The one number
440.
Not 250, which is the number in every headline. Not $300 million, which is the number in the filing's first item. 440 is the number of net new coffeehouses Starbucks now expects to open worldwide in fiscal 2026, and it appears in Item 7.01 of Monday's 8-K, underneath the part everyone read. The prior guidance was 600 to 650.
That is a cut of roughly a third to the rate at which this company adds stores, disclosed on the same page as a store-closure announcement written in the language of housekeeping.
What was actually announced
On 24 September the chief operating officer published a letter saying Starbucks would close about 250 North America coffeehouses "later this week," in locations where the company does not "see a path to acceptable financial performance." The same day, an 8-K reporting a 22 September board approval put numbers to it:
- ~$300 million in total restructuring charges
- of which ~$200 million cash, primarily lease exit costs and employee separation benefits
- and ~$100 million non-cash, from disposal and impairment of company-operated coffeehouse assets
- with the majority of closures complete by the end of fiscal 2026 - which ends 27 September, two days from now
No headcount was disclosed in either document. Anyone publishing one is not getting it from the company.
The framing deserves a note. Both the letter and the filing describe the closures as "approximately 1% of our more than 18,000 North America coffeehouses." Against the 18,371 North America stores reported at 28 June, 250 is 1.36%. Against the 11,149 of those that Starbucks actually operates itself - the only ones it can unilaterally close - it is 2.2%. The company chose the denominator that produces the smallest number. That is not a scandal; it is worth noticing.
It is also the second wave, not the first. Starbucks closed 247 stores in the first three quarters of fiscal 2026 under the previous plan, and its North America store count was already down about 2% year over year before this week.
Where it sits
At 94.29 the stock is below every average anyone watches: the 20-day at 99.74, the 200-day at 98.49, the 100-day at 102.97 and the 50-day at 103.24. It lost the 200-day nine sessions ago. The weekly support band sits overhead at 102.92 and 101.33. RSI is 28.8, the first reading under 30 in this decline.
Against its own year, the picture is less dramatic than that list sounds. Starbucks is down 14.7% from the 110.51 high of 13 August, but still up 20.9% from the 77.99 low of last November and up 11.9% year to date. This is a drawdown inside a positive year.
The retracement of the decline is the cleanest read on the chart. Measured from 110.51 down to 92.76 on 24 September, the levels fall at 96.95, 99.54, 101.64, 103.73 and 106.71 - and not one of them has been touched since the low was set. The stock has not yet reclaimed even the shallowest of them.
The tape says the news is not the story
Here is what makes this one interesting.
The sessions on either side of the announcement moved -0.5% and +0.7%, on volume no heavier than the ten-day average of 7.8 million shares. A $300 million charge and a one-third cut to store growth produced, so far, nothing.
That is because the repricing already happened. By the last close before any of this was public - 94.14 on 23 September - the stock had given up 14.8% from its 13 August high, with no session worse than -2.5%, no gap larger than -1.3%, and twelve of the last sixteen sessions closing red. It slipped further intraday on the day of the announcement, to 92.76, which takes the full decline to 16.1%. Either way there is no event in this chart. It is six weeks of continuous supply.
Two readings fit. Either the market worked out what was coming and the announcement was confirmation, or the selling was about something else entirely and the closures are incidental to it. The tape cannot tell you which, and neither can I. What it does rule out is the simplest story - that a headline knocked the stock down. The sequence is the wrong way round.
The catch
The demand numbers are good, and that cuts against the gloom. Third-quarter comparable sales were +7.9% globally, +7.9% in the US and +8.1% across North America, and the company described the growth as transaction-led - more visits, not just higher prices. Revenue of $9,322.7 million was down 1.4%, but that is the China joint venture converting to a licensed model, not a sales decline. Closing unprofitable stores while comparable sales run near 8% is a defensible use of capital, and a company with a demand problem does not print those comps.
The charge may be larger than it looks against what was already flagged. The July 10-Q said roughly $120 million and $110 million remained to be incurred under the two existing restructuring plans across the rest of fiscal 2026 and the first half of fiscal 2027 - about $230 million in total. This week's $300 million exceeds that entire runway. The 8-K does not say whether the new figure is incremental to those plans or inclusive of them, so the honest position is that this is unresolved, not that it is a step-up. It is the first thing worth checking in the next filing.
China is no longer in the comparison. The joint venture with Boyu closed in April 2026 and 7,991 stores moved to a licensed model, which removes China from the comparable-sales base entirely. The last disclosed China comp was +0.5% in the second quarter. Anyone comparing this year's growth to last year's is not comparing the same company.
And the growth framework has not been restated. The January Investor Day laid out a fiscal 2028 framework built on more than 2,000 net new stores, alongside 13.5-15% non-GAAP operating margins. Nothing in this week's filing reconciles a year of roughly 440 against that. It may be a pause; it may be a reset. The company has not said.
One practical note: Starbucks' fiscal year ends 27 September, and as of today its investor relations calendar shows no scheduled date for the fourth-quarter report. If you see a date published somewhere, it did not come from the company.
The question
The stock fell 15% before the announcement and moved less than one percent on it. In November, Starbucks reports a fiscal year that ends this weekend, and the openings number stops being guidance and becomes a fact. Does 440 turn out to be the bottom of the cut, or the first print of a smaller company?