Fuel Took the Year's Growth, Carnival Raised Anyway: A Record Q3, Europe Confirmed Transitory, and a 2027 Book at Records

Results vs. Consensus

Carnival reported fiscal third-quarter results before the open on September 29 for the quarter ended August 31, 2026, and held its call at 10:00 a.m. ET the same morning. This is a seasonal peak quarter: Q3 carries the Alaska season, the European summer and the North American school holidays, and it produced $1,963M of the ~$3,080M of adjusted net income now guided for the full year.

Q3 FY2026 Scorecard

The notable feature of this scorecard is that consensus and management's own June guide sat on exactly the same number. The $0.08 of adjusted EPS upside is therefore both the beat versus the Street and the beat versus guidance, and the three sources of it are disclosed rather than inferred.

Against the June Guide, Line by Line

Management's decomposition of the $0.08: $0.05 from revenue, where yields landed 1.2 points above guidance on strong close-in demand and onboard spending; $0.02 from cost, where unit costs excluding fuel came in a point better; and $0.01 from fuel consumption, which fell nearly 4% year over year on top of a greater-than-5% reduction in the prior-year quarter. Notably, fuel price was a small negative inside the quarter ($826 per metric ton against $812 guided), so none of the beat came from the commodity moving the company's way.

Year-over-Year Detail

Sequential Comparison (Q3 vs. Q2 FY2026)

Carnival's quarters are not comparable in dollars. Q3 is the summer peak and Q2 is a shoulder quarter, so the step-up below measures seasonality, not momentum. The two rows worth reading sequentially are the per-ALBD cost line, which is capacity-normalized but still seasonal, and the occupancy line, which shows how much of the peak was actually filled.

Revenue

Total revenues of $8,435M (+3.5%) were produced on 1.2% more capacity, which means roughly two-thirds of the growth is rate rather than volume. The split inside the top line is more informative than the total: passenger ticket revenue grew 1.8% while onboard and other revenue grew 6.7%, lifting onboard to 34.5% of revenue from 33.4%. Occupancy at 111.8% was essentially unchanged from 111.7%, so the ship was as full as it was a year ago and the incremental dollar came from what the guest paid, not from finding another guest. Constant-currency net yields of $255.09 per ALBD (+2.4%) set a record on top of a prior-year quarter that itself grew almost 5%, which is the harder comparison and the reason the 1.2-point beat against guidance matters more than its size suggests.

Margins

The margin lines look worse than the business performed, and the reconciliation is clean. Operating margin fell 154bps and adjusted EBITDA margin fell 123bps, but fuel alone added $164M of cost on a $282M increase in revenue. Gross margin yields, which are the GAAP-comparable measure and therefore carry fuel, fell 1.3%; net yields, which exclude it, rose 2.4%. That 3.7-point gap is mostly the fuel bill. Below the fuel line, the cost program delivered another quarter of outperformance: adjusted cruise costs excluding fuel per ALBD rose 1.8% in constant currency against a ~2.8% guide. That year-over-year comparison is the capacity-normalized evidence; the sequential decline from Q2 is seasonal and matches last year's. Selling and administrative expense is the one line growing faster than revenue (+7.1%), which bears watching but is small relative to the unit-cost win.

EPS

Adjusted EPS of $1.43 is flat year over year while adjusted net income fell 1.0%, and the gap between those two facts is share count. Diluted shares fell 2.4% to 1,368M, which closed a $19M earnings decline into a flat per-share result. The honest footnote is that most of that reduction is not the buyback. The prior-year diluted count of 1,402M included convertible-note dilution (the prior-year figures carry an $18M interest add-back that the current year does not), and those notes have since settled into equity. Basic shares rose 3.8% to 1,363M over the same period. Shares actually outstanding went from 1,312M at the November 2025 year-end to 1,352M at August 31, up about 40M despite roughly 45M repurchased, because convert settlement issued more than the buyback retired. GAAP diluted EPS rose 5.3% to $1.40 from $1.33, and GAAP net income of $1,920M is an all-time high, both helped by the $88M year-over-year drop in debt-extinguishment costs.

Revenue Composition, Yields and Deployment

Carnival does not report brand-level segment financials, so the operating structure of the quarter lives in three places: the two revenue streams, the per-ALBD yield and cost metrics that normalize for capacity, and the deployment mix that management discussed qualitatively on the call. All three are below.

Revenue Streams

Contribution is calculated as each stream less its own directly attributable cost line (commissions and transportation for ticket, onboard and other costs for onboard), which is how the company itself builds the net-yield bridge. The result is the single most useful operating fact in the release: ticket contribution margin was flat at 82.1%, while onboard contribution margin expanded 139bps to 69.0%. The onboard engine is not only growing four times as fast as the ticket, it is growing more profitably.

Per-ALBD Metrics (the capacity-normalized view)

Read down that column and the quarter explains itself. Every metric that includes fuel is negative. Every metric that excludes it is positive, with a yield-to-cost spread of roughly 60bps in the company's favor. Carnival's operating model worked in Q3; the commodity did not.

Key Operating Indicators

Passenger Ticket

Ticket revenue grew only 1.8%, the slowest of the two streams, and that is the one line a bear can point at. The explanation is mix and comparison rather than price weakness: the prior-year third quarter was itself a near-5% yield grower, occupancy was already above 111%, and the European itineraries that would normally carry the richest ticket pricing were working through the tail of the spring disruption. Contribution margin after commissions and transportation held exactly flat at 82.1%, so nothing was bought with discounting.

Assessment: Ticket growth of under 2% is the weakest operating statistic in the release, and it is the number to watch in the fourth quarter and into the first half of 2027. It is a comparison and deployment artifact this quarter rather than evidence of pricing erosion, but it leaves the yield story more dependent on onboard monetization than it was a year ago.

Onboard and Other

Onboard revenue grew 6.7% with contribution margin up 139bps, and management was unusually specific about why. Growth was broad rather than concentrated, and more than half of onboard revenue is now pre-booked, which both pulls the spending decision forward and raises the attach rate.

"So the third quarter onboard and revenue was really broad-based across all categories, and it was also across all brands on both sides of the Atlantic."
— David Bernstein, CFO

"We're seeing more than 50% of our revenues prebooked. So we're getting the benefit of the second wallet meaningfully onboard."
— David Bernstein, CFO

Asked for a real-time read on whether consumer behavior had changed in the three months since the June call, the CEO's answer was unqualified.

"Yes. No, if anything, onboard has just accelerated. So we see continued strength from the consumer."
— Josh Weinstein, CEO

Assessment: This is the highest-quality growth in the business and the piece of the thesis that has been confirmed most consistently. Pre-booking shifts onboard spending from a discretionary in-voyage decision to a committed one made at the same moment as the ticket purchase, which makes it both more predictable and less exposed to a soft consumer mood at sailing. The margin expansion says the mix is moving toward the higher-contribution categories rather than toward discounted bundles.

The Caribbean and the Paradise Collection

The Caribbean faces an industry capacity wave, and Carnival's answer remains exclusive destinations rather than price. Celebration Key completed its first year having welcomed almost 2.5 million guests, and the second pier is now open. RelaxAway, Half Moon Cay and Isla Tropicale each welcomed more than 250,000 guests following their enhancements.

"With the second pier now open, Celebration Key is expected to welcome approximately 3.5 million guests next year with 31 ships calling versus 26 this year."
— Josh Weinstein, CEO

The reach is also widening beyond the flagship brand: Princess joins next month, with select AIDA and Costa calls late next year, and roughly 35% of Carnival Cruise Line's 2027 Caribbean capacity will sail itineraries calling at both Celebration Key and RelaxAway. On the competitive backdrop, management was blunt about preferring less industry supply rather than claiming immunity to it.

"Option #1 is no growth in the Caribbean. Option #2 is, if you look at 2027, something like 37% growth over a 3-year period. I'm going to go with option A."
— Josh Weinstein, CEO

Assessment: The destination strategy is the mechanism that converts owned real estate into pricing power, and it is now large enough to measure: guest throughput at the three Paradise Collection assets rises materially next year on a fleet that is barely growing, which is the definition of sweating an existing asset base. The candid acknowledgement that a 37% three-year industry capacity increase is unwelcome is more useful than a claim of indifference would have been, and it frames the Caribbean as the segment where yield has to be defended rather than harvested.

Europe

Europe caused the June guidance cut and is now the clearest evidence that the cut was transitory. Bookings inflected in June and built through the quarter, European bookings for 2027 rose sharply during the same window that 2026 Europe paused, and the 2027 deployment book has been shifted toward the region.

"As a result, in 2027, Europe will, for the first time, tie with the Caribbean as our largest deployment region, each representing 34% of our mix."
— Josh Weinstein, CEO

Management emphasised Northern Europe specifically, framing cooler-weather itineraries as a distinct demand pool rather than a substitute for the Caribbean, and noting the company has more European sailings outside the Mediterranean than inside it. The one unresolved piece is the first quarter of 2027, which still carries spillover from the spring disruption through long-haul and exotic itineraries.

Assessment: Europe has moved from the quarter's problem to the quarter's proof. Shifting a third of 2027 capacity into the region that just absorbed a geopolitical shock is a capital-allocation vote, not a talking point, and the longer shoulder seasons in Northern Europe extend the revenue-producing calendar on the same hulls. The residual Q1 2027 drag is real and management flagged it unprompted, which is the right way to handle it.

Alaska and the Land-and-Sea Platform

Alaska again went unmentioned as a problem, which given that Q3 contains the entire Alaska season is itself the signal. Management described an integrated platform of thousands of hotel and lodge rooms, 20 glass-domed railcars and the largest motorcoach fleet in Alaska, and characterised the combined access as difficult to replicate. Tour and other revenues of $191M (up from $179M) are the small reported slice of this, carrying $109M of tour and other expenses.

Assessment: This is the quietest durable advantage in the portfolio: a seasonal, high-yielding franchise where the barrier to entry is land-side logistics that no competitor can order from a shipyard. It also diversifies the deployment book away from both the Caribbean capacity wave and the European geopolitical exposure, which is worth more in a year like this one than its revenue share suggests.

Key Topics & Management Commentary

Overall Management Tone: Confident and comfortable, and the most relaxed posture of the past four calls. Management led with records, volunteered the one soft spot (first-quarter 2027) before anyone asked, and spent the bulk of its energy on 2027 and 2028 positioning rather than defending 2026. The tell was the handling of fuel: rather than minimising it, management stated the headwind plainly, pointed at the offsetting operational improvement, and then explicitly declined to hedge it, which is a position taken from strength rather than necessity. Where the call was least forthcoming was 2027 quantification, with at least four separate attempts to extract a yield or cost number deflected to the December call.

1. The Beat, Decomposed

The $0.08 of upside against the June guide was attributed cleanly and in order of size: revenue first, then cost, then fuel consumption. Yields landed 1.2 points above guidance on close-in demand and onboard spending, worth $0.05; unit costs excluding fuel came in a point better, worth $0.02; and consumption efficiency added the remaining $0.01.

"Once again, we closed out another excellent quarter with revenues, yields and reported net income all reaching new highs, while customer deposits once again set a record."
— Josh Weinstein, CEO

The CFO added the detail that matters for the full year: "Importantly, nearly all the third quarter cruise cost savings flow through to our full year September guidance." That is the difference between a timing beat and a structural one, and it is the second consecutive quarter in which management has drawn that distinction explicitly rather than leaving it to be inferred.

Assessment: A beat led by revenue rather than cost is a better beat, and this is the first quarter of the year where that was true. Last quarter's $0.07 beat was mostly cost; this one is majority yield. The sequencing matters because it shifts the burden of the thesis back onto demand, where it belongs, and away from the finite cost lever.

2. Fuel: The Headwind That Took the Whole Year

This is the central fact of Carnival's 2026 and the report's organising idea. The September guide carries $2.25B of fuel expense at $768 per metric ton. FY2025 actually spent $1,808M at $610. By our calculation that $442M increase is worth roughly $0.32 per share on the guided 1,376M diluted shares, and Carnival pays almost no tax, so the pre-tax and after-tax figures are effectively the same. Full-year adjusted net income is guided to ~$3,080M against FY2025's $3,079M, and adjusted EPS to ~$2.24 against $2.25. Add the fuel back and adjusted net income compounds about 14%.

Against the original December 2025 plan the arithmetic is starker. That guide assumed 2.8M metric tons at $524, for $1.63B of fuel expense including allowances. The fuel line is now $620M worse than the original assumption, roughly $0.45 per share, while the adjusted net income guide has fallen only $370M, from ~$3,450M to ~$3,080M. The operating business has therefore added about $250M versus the plan it was given nine months ago while absorbing a $620M commodity shock.

"Taken together, we've generated more than $150 million of operational improvement since our June guidance, fully offsetting the impact of higher fuel prices that we now expect."
— Josh Weinstein, CEO

"Yes, fuel can be a volatile input cost with a track record of prices going up and down. But amidst that noise, let's not lose sight of our underlying operational improvement."
— Josh Weinstein, CEO

Assessment: Two readings are both correct and the rating depends on which one the market is pricing. The bull reading is that 2026 is a trough year created by an input price, the operating business grew double digits underneath it, and the stock trades on the suppressed number. The bear reading is that an unhedged operator has now spent two consecutive years having its earnings growth taken by the same line item, and "ex-fuel" is not a currency shareholders can spend. Our judgement is that the bull reading is better supported on the operating business, because the offset was delivered rather than promised: management has now covered a fuel shock with operational improvement in two successive guides. It does not carry the rating, because at the fourth quarter's $896 fuel price the offset only holds 2027 earnings flat.

3. The Decision Not to Hedge, Restated Under Pressure

Six months into a conflict-driven fuel spike, and with the company the only major cruise operator that carries no fuel hedges at all, management was asked directly whether the philosophy had changed. The answer was a flat no, with reasoning rather than reflex, and it was paired with the consumption statistics that are the actual policy.

"It's a very fair question to ask because it does have the ability to reduce volatility in any given year. As we've looked at our business and we've looked at this financially as well, we do believe that, that's a short-term band-aid that sometimes pays off, sometimes it doesn't. But the reduction in the volatility to us isn't worth paying banks or counterparties to effectuate those types of trades."
— Josh Weinstein, CEO

"And the best way you can combat the input cost is to use less of it."
— Josh Weinstein, CEO

The scale of the consumption program is now material enough to quantify. Consumption per ALBD is down 26% since 2019 and 13% in just three years, and the CFO put a dollar figure on it.

"The lower consumption represents savings of nearly $750 million at September guidance fuel prices."
— David Bernstein, CFO

Assessment: The intellectual case against hedging is sound: hedges smooth a line without changing its long-run level, and they cost a spread. The practical consequence is that Carnival's reported earnings will carry full commodity beta in both directions, which is a permanent widening of the distribution of outcomes and a legitimate reason for a lower multiple than a hedged peer. The $750M of structural consumption savings is the better answer, and it is a real one, but it does not convert a $442M annual price swing into a non-event. This is the pillar of the thesis that got worse this quarter, and it got worse by policy rather than by accident.

4. Bookings Inflected, Then Accelerated

June was identified on the last call as the turn. This quarter management described that turn continuing and compounding, with volumes "meaningfully ahead of last year and far outpacing capacity growth," and the forward book now at records across two separate years.

"For full year 2027, we are already half booked with both occupancy and pricing at record levels."
— Josh Weinstein, CEO

"2028 is also off to an excellent start at higher occupancy and even higher prices year-over-year, and our booking curve is further out than it has ever been at this point in the year."
— Josh Weinstein, CEO

Customer deposits reached a third-quarter record of $7,639M, up 6.9% against the prior-year record and ahead of the prior-year high by about $0.5B, achieved on flat capacity over the next twelve months. The CEO also clarified in Q&A, after a question noting the release said "record levels" rather than "higher," that 2027 occupancy and pricing are in fact both higher year over year.

Assessment: This is the strongest single data set in the release and it directly resolves the question the June quarter left open. A deposit book growing 7% against zero capacity growth is pure price and attach rate, and a booking curve extending further out than it ever has is the behaviour of a customer base with conviction rather than one being pulled forward by discounts. Half of 2027 sold at record price and occupancy sets a floor under next year's yield that no amount of fuel volatility can touch.

5. The One Acknowledged Soft Spot: First-Quarter 2027

Management volunteered the residual damage before it was asked about, and repeated it in three separate answers. The spring disruption hit long-haul and exotic itineraries hardest, and those bookings sit disproportionately in the fiscal first quarter.

"And while the booking disruption we experienced this spring extended into the first quarter of 2027, Q1 bookings have also rebounded meaningfully over the past 3 months, reinforcing our view that the impact is temporary."
— Josh Weinstein, CEO

The CFO was more precise about the shape: 2027 strength is "weighted to quarters 2 through 4" because the disruption "also carried heavily into bookings for the first quarter of 2027." Attempts to extract a quantified Q1 yield impact, or to use the Q3 and Q4 exit rate as a Q1 proxy, were declined.

Assessment: Flagging a soft quarter two calls in advance, unprompted, is the behaviour of a management team that expects to be believed on the rest. The cost is that the first reportable quarter of 2027, which lands in late March, is pre-seeded as the weak one, so the stock has a known soft print ahead of it. The benefit is that nobody can be surprised by it. We would treat a Q1 2027 yield miss as already disclosed rather than as new information.

6. Carnival Rewards: Cash Positive, Optically Negative Until 2028

The new loyalty program launched September 1 with strong early engagement, and co-branded credit-card issuances have accelerated sharply.

"And since the program went live September 1, issuances have accelerated significantly, more than tripling from pre-announcement levels."
— Josh Weinstein, CEO

The accounting, however, defers revenue equal to the value of benefits earned, which creates a reported-yield headwind until redemptions outrun new deferrals. The CFO quantified the drag precisely across three years.

"We previously explained a 0.2 point impact for the full year 2026, all of which is from the 0.6 point impact in the fourth quarter of 2026. And for full year 2027, there will be an additional 0.4 point impact."
— David Bernstein, CFO

"Until 2028, when the impact turns positive, we do expect to have accounting-driven yield headwinds."
— David Bernstein, CFO

This is why the fourth-quarter yield guide reads two ways: approximately 1.7% in constant currency as reported, or approximately 2.3% normalised for the loyalty deferral, the latter in line with the 2.4% growth Q3 just delivered.

Assessment: A program that is cash-positive from launch but revenue-negative for two reporting years is exactly the kind of disclosure that gets mis-modelled, and management deserves credit for sizing it to the decimal before anyone had to ask. Practically, it means reported 2027 yields will understate the underlying rate by 0.4 points, and anyone comparing Carnival's headline yield growth to a peer's in 2027 will be comparing different things. It is a modelling adjustment rather than an economic cost.

7. The Cost Program Delivers a Second Time

Last quarter's cost outperformance was the thing we flagged as needing a repeat before it could be called structural. It repeated. Unit costs excluding fuel came in a point better than guidance in the quarter, and the full-year expectation has improved by more than a point cumulatively even after absorbing higher-than-expected inflation. On a normalised basis, stripping expense timing, the partial-year operating costs of two new exclusive destinations and elevated logistics costs from the Middle East disruption, full-year adjusted cruise costs excluding fuel per ALBD are up only approximately 1.1%.

The CEO tied the next leg of the program explicitly to technology, citing AI across commercial systems, shoreside automation, and vessel management. Asked about the multi-year cost trajectory embedded in the PROPEL plan and whether its promised deceleration had a turning point, the answer was that some of it has already arrived early.

"And while it's still early, we are beginning to capture opportunities embedded in our PROPEL targets sooner than expected, leveraging our unmatched scale, sharpening commercial execution, investing where we generate outsized returns, and advancing technology to enhance revenue and control costs."
— Josh Weinstein, CEO

Assessment: Two consecutive quarters of 100bps-plus cost beats, with the savings carried into the full-year guide rather than banked as timing, moves this from a claim to a demonstrated capability. The caveat we raised in June still stands: flattening unit costs is a level reset rather than a renewable tailwind, and the 2027 setup includes offsetting pressures discussed below. But the lever has now been pulled successfully through both a fuel shock and a demand shock, which is what "flexibility" is supposed to mean.

8. The 2027 Cost Setup Is More Balanced Than It Looks

Two forces work against each other next year, and management quantified both. Roughly a point of 2026's cost base is non-recurring, which would be a tailwind; but the fleet modernization program steps up materially, which is a headwind.

"Probably a little over, yes, a little over 1 point. But keep in mind that this year's Celebration Key operated the full year. Half Moon Cay had a half a year of operation."
— David Bernstein, CFO

"Next year, we will have 5 Evolution ships in dry dock compared to 2 this year."
— David Bernstein, CFO

Capital spending moves the same way. Non-newbuild capex is $2.4B in 2026, and the CFO signalled a step-up: "we do expect to see somewhat higher than the $2.4 billion a year for the next few years." Formal 2027 guidance, for costs and capex alike, was deferred to the December call, with the CFO offering only that "there's nothing in particular too meaningful to drive the cost one way or the other."

Assessment: Anyone modelling 2027 off a clean one-point roll-off of 2026 one-timers will be disappointed. Five dry docks against two consumes available ship days and carries direct cost, and a capex step-up above $2.4B on a fleet with few newbuilds means the free-cash-flow improvement next year is more modest than the earnings improvement. This is the most underappreciated disclosure on the call and the reason we are not reaching for an aggressive 2027 estimate.

9. Investment Grade at a Second Agency, and the End of Secured Debt

S&P upgraded during the quarter, making Carnival investment grade at two agencies, and the upgrade released the remaining collateral.

"During the quarter, S&P upgraded the company's credit rating, making it the second rating agency to award the company an investment grade rating. Following this upgrade, the company does not have any remaining secured debt."
— Q3 FY2026 earnings release

Total debt fell to $23,912M from $26,640M at the November 2025 year-end, and the company used cash on hand to redeem $500M of 7% notes, among its highest-coupon instruments. Interest expense fell 10.1% year over year to $285M, and debt-extinguishment costs fell from $111M to $23M. The CFO framed the position against history: total debt is "a far cry from our $36 billion peak in 2023."

Assessment: The four-year balance-sheet repair is finished. An unsecured, investment-grade capital structure at two agencies with debt below $24B is the end state the 2022 thesis was underwriting, and the cost of capital benefit is now permanent rather than prospective. The corollary, which the market has not fully absorbed, is that this pillar can no longer be a source of upside surprise. From here the equity has to be driven by EBITDA growth, not by refinancing arithmetic.

10. Capital Returns Accelerate, but the Deleveraging Clock Has Paused

Buyback execution ran well ahead of the cadence management guided to in June, when it cautioned against annualising a $450M quarterly pace.

"To date, we have opportunistically repurchased nearly $1.2 billion of stock, representing 45 million shares."
— David Bernstein, CFO

By the company's own figures that is nearly $800M since the start of the third quarter alone, against $450M through the first half, and it was spent into a falling share price. Dividends added $204M in the quarter and $618M year to date, bringing total expected fiscal-year shareholder returns to nearly $2B. The CEO was explicit that all three uses of cash are running simultaneously.

"And yes, we are doing all 3. We continue to invest in the highest return opportunities across our brands and destinations. We continue to reduce debt and strengthen our financial position."
— Josh Weinstein, CEO

The arithmetic underneath that is less comfortable. By our calculation net debt was $22,692M at August 31 ($23,912M of debt less $1,220M of cash) against $22,646M at May 31, so net debt rose $46M sequentially. Applying management's own definition (total debt less cash, divided by twelve-month adjusted EBITDA) gives roughly 3.10x, essentially unchanged from the 3.1x the company disclosed in June. The method reproduces June's disclosed figure exactly, which is why we have confidence in it. Cash fell $708M year to date. Management disclosed the leverage ratio in the June release and did not repeat it this quarter, offering instead that it continues to expect "year-over-year improvement in our balance sheet and leverage metrics," which is true against the 3.6x of a year ago and silent on the sequential flatline.

Assessment: The buyback has been well executed, and buying roughly 45M shares at an average we calculate near $26 while the stock traded down to $21.79 is a reasonable use of capital for an operator that believes its own 2027 book. But investors should be clear about what changed: Carnival is no longer deleveraging on a ratio basis. Free cash flow now funds capital returns, and with fuel holding trailing EBITDA flat, the leverage ratio has stopped falling. That is a defensible choice at 3.1x with two investment-grade ratings in hand. It is also why the credit narrative is finished as a driver, and it deserved a sentence from management rather than an omission. Separately, the June signal that "a moderate increase" to the dividend was "rational and reasonable" has not yet been acted on; the payout remains at its prior rate.

11. PROPEL: Pulled Forward on Operations, Still Hostage to the Commodity

The multi-year framework was reaffirmed, with management claiming early capture of embedded opportunities and pointing to broadening returns on invested capital.

"Despite the significant fuel price headwind this year, we expect to finish 2026 with even more brands generating mid-teens or higher returns on invested capital than last year."
— Josh Weinstein, CEO

The arithmetic of the plan has become harder, though, and in a way worth stating plainly. PROPEL targets more than 50% adjusted EPS growth from 2025 by 2029. FY2025 adjusted EPS was $2.25, so the target is more than $3.38. FY2026 is now guided to ~$2.24, a penny below the starting point, which means the entire required increase has to be delivered in the three years from 2027 to 2029. By our calculation that is a compound annual growth rate of about 14.7% off the 2026 base, against roughly 10.7% if 2026 had grown as originally planned. Fuel has not reduced the target; it has compressed the runway. Management did not dispute the framing when it was put to them in Q&A, and was candid about the dependency.

"So I'm not pretending it's not going to be hard work, but it's hard work that I think we're all excited to take on, and we've got pretty good road maps to get there. Now if fuel would cooperate, that would make it a hell of a lot easier. But clearly, we can't count on that."
— Josh Weinstein, CEO

Assessment: The honest reading is that PROPEL is now a stretch rather than a glide path, and that the stretch is a commodity problem rather than an execution problem. Three years of 14.7% adjusted EPS growth is achievable on the stated levers (a positive yield-to-cost spread, falling interest expense, a shrinking share count, rising destination throughput) only if fuel is roughly flat. The target's credibility therefore rests on the same unhedged assumption as the rest of the model. We treat PROPEL as a direction of travel with a real road map, not as an underwriting input.

Guidance & Outlook

Full-Year 2026: Raised, Against Expectations of a Cut

Management's own bridge from $2.22 to $2.24 is worth reproducing because it shows how much work was done to move so little: improved operating performance added $0.12 ($0.08 from yields, $0.02 from costs excluding fuel, $0.02 from fuel consumption and other items), share repurchases added $0.01, and higher fuel prices took $0.11 away. The $0.08 of yield improvement splits into $0.05 flowed through from the third quarter and $0.03 of additional fourth-quarter expectation, so the fourth-quarter yield outlook was raised, not lowered.

Fourth Quarter 2026: The Optics Get Worse Before They Get Better

Implied fourth-quarter math, and what the full-year raise conceals. In June, the $2.22 full-year guide less $0.61 of first-half actuals less the ~$1.35 third-quarter guide left approximately $0.26 implied for the fourth quarter. The explicit guide is now ~$0.20. So Carnival beat the third quarter by $0.08, raised the full year by $0.02, and the fourth quarter absorbed the difference. The reason is entirely fuel: June's full-year fuel assumption implied roughly $508M for the fourth quarter after the Q3 guide and first-half actuals; the figure is now $0.64B, about $132M worse, or roughly $0.10 per share. The ~$0.04 of yield and cost improvement management flowed into the fourth quarter offsets part of that, leaving the $0.06 net reduction. A $215M year-over-year increase in a single quarter's fuel bill is what turns a controlled operating quarter into a 41% adjusted-EPS decline.

Street at: Full-year adjusted EBITDA guidance of ~$7.14B was described as in line with analyst expectations, and the ~$2.24 adjusted EPS guide is a modest raise against where the Street had settled. The more relevant fact is the direction: price targets had been cut almost uniformly through mid-September on the expectation of a fuel-driven guidance reduction, and the company raised instead.

Guidance style: Carnival has now beaten its own quarterly adjusted EPS guide in each of the last several quarters, by $0.11 in the fourth quarter of 2025, by $0.07 in the second quarter of 2026 and by $0.08 here, and the full-year guide has moved up in each of the last two revisions, the latest despite a deteriorating fuel assumption. That pattern argues the ~$0.20 fourth-quarter figure is set conservatively, with the yield assumption already raised and the cost assumption already proven beatable twice. The one variable genuinely outside management's control is the $896 per metric ton embedded for the quarter, against sensitivity disclosure of $59M of adjusted net income for every 10% move in the fuel price.

Analyst Q&A Highlights

Whether the Fuel-Hedging Policy Survives a Spike

The most consequential exchange of the call. With the quarter's fuel bill up 36% and the full-year guide absorbing a $442M increase against the prior year, the question was whether a company that hedges nothing would revisit the policy. Management did not hedge the answer either: it restated the decision, gave the financial reasoning, and redirected to consumption as the real instrument.

Q: "So I'll just sort of ask the quarterly kind of question here, but it's 6 months into this conflict and fuel kind of doing what it's doing, any change or update to the philosophy longer term or the conversation internally thinking about potentially looking at hedging at some point in the future?"
— Brandt Montour, Barclays

A: "Yes. Thanks for getting that out of the way, Brandt. So only when fuel is up, right? That's the only time people ask. It's a very fair question to ask because it does have the ability to reduce volatility in any given year. As we've looked at our business and we've looked at this financially as well, we do believe that, that's a short-term band-aid that sometimes pays off, sometimes it doesn't. But the reduction in the volatility to us isn't worth paying banks or counterparties to effectuate those types of trades. And that's why we have been maniacally focused on our consumption rates."
— Josh Weinstein, CEO

Assessment: Management answered fully and without defensiveness, which is more than the question required. The position is intellectually coherent, but it is also the single structural reason to apply a discount to Carnival's earnings multiple relative to a hedged operator: shareholders own the full commodity distribution in both directions, permanently and by design. Investors who dislike that should not own the stock, because this answer will not change.

Whether the 2029 Earnings Target Survives a Flat 2026

A direct challenge to the multi-year framework, built on the observation that if 2026 lands near $2.20 and the 2029 destination is near $3.50, the implied growth rate is far above anything the company has guided. Management conceded the dependency rather than defending the number, separating what it controls from what it does not.

Q: "Josh, so as we think about the PROPEL targets, that would indicate you guys are targeting about, let's say, about $3.50 a share in earnings as we go out to 2029. 2026, we kind of know now is going to be what, $2.20 a share in earnings. So that's implying about 20% earnings growth a year through 2029... as we sit here today, is that 20% earnings growth still pretty realistic in your view?"
— Steven Wieczynski, Stifel

A: "And operationally, we are doing... I would think this was a pretty good year against the backdrop that we were given. And so yields up almost 3% on a normalized basis in a year where there is the biggest geopolitical crisis we've seen in, I would argue, decades... Fuel prices, we do not know. You're right. We don't know what's going to happen in the future... So I'm not pretending it's not going to be hard work, but it's hard work that I think we're all excited to take on, and we've got pretty good road maps to get there. Now if fuel would cooperate, that would make it a hell of a lot easier. But clearly, we can't count on that."
— Josh Weinstein, CEO

Assessment: The exchange is valuable precisely because management did not claim the target was easy. Our own arithmetic off the disclosed FY2025 base of $2.25 puts the required growth at about 14.7% compounded over 2027 to 2029 rather than 20%, so the question overstated the hill slightly, but the direction is right: a flat 2026 has compressed four years of required growth into three. The candour is a credibility asset. The dependency on an unhedged input is a real limitation on how much weight the 2029 target can bear.

Why Bookings Accelerated While the Macro Headlines Did Not Improve

The puzzle of the quarter: bookings inflected and then accelerated through a period in which neither the geopolitical news flow nor airfares improved. The question offered three candidate explanations (drive-to port proximity, European sourcing, and headline fatigue) and management accepted all of them, then added a value argument.

Q: "It sounds like the last 3 months of bookings were really good for you guys, which is particularly noteworthy given the fact that the macro and the geopolitical headlines haven't really gotten any better... do we think there's just fatigue with all of these headlines and people are saying, 'I'm going to go on my vacation, hell or high water,' and we're now beginning to see that."
— James Hardiman, Citi

A: "I mean, taking a step back, June was an inflection point in the booking momentum, and it was positive year-over-year, and then we saw a nice acceleration in both July and August... There is a growing change in the mentality of Americans, which is just catching up to Europeans, as we've talked about before, that vacations are sacrosanct. And they will take them in good times and in bad... And so we are an amazing value for people if the consumer confidence isn't great, if there is pressure from things like price of gasoline, if there's other inflation, if there's concerns about fill-in-the-blank because that's all we hear about nowadays."
— Josh Weinstein, CEO

Assessment: The month-by-month specificity (positive in June, accelerating in July and August) is the part that matters, because it is falsifiable and it matches the record deposit balance. The value argument is the right structural one for a cruise operator in a soft-consumer environment: the product competes against land-based vacations on price per day, and it gains share when discretionary budgets tighten. That makes the demand pillar more defensive than a conventional consumer-discretionary read would assume.

How Much of the 2026 Cost Base Actually Rolls Off

The most useful modelling exchange on the call. A question proposing that 2026 contained 130 to 150 basis points of non-recurring cost drew both a confirmation and an immediate, unprompted offset that most of the sell-side framing had not included.

Q: "If I'm not mistaken, I think '26 had a few different onetime costs in it. There was some logistics moving crew around. I think you had kind of some Celebration Key overflow from FY '25. Is it fair to say that this year had something in the range of, I don't know, 130 to 150 basis points of one-timers in the number? That may not occur next year."
— Benjamin Chaiken, Mizuho

A: "Probably a little over, yes, a little over 1 point. But keep in mind that this year's Celebration Key operated the full year. Half Moon Cay had a half a year of operation. And so we're also in a situation, but we're talking just a couple of tenths of a point. Next year, we will have 5 Evolution ships in dry dock compared to 2 this year. But overall, that may shift some of the costs around, but we are very confident in next year and our cost structure."
— David Bernstein, CFO

Assessment: The confirmation of "a little over 1 point" of one-timers is a genuine 2027 tailwind, but the dry-dock disclosure in the same breath is the part to carry into a model. Five Evolution ships in dry dock against two is a meaningful step-up in both direct cost and lost revenue-producing ship days, and it arrives alongside a capex increase above the $2.4B non-newbuild run rate. The net 2027 cost picture is closer to neutral than the one-timer roll-off alone implies, and management chose to say so rather than let the favourable half stand on its own.

Absorbing the Caribbean Capacity Wave While Pivoting to Europe

A recurring thread across several exchanges probed whether the deployment shift toward Europe is a strategic conviction or a retreat from Caribbean price competition, given the industry supply arriving there. Management separated the two, then declined to pretend the capacity wave is welcome.

Q: "In your prepared remarks, you talked about the fact that Europe for the first time will tie the Caribbean for the largest deployment region... So maybe speak to what you're seeing on both sides of the Atlantic. Are you concerned about what's going on with respect to Caribbean capacity into 2027?"
— James Hardiman, Citi

A: "The Europe question is this isn't new for us... We actually have more European sailings outside of the Med than in the Med, and we love that position. It's also longer seasonality... With respect to the Caribbean... Option #1 is no growth in the Caribbean. Option #2 is, if you look at 2027, something like 37% growth over a 3-year period. I'm going to go with option A. Option A would make my life easier, make a lot of people's lives easier. But it's not taking away from the fact that the Caribbean is an amazing part of our portfolio... Does it face pressure when there's that kind of capacity coming at one time? Yes, it does. And we've seen it before, and we get through it."
— Josh Weinstein, CEO

Assessment: Admitting that a 37% three-year industry capacity increase creates pressure, while pointing to a quarter in which Caribbean-inclusive yields still rose, is the most credible way to handle the question. The Europe pivot reads as opportunity-led rather than defensive: longer shoulder seasons on the same hulls is a revenue-day argument, not a price-avoidance argument. The honest residual is that Carnival's Caribbean yield in 2027 will be defended by owned destinations rather than by scarcity, and that defence has not yet been tested against the full wave.

What Is Actually Driving Onboard Revenue, and Why It Will Not Be Split

A request to decompose the 6.7% onboard growth between the new exclusive destinations and the core onboard business, and to do the same for the fourth-quarter yield guide, was declined on the grounds that bundling has made the ticket-versus-onboard split less meaningful.

Q: "Could you give us a little bit more color on the 7% onboard and other revenue growth in the quarter? Specifically, within that growth, how much did you see from last year's open Celebration Key and how much was more from the actual onboard itself? And how would we -- within your 4Q net yield guide, how much would you attribute from Celebration Key and how much from the core -- ex-Celebration Key?"
— Charles Scholes, Truist Securities

A: "So the third quarter onboard and revenue was really broad-based across all categories, and it was also across all brands on both sides of the Atlantic... I think our bundled packages are helping to contribute to the onboard revenue. We're seeing more than 50% of our revenues prebooked. So we're getting the benefit of the second wallet meaningfully onboard... And as far as the fourth quarter is concerned, I mean, we give guidance in total and yields... breaking down the 2 components isn't as meaningful as it used to be."
— David Bernstein, CFO

Assessment: The refusal is defensible on the merits, because bundling genuinely does move revenue between the two lines, and the company has been consistent in directing attention to total yield. It is nonetheless the most significant disclosure gap in the quarter. Without a destination-versus-core split there is no way to verify from the outside how much of the yield premium the Paradise Collection is actually generating, which is precisely the return question the capital spending has to answer. The "broad-based across all categories" framing is reassuring but unquantified.

Whether an Investment-Grade Balance Sheet Changes the Newbuild Cadence

With the credit now investment grade at two agencies and leverage near management's target, the question was whether the capacity-growth discipline that underpins the pricing story might loosen. The answer closed the door for roughly five years.

Q: "Now that you guys are investment grade, any thoughts as we think longer term? I know maximizing ROIC investments is the priority. But with the balance sheet in such good shape, is there any thought around just trying to expand kind of the capacity growth slightly?"
— Raymond Bowers, Wells Fargo

A: "Yes. I mean, look, our capacity growth is pretty fixed for the next half decade. So that's where we are. If there were opportunities that came around that are one-off unique things, I'd always look at it, and that's part of the job, but I think that's where we are. Is there more opportunity as we get further into the 2030s for more ships? Yes. I still think it will fit within our 1 to 2 ship a year construct."
— Josh Weinstein, CEO

Assessment: This is the answer the thesis needs. Measured capacity growth is not a constraint the company is waiting to escape; it is the strategy, and it is now locked in by an order book that cannot be accelerated. Capacity rises 0.5% in 2027 on top of 1.0% in 2026, against an industry adding supply into the Caribbean at multiples of that rate. Every incremental dollar of demand therefore has to clear through price or through onboard monetization rather than through new berths, which is exactly the mechanism the yield thesis depends on.

What They're NOT Saying

1. No 2027 guidance of any kind, and at least five attempts to extract it: questions on the Q1 exit rate, the yield-to-cost spread, the cost algorithm, the capex profile and whether 2027 clears the historical 2% yield benchmark were all deferred to the December call. One was met with "I appreciate you giving it one last shot for the team on trying to get us to give more guidance, but we're not going to." That is normal timing, but it means the entire 2027 case currently rests on booking-curve adjectives rather than a framework.
2. The net debt to adjusted EBITDA ratio was disclosed in June and omitted in September: the release and the call both stop at expecting "year-over-year improvement in our balance sheet and leverage metrics." Our calculation, using management's own definition and a method that reproduces June's disclosed 3.1x exactly, puts the ratio at roughly 3.10x, unchanged sequentially, with net debt up $46M quarter on quarter. The omission is consistent with a number that stopped improving.
3. The share-count arithmetic is presented gross, not net: $1.2B and 45M shares repurchased is disclosed prominently. That shares actually outstanding rose about 40M since the November year-end, because convertible-note settlement issued more equity than the buyback retired, is visible only by differencing the balance-sheet share lines. Anyone modelling share count off the buyback alone will be wrong in the wrong direction.
4. No dividend increase, three months after signalling one: the June call described "a moderate increase as we look forward" as "rational and reasonable." The quarterly payout is unchanged ($204M paid in the quarter, $618M year to date), and the subject did not come up. The signal has slipped at least one quarter.
5. No quantification of the exclusive destinations' yield contribution: management declined to split onboard growth or the fourth-quarter yield guide between Celebration Key and the core business. Guest counts are disclosed generously; the revenue and return per guest are not. For a strategy absorbing a rising share of a $2.4B-plus annual non-newbuild capex budget, that is the measurement that is missing.
6. Weather was never mentioned: the quarter contained publicly reported itinerary disruptions, including a delayed return and a shortened Bahamas voyage on one ship and a hurricane-driven diversion away from Mexican ports on another, with shore-excursion refunds. Neither prepared remarks nor Q&A touched weather, so the cost of in-quarter disruption is unsized even though it sits inside the yields being celebrated.
7. No sizing of the first-quarter 2027 drag: the residual booking impact was flagged repeatedly and qualitatively, and the request to triangulate it from the Q3 and Q4 exit rate was explicitly refused. Investors know a soft quarter is coming and cannot size it.
8. Selling and administrative expense grew 7.1%, twice the rate of revenue, with no commentary: the cost narrative was entirely about cruise costs per ALBD. The S&A line rose $55M year over year and received no mention in prepared remarks or Q&A.
9. No formal PROPEL baseline, still: the target of more than 50% adjusted EPS growth from 2025 has now gone three quarters without management stating the starting number. FY2025 adjusted EPS of $2.25 is public, and an analyst used a 2029 destination near $3.50 on this call without being corrected, which is tacit confirmation rather than disclosure.

Market Reaction

• Pre-print setup: CCL closed at $22.14 on September 28, down 27.5% year to date against an S&P 500 up 12.2%, down 10.6% over the trailing 30 days and down 24.7% over the trailing twelve months. The stock had closed at a 52-week low of $21.79 two sessions earlier, at the bottom of a $21.79 to $33.99 52-week closing range, after six consecutive down weeks. Short interest had climbed to 4.4% of float, up 35% over the prior month, with borrow costs falling rather than rising. Price targets had been cut almost in unison through mid-September, leaving a mean target near $34 against a share price near $22, and at least one published view expected the company to cut its full-year guide on fuel.
• Reaction session (September 29, reported before the open): shares gapped up 10.2% to open at $24.39, traded between $24.00 and $25.29, and closed at $25.11, up 13.4% ($2.97) on volume of 75.5M shares against a 20.9M 30-day average (3.6x). The S&P 500 fell 0.2% the same day.
• Follow-through (September 30): the stock opened higher at $25.39, faded through the session and closed at $24.54, down 2.3%, holding 10.8% above the pre-print close on volume of 36.0M.

The size of the move is a function of the setup, not of the news. A 13.4% single-session gain on a company whose full-year adjusted EPS guide rose two cents is not a repricing of long-term earnings power; it is the unwinding of a position built on the expectation of a cut. Six down weeks into a fresh 52-week low, a third more short interest than a month earlier, and a sell-side community that had marked targets down across the board had compressed expectations to the point where a modest raise became a violent surprise. The mechanism is visible in the volume: 3.6 times the 30-day average is covering, not accumulation.

What the session does not tell us is how much of the move is durable. The stock gave back a fifth of the gain the following day on still-elevated volume, which is the signature of short-term holders taking the gap rather than long-term holders establishing positions. At $24.54 the stock trades near 11x the ~$2.24 full-year guide and about 7.8x enterprise value to guided adjusted EBITDA on our calculation, having been near 9.9x and 7.4x respectively the day before the print. In other words the rally recovered a derating rather than creating a premium: the stock is still roughly 28% below its 52-week closing high and still down about 20% year to date.

Street Perspective

Debate: Is 2026 a Trough Year or the New Run Rate?

Bull view: Adjusted net income is guided flat against 2025 only because the fuel bill rose $442M. The operating business grew roughly 14% underneath it, yields set another record, unit costs excluding fuel beat guidance again, and interest expense fell 10%. Normalise fuel at any level and 2027 earnings power is materially above the $2.24 the stock is being valued on, before counting a 2027 book already half sold at record price and occupancy.

Bear view: "Ex-fuel" is not an earnings stream. An operator that refuses to hedge has now had two consecutive years of growth consumed by the same line, management has said plainly that it cannot count on fuel cooperating, and the fourth quarter is guided to a 41% adjusted-EPS decline. A business whose reported earnings have been flat for two years deserves to be valued on reported earnings, and the market is doing exactly that.

Our take: The bull side has the better evidence, with one concession. The concession is that commodity beta is permanent here and justifies a structurally lower multiple than a hedged peer would carry. The evidence is that Carnival has now absorbed a $620M deterioration in its fuel assumption against the original 2026 plan while reducing its adjusted net income guide by only $370M, which is a demonstrated offset rather than a promised one. Two successive guides have been defended through external shocks, and the forward book is at records. We would underwrite normalised rather than reported earnings if the normalisation were in the company's control. It is not, and at the fourth quarter's $896 fuel price our normalised 2027 estimate is no higher than the reported 2026 guide.

Debate: Has the Deleveraging Story Been Replaced by a Capital-Return Story, and Is That an Upgrade?

Bull view: Two investment-grade ratings, no secured debt, total debt below $24B against a $36B peak, and a 10% year-over-year reduction in interest expense mean the balance sheet is repaired and the cost of capital gain is locked in. Free cash flow is now available for shareholders, and the company is using it: nearly $1.2B of repurchases at an average we calculate near $26, roughly $2B of total fiscal-year returns, and a dividend with room to grow. Every share retired below intrinsic value compounds the recovery.

Bear view: The leverage ratio has stopped falling. Net debt rose $46M sequentially, cash is down $708M year to date, and the company quietly stopped disclosing the ratio it was advertising three months ago. Buying back stock at an average above the current price while trailing EBITDA is flat on fuel is a worse use of capital than retiring debt, and the dividend increase that was signalled in June has not arrived. The easy part of the balance-sheet story is over and the remaining upside has to come from EBITDA that fuel is suppressing.

Our take: The bear has the better observation and the bull has the better conclusion. The observation is correct and under-discussed: this is the first quarter in years in which Carnival did not deleverage, and the omission of the ratio from the disclosure is a tell. But at 3.1x with two investment-grade ratings secured, shifting the marginal dollar from debt reduction to repurchases below intrinsic value is the right trade, not a retreat. The legitimate criticism is of the communication rather than the capital allocation. Management should have said what we have had to calculate.

Debate: Does the Europe Pivot De-Risk the Franchise or Concentrate It?

Bull view: Europe tying the Caribbean at 34% of the 2027 mix is a deliberate reallocation toward longer shoulder seasons, higher-rate itineraries and a region whose 2027 book grew sharply during the very window that 2026 Europe paused. It also reduces exposure to the industry capacity wave arriving in the Caribbean, where supply grows roughly 37% over three years while Carnival adds 0.5% of fleet capacity. Shifting deployment rather than building ships is the highest-return way to raise yield.

Bear view: Carnival just spent two quarters explaining that a Mediterranean geopolitical event cost it a full point of annual yield, and the response is to put a third of next year's capacity into the region where that happened. The 2026 episode is evidence that European demand is the more fragile of the two books, dependent on long-haul airfare and transatlantic flight capacity that Carnival does not control. Concentrating into it trades a known capacity problem for an unknowable political one.

Our take: The bear's framing is the sharper risk identification and the bull's is the better economics. Northern Europe, which management emphasised repeatedly and which carries more of Carnival's European sailings than the Mediterranean does, is not the geography that broke in 2026, and the longer shoulder seasons genuinely extend revenue-producing days on a fixed fleet. On balance this is diversification rather than concentration, because the alternative is leaning harder into a Caribbean absorbing a 37% three-year supply increase. But it does mean the next Mediterranean disruption costs more than the last one, and that is a real change to the risk profile that deserves to be priced.

Model Update Needed

Valuation impact: We cut our FY2027 adjusted EPS estimate to roughly $2.20 to $2.30 from ~$2.45 to $2.55. The June estimate assumed fuel flat at the June full-year assumption of $713 per metric ton. The fourth-quarter guide now embeds $896. On roughly 2.65M metric tons in 2027 (capacity up 0.5%, consumption per berth day still falling), that price is about $0.34B of cost above the $768 full-year average, or about $0.26 a share, and Carnival pays almost no tax, so the pre-tax and after-tax figures are effectively the same. Applied to the ~$2.24 FY2026 guide with the same operating growth we assumed in June, the estimate is roughly flat year over year. On that estimate, a 12x to 13x multiple supports $26.40 to $29.90. Cross-checking on enterprise value, an FY2027 adjusted EBITDA estimate near $7.0B (our prior $7.35B less the fuel repricing) at 8.0x to 9.0x, less net debt we model continuing to fall toward $21.5B, supports $26 to $31 per share. Taking the two methods together we set a 12-month fair-value range of $26 to $30, with a $28 midpoint. That is below the ~$33 to $36 range we set in June for two reasons: the multiple moves to 12x to 13x from ~14x to 15x, because the fuel exposure that has now materialized is, as argued above, a structural reason for a lower multiple, and the estimate falls, because fuel is now priced at the run rate rather than assumed flat. Against the September 30 close of $24.54, the midpoint implies roughly 14% upside, with the range spanning about 6% to 22%. The asymmetry sits in what we have not underwritten, namely a lower fuel price, the PROPEL cost program running ahead of schedule as management says it already is, and a buyback that at these prices retires roughly 4% of the shares a year.

Thesis Scorecard Post-Earnings

Overall: Thesis strengthened on the operating pillars and weakened on the one input the company does not control. Six of the seven points improved or held, including the decisive one: the Europe air pocket that caused June's guidance cut has closed, and the 2027 book is strong enough that management is reallocating a third of next year's capacity into the region. The fuel bear point, which we marked CONTAINED in June on the strength of a sub-$0.01 net impact, has now materialised in full and taken an entire year of earnings growth. The balance-sheet pillar reached its end state and, in the same quarter, stopped generating improvement. Net of all that, the business is compounding and the reported numbers are not, and at today's fuel price they will not compound in 2027 either. That is a gap we can no longer take an Outperform position on.

Action: Downgrade to Hold from Outperform; conviction 6. We upgraded in June on an explicit test: that the European weakness was transitory and the full-year cost offset was structural. Both were answered affirmatively, the second for a second consecutive quarter, and the company raised a guide the Street had marked down on the expectation of a cut. What changed is the input we do not underwrite. Repriced at the $896 per metric ton the fourth-quarter guide embeds, fiscal 2027 earnings land near $2.25, flat on 2026, and our fair value falls to $26 to $30. A 14% midpoint is not enough compensation for an unhedged commodity that has now taken two consecutive years of growth, and the $25 stock is not the $22 stock: the +13% session spent most of the post-print asymmetry. This is the condition that drove our March downgrade, a fuel bill absorbing the operating improvement, and the June upgrade rested on that condition having passed. We would return to Outperform on either of two things: fuel back toward the mid-$700s per metric ton, which restores growth to the 2027 estimate, or a December 2027 guide whose yield-to-cost spread grows earnings at the $896 run rate. We would move to Underperform on a fourth-quarter yield print below the ~1.7% constant-currency guide together with evidence that the Q1 2027 residual is broader than long-haul itineraries. Fuel will do what it does, and by management's own choice shareholders own all of it.