Remembering Kioxia - Rating Update

Have I remembered why I got into this memory play, or am I paying to forget?

I first invested in Kioxia on April 7th (@ 14.59) because I believed the market hadn’t begun to run in an important piece of the AI infrastructure story. A leader in NAND, the former Toshiba entity, was not “hot” like Micron, and not as volatile as Samsung and SK hynix.

Since then, the business has reported stronger financial results, my understanding of the company has matured, and the great unwinding of investor confidence in the AI play has settled in among the myriad of concerns plaguing tech. My initial entry has averaged up. This looked ok as Kioxia peaked just under $71 a month ago, but what about now as we dip deep to $32?

Created with TradingView

The burning question I had on Friday as we experience another chip sector sell off became: Would I allocate new capital to Kioxia today?

This feels like a very different exercise than simply defending my existing position.

The purpose of this review is to revisit my “why” for getting in this trade in the first place, challenge it with new data, and determine whether the facts support increasing, maintaining, or reducing my exposure.

I know, Cynce is not known for deep due diligence but I really have been biased on this place and I need to get over myself. When I initiated this position, the theory was straightforward. I believed the market was overlooking a company with meaningful exposure to enterprise storage, improving memory fundamentals, and one of the most important secular trends in technology: AI infrastructure.

Rather than adding my position immediately, I started to think about things differently.

If I had no position today, would I buy more at today’s price?

This was important as I try to be better and more disciplined, but there are also no US options on Kioxia, so without any fancy mischief, I resorted to actually auditing every assumption behind my original investment. Is this to say I would have just bought Jan 2028 calls and left it be if they were available? Probably. But that’s not necessarily the right thing to do.

Over the past few weeks, I began reviewing Kioxia’s financial performance, competitive position, valuation, industry structure, and the broader memory cycle. My goal wasn’t to defend my investment. It was to determine whether the evidence justified increasing my position to an already set limit I haven’t reached.

I’ll admit, after my analysis, some parts of the my original theory strengthened.

Fiscal 2025 results (which ended this Spring of 2026) were exceptional. Revenue reached ¥2.338 trillion (~$14B), operating profit nearly doubled, net income more than doubled, and the balance sheet continued to improve. Enterprise SSD adoption, AI infrastructure spending, and hyperscale storage demand remain compelling long-term tailwinds.

However (there is always a however, isn’t there), other parts of the story had more nuance than I had hoped.

Memory remains one of the semiconductor industry’s most cyclical businesses. Strong results today don’t eliminate the possibility of pricing pressure tomorrow. There is quite a bit of commodity to this space. Competition from Samsung, SK hynix, and Micron remains intense, while capital intensity and execution risk continue to shape returns throughout the cycle.

These factors influenced how I am thinking about them today, in view of the sector and the AI capital trade risks today, not in the rearview mirror of last Summer or April. While my confidence in the business has increased, my confidence that today is the right time to materially increase my position has not.

Those are two different decisions.

Kioxia continues to look like a business I’d be comfortable owning for years. The company appears to be executing well and participating in attractive long-term trends. However, the evidence today supports patience rather than additional capital deployment.

For now, I’m comfortable remaining a shareholder while continuing to monitor execution, margins, enterprise SSD adoption, and the durability of the current recovery.

Portfolio View: Hold & Monitor.

If you think I should publish the 20 or so pages I pulled together on this, I will after their earnings this week. I would rather have a more comprehensive view with updated guidance before I put anything like that forward for broader consumption.

As usual, thanks for reading. If you found value in my perspectives, please tell your friends. I write to think out loud and to help myself figure out this wild hobby.

Also… I’m not a financial professional. Do your own research and remember that our risk tolerances are probably not the same. Please don’t do what I do, its for entertainment purposes only :)