google more important than the previous semiconductor earnings?

Because $GOOG is the first true “buyer of semiconductors” to report this earnings season.

After Micron’s earnings, the market already knows how strong memory pricing, shipments, and profits are. Even if the next few memory manufacturers continue to beat expectations, they would largely be reconfirming a reality that has already been priced in.

That is why we often see excellent earnings followed by a falling stock price.

Google is different.

The market is no longer worried about whether memory chips can be sold. It is worried about whether Google, Microsoft, Meta, and Amazon are still willing to build AI infrastructure at such an aggressive pace.

Alphabet’s current 2026 Capex guidance is $180 billion to $190 billion—nearly double its 2025 spending—and the company expects Capex to “increase significantly” again in 2027.

Meanwhile, Google Cloud revenue grew 63% last quarter to $20 billion, operating income reached $7 billion, and its backlog climbed to $462 billion. Alphabet investor presentation

So the real question tomorrow is:

Can this enormous investment continue translating into cloud growth, new orders, and profits?

Three possible scenarios for Alphabet’s earnings

1. The neutral scenario

Alphabet maintains its 2026 Capex guidance of $180 billion to $190 billion, continues to expect higher spending in 2027, and Google Cloud revenue and profitability broadly meet expectations.

This would not create a meaningful headwind for semiconductors, but it would probably not be enough to trigger another round of earnings upgrades either.

Memory stocks could continue their oversold rebound, but neither the technical setup nor fund flows would be strong enough to confirm a genuine reversal.

2. The bullish scenario

Alphabet raises CapEx again or provides a more aggressive 2027 investment outlook, while Google Cloud growth, backlog, and AI monetization all significantly exceed expectations.

That would confirm that the cycle remains intact:

Financing → data-center construction → AI revenue generation → further investment.

If this happens, the market will begin raising its demand expectations for HBM, server DRAM, enterprise SSDs, and the entire AI infrastructure supply chain.

Investors will suddenly remember SK Group Chairman Chey Tae-won’s recent bullish comments on memory demand. Underinvested funds could start chasing the rally again, pushing the market back into full FOMO mode.

Only this scenario would be strong enough to upgrade today’s rebound into a genuine reversal.

If it happens, today could ultimately prove to have been the reversal point.

3. The bearish scenario

Google does not even need to cut its 2026 CapEx guidance.

A weaker tone on 2027 spending—or signs that depreciation and data-center costs are beginning to materially erode Google Cloud’s profitability—would be enough to put semiconductors under pressure again.

The market is never trading only on whether companies are still spending this year. It is trading on whether the growth rate of future CapEx is beginning to peak.

Once investment growth starts slowing, even today’s exceptional memory pricing and profitability will be treated as peak-cycle numbers.

My view

Today’s broad rally across Korean, Japanese, and U.S. memory stocks suggests that the market is already positioning for a neutral-to-bullish outcome:

Alphabet probably will not cut AI CapEx, and Google Cloud is unlikely to deliver a major disappointment.

The problem is that a neutral result has already been partially priced into today’s rebound.

The market is always smarter than any one of us. Never assume you can outsmart it—I learned that lesson the hard way. 😭[Cry][Facepalm][Heartbreak]

With leverage still elevated and forced deleveraging not yet fully over, I am not willing to bet on the most bullish scenario.

A neutral outcome may only sustain the rebound, while a bearish outcome could trigger a second round of valuation compression. The risk-reward is therefore not as attractive as it may appear.

So my conclusion remains unchanged:

What rallied today was the expectation that “Google probably will not cut Capex.”

The market is not yet pricing in a renewed acceleration in AI spending.

Google is the first test tomorrow. Whether Microsoft, Meta, and Amazon continue raising CapEx later this month will ultimately determine whether memory stocks can move from a rebound into a genuine reversal.

Intel $INTC and Nokia $NOK also report on Thursday. Their results will be worth watching for further evidence on demand for CPUs, photonics, and optical communications within the broader AI infrastructure narrative.