To The Moon

Memory investors are getting nervous, and understandably so.

Micron is now around $1,003, more than 18% below its recent high. SanDisk has fallen toward $1,548. SK Hynix remains under pressure, while leveraged memory trades have been hit much harder.

After such a powerful run, an 18% drop can feel like the market is telling us something serious.

Maybe it is.

But I think we need to separate three very different things:

A stock correction.

A conventional DRAM downturn.

A breakdown of the AI memory thesis.

They are not the same.

Right now, I see evidence of the first two concerns.

I do not yet see enough evidence for the third.

Why Has MU Fallen So Much?

The uncomfortable answer is that the memory story became almost too successful.

Investors spent months buying the same thesis:

AI demand is exploding.

HBM is scarce.

Memory pricing is rising.

Margins are expanding.

Supply cannot keep up.

Eventually everyone knows the story.

At that point, excellent fundamentals stop being enough. The market starts asking what happens after excellent.

That is where we are now.

Micron does not merely need strong earnings anymore. Investors want evidence that extraordinary memory economics can persist.

And conventional DRAM is creating doubt.

The Weakness Is Not Imaginary

I do not want to dismiss the bearish argument simply because I remain constructive on MU.

There is a genuine imbalance developing within memory.

HBM demand remains heavily driven by AI accelerators and high-end data-centre infrastructure.

But conventional DRAM has a broader customer base.

PCs.

Smartphones.

Consumer electronics.

Traditional servers.

Industrial applications.

These markets are not experiencing anything close to the same demand explosion.

At the same time, memory manufacturers naturally prioritise the products offering the highest returns.

That created an unusual cycle where AI memory became extremely valuable while parts of conventional memory remained much less exciting.

Now investors are asking whether improving supply will eventually expose weakness elsewhere.

That is a legitimate concern.

But Here Is the Important Distinction

DRAM weakness does not automatically mean HBM weakness.

And HBM is increasingly important to the Micron thesis.

HBM consumes enormous manufacturing resources relative to conventional DRAM.

So even if conventional DRAM demand is softer, strong HBM production can absorb capacity that otherwise would have flooded traditional markets.

That is why simply saying DRAM is oversupplied does not tell me enough.

I want to know:

Which DRAM?

Which customers?

Which capacity?

Which pricing contracts?

Which end market?

The memory market has become much more segmented.

Nvidia Is Both the Strength and the Risk

There is another issue investors should take seriously.

AI memory demand is highly concentrated around Nvidia and the broader AI accelerator ecosystem.

That is wonderful while AI infrastructure spending is accelerating.

It becomes dangerous if the spending cycle slows.

This is why I would not describe MU as a low-risk investment simply because HBM demand remains strong.

HBM concentration has created extraordinary earnings power.

It has also created dependency.

If hyperscalers continue building AI infrastructure aggressively, Micron should benefit enormously.

If AI capex expectations roll over, memory stocks could derate before actual HBM orders decline.

That is probably part of what we are seeing now.

An 18% Correction Is Not Extraordinary for MU

This is psychologically important.

When a stock has risen dramatically, investors become anchored to the high.

Then an 18% decline feels catastrophic.

But the correct question is not:

How far has MU fallen from its high?

It is:

How much has the earnings thesis deteriorated relative to the share price?

If earnings expectations decline 5% while the stock falls 20%, valuation improves.

If earnings expectations decline 30% while the stock falls 20%, the stock may actually have become more expensive.

That is the calculation I care about.

So far, I think MU price deterioration has been faster than fundamental deterioration.

That makes me interested.

Not reckless.

Interested.

$1,000 Is Now the Battlefield

Micron around $1,003 is sitting almost exactly where I wanted the next major test.

The psychological $1,000 level matters.

If buyers defend it and MU begins forming higher lows, this correction may simply be resetting expectations.

But if $1,000 breaks decisively, I would not assume $990 is automatically the bottom.

There is relatively little reason for momentum traders to defend a round number once it fails.

That opens the door to the next support zones.

My Updated MU Pick Levels

$990-$1,020: First Pick

We are here now.

I am comfortable beginning a position, but I would deliberately keep ammunition.

I would not go all-in at $1,000.

$940-$975: Preferred Pick

This is my favourite zone.

If MU reaches here while HBM demand and forward customer commitments remain intact, the risk/reward becomes much more attractive.

$880-$925: Strong Pick

This is where fear would probably be considerable.

If the decline is driven primarily by semiconductor derating and conventional DRAM concerns rather than collapsing HBM demand, I would become significantly more aggressive.

$1,050-$1,080: First Recovery

I want MU to reclaim this area before declaring that the correction has ended.

$1,120-$1,150: Bullish Confirmation

Above here, momentum begins turning meaningfully positive again.

$1,200+: Next Leg

I would only expect this if the market becomes convinced that HBM strength can offset conventional memory normalisation.

Below $880: Thesis Check

At that point I would stop treating every red candle as a bargain and investigate whether earnings expectations are changing materially.

SanDisk Is More Sensitive to This Debate

SNDK around $1,548 is already entering an area I find interesting.

But I would give SNDK slightly less room for error than MU.

Its NAND exposure means pricing and supply discipline matter enormously.

My updated levels:

$1,525-$1,575: First Pick

We are essentially here.

$1,450-$1,500: Preferred Pick

This is where I would be happier accumulating.

$1,350-$1,425: Strong Pick

At this point I would want to confirm that NAND fundamentals remain intact before adding aggressively.

$1,625-$1,675: First Recovery

Above $1,750: Bullish Confirmation

Below $1,350: Thesis Check

For me, MU remains the cleaner AI-memory trade.

SNDK offers more cyclical upside, but also greater sensitivity if conventional memory conditions weaken.

What Actually Calls the DRAM Bottom?

I do not think one green day will do it.

I want a sequence.

First, memory pricing expectations stop deteriorating.

Second, MU stops making lower lows.

Third, HBM commentary remains strong.

Fourth, AI capex remains intact.

Fifth, conventional DRAM inventory begins stabilising.

Then I want price confirmation.

For MU specifically:

Hold $940-$1,000 → reclaim $1,050 → reclaim $1,100.

That would make me much more confident that the market has absorbed the bad news.

My 2-4 Week Scenarios

My base case, 50% probability, is that MU forms a volatile bottom somewhere between $940 and $1,050.

There may be several false starts. Investors are nervous and semiconductor positioning remains crowded.

My bull case, 30% probability, is that $1,000 approximately holds, HBM commentary stays strong and MU reclaims $1,080.

That could quickly bring $1,120-$1,150 back into play.

My bear case, 20% probability, is the one investors should prepare for rather than pretend cannot happen.

MU breaks $940 and falls toward $880-$925 as investors price a more conventional memory downturn.

I would not automatically sell there.

I would check the fundamentals first.

If HBM demand remains strong, customer commitments remain intact and AI infrastructure spending continues, $900 MU could become considerably more attractive than $1,200 MU ever was.

What Would Make Me Genuinely Worried?

This is the section I think nervous shareholders should focus on.

I would become materially more bearish if several things happen together:

HBM orders weaken.

AI hyperscalers reduce capex.

HBM pricing begins falling sharply.

Conventional DRAM inventories rise.

NAND pricing rolls over.

Micron lowers forward expectations.

New supply arrives faster than demand.

Gross margins begin contracting much faster than expected.

If we get three or four of those simultaneously, I change my thesis.

A falling share price alone is not enough.

There Is Also a Bullish Outcome Few Are Discussing

What if conventional DRAM weakens, but HBM remains exceptionally strong?

Then memory manufacturers continue shifting capacity toward higher-value products.

That can gradually rebalance conventional DRAM supply.

In other words, the weakness itself can eventually create the conditions for recovery.

Memory has always been cyclical.

High prices encourage supply.

Oversupply crushes pricing.

Weak pricing discourages investment.

Supply tightens.

Pricing recovers.

The difference this cycle is that AI has inserted a huge structural source of demand into that traditional loop.

That does not eliminate the cycle.

It may change its shape.

For Those Worried About the Drop

I would not tell anyone that MU cannot fall another 10%.

It absolutely can.

I would also not interpret an 18% correction as proof that the memory thesis has failed.

Instead, I would ask one question every week:

Is the stock falling because earnings are deteriorating, or because expectations are resetting?

Right now, I think it is more the second than the first.

But the gap is narrowing, which is why I am watching the data more carefully.

My Pick

I still prefer MU over the broader memory group.

But my strategy has changed from momentum to accumulation.

$990-$1,020: Start

$940-$975: Preferred

$880-$925: Strong

I would deliberately split my capital.

For example, rather than deploying 100% at $1,003, I would think in tranches.

25% around $1,000

35% around $940-$975

40% around $880-$925

If MU reverses before reaching the lower levels, I still participate.

If the correction deepens, I have capital available when fear is greater.

And if fundamentals actually deteriorate, I am not trapped with a full position purchased too early.

That is how I would handle this correction.

Not by pretending the risks do not exist.

And not by panicking because the chart is red.

The memory bull case has changed.

The easy stage was recognising the shortage.

The harder stage is determining whether HBM structural demand can overpower conventional DRAM cyclicality.

That is the battle I am watching now.

Until the fundamental evidence tells me otherwise, I see this as a dangerous correction inside a still-intact AI memory cycle, not proof that the entire supercycle is over.

But $1,000 is no longer a level I take for granted.

It needs to prove itself.

I am not a financial advisor. Trade wisely, Comrades!

Are you buying MU around $1,000, waiting for $950, or worried this memory correction has much further to go?

If you find this post interesting, give it wings! 🪽 Repost and share the insights.

Do consider ‘Follow me’ and get firsthand read of my daily new post. Thank you. @Daily_Discussion  @TigerPM  @TigerStars  @TigerEvents  @Tiger_SG