Kioxia (285A): up 153% since February, down 58% from its high—why the starting price matters

One closing price, three comparisons

This article uses the August 7 close and the historical prices below, before the stock split scheduled for October. The “about six months” interval runs from February 10 to August 7.

Starting pointStarting priceAugust 7 closeChange
February 10, 2026 close¥18,845¥47,730+153.28%
June 22, 2026 52-week high¥112,700¥47,730−57.65%
July 21, 2026 close¥61,060¥47,730−21.83%

The increase from February was ¥28,885 per share, leaving the stock at approximately 2.53 times that starting price. From February to the June high, it had risen about 498%, or to roughly 5.98 times the starting level. Much of that extraordinary rise had therefore reversed by August without eliminating the earlier gain.

These figures compare historical prices. They do not identify a typical investor’s purchase price, show the performance of all shareholders, or include dividends and trading costs. The peak comparison also uses a 52-week high rather than a closing-price-to-closing-price interval.

What the entry price changes for 100 shares

Consider two hypothetical purchases of the same quantity, one at the February close and one at the June high. Taxes, fees and dividends are excluded, and the table shows an unrealized price change rather than cash from a sale.

Hypothetical entryPurchase amountValue on August 7Unrealized gain or loss
100 shares at ¥18,845¥1,884,500¥4,773,000+¥2,888,500
100 shares at ¥112,700¥11,270,000¥4,773,000−¥6,497,000

One position has gained about ¥2.89 million; the other has lost about ¥6.5 million. That difference comes from the price paid, despite identical holdings at the end of the comparison. Identifying a promising business and deciding what price to pay for its shares are separate judgments.

Regaining the high requires a larger percentage rise

Returning from ¥47,730 to ¥112,700 would require a gain of ¥64,970 per share, or approximately 136.12%. A fall of nearly 58% does not reverse with a subsequent gain of 58%, because the recovery percentage is calculated from the smaller remaining value.

The calculation describes the distance to a previous price, not a forecast that the price will be regained. A partial rebound, a return to an investor’s entry price and a return to the market peak are different outcomes.

The daily range adds another measure of risk

On August 7, the stock opened at ¥49,330, reached ¥50,060, fell as low as ¥44,170 and closed at ¥47,730. The ¥5,890 intraday range was equivalent to ¥589,000 for 100 shares. That is the difference between the day’s high and low, not the loss every holder incurred.

The stock-information snapshot used for this comparison showed market capitalization of approximately ¥26.16 trillion. That valuation needs an earnings and cash-flow assessment; the daily trading range says more about price volatility than about business value.

A rising chart can become its own reason to buy

A rapid rally attracts attention and can make waiting feel costly. Investors may begin reacting to the fear of missing out rather than to a fresh estimate of earnings. Expectations for AI storage and a stronger NAND market can also be reflected in the price faster than the business realizes them.

Under those conditions, good results can disappoint if the price already assumes something better. Shareholder sales or reductions in margin-financed positions may affect trading, but this price history does not establish how much either contributed to the decline. Assigning the move to a specific participant would require additional evidence.

Kioxia’s core businesses are NAND flash memory and SSDs. Demand matters alongside memory prices, shipment volumes, product mix, factory utilization and investment. A price below half its high does not by itself show that the business is broken or that its current valuation offers a bargain.

Evidence that would change the outlook

MeasureWhat it helps assess
NAND prices and shipmentsWhether the industry recovery is continuing
Enterprise and data-center demandWhether AI-storage expectations are becoming sales
Operating margin and operating cash flowWhether stronger sales produce profit and cash
Capital spending and competitors’ capacityThe cost of growth and risk of future oversupply
Major-shareholder holdings, margin balances and volumeThe scale of potential supply and the market’s ability to absorb it

A favorable scenario would see firm NAND and AI-SSD demand produce stronger profit and cash than expected. A weaker scenario would combine softer demand or expanding capacity with lower memory prices and reduced valuations for future growth.

The three-for-one split scheduled for October 1 would lower the minimum purchase amount. It would not directly change company value; historical prices and per-share earnings would need consistent adjustment after it takes effect.

The useful decision question is what future cash generation supports the current price. Past gains, the distance from the peak and a personal entry price provide context, but none answers that valuation question alone.

Related pages

Sources

Price comparisons use the August 7, 2026 snapshot and the stated historical dates. Returns are rounded and exclude dividends, fees and taxes.