Intel’s Weirdest Spending Shift 😕

I’ve been reading The Little Book of Common Sense Investing again.

It’s a great on-ramp into investing.

If you’re a beginner, it explains things without jargon or hype.
If you’ve been investing for a while, it’s a useful reminder of what actually matters (and what doesn’t).

No strategies. No predictions.

Just the core ideas most people forget once markets get noisy.

If investing ever felt overwhelming — or overcomplicated — this is a great reset which I why I re-read it from time to time.

I have a physical copy. It works great as an audio book,

too. You can get it however you prefer by clicking here.

For months, Intel told investors to expect lower capital spending.

Now? That story’s getting a rewrite.

On the earnings call, CFO David Zinsner said 2026 CapEx will likely be flat to slightly down vs. 2025 — but here’s the kicker:

Way more money is going into tools.

Think fewer warehouse expansions… and way more shiny, expensive chip-making machines.

“We’re spending a lot less on space… and ramping up tool spending quite a bit in ’26 relative to ’25.”

In other words: The factory’s built. Now it’s time to plug it in.

Tool spending = actual production capacity.

Intel says 2026 CapEx is really about supporting demand in 2027 and beyond.
So this isn’t panic spending — it’s forward-looking muscle-building.

But the market didn’t clap.

Intel stock dropped ~16% on Friday after earnings.

Why?

Wall Street wanted results now. Intel’s talking about results later.

Classic mismatch.

From Intel’s 10-K:

So yes — spending is lower overall. But the mix has changed.

Intel is going hard on:

Zinsner confirmed they’re aggressively securing tools and ramping wafer starts on those nodes.

But Intel 14A? That one’s on a leash.

Why?

Because it’s tied to external foundry customers.

No customers = no massive build-out.

And Intel didn’t sugarcoat the risk.

If Intel can’t land a major external customer, 14A may not be economical — and could be paused or scrapped entirely.

That’s a real line in the sand.

While Intel tightens the belt, Taiwan Semiconductor Manufacturing is doing the opposite.

TSMC’s 2026 CapEx plan:

For context:

TSMC CFO Wendell Huang summed it up: New nodes are brutally expensive.

Every shrink comes with a bigger bill.

All this tool spending is music to the ears of equipment makers:

No matter who wins the chip race… the shovel sellers get paid.

The chip war isn’t slowing down. It’s just getting more expensive.

Chipmakers are fighting over advanced nodes, but tool makers get paid either way. Intel is ramping tool spend, and TSMC is throwing $50B+ at equipment. Rising node complexity = rising tool demand.

📌 Action: Accumulate semiconductor equipment leaders like $ASML ( ▲ 0.8% ), $AMAT ( ▲ 1.17% ), $LRCX ( ▲ 0.82% ), or $KLAC ( ▲ 1.28% ) on market pullbacks. Hold through the multi-year CapEx cycle.

Intel’s stock dumped ~16% on weak near-term guidance — but the spending shift signals capacity coming online later, not retreat. This sets up classic “ugly now, better later” price action.

📌 Action: Buy $INTC ( ▲ 5.71% ) in tranches near major support zones. Treat it as a slow rebuild, not a moonshot. Trim into sharp relief rallies. Wall St hates waiting — you don’t have to.

TSMC isn’t slowing — it’s accelerating. $52B–$56B CapEx means years of guaranteed ecosystem demand, even as costs rise. If advanced chips grow, TSMC stays central.

📌 Action: Hold or add $TSM ( ▲ 4.25% ) as a core long-term position. Expect volatility, not collapse. Pair with patience. When everyone needs chips, they need TSMC.

There’s chatter flying around that Samsung Electronics slapped an 80% price hike on its memory products.

Samsung’s response? Yeah… no.

According to local reports, that headline-grabbing number is not accurate.

This is where things get spicy 🌶️

Samsung’s distributors have sent out notices about price increases, but not because Samsung woke up and chose violence.

The reasons:

Costs are up, supply is tight, and someone has to pick up the bill.

Memory module manufacturers say:

In other words: The vibes say “higher prices.” The paperwork says “not yet.”

One manufacturer hinted prices are going up — just didn’t say how much. Which is finance-speak for “brace yourself.”

Earlier this month, reports suggested SK hynix and Samsung are considering up to 70% increases for server memory in Q1.

Why? Because AI workloads are memory-hungry monsters

And South Korea basically owns the memory game:

They’re the Big Three.

Here’s the nerdy-but-important bit:

Memory suppliers have been cutting production of older DRAM to focus on HBM (High-Bandwidth Memory) — the premium stuff needed for AI accelerators.

HBM =

And SK hynix? They’re a major HBM supplier to Nvidia — aka the final boss of AI chips.

So yes, shortages are real. And yes, prices eventually move higher.

Just… not 80% overnight.

The memory market is tightening. Prices are creeping up. But the “80% jump” headline?
That’s clickbait, not a price list.

Short sellers are playing chicken with a freight train… and the train’s name is SanDisk.

While the stock has been ripping higher, bears have been doubling down instead of tapping out. According to Bloomberg, short interest in Sandisk has quietly crept from ~4% to 7.5% of the float since November.

People keep betting against the rally. And the rally keeps saying, “lol.”

S3’s short-squeeze risk score just hit 82.5… a level they straight-up label “extreme.”
That’s not a yellow light. That’s the dashboard flashing red while smoke comes out of the hood.

Why? Short sellers are now sitting on roughly $3B in mark-to-market losses.

That’s not paper cuts — that’s financial blunt-force trauma.

SanDisk shares are up 112% this year.

For context:

It’s now one of the top performers of 2026, and it’s doing it while bears are still stubbornly holding the rope.

This isn’t meme-stock chaos. There’s actual fuel here:

And then came the catalyst.

At CES earlier this month, Nvidia CEO Jensen Huang casually tossed a grenade into the market by calling storage a “completely unserved market.”

Short sellers heard that and said: “Nah, we’re good.”

The stock heard that and said: “Vertical line.”

SanDisk isn’t some overnight hype story:

Since that re-listing? The stock is up nearly 1,200%.

Yes. Twelve. Hundred. Percent.

Sometimes the market doesn’t ask shorts to leave politely. It escorts them out through the window

Short interest is climbing while price keeps ripping — that’s fuel, not noise. If shorts are forced to cover, moves tend to be fast and violent.

📌 Action: Build a starter position in $SNDK ( ▼ 4.2% ), then add on shallow pullbacks (2–5%) while short interest remains elevated. Trail a stop once momentum accelerates. You’re letting other people’s bad bets power your upside.

AI spending isn’t just chips anymore — data has to live somewhere. Storage is quietly becoming the bottleneck.

📌 Action: Pair $SNDK ( ▼ 4.2% ) with semiconductor ETFs like $SMH ( ▲ 1.52% ) or $SOXX ( ▲ 1.46% ) to capture the broader AI → memory → storage flow. Rebalance monthly toward the stronger relative performer. This isn’t a single-stock YOLO. It’s a theme shift.

Flash memory shortages = price hikes + margin expansion. That’s the kind of fundamental tailwind institutions love to hold through volatility.

📌 Action: Treat $SNDK ( ▼ 4.2% ) as a core growth holding, not a flip. Size it like a conviction play and hold through earnings as long as pricing strength persists. Scarcity + demand = earnings surprises. Markets pay up for that.