Has the September curse in the storage sector disappeared?

The rally is no longer being driven by just one company.

Micron broke through $1,000. SanDisk rose 11.75% in a single day, and the rally continued on-chain after U.S. stocks closed. SK Hynix, Samsung, NAND, DRAM, HBM, and enterprise SSDs are forming sector-wide resonance.

This shows that capital has begun to reprice the entire storage industry chain.

AI servers consume more than just GPUs.

HBM handles high-speed computing. DRAM provides server memory. NAND and enterprise SSDs handle massive data storage. As AI data centers continue to expand, storage is shifting from a supporting role in past cycles to a core asset of AI infrastructure.

Three forces are behind this round of gains.

First, AI demand continues to exceed supply.

Second, storage prices are rising, rapidly releasing corporate profits.

Third, with the U.S. midterm elections approaching, capital has begun trading AI, semiconductors, U.S. manufacturing, and the wealth effect of technology stocks.

But the September curse has not completely disappeared.

The storage sector already underwent a deep correction earlier, so some of the risks were absorbed in advance within the sector. However, the S&P and Nasdaq remain at high levels, while U.S. Treasury yields continue to rise.

The 10-year Treasury yield reached as high as 4.82%, while the 30-year yield exceeded 5.2%. The economy is not in recession, and corporate profits can still grow, but high interest rates are compressing valuations across the technology sector.

A soft landing can support storage demand, but it cannot prevent a market correction.

The most likely market path is:

Use September’s CPI release, Federal Reserve meeting, and high-level profit-taking to complete a shakeout.

Resume pushing up AI, semiconductors, storage, and U.S. manufacturing in October.

Maintain the indices and market sentiment ahead of the midterm elections in November.

So this September looks more like sector rotation and a high-level shakeout. The probability of a broad, trend-driven collapse is temporarily not high.

The period that truly requires caution may be March–April 2027.

By next spring, the midterm elections will have ended. The market will begin to test whether AI capital expenditure can continue, how long higher storage prices can be sustained, and whether new capacity from manufacturers such as Micron, SK Hynix, Samsung, and Kioxia will once again alter the supply-demand relationship.

Stocks always trade the future in advance.

Even if new capacity is not released until the second half of 2027, capital may begin trading a peak in storage prices as early as March–April.

So my judgment is:

The main uptrend in the storage sector has begun.

The September curse has been partially absorbed in advance, with the remainder potentially being resolved through consolidation and sector rotation.

The midterm elections may continue to push the real major correction further back.

If this rally continues through the end of the year and then gets another boost from earnings in the first quarter of next year, March–April 2027 will become the most dangerous correction window for the entire storage sector.

Spot holdings can continue to be held.

Leverage must be controlled in futures contracts.

The biggest risk to the storage bull market is not a lack of upside.

It is that after everyone understands the logic, capital will use one sharp correction to wipe out all the highly leveraged positions.$MU $SNDK