The Best Print of the Turnaround Bought Exactly Zero Alpha: Comp +7.9%, EPS Beats by 29%, and the Stock Finishes the Day Behind the Market

Results vs. Consensus

Year-over-Year Detail (from the 8-K Exhibit 99.1 income statement)

Sequential Detail (Q3 FY26 vs. Q2 FY26)

Revenue

The headline revenue decline of 1.4% is a reporting artifact and should be read as such: China retail was deconsolidated into the Boyu joint venture during the quarter, removing roughly $700M of quarterly company-operated revenue from the International segment and replacing it with $53M of licensing and product-supply revenue plus 40% equity-method economics. Strip that mechanical change and the underlying business grew. Licensed store revenue rose 8.6% and "other" revenue (largely Channel Development) rose 14.5%, both of which are the post-transition growth engines.

The revenue miss against a $9.44B consensus is likewise mostly a modeling artifact of how individual contributors handled the JV transition date rather than a demand signal. We would not spend much time on it, and neither did the call. What matters far more is that the comp base itself changed shape: North America store count fell 2% year-over-year to 18,371, and the consolidated comparable-store revenue base of $6,889.3M sits across 16,933 stores versus 17,230 a year ago. Fewer stores carrying more volume is exactly how closure-driven sales transfer inflates a comp.

Margins

Two things are simultaneously true and the market spent Thursday sorting them out. First, the milestone is genuine: North America operating margin expanded year-over-year for the first time since Q1 FY2024, and management stated it expanded "even without the impact of tariff refunds." Second, the magnitude is heavily flattered. On a non-GAAP basis North America margin rose approximately 280 bps, but the CFO volunteered that excluding tariff refunds the improvement was "more than 100 basis points." That means roughly 180 bps of the 280 bps, close to two-thirds, was a refund of tariffs the company had already paid in the first three quarters of the year.

Underneath the refund noise, the operating story holds up. Store operating expenses fell to 45.0% of revenue from 45.9%, and as a percentage of company-operated store revenue they were 55.9% versus 55.6%, essentially flat despite a full quarter of Green Apron labor investment still in the base. Consolidated G&A fell 11.6% as reported, or roughly 20% excluding the $44.1M of transaction costs and $15.3M of transformation costs carried in the line this year, on cost savings, the China deconsolidation, and the lapping of prior-year leadership-experience costs. The $2B gross cost-savings program running through FY2028 is showing up first in G&A and other operating expenses exactly as management framed it.

EPS

GAAP EPS of $0.91 exceeds non-GAAP EPS of $0.85, which is unusual and worth explaining. The $536.3M pre-tax gain on the China divestiture contributed $0.47 to GAAP EPS and is excluded from non-GAAP; running the other way, $302.6M of restructuring and impairments ($0.26), $44.1M of transaction costs ($0.04), and $18.1M of transformation costs ($0.02) are added back. The net effect flips the usual ordering. Neither figure is the run-rate. The relevant number for forward modeling is the normalized ~$0.70.

There is no buyback support in the EPS line. Diluted shares rose to 1,143.8M from 1,139.8M, and the cash flow statement carries no share-repurchase line at all for the nine months. Every dollar of the $2,544.2M in net divestiture proceeds went to debt reduction and the ordinary dividend.

Segment Performance

Segment operating margins above are GAAP and therefore absorb the quarter's $302.6M of restructuring and impairments. On a non-GAAP basis, North America margin was 16.1% (versus 13.3%, a gain of approximately 280 bps) and International operating income was $300.9M on a 22.7% margin. Both figures were confirmed by management on the call.

North America: The Milestone Quarter, With an Asterisk on Its Size

North America revenue rose 6.8% to $7,395.1M on a comp of +8.1% (transactions +4.5%, ticket +3.5%), with the US at +7.9% (transactions +4.2%, ticket +3.6%) and Canada described as stronger still. The segment delivered the quarter's headline achievement, and management led with it.

"We also reached a milestone, with North America operating margin growing year-over-year for the first time since Q1 fiscal 2024. This was driven by operational improvements across both our company-operated and licensed businesses." — Brian Niccol, Chairman and CEO

The operational substrate is credible and improving on measurable lines. Two-thirds of North America company-operated coffeehouses now sit at four or more shots on the internal GROW ranking, up more than five points sequentially and more than forty points since the system launched last October. Food availability is close to 99%, roughly ten points better than a year ago. The share of coffeehouse leaders with two or more years of tenure improved about seven points year-over-year, which matters because management has repeatedly tied leader stability to store performance. Target service times were achieved across every access point in the quarter even as transactions grew.

The asterisk is size, not direction. Of the roughly 280 bps of non-GAAP margin expansion, management attributed "more than 100 basis points" to the underlying business and the balance to tariff refunds. Separately, allocated G&A within the segment fell 43.6% to $95.9M while Corporate G&A rose 9.3% to $463.0M, so a portion of the segment's optical improvement reflects where costs are being carried rather than whether they were eliminated. Consolidated G&A did fall a real $78.4M, so this is not purely a shuffle, but the North America segment line flatters the underlying improvement by a meaningful margin.

Assessment: The inflection is real and arrived on the schedule management laid out, which is a genuine execution win. Sizing it honestly at "more than 100 bps" underlying rather than 280 bps changes the FY27 margin path materially, and that recalibration is the single largest driver of our EPS revision below. We would want to see one clean quarter, without refunds and without allocation shifts, before extrapolating the mid-teens segment margin framework from the January Investor Day.

International: The First Clean Look at a Post-China Portfolio

International revenue fell 34.2% to $1,322.6M as China retail moved to the joint venture, with company-operated revenue down 50.8% and licensed revenue up 20.7%. The comp of +5.7% (transactions +2.6%, ticket +3.1%) was led by Japan, now the largest international company-operated market, on a 30th-anniversary innovation program and what management acknowledged was "a favorable prior year comparison." The UK also contributed. The segment posted its sixth consecutive quarter of positive systemwide comps across 90 markets.

China's contribution is now disclosed on entirely different terms, and the disclosure is thin.

"In the third quarter, we reported $53 million of net revenues attributable to China within our international P&L and operating margin above 100%, reflecting the structure's margin accretive nature. As the joint venture moves beyond this transitional period and scales, we expect our economics to build over time." — Cathy Smith, Executive Vice President and CFO

An operating margin above 100% is arithmetic, not performance: royalty and product-supply revenue of $53M is booked alongside $22.8M of equity-method income that carries no associated revenue line, so the ratio exceeds unity by construction. It tells a reader nothing about how the China business is actually trading. For the first time in years there is no China comparable-sales figure, no China revenue trend, and no unit economics. Roughly 90% of the international portfolio now runs through a licensed structure, and management framed the long-run JV ambition at up to 20,000 coffeehouses.

Assessment: Structurally this is the right model and the capital-light shift is genuinely value-accretive over time. Practically, investors have traded a China business they could measure for one they cannot. The $22.8M of quarterly equity income annualizes to under $100M against a company guiding to roughly $3.0B of non-GAAP operating income, so the near-term earnings contribution is small and the "build over time" language is doing a lot of work. We are modeling China conservatively until the JV discloses its own operating metrics.

Channel Development: Margin Snaps Back, But Coffee Prices Are Doing Some of the Work

Channel Development revenue rose 21.5% to $587.9M and operating income rose 40.2% to $306.2M, with margin expanding 700 bps to 52.1%. That is a striking reversal from Q2, when revenue grew 39% but margin compressed 680 bps to 40.5% on mix. Product and distribution costs fell to 54.6% of segment revenue from 63.4%. The multi-serve Refreshers concentrate and sweet cream platforms are generating trial and repeat rates management described as more than twice the segment norm, and a zero-sugar Doubleshot energy extension launched in North America.

Management was candid that the revenue line is not purely volume-driven, noting the segment grew 22% "helped by coffee inflation" and later warning that "our channel development revenues can move with coffee price trends, given the structure of our CPG business." That cuts both ways. Coffee prices are easing, which management expects to make coffee "largely immaterial to the year-over-year margin comparisons" in Q4, and the same easing removes a revenue tailwind from this segment.

Assessment: Still the most underrated asset in the portfolio and still an asset-light, high-margin annuity that is structurally separable from store-operations risk. But the quarter-to-quarter margin swing from 40.5% to 52.1% should temper anyone extrapolating either endpoint. We model the segment at a mid-to-high-40s margin through FY27 with high-single-digit to low-teens revenue growth once the coffee-price tailwind neutralizes.

Key Operating KPIs

Key Topics & Management Commentary

Overall Management Tone: Assured and, for the first time in this turnaround, unhedged about the near term. Where the Q2 call framed the quarter as "the turn," this one asserted a completed proof ("It's clear proof that our Back to Starbucks plan is working," "we remain ahead of schedule") and spent the prepared remarks on operational evidence rather than on defending the thesis. The notable feature was candor under questioning: the most consequential disclosures of the quarter, that a little under half the comp came from closures, sales transfer, and delivery, and that North America margin expanded "more than 100 basis points" excluding tariff refunds, were volunteered readily rather than resisted. Where the posture thinned was forward framing: FY2027 was referenced repeatedly as an acceleration year for uplifts, technology, and international units, but no financial contour was offered for it.

Topic 1: The Comp Is Excellent and Roughly Half of It Is Not Demand

This is the quarter's most important disclosure and it did not appear in the press release, the prepared remarks, or any of the media coverage. It surfaced only because an analyst asked how much of the comp came from operational mechanics rather than brand momentum.

"With a 7.9% comp that we drove, about half of it or a little bit less than half of it was due to closures, sales transfer, delivery growth. The rest of it is all that great store performance and menu and innovation. The hours of operation piece was actually very small, just a couple of basis points." — Cathy Smith, Executive Vice President and CFO

Taken at face value, roughly 3.5 to 3.9 points of the 7.9% comp came from closing stores and redistributing their volume to surviving ones, plus delivery channel growth. That leaves underlying store-and-menu comp near 4%. Sales transfer is not incremental system revenue; it is the same customers walking into a different door. Delivery is incremental revenue but is channel shift rather than brand-driven traffic, and management confirmed it currently carries no margin trade-off and no measurable cannibalization.

The composition detail reinforces the point. Global comp accelerated from +6.2% to +7.9% on modestly better transactions (+3.8% to +4.2%) and a much larger ticket step (+2.3% to +3.6%). In the US the split is starker still: ticket went from +2.6% to +3.6% while transactions edged down from +4.4% to +4.2%, so ticket more than accounted for the entire US comp acceleration. Since pricing contributed less than a point of the ticket growth, the ticket gains are attach, modifications, and delivery mix. Food attach set a Q3 record with the strongest gains in the afternoon.

Assessment: A 4% underlying comp on positive transactions is a genuinely good result for a business that was comping negative a year ago, and we do not want to be churlish about it. But 4% is a different investment case than 7.9%, and with management now guiding to further closures and modest North America unit growth through FY2027, the closure-transfer contribution is becoming a recurring feature of reported comp rather than a one-time cleanup. Investors underwriting the headline number are underwriting a construct.

Topic 2: North America Margin Inflects, Finally

The first year-over-year North America operating margin expansion since Q1 FY2024 is the datapoint the entire bull case has been waiting on, and it closes the loop on the sequencing management promised: revenue first, earnings second.

"We said we would drive sales growth first and earnings would follow. Our results show we're walking the talk. We're on the right path, we remain ahead of schedule." — Brian Niccol, Chairman and CEO

The CFO's own sizing is the more useful number. Non-GAAP North America margin expanded approximately 280 bps, of which "more than 100 basis points" survives the removal of tariff refunds. Store operating expenses as a percentage of company-operated store revenue were 56.1% versus 56.5% a year ago, a genuine 40 bps of leverage while Green Apron labor investment remains fully in the base. That August anniversary matters: from Q4 forward the company laps the launch, and the labor-investment headwind converts to a neutral comparison.

Assessment: Directionally this is the confirmation the thesis needed, and the labor lap should extend it through Q4 and into FY27. The magnitude is the issue. A run-rate of "more than 100 bps" of annual North America margin expansion is a respectable trajectory toward the mid-teens framework, but it is a multi-year path, not the one-year step the 280 bps optic implies.

Topic 3: Tariff Refunds Are a Timing Benefit and Management Said So

The most analytically honest moment of the call came unprompted in the CFO's prepared remarks, where she pre-empted the Street from annualizing the Q3 COGS rate.

"More precisely, in the quarter, our consolidated product and distribution costs were 30.3% as a percentage of net revenues. We believe the better proxy for a more normalized Q3 COGS rate is the year-to-date metric of 32.3%." — Cathy Smith, Executive Vice President and CFO

The refunds received in Q3 largely offset tariffs the company had already incurred across the first three quarters of FY26, so the year-to-date view nets them out. On $9,322.7M of revenue, the 200 bps gap between the reported 30.3% and the normalized 32.3% is roughly $186M of operating income, or about $0.13 per share after tax. Asked directly whether more refunds sit in the Q4 guide, management said it believes it has "largely gotten the refunds we're entitled to at this point."

Assessment: Credit where due: management flagged the distortion twice without being asked and gave the Street the normalized denominator. That is the behavior of a team confident in the underlying trend. The corollary is that anyone modeling forward off a 14.4% consolidated margin or an $0.85 quarterly EPS is modeling a number the company has explicitly disowned.

Topic 4: The Guide Raise Is Large, and Q4 Implies a Sequential Step Down

FY26 non-GAAP EPS guidance moved to $2.55 to $2.65 from $2.25 to $2.45, raising both ends and lifting the midpoint roughly $0.25. Consolidated operating margin guidance went to greater than 11.0% from "slightly higher year-over-year." US comp guidance moved to slightly greater than +6.0% with Q4 guided at 6.5% or better.

Running the arithmetic: nine-month non-GAAP EPS is $1.91 ($0.56 in Q1, $0.50 in Q2, $0.85 in Q3), so the full-year range implies Q4 non-GAAP EPS of $0.64 to $0.74. Against Q3's reported $0.85 that reads as a 13% to 25% sequential decline, which looks alarming until it is set against the normalized Q3 of roughly $0.70. On that basis the implied Q4 straddles the normalized Q3 almost exactly, and against Q4 FY25's actual $0.52 it represents 23% to 42% year-over-year growth. The guide is not a deceleration call; it is what a clean quarter looks like.

Assessment: The raise is real and it exceeds our prior $2.42 estimate by a wide margin. But it is worth naming precisely what it implies, because the difference between "Q4 EPS falls 25% sequentially" and "Q4 EPS is flat on a normalized basis" is the whole argument, and only the second reading is correct.

Topic 5: $2.54B of Proceeds, and Not a Dollar of It Went to Shareholders

Last quarter's single largest open question was the size of the Boyu cash inflow. The answer sits in the nine-month cash flow statement: net proceeds from the divestiture of certain operations were $2,544.2M. Long-term debt fell to $11,780.2M from $14,575.9M at the FY25 close, the CFO cited roughly $1.8B of debt repaid in the quarter alone, and leverage now stands at 2.9x. Nine-month operating cash flow rose 7.1% to $3,604.1M while capex was cut 52.0% to $887.8M, producing free cash flow of roughly $2.72B against $1.52B a year ago.

What did not happen is equally notable. There was no share repurchase in the nine-month period, the quarterly dividend rose one cent to $0.62 from $0.61, and diluted shares drifted up to 1,143.8M. A company that generated $2.5B of asset-sale proceeds and nearly doubled its free cash flow returned no incremental capital to shareholders.

Assessment: Defensible, and probably correct. Starbucks carries a $(7.7)B shareholders' deficit and had $14.6B of long-term debt entering the year; protecting the investment-grade profile before buying stock near a 52-week high is the disciplined choice. But the FY26 setup no longer contains a capital-return catalyst, and the "Boyu proceeds unlock a buyback" thesis that circulated after Q2 should be retired. A 1.6% dividend increase against a historical dividend CAGR management still cites at 17% is its own quiet signal about near-term capital priorities.

Topic 6: The Uplift Program Is Working and Is Being Accelerated

Cumulative North America coffeehouse uplifts passed 1,000 in the quarter, hitting the FY26 goal ahead of plan. The pace is the story: the base stood at roughly 300 at the end of Q2, meaning more than 650 were completed in Q3 alone. Management raised the FY26 target to at least 1,500 and signaled a further step up in FY27.

"My last plug is the team's doing a great job of not closing stores. They do the coffee house uplifts overnight, so we don't change the routine for our customers. The fact that we don't take a coffee house offline, I think is incredibly important. They're a very low level of investment, as we've shared, $150,000 on average. It's a great return." — Cathy Smith, Executive Vice President and CFO

Management reports transaction lift across all formats, channels, dayparts, and urbanities, and pairs it with brand health metrics at five-year highs on affinity, consideration, and purchase intent. At roughly $150,000 per uplift executed overnight without closing the store, the capital intensity is a fraction of the full remodels that preceded this program, which is precisely why nine-month capex could fall by half while the customer-facing investment accelerated.

Assessment: The most capital-efficient initiative in the plan and the clearest evidence that the current management team learned from the prior remodel cycle. This is a durable FY27 tailwind and one of the stronger arguments for the bull case. The caveat is that early-cohort uplift lift is measured against the worst stores in the fleet; the incremental return on uplift number 3,000 will not match uplift number 300.

Topic 7: North America Unit Growth Goes Quiet Through FY2027

Buried in the prepared remarks is a genuinely new and underappreciated disclosure about the growth algorithm.

"While net new company-operated unit growth in North America may remain modest through fiscal 2027, we expect international to be a meaningful contributor to unit growth." — Brian Niccol, Chairman and CEO

The current numbers already reflect it. North America added just 27 net new company-operated stores in the quarter against 41 net licensed closures, for a segment total of negative 14. Total North America store count is down 2% year-over-year. Every unit of the company's 0.5% total store growth came from international, where 189 net new coffeehouses opened. The CFO added that the company is "gaining deeper visibility into some underperforming coffee houses, which could result in some closures."

Management framed the cause as inherited rather than current, pointing to a development strategy that put the wrong stores in the wrong places two to three years ago and now requires cleanup before the pipeline can be rebuilt with discipline.

Assessment: Strategically correct and the right long-run trade. Financially it means that through FY27 the North America earnings algorithm is comp plus margin with no unit contribution, while the comp itself is partly manufactured by the very closures being described. That is a narrower path than the January Investor Day framework assumed, and it is the second-largest driver of our FY27 revision.

Topic 8: Loyalty Finally Gets a Number

The redesigned three-tier program launched in March, and last quarter we flagged the absence of any hard metric as an information gap. It has been filled: 35.8 million 90-day active members in the US, growing both sequentially and year-over-year, with members visibly migrating from Green to Gold and Gold to Reserve four months after launch. Average stored-value-card reload amounts continue to grow. Free Mod Mondays produced a repeat behavior management quantified, with one in three members who tried a new modification reordering it in subsequent weeks.

The S'mores lineup, offered to Rewards members first, is tracking as the strongest summer coffee limited-time offer in several years and is resonating particularly with Gen Z. Management noted sales growth was broad-based across both members and non-members, which is the harder test.

Assessment: The cannibalization risk we flagged after Q2 has not materialized; premium attach appears to be winning. Still missing are the Reserve-tier mix and Rewards share of tender, which are the metrics that would show whether the tier structure is driving incremental spend or simply reclassifying existing behavior. The +3.6% ticket with less than a point of pricing is supportive circumstantial evidence.

Topic 9: FY2027 Is a Spending Year

Management previewed three FY27 investment vectors without sizing any of them: an accelerated uplift pace, a technology modernization cycle covering new inventory ordering, staffing and scheduling, and point-of-sale systems, and an international support-organization rebuild. The Best of Starbucks Reward partner incentive launched at the close of the quarter, offering eligible Green Apron partners up to $300 per quarter against coffeehouse performance goals, and carries no disclosed cost.

Offsetting these is the $2B gross cost-savings program running through FY2028, which management confirmed is on track and balanced across product and distribution costs, operating expenses, and G&A. FY26 consolidated G&A dollars are expected to run below FY2023 levels.

Assessment: The savings program is credible and already visible in the G&A line. But a technology modernization year, an accelerating remodel program, and a new partner incentive layered together create real FY27 cost pressure that no one has quantified, at the same time as the tariff refund does not repeat and North America adds no units. This is the specific combination that makes us uncomfortable underwriting a 40%-plus FY27 EPS growth rate.

Topic 10: The Category Is Getting Crowded and Management Likes Its Hand

Refreshers delivered double-digit year-over-year revenue growth in the US and were the platform most cited on the call, at a moment when large quick-service competitors are advertising aggressively into the same cold-beverage space. Management's position is that category entry by non-specialists expands the occasion rather than splitting it.

"I think that was my going in belief, which is, if you're the category leader in the space, when other people start advertising in the space, you continue to execute really well, our plan is to get not only our fair share, but more than our fair share." — Brian Niccol, Chairman and CEO

The innovation pipeline behind it is dense: blended Refreshers, a sparkling "Spritzers" test in select markets, the returning Pumpkin Spice Latte and Peppermint Mocha seasonal anchors, orange cream beverages, a matcha platform positioned for the afternoon, and food tests including wraps. Digital menu boards reaching 80% to 90% penetration by September enable dayparted merchandising, which management ties directly to building the afternoon.

Assessment: The competitive read is reasonable and the innovation cadence is the best it has been in years. The afternoon daypart remains the largest unmonetized opportunity in the North America business, and it is also the one where a resurgent competitive set has the most room to take share. This is a watch item for FY27, not a Q3 concern.

Guidance & Outlook

Implied Q4 ramp: Nine-month non-GAAP EPS of $1.91 against the $2.55 to $2.65 range leaves $0.64 to $0.74 for Q4, versus $0.52 in Q4 FY25 (growth of 23% to 42%) and versus a normalized Q3 of approximately $0.70 (roughly flat to modestly up). The supports are the August anniversary of Green Apron Service, which converts the labor-investment headwind to a neutral comparison; coffee pressure that management expects to become "largely immaterial" to year-over-year margin comparisons; continued cost-savings realization; and a full quarter of lower interest expense on the reduced debt load. The offsets are the tax rate reverting to the mid-20s and the absence of further tariff refunds.

Street at: Pre-print FY26 consensus sat near the prior guide midpoint of $2.35; the new midpoint of $2.60 is roughly 11% above it and above our own prior $2.42 estimate. Post-print price targets moved up across the board with unusually wide dispersion, spanning roughly $92 to $120 against a $105.85 close. That spread is itself informative: after four consecutive positive-comp quarters the Street still cannot agree on what this business earns in FY27.

Guidance style: Conservative-realistic, unchanged from the CFO's posture since she took the seat. She has now raised the FY26 range twice and both raises proved warranted within a quarter. The Q4 US comp guide of "6.5% or better" was paired with an explicit caution about lapping tougher year-over-year traffic comparisons and "continued variability in the broader consumer landscape," which is the same hedging pattern that preceded the last two beats.

Analyst Q&A Highlights

How Much of the Comp Is Actually Demand

A housekeeping question about the mechanical contributors to comp produced the single most consequential disclosure of the quarter, and one that appeared nowhere in the press release or prepared remarks. The question separated coffeehouse initiatives (extended hours, store closures, delivery) from brand-driven performance and asked management to size each. The answer conceded that a little under half of the headline number came from the mechanical side.

Q: "First, a housekeeping question on the impact of coffee house initiatives like extended hours, closed stores, and delivery, and how these factored into the Q3 comp relative to your brand initiatives. Second question, unrelated on tariff refunds and whether you're anticipating more in Q4 and to what extent the guide contemplates anything there."
— Zach Fadem, Wells Fargo

A: "With a 7.9% comp that we drove, about half of it or a little bit less than half of it was due to closures, sales transfer, delivery growth. The rest of it is all that great store performance and menu and innovation. The hours of operation piece was actually very small, just a couple of basis points."
— Cathy Smith, Executive Vice President and CFO

Assessment: Management answered fully and without deflection, which is to its credit, but the disclosure reframes the entire print. It puts underlying comp near 4% rather than 7.9% and means the sell-side framing of a demand blowout is not supported by the company's own decomposition. The same exchange also closed the door on further tariff refunds in Q4, removing the other pillar of the margin beat from the forward model. Two of the quarter's three headline achievements were materially qualified in a single answer, and the stock's subsequent fade suggests the buy side heard it.

Durability of the Comp Against a 3% Long-Term Algorithm

The opening question of the call went straight to the tension every holder is sitting with: the long-term plan calls for roughly 3% annual comp, the business is printing more than double that, and investors want to know which number is the real one. Management declined to reconcile the two directly, answering instead on the operational sources of momentum and the runway still available in both dayparts.

Q: "How long can you keep up the same store sales momentum that you're seeing currently? I know the long-term plan calls for 3% annually, but I suspect you're aiming for something higher. I guess could you just give us your thoughts on where you are in the journey of recovering the sales volumes and what are the building blocks to keep this type of momentum or something above the 3% level going for the next few years?"
— David Tarantino, Baird

A: "To answer your question on the momentum, the good news for us is, I think this is driven by just better operating practices. Thinking about the Green Apron Service model, meaning we now, I think, are staffed better. We think we have the right routines, the right coaching taking place, so that we're giving better customer experiences for all of our customers every single day. If you look at where the business was, there's still lots of space to add more transactions both in the morning and in the afternoon."
— Brian Niccol, Chairman and CEO

Assessment: A qualitative answer to a quantitative question. The response is credible on the operational drivers and consistent with the GROW, service-time, and availability metrics disclosed in the prepared remarks, but it did not engage with the 3% algorithm at all. Read alongside the comp-composition disclosure, the honest synthesis is that underlying comp is running around 4% against a 3% long-term plan, with closures and delivery carrying the rest. That is a beat against the algorithm, not a doubling of it.

Why Slow Development Now, With the Business Working

A recurring line of questioning pressed on the apparent contradiction between the strongest operating momentum in years and a decision to slow North America development while stepping up closures. Management located the cause squarely in decisions made before the current team arrived, and framed the cleanup as a precondition for rebuilding the pipeline rather than a retreat from the opportunity.

Q: "You guys talked about the opportunity in 2027 to slow development, step-up closures of underperforming stores. I'm curious kind of why now is the time when the tide is clearly being lifted for Starbucks."
— Andrew Charles, TD Cowen

A: "Unfortunately, we did not have a great development strategy, when you kind of go back two or three years ago. It is just one of those things where we either did really difficult remodels on stores or we potentially put the wrong store in the wrong place. We're having to clean that up and fix that up. The good news is, as the business responds, it becomes more clear where you have the true problem stores."
— Brian Niccol, Chairman and CEO

Assessment: A direct and unusually blunt answer that reflects well on management's willingness to name inherited problems. The strategic logic is sound. The modeling consequence is that North America contributes no unit growth through FY2027 while the closures themselves feed the sales-transfer component of comp, which means the reported comp gets structurally better as the store base gets structurally smaller. Both effects need to be carried in a forward model, and most are carrying only the first.

Uplift Economics and What Actually Gets Closed

Questioning on the remodel program sought both the measured return per store and the profile of closure candidates. The response paired a strong qualitative read on transaction lift with the specific unit economics, and management was notably eager to volunteer the cost figure and the no-downtime execution model.

Q: "I'm curious how much you think the uplifts are helping you right now in the U.S., and how you're measuring the success of that on a per store basis. Relatedly, when you talk about maybe some more closures, what are some of the characteristics of those stores?"
— Brian Harbour, Morgan Stanley

A: "It's really about performance and location and sometimes where the actual asset is from a standpoint of remodeling versus we'd be better off just building a new store. It's not a question of whether or not we think Starbucks can work in the trade area. It's more to do with, is this the right representation of Starbucks and are these the economics that frankly Starbucks should earn?"
— Brian Niccol, Chairman and CEO

Assessment: The most encouraging exchange on the call. A $150,000 overnight uplift that lifts transactions across every format and daypart is a genuinely high-return use of capital, and the acceleration from roughly 300 cumulative to more than 1,000 in a single quarter demonstrates the operational capacity to scale it. What was not provided is a measured transaction lift percentage or a payback period, which leaves the return claim qualitative. The closure logic is sound and confirms these are asset and location decisions rather than trade-area exits.

Whether Green Apron Labor Hours Are Now Fixed or Still Flexible

A question aimed at the FY27 margin bridge asked whether the labor hours added under Green Apron Service represent a permanent cost layer or whether traffic growth can now be absorbed without proportional hour additions. Management's answer implied operating leverage in the morning peak and a willingness to add hours later as demand justifies them.

Q: "In terms of Green Apron Service, did we absolutely nail it in terms of labor hours, in terms of not only your current traffic, but potentially increased traffic going forward? In other words, could we expect the number of labor hours to stay relatively steady or perhaps have some efficiency opportunity as traffic has grown?"
— John Ivankoe, JPMorgan Chase & Co.

A: "Obviously in the morning, that's where you could really see more efficiency gains out of the labor that we've got there at peak. As you kind of move through the day, I think we're going to start to see the same opportunity and eventually we'll earn our way into having the additional hours necessary to match the growth of the business."
— Brian Niccol, Chairman and CEO

Assessment: The most important answer for the FY27 margin model and a constructive one. "Earn our way into having the additional hours" signals that incremental labor is gated on demonstrated demand rather than deployed ahead of it, which is the disciplined posture. Combined with the August anniversary of the Green Apron launch, this supports continued North America margin expansion into FY27, though at the "more than 100 basis points" underlying pace rather than anything resembling the tariff-flattered 280 bps.

Delivery Incrementality and Margin Structure

Given that delivery was named as a component of the comp, questioning turned to whether the channel is additive or merely displacing in-store visits, and what it does to unit margins as penetration rises. Management's answer was unqualified on both counts and previewed an in-app delivery integration that would extend loyalty economics into the channel.

Q: "I had a question on delivery. As that scales at a higher penetration rate, how should we think about incrementality and the trade-off versus in-store visits? Is there any color you could share around channel margin implications?"
— Steve McManus, BNP Paribas

A: "Right now there's no trade-off on margins, and we have yet to see any meaningful cannibalization from the delivery channel. We're really excited about it, and pretty soon we'll have it also in our app, with obviously a white label partner, which I think will also make the platform that much more attractive because then you'll be able to get rewards or Stars as part of the program."
— Brian Niccol, Chairman and CEO

Assessment: A clean answer that removes one bear concern. If delivery is genuinely incremental at neutral margin, the portion of comp it contributes is higher quality than the closure-transfer portion, and the pending in-app integration with Stars eligibility should accelerate it further. The claim of zero cannibalization is asserted rather than evidenced, and it is the kind of statement that tends to soften as penetration rises, but at current levels there is no reason to doubt it.

Company-Operated Versus Licensed Performance in North America

With North America licensed revenue roughly flat and 41 net licensed closures in the quarter, questioning probed whether a performance gap exists between the two ownership models and what levers management holds over the licensed base. The response reported convergence and described extending the company-operated management system into the licensed portfolio.

Q: "What's your assessment on the same-store sales and operational performance of your stores versus those of the licensed stores in North America, right? What are some of the actions within your control to close that gap?"
— Danilo Gargiulo, Bernstein

A: "In general, what's happening in our licensed business is we're putting the same rigor in place that we've got in our company store. We're putting in place the GROW report, obviously tweaked to recognize that it's a licensed model... Because our goal is there should be no real difference between a Starbucks experience at a company-owned store or a Starbucks experience at a licensed store."
— Brian Niccol, Chairman and CEO

Assessment: Relevant well beyond North America, because roughly 90% of the international portfolio now runs on a licensed structure and the China joint venture is the largest licensee of all. Exporting the GROW system and explicit experience standards into licensed operations is the mechanism by which brand quality gets defended in a capital-light model, and it is the right answer to the central risk of the licensing pivot. Execution here is a multi-year project with no disclosed metrics yet.

What They're NOT Saying

1. The underlying comp, anywhere except under direct questioning. The 4%-ish demand-driven comp appears in no press release, no prepared remark, and no slide. It exists solely because one analyst asked for the decomposition. A company confident in a 4% underlying comp on positive transactions could lead with it; leading with 7.9% instead is a choice.
2. Any FY2027 financial framework. FY27 was referenced repeatedly as an acceleration year for uplifts, a modernization year for technology, and a growth year for international units, with no revenue, margin, or EPS contour attached to any of it. The January Investor Day multi-year framework was not refreshed or reaffirmed despite three quarters of results running ahead of it.
3. China, in any measurable form. No China comparable sales, no JV revenue, no JV unit economics, no store count trajectory beyond an aspirational "up to 20,000 coffeehouses over time." The disclosed "$53 million of net revenues and operating margin above 100%" is an accounting ratio, not a performance metric. Investors have lost the ability to track a business that was recently a quarter of the store base.
4. The cost of the Best of Starbucks Reward. Up to $300 per quarter per eligible Green Apron partner across a North America base of roughly 11,100 company-operated stores is a potentially material cost layer, launched at the close of the quarter and left unquantified. This is the second consecutive quarter a new partner incentive has been announced without a dollar figure.
5. The size and scope of the remaining restructuring program. Restructuring and impairments hit $302.6M in the quarter and $415.8M year-to-date against $137.0M in the prior-year nine months, with $174.3M of the quarterly charge landing in North America. No total program size, closure count, or completion date has been provided, even as management signals more closures ahead.
6. Reserve tier mix and Rewards share of tender. The 35.8M active-member figure is a genuine improvement in disclosure and answers last quarter's gap, but the tier composition and share-of-tender metrics that would prove the redesign drives incremental spend rather than reclassifying existing behavior remain unpublished.
7. Q4 global comp. Only US comp was guided for Q4 at "6.5% or better." The global figure is inferable from the full-year "nearing 6%" language but was not stated, which is a curious omission in the first quarter where international is carrying all of the unit growth.
8. Why there is no buyback. Neither the release nor the call addressed capital return priorities beyond the dividend and deleveraging. With leverage at 2.9x, free cash flow near $2.7B over nine months, and $2.5B of proceeds banked, the absence of any framing on when repurchases resume is a notable silence.

Market Reaction

• Pre-print setup: SBUX closed at $104.14 on July 29 ahead of the after-close release, up 23.7% year-to-date against an S&P 500 up 6.9%, up 12.0% over trailing twelve months, and essentially flat over the trailing thirty days (+0.1%). The 52-week closing range entering the print was $78.46 to $108.37, placing the stock within 4% of its high.
• After-hours move: Shares traded as high as roughly $112.35 after the release and call, up approximately 7.9% from the close, on the EPS beat, the comp acceleration, and the guide raise.
• Next-day session: The stock gapped open at $107.22 (+3.0%), reached an intraday high of $108.46, traded as low as $103.63 (below the pre-print close), and settled at $105.85, up 1.6%. Volume of 23.4M shares was 3.4x the 30-day average of 7.0M.
• Relative performance: The S&P 500 closed up 1.7% the same session. On a roughly 29% EPS beat and a $0.25 raise to the FY26 midpoint, SBUX generated no relative alpha.

The fade is the most informative thing that happened. The stock gave back roughly three-quarters of its after-hours gain between the post-call print and the closing bell, on 3.4x normal volume, and finished the day marginally behind the index. That is not the signature of a market that disbelieves the quarter. It is the signature of a market that read the transcript.

The sequence is legible. The press release and the prepared remarks supported the after-hours move: comp +7.9%, EPS +70%, margin +430 bps, guidance raised at both ends. The Q&A then supplied three qualifications in short order. Roughly 200 bps of the consolidated margin gain was a tariff refund the company does not expect to repeat. The 21.8% tax rate carried a one-time catch-up and reverts to the mid-20s. And a little under half the comp came from closures, sales transfer, and delivery rather than demand. By Thursday's open the gap had compressed from +7.9% to +3.0%, and it kept compressing all session.

Positioning did the rest. The stock entered the print up nearly 24% year-to-date against an index up under 7%, within 4% of a 52-week high, and flat over the prior month, which is the classic profile of a name where the good news is already owned. A beat of this magnitude producing zero relative return says the expectations bar had already been raised to roughly where the quarter landed. Post-print price targets moved up broadly but with unusual dispersion, spanning roughly $92 to $120 around a $105.85 close, which is another way of saying the Street agrees the operating story is working and disagrees entirely about what it is worth.

Street Perspective

Debate: Is a 7.9% Comp the Run-Rate or a Construct?

Bull view: Four consecutive quarters of positive global comps, transactions positive at +4.2%, pricing contributing less than a point of ticket, brand health metrics at five-year highs, and target service times achieved across every access point. The operational machinery is demonstrably working, and management is guiding Q4 US comp to 6.5% or better with what it called a strong start to the quarter. Underlying momentum is durable well above the 3% long-term algorithm.

Bear view: The company's own CFO put a little under half the comp on closures, sales transfer, and delivery. Underlying demand comp is closer to 4%. Sequentially, ticket more than accounted for the whole US comp acceleration while US transactions decelerated from +4.4% to +4.2%. With more closures explicitly signaled and North America unit growth flat through FY2027, the reported comp will keep being flattered by a shrinking denominator, which is not a quality of earnings any investor should pay a premium multiple for.

Our take: The bear framing is factually stronger and the bull framing is directionally right. A 4% underlying comp on positive transactions is a genuinely healthy result and comfortably beats the long-term algorithm. But the gap between 4% and 7.9% is the gap between a good business recovering and a business in a boom, and the multiple currently embeds the second reading. We underwrite 4% to 4.5% underlying and treat the closure-transfer contribution as a real but low-quality supplement that will persist through FY27 and then fade.

Debate: How Much of the 14.4% Margin Is the Business?

Bull view: Second consecutive quarter of consolidated margin expansion, North America margin positive year-over-year for the first time since Q1 FY2024, and management stated both expanded even excluding tariff refunds. Store operating expenses improved as a percentage of company-operated revenue while Green Apron labor investment was still fully in the base. The $2B cost-savings program has two more years to run and FY26 G&A will land below FY2023 levels. The operating leverage is structural.

Bear view: Roughly 200 bps of the consolidated gain and roughly 180 of North America's 280 bps came from a one-time refund of tariffs previously paid. Management said it has largely collected what it is owed. Normalize for that and the consolidated margin is about 12.4%, not 14.4%. Layer in an FY27 technology modernization cycle, an accelerating uplift program, and an unquantified partner incentive, and the FY27 margin path is far shallower than the Q3 optic suggests.

Our take: The bear arithmetic is correct and the bull conclusion still mostly survives it. A normalized 12.4% consolidated margin against 10.1% a year ago is roughly 230 bps of genuine expansion, which is a strong result and validates the sequencing management promised. The error is extrapolation, not direction. We model FY27 non-GAAP operating margin in the 13% to 14% range rather than the 15%-plus that annualizing this quarter would imply, and that single assumption accounts for most of the gap between our FY27 estimate and the more bullish sell-side models.

Debate: Does the Capital-Light Pivot Create or Transfer Value?

Bull view: Roughly 90% of the international portfolio now runs through licensing, capex fell 52% year-over-year while customer-facing investment accelerated, nine-month free cash flow nearly doubled to $2.72B, and leverage came down to 2.9x with $2.5B of proceeds applied to debt. Starbucks is converting a capital-intensive retail footprint into a royalty stream on a global brand, which deserves a higher multiple, not a lower one.

Bear view: The company gave up 60% of China's economics and the visible replacement is $53M of quarterly revenue and $22.8M of equity income, annualizing to well under $100M against roughly $3.0B of non-GAAP operating income. China is now unmeasurable from the outside. North America adds no units through FY2027. Growth reduces to comp plus margin, and the comp is partly manufactured by closures. This is a smaller, less transparent company with a better balance sheet.

Our take: The bull case is right on the long arc and the bear case is right on the next eight quarters. Capital-light licensing is the correct structure for a global brand, and the balance sheet transformation is real and underappreciated. But the earnings contribution from the restructured China business is currently immaterial, the disclosure is genuinely worse, and the near-term algorithm is narrower than it was a year ago. That combination argues for owning the business and not overpaying for the transition, which is precisely where our rating lands.

Debate: Does $2.5B of Proceeds and No Buyback Matter?

Bull view: With a $(7.7)B shareholders' deficit and $14.6B of long-term debt entering the year, deleveraging to 2.9x is the highest-return use of the proceeds and protects the investment-grade rating that underpins the cost of capital. Buying stock within 4% of a 52-week high would have been poor discipline. The capacity is preserved and can be deployed later at better prices.

Bear view: A dividend increase of one cent against a business that nearly doubled free cash flow, alongside a diluted share count that drifted higher, means shareholders saw no incremental return from a $2.5B asset sale. The Boyu-funds-a-buyback thesis that supported the stock after Q2 is dead, and management offered no framework for when repurchases resume.

Our take: The bull case wins on the merits and the bear case wins on the setup. Debt paydown was the right call and we would have made the same one. But it removes a catalyst that was in the stock, and the silence on when capital return resumes is a real gap given the balance sheet is now in acceptable shape. We would treat a buyback authorization at the Q4 print as genuine upside to our rating, not as an expected event.

Model Update Needed

Valuation: At the July 30 close of $105.85 and our revised estimates, SBUX trades at 40.7x FY26 ($2.60), 32.1x FY27 ($3.30), and 26.1x FY28 ($4.05 mid). Applying a 30x to 33x multiple to FY27, which is a fair range for a quality consumer compounder with proven operational momentum, a de-risked balance sheet, and a narrower near-term growth algorithm, supports $99 to $109 of fair value on FY27 alone. Adding roughly a 2.3% dividend yield and some credit for the FY28 path takes the 12-month price target to $105 to $115, down from $110 to $120. From $105.85 that is a total return of roughly 1% to 11%, centered near market-like. The FY27 estimate cut, not the FY26 raise, is what moves the target: the prior $3.55 to $3.70 range required an FY27 operating margin above even this quarter's tariff-flattered peak, and the first clean look at the post-China revenue base does not support it.

Thesis Scorecard Post-Earnings

Overall: Thesis operationally strengthened, valuation thesis broken. Five of six bull pillars remain confirmed, four of six original bear points stay refuted, and management delivered the North America margin inflection that was the central open question of the last four quarters. Not one of the four downgrade triggers we set after Q2 fired. What changed is not the business; it is the price against a forward estimate we are now cutting. The FY27 EPS revision from $3.55 to $3.70 down to $3.15 to $3.40 is the entire rating action, and it rests on two disclosures from this specific call: that underlying North America margin expansion is "more than 100 basis points" rather than 280, and that North America will add no meaningful units through FY2027.

Action: Hold. Existing positions should be retained rather than sold. The operating trajectory is intact, the balance sheet is materially de-risked, the uplift program is a genuine FY27 tailwind, and the cost-savings plan has two years left to run. But we would not add at $105.85 into 32x FY27 on an estimate we just reduced, in a name that just delivered its best print of the cycle and finished the session behind the index. We would revisit Outperform on a pullback toward the mid-$90s, on a buyback authorization at the Q4 print, or on evidence in Q4 that closure-adjusted underlying comp is holding above 4%. We would move to Underperform if underlying comp decelerates below 3% while reported comp stays elevated on closures, or if the FY27 spending program lands materially above the cost-savings run-rate.