To The Moon

Before discussing anything else, let me start with a simple question about the memory sector.

From their recent highs to where they are now, stocks such as [$SK hynix (SKHY)$](https://ttm.financial/S/SKHY) and [$SanDisk (SNDK)$](https://ttm.financial/S/SNDK) have fallen sharply in less than two weeks.

Have HBM orders declined during this period? Have NAND flash and DRAM prices fallen?

Clearly, the answer is no.

This suggests that the fundamentals have not changed. What we are seeing is simply a short-term leverage unwind.

Below is my detailed analysis and personal view based on the available data.

First: SK hynix’s fundamentals have not been disproven, but the short-term bottom has not been confirmed

SK hynix’s Korean-listed shares plunged by more than 15% at one point today. The KOSPI fell nearly 9% and triggered a temporary trading halt. Meanwhile, SK hynix’s ADR had just risen 12.8% on its first day of trading in the U.S.

Such an extreme divergence obviously cannot be explained away with a simple statement like “earnings missed expectations.”

I am more inclined to break the sell-off down into three components.

First, market expectations were too high.

Everyone now knows that HBM demand is strong, margins are high, and SK hynix is one of Nvidia’s most important suppliers.

Once all of this becomes market consensus, “strong earnings” alone are no longer enough to move the stock higher. What the market really wants to see is:

Can HBM4 shipments continue to exceed expectations?

Can profit margins continue to improve?

Can order growth remain at its current pace through 2027?

If even one of these falls slightly short of the most optimistic expectations, investors will choose to take profits.

Second, leverage in the Korean market is too high.

A large number of Korean retail investors have gained exposure to SK hynix through margin financing and 2x leveraged long ETFs.

Once the underlying stock begins to fall, leveraged ETFs are forced to rebalance their positions, while margin investors face collateral calls. This ultimately creates a feedback loop:

The stock falls → leveraged products are forced to sell → the stock falls further.

This is not ordinary profit-taking. It is a classic liquidity stampede.

Third, the long-term demand outlook has not fundamentally changed.

SK hynix’s CEO still believes that the industry could face the most severe memory shortage in its history by 2027, with supply potentially remaining insufficient even beyond 2030.

My conclusion is therefore straightforward:

SK hynix’s long-term investment thesis has not been invalidated, but a short-term bottom has not yet been confirmed.

A great company can still see its stock price fall further.

Fundamentals determine where the stock can ultimately recover to; leverage and liquidity determine how far it can fall in the meantime.

Second: Why do Micron and SK hynix have different earnings sensitivity?

If both companies benefit from rising memory prices, why might Micron benefit more than SK hynix in the short term?

Many investors have a misconception:

Because SK hynix has the strongest HBM technology, it must also have the greatest earnings sensitivity in every memory upcycle.

That is not necessarily true.

Although [$Micron Technology (MU)$](https://ttm.financial/S/MU) and SK hynix are both memory companies, the sources of their near-term earnings growth are not exactly the same. Their levels of leverage exposure also differ, as shown clearly in the chart I provided.

Micron has relatively greater exposure to conventional DRAM.

Contract prices for products such as DDR5 and server DRAM are generally renegotiated periodically. Once market prices rise, the new contract prices can flow through to the income statement relatively quickly.

Therefore, the faster conventional DRAM prices increase, the more visible Micron’s near-term earnings upside may become.

SK hynix is different.

HBM is not ordinary, standardized DRAM. It is deeply integrated with the platforms of customers such as Nvidia and AMD and requires lengthy qualification processes, joint development, and advanced packaging.

Its supply volumes, pricing, and delivery schedules are often determined earlier through long-term agreements.

This means that although HBM offers higher technological barriers, stronger customer stickiness, and a longer growth runway, it may not capture the full benefit of rising conventional DRAM spot prices in every quarter.

In other words:

Micron is a more direct beneficiary of rising memory prices.

SK hynix is a core long-term beneficiary of the expansion in AI computing infrastructure.

Micron’s advantage is greater near-term earnings sensitivity. SK hynix’s advantages are its technological moat and the stronger long-term visibility of its orders.

The market’s current concern about SK hynix is not that HBM demand has disappeared. Rather, it is that the pace at which HBM4 shipments and profits are being realized has temporarily fallen short of the most aggressive expectations.

This is also why I believe Micron and SK hynix should not be compared using the same simple price-to-earnings multiple.

SK hynix needs to be valued on a sum-of-the-parts basis:

Conventional DRAM and NAND remain cyclical businesses.

HBM is closer to a growth business tied to AI infrastructure.

What will truly determine the next upward re-rating of SK hynix is not how much DRAM prices can rise this year, but whether HBM can sustain high growth and strong profitability through 2027 and beyond.

Third: This is not an ordinary correction—it is a deleveraging event in the Korean market

Why did SK hynix fall so dramatically today?

Because the market is no longer trading on fundamentals. It is trading on leverage.

SK hynix’s Korean-listed shares plunged by more than 15% today, while the KOSPI fell nearly 9% and even triggered a temporary intraday trading halt. Some leveraged SK hynix products lost more than one-third of their value in a single day.

This is not how a normal market should behave when digesting an ordinary piece of bad news.

Over the past year, Korean AI and memory stocks have risen too far and too quickly.

More and more retail investors were no longer satisfied with owning the underlying shares. Instead, they amplified their returns through:

• Margin accounts

• 2x leveraged long ETFs

• Over-the-counter leveraged products

• Derivatives and on-chain contracts

During the rally, these tools accelerated the stock’s climb toward its peak.

During a sell-off, however, they have exactly the opposite effect.

Leveraged ETFs must rebalance daily. Margin investors must meet collateral calls. Institutional risk models require exposure to be reduced.

As a result, everyone is forced to sell at the same time.

This explains how SK hynix could suffer a historic one-day decline even though its fundamentals had not suddenly deteriorated.

The most frightening thing about deleveraging is that it does not automatically stop simply because valuations have become cheap.

As long as margin positions have not been fully cleared and trading volume in leveraged ETFs remains far above normal levels, the stock may continue falling below what its fundamentals would suggest is fair value.

To determine whether the Korean market has genuinely bottomed, I will be watching four indicators:

• Whether trading volume in leveraged ETFs declines significantly

• Whether Korean retail margin balances begin to contract

• Whether foreign investors stop being net sellers of SK hynix and Samsung Electronics

• Whether the Korean won stabilizes

If these indicators do not improve, a single intraday rebound does not mean that the deleveraging process is over.

Do not look for an absolutely “reasonable” valuation in the middle of a liquidity stampede.

When the market is clearing out leverage, its first concern is not what a company is worth. Its first concern is who still has not sold.

And do not try to predict the short-term direction of the market. Short-term market movements are inherently unpredictable—you are simply betting on your luck.

The declines over the past few days have been exceptionally severe, so I will publish this article in two parts.

Let me reiterate my view once again:

The memory upcycle remains intact. The leverage unwind will be extremely painful, and judging by current market sentiment, it is not over yet.