The rise in SanDisk's stock price is a valuation re-rating in the AI era!
$Sandisk(SNDK.US) $Micron Tech(MU.US) Trend is bullish!
Today, I'd like to share with you SanDisk's performance and valuation. I've mentioned many times in my previous articles that SanDisk is undergoing a valuation re-rating in the AI era. The rise in SanDisk's stock price is not driven by hype. SanDisk is still on track to make history by winning the title of the S&P 500's top performer for two consecutive years (see my articles from December 21st last year and New Year's Day 2026).
To put it simply, the estimated calendar year 2026 EPS is around $90-$110, with a median of $100. Note this is the calendar year. SanDisk's fiscal year 2025-2026 runs from July 1, 2025, to June 30, 2026, and the next fiscal year follows suit. So, for the convenience of stock valuation, I use the calendar year. The P/E ratio on the App is not a meaningful reference. Once SanDisk releases its earnings report and guidance on April 30th, everyone will be able to calculate the 2026 calendar year EPS (it always exceeds my estimates). It's crucial to note that our price target for this year is based on this year's performance expectations, not 2027's. This isn't like the 'Mian A' market where they hype up things that won't happen for the next decade.
Next is valuation. How SanDisk and Micron Technology are priced depends entirely on whether the market perceives memory products as cyclical stocks or AI computing power stocks. If valued as an AI stock with a 25x P/E ratio, SanDisk's target this year would be above $2500. Considering the market needs time to understand, I give a 15-20x valuation, which is $1500-$2000.
Looking at 2027, the biggest value driver for SanDisk to become a memory giant is HBF, SanDisk's original invention. Its role in matching HBM during the inference phase is irreplaceable; it's the memory middle layer for NVIDIA's RUBIN platform. In 2027, HBF will significantly contribute to earnings. It's foreseeable that SanDisk can achieve over 100% EPS growth (not just from HBF, bit shipments will also grow), meaning EPS above $200. If the market further recognizes SanDisk as a leading growth stock in the AI supply chain, a 25x P/E ratio isn't excessive, right? That's a $5000 stock price.
SanDisk's stock price rises as the market continuously debates whether it's a cyclical or an AI growth stock. Some things are like that; when you look back, you think, 'Ah, how did I miss such a simple question?!'
Is it really that simple? The cyclical theory, moat theory, AI bubble theory, 'ChangXin Memory and Yangtze Memory are about to massively expand production, memory will be dirt cheap again'—these arguments are constantly tools for short sellers. The stock price rises in a winding path.
First, the AI bubble theory is just a bull-bear debate in the capital markets, unrelated to strategy. AI is the US's winning strategy in the new round of tech competition. Although the Eastern power lags far behind, it must try hard to respond. Due to the blockade of lithography machines, the 'East' can only use mature processes to solve basic problems and buy time; anything below 28nm is off the table. SMIC's multiple patterning process can barely supply a few essential clients like Huawei at a huge cost. ChangXin Memory's DRAM expansion isn't a money issue; it's a lack of lithography machines. Even at full capacity, its Shanghai fab's market share won't exceed 5%, and it's currently two generations behind. Yangtze Memory has lower requirements for lithography machines, but it doesn't have them either. The bit density of its latest 3D Flash product is only a little over half of Kioxia/SanDisk's. It's okay for consumer-grade use, but is it competitive for enterprise-grade? They can only make do with what they have; something is better than nothing! By the end of 2028, Yangtze Memory aims for a 15% market share. Note the technological gap; don't just look at quantity. It's not competitive in the AI field, or rather, it has no impact on the original memory manufacturers (reality might be worse, not deliberately disparaging). Meanwhile, the demand for memory in the AI era is growing exponentially. Besides HBM and HBF, the demand for enterprise SSDs is surging now. The expansion speed of original manufacturers is far behind the growth of AI demand. Micron's last earnings report stated capacity can only meet 50% to two-thirds of major customers' demand. SanDisk's report and outlook on Thursday will also indicate the degree of NAND demand growth. 'Forever short on memory' might not last forever, but a shortage for three to five years is visible. Note that this demand discussion has nothing to do with consumer electronics. Where's the cycle? At this point, I've addressed both the moat theory and the cyclical theory.
///
In summary, the rise in SanDisk's stock price is the process of market understanding. The cyclical theory, moat theory, and AI bubble theory will be disproven one by one in the AI era!
I must especially explain that my articles are written freely, based on my conservative estimates, without listing specific data. This is because most of the data in the media and from investment banks is not real (real data from big investment banks isn't given to you for free). The thesis format is all about copying; I disdain writing theses anymore (I wrote too many for my degree before), as they have no practical reference value.
Finally, I wish you all good luck and prosperity!
The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.
