Compass Group: A Fifth Year of 4–5% Net-New Business and 96% Retention Make the Compounding Real, but at $31.10 a 5.1% Owner-Earnings Yield and 1.7x Leverage Leave No Conservative-Case Margin of Safety

The margin is the second decisive variable. At approximately 7.4%, every ten basis points on FY2026 consensus revenue of $50.7bn is worth roughly $51m of operating profit. After a 25.5% tax rate and using roughly 1.70bn shares, that is approximately 2.2 cents of EPS. A 50-basis-point miss therefore moves operating profit by roughly $254m before considering financing effects, far more consequential than a modest shift in like-for-like volume. Compass has protection: management says around two-thirds of contracts now contain dynamic pricing, while the balance generally has indexation for food and labour. In the Q3 call, it said North American food inflation was running at roughly 3.5% and that pricing normally follows food costs with a lag instead of being taken as a margin enhancer. Those contract changes make the model more inflation-resistant than it was before the pandemic, but they do not eliminate timing risk.

North America remains the economic engine. It generated $16.719bn of first-half revenue and $1.411bn of underlying operating profit, an operating margin of about 8.4% on the disclosed numbers. International generated $8.264bn and $504m, about 6.1%. Fiscal 2025 showed the same structural gap: 8.2% versus 6.1%. The International opportunity therefore includes both contract growth and margin convergence. But investors should resist assuming Europe simply becomes North America. European labour rules, local procurement, contract conventions and fragmented customer behaviour differ, which is precisely why Compass has recently spent heavily on regional platforms.

The acquisition cycle is the clearest change in financial risk. Compass spent around $2.3bn on net M&A in the first half of FY2026 and $2.4bn year to date by the June quarter. Vermaat alone consumed $1.624bn of first-half acquisition cash; the agreed enterprise price had been about €1.5bn. Pro Care Management added $270m, following FY2025 acquisitions including Dupont Restauration and 4Service. Vermaat was the largest acquisition in Compass's history and was expected to produce roughly €700m of annual sales, implying a purchase price of about 2.1 times sales. Compass has not disclosed enough clean standalone earnings information for me to calculate a defensible acquisition EV/EBIT or prospective IRR, which is a significant analytical limitation.

The balance sheet can absorb it, though the cushion has narrowed. Net debt was $8.6bn at 31 March 2026, equivalent to 1.7 times underlying EBITDA, above management's 1.0–1.5-times target. Liquidity was $3.6bn, including a fully undrawn $3.2bn revolving facility, and covenant headroom remained wide. Compass refinanced Vermaat with €1.5bn of bonds maturing in 2029 and 2035. The consequence for shareholders is straightforward: buybacks have disappeared because debt capacity is being used for European acquisitions. The previous $500m repurchase programme ended in December 2024. Under Compass's stated capital-allocation order (organic investment, bolt-on M&A, a strong investment-grade balance sheet, roughly 50% dividend payout and only then surplus returns), the absence of a fresh buyback is economically consistent with leverage at 1.7 times.

The quality of "underlying" profit deserves more scrutiny than it often receives. Fiscal 2025 underlying operating profit of $3.335bn exceeded statutory operating profit of $2.964bn by $371m. Of that, $357m was acquisition-related. The acquisition charge included $226m of amortisation of acquired intangibles, $88m of transaction costs, $27m of contingent-consideration adjustment and other items. H1 FY2026 contained another $234m of non-underlying operating charges, $223m acquisition-related. Acquired-intangible amortisation is non-cash and should not be treated like food or labour expense; transaction and integration costs are more problematic because a company acquiring every year cannot persuasively call every acquisition cash cost economically exceptional. For that reason I value Compass using normalized owner cash earnings alongside underlying profit instead of simply capitalising management's underlying EPS.

Fiscal 2025 statutory free cash flow was $1.865bn; underlying free cash flow was $1.975bn. Capex was $1.514bn, or 3.3% of revenue. The annual report splits that expenditure into $347m of intangibles, $492m of contract-fulfilment assets, $545m of property, plant and equipment, $197m of contract prepayments, less $67m of asset disposals. Compass does not disclose a clean "maintenance versus growth" split. Because fulfilment assets and prepayments are heavily associated with contract mobilisation and retention, I assume 45% of total capex is economic maintenance and 55% is growth or contract-acquisition capital. That is an analytical assumption, not company guidance. On this basis, FY2025 owner earnings are roughly $2.70bn: statutory FCF plus about $0.83bn of estimated growth capex. At $52.9bn equity value, that is a 5.1% owner-earnings yield, or around 19.6 times owner earnings.

Qualitative portrait: high-quality compounding growth. The label fits because Compass has produced four completed years of target-range net-new business, 95–97% regional retention, expanding operating margin, high cash generation and 18.2% ROCE while operating in a market where most potential demand remains outside large global providers. The label requires one qualification: today's growth algorithm is being supplemented with substantially more acquisition capital than it required several years ago.

The market's debate has shifted accordingly. During the recovery from COVID, the question was whether volumes and margins would normalize. By 2024–2025 the market increasingly accepted the structural share-gain story. In July 2025, shares jumped sharply when Compass announced Vermaat alongside a raised profit outlook. In February 2026, the stock fell sharply despite an organic-growth beat because AI-generated workplace disruption suddenly became a valuation concern. By July, a perfectly respectable 7.1% Q3 organic print and unchanged above-11% profit-growth guidance were no longer sufficient to trigger another re-rating. That progression tells us that the market now expects good execution; the marginal questions are duration, AI/workplace exposure, acquisition returns and how much margin expansion remains.

The central bull/bear disagreement is whether 4–5% net-new business is a durable structural capability or an unusually favourable post-pandemic regime. If it is durable, Compass can grow organic revenue around 6–8% even with low single-digit inflation and modest volumes, while incremental margin gains turn that into high-single- or low-double-digit earnings growth. If it falls back towards the old roughly 3% regime while pricing normalises, organic growth can fall towards 4–5%; at the same time, the current valuation and recent M&A returns leave considerably less room for execution errors. Management's retention statistics and first-time-outsourcing mix support the structural view, but only another few years through a weaker macro environment can prove it.

At $31.10, Compass is priced somewhere between a distressed caterer and an extravagant consumer-growth stock. Using FY2025 underlying EPS of $1.319 gives a trailing underlying P/E of roughly 23.6 times. Using the company's FY2026 operating-profit consensus, management's finance-cost guidance and a 25.5% tax rate gives an indicative underlying EPS close to $1.49 and a forward P/E around 21 times. My FY2025 owner-earnings calculation gives roughly 19.6 times. Those are defensible multiples for a business able to sustain high-single-digit organic-plus-margin compounding, but they leave little valuation protection if growth normalises quickly.

Vertical Analysis: History, Financial Development and Capital-Market Narrative

Compass's origins are unusually relevant to the current strategy because the company has repeatedly returned to the same basic formula: specialise in feeding people at somebody else's site, add purchasing scale, decentralise execution, and use acquisitions to fill geographic or sector gaps. Compass's own heritage material traces the business to wartime factory feeding in the 1940s. Historical records identify Jack Bateman's Factory Canteens Limited, founded in 1941, as a root of the present group. Bateman Catering and Midland Catering entered Grand Metropolitan in the late 1960s; management bought the catering interests out in 1987, creating Compass Group, and the company first listed in London in 1988. I could verify the 1988 listing, but not a sufficiently reliable primary archival source for the original IPO price, proceeds or flotation valuation; those figures are therefore deliberately omitted.

The first stage, from factory catering through the 1987 buyout, established the economic proposition: employers and institutions could hand over a non-core but operationally important activity to a specialist. Food service had little technology moat and modest unit margins, but aggregating many sites created purchasing leverage and operating know-how that individual factories or institutions could not easily replicate. That logic remains essentially unchanged. What changed over subsequent decades was the addressable customer set and the scale at which procurement, culinary brands and management systems could be applied.

The second stage, from the 1988 listing into the early 2000s, was global consolidation. Compass expanded geographically and through sector brands, including the development of Eurest in workplace dining and a succession of acquisitions. The group also went through the Granada combination and subsequent demerger around the turn of the century. The period built the scale that still matters today, yet it also illustrated the dangers of excessive organisational complexity and acquisition velocity. Large outsourcing groups can grow faster than their control systems because hundreds of thousands of local transactions are being consolidated through a relatively thin corporate centre.

That experience shaped the third stage: the long move towards a simpler, more decentralised, food-focused model. The strategy eventually emphasised local operating accountability, sectorisation and disciplined contract economics instead of conglomerate breadth. The current annual report still presents decentralisation as a competitive advantage: local management remains responsible for clients and operations while the group provides purchasing, sector brands, culinary technology and financial capacity. Central cost in FY2025 was only $151m against $46.1bn of underlying revenue, illustrating how thin the corporate layer is relative to the operating network.

Dominic Blakemore's era, beginning as group chief executive in 2018, brought Compass into the fourth stage just before the largest disruption in the industry's modern history. COVID temporarily broke the volume assumptions behind workplace, education and sports catering because customers could not use their sites. The company protected liquidity, reconsidered its geographic footprint and emerged with a sharper emphasis on contract quality and retention. The significance of COVID extended beyond the revenue shock. Clients also became more conscious of labour availability, food inflation, health and safety, digitisation and the difficulty of operating catering in-house. Those forces helped increase outsourcing interest after reopening. The company now explicitly argues that contract structures are more inflation-responsive than they were before the pandemic.

The fifth stage began around FY2022 and is the stage investors are buying today: structurally stronger contract growth paired with active portfolio reshaping. Compass reports four completed years of 4–5% net-new business through FY2025 and says FY2026 is on course to be the fifth. Client retention is around 96% at group level, 97% in North America and 95% internationally. That combination means the gross sales organisation can produce meaningful share gains even before price or volume.

The numbers show how different the current business is from the pandemic trough. The 2025 annual report reports $46.1bn of underlying revenue and a 16.8% revenue CAGR from 2021 to 2025. The same document shows 86% of revenue from food and 14% from support services; geographically, 68% was North America and 32% International. Sector mix was 39% Business & Industry, 23% Healthcare & Senior Living, 18% Education, 14% Sports & Leisure and 6% Defence, Offshore & Remote. This diversification matters more than a simple country count because different client populations move differently through the cycle.

The operating trend is stronger than the revenue table alone suggests. FY2025 organic growth of 8.7% already contained less inflation than the immediate post-COVID years, yet Compass still expanded margin. H1 FY2026 then maintained 7.2% organic growth with only 0.7% like-for-like volume because contract wins did most of the heavy lifting. This is the evidence behind the structural-compounder case.

Cash generation remains good, though seasonal. FY2025 underlying free cash flow of $1.975bn represented 87.8% conversion of underlying profit, versus 66.2% in the first half of FY2026. The first-half deterioration is consistent with Compass's seasonal working-capital pattern and mobilisation schedule; by itself it is not evidence of deteriorating earnings quality. The more important test is full-year conversion after the heavy mobilisation cycle.

The balance sheet tells a different story because acquisition capital is now being deployed faster than cash can replenish it. Net debt rose from $6.418bn in September 2025 to $8.6bn by March 2026. Leverage increased from 1.4 to 1.7 times underlying EBITDA, outside the company's 1.0–1.5-times target. The increase was purposeful, not distressed: Vermaat, Pro Care and bolt-ons accounted for about $2.3bn of first-half net M&A. At 31 March Compass still had roughly $3.6bn of liquidity, and its disclosed private-placement covenant metrics were 1.6 times debt/EBITDA against a 3.5-times ceiling and 17.7 times interest cover against a three-times minimum.

ROCE fell from 19.0% in FY2024 to 18.2% in FY2025. The 2025 annual report calculates the latter from $2.476bn of net operating profit after tax and average capital employed of $13.572bn. Management attributes dilution partly to acquisitions. This is a real warning signal because a serial acquirer can keep EPS growing while lowering the return earned on each incremental dollar of capital. The group-wide return remains attractive; the question is whether the European portfolio will eventually earn anything close to it.

A rough near-term test is sobering but not conclusive. Management says acquisitions should add about two percentage points to FY2026 operating-profit growth. Two percent of FY2025's $3.335bn underlying operating profit is only around $67m. Against roughly $2.4bn of FY2026 year-to-date net M&A, that represents less than 3% of acquisition expenditure as first-year incremental operating profit. That is not an acquisition IRR: Vermaat closed partway through the year, other acquisitions have different closing dates, and synergies have not matured. It does show why the next three years matter more than the first reported year. The European platform needs a substantial profit ramp to justify the capital deployed.

Compass's capital-market narrative has moved through the same stages. The company reports total shareholder return of roughly 197% over the ten years to FY2025, reflecting the long-term value created by outsourcing growth, margin discipline and distributions. COVID created the obvious break. The post-pandemic recovery then became a rerating as investors saw retention improve and net-new business settle above its earlier norm.

The 2025–2026 price action exposes what currently moves the multiple. In July 2025, the Vermaat announcement arrived alongside a higher profit outlook, and Compass shares jumped sharply. Investors interpreted the combination as evidence that organic execution remained strong enough to finance a more aggressive European expansion. In February 2026, the exact opposite happened: Q1 organic revenue growth of 7.3% beat expectations of 7.1%, yet the shares fell as much as 8% because investors focused on AI's potential effect on office employment. In July 2026, Q3 organic growth of 7.1% and unchanged guidance failed to create a positive surprise. These episodes say the valuation now reflects confidence in ordinary execution; new upside requires evidence that growth persists longer, margins rise further or acquisition returns surprise positively.

The April 2026 trading-currency change complicates long-term charts. Compass shares moved from sterling pence to dollars on the LSE on 1 April. Any vendor series that simply connects GBp observations before that date with dollar observations afterwards produces a meaningless apparent price collapse. So I use percentage reactions for the historical discussion and the verified $31.10 dollar close on 17 September 2026 for current valuation.

Business Model, Margin Economics, Moat, Industry and Governance

Compass's revenue machine can be reduced to three variables: retained prior-year business, newly mobilised contracts net of lost contracts, and changes in revenue at existing locations through price and volume. That arithmetic is more useful than restaurant-industry concepts such as comparable-store sales. In H1 FY2026, 3.8 points of growth came from net new business, 2.7 from pricing and 0.7 from like-for-like volume. In Q3 the net-new component accelerated back into the 4–5% target range.

Gross wins explain why retention alone is insufficient. Compass signed $4.3bn of annualised new business during the trailing twelve months to June, up 16% year on year. Half represented first-time outsourcing. With retention at 96%, those wins translate into a sizable positive net contribution after lost business. A pure share-shift market would eventually become a zero-sum contest between Compass, Sodexo, Aramark and regional operators; first-time outsourcing enlarges the accessible pool for all of them.

First-time outsourcing is also where recession sensitivity becomes nuanced. Cost pressure can persuade a hospital, university or company to replace an internal food operation with a specialist. Conversely, some Business & Industry opportunities are linked to new campuses, buildings and infrastructure. Management indicated in its half-year discussion that a large part of such activity has been connected to clients' new investment. A downturn can do both: improve the outsourcing proposition and delay physical mobilisation. The net effect depends on sector mix.

The sectors illustrate that diversification. Business & Industry is presently the growth leader, with double-digit organic growth and more than $2bn of trailing new wins, including data-centre construction and operations. Education is benefiting from a strong selling season, including Compass's largest sector win to date at the University of Kentucky. Healthcare and Senior Living has structural first-time-outsourcing potential because much of the sector remains self-operated. Sports & Leisure offers another underpenetrated pool as stadiums and venues try to increase food spend per attendee. Defence, Offshore & Remote benefits when governments and energy/infrastructure clients invest in sites that require complex logistics.

Canteen broadens Business & Industry beyond traditional cafeterias. Its more than $4bn of sales across unattended formats allows Compass to monetise smaller populations with lower labour intensity. This is strategically useful if hybrid work reduces average office occupancy: the same building may no longer support a full canteen but can still support a micro-market, pantry or automated retail offering. That is one reason the AI/office bear case should be analysed as a mix shift as well as a volume loss.

Contract type determines who carries inflation risk. Compass's accounting disclosure says transactions may be priced per meal, vary with volumes or reimburse costs plus an agreed management fee. Management says approximately two-thirds of the current contract base has dynamic pricing and the rest generally contains indexation. In Q3 it described even historically less-flexible contracts as having become more agile since the inflation shock. Food inflation usually becomes price with a lag, not a permanent margin loss, provided Compass retains the client after repricing.

Food purchasing scale is a genuine economic advantage. Food is roughly one-third of costs, which means a procurement saving of 1% on the food basket is equivalent to around 0.33% of revenue before sharing with clients or reinvestment. Compass increasingly supplements direct purchasing with group-purchasing organisations; Pro Care Management expanded that capability in Germany and gave Compass purchasing organisations in five of its ten largest markets. The company can spread supplier negotiation, product specification and menu engineering over tens of billions of dollars of sales in a way an individual client cannot.

Labour is larger and more difficult. The annual report identifies labour as the predominant in-unit cost but does not disclose a clean group or regional labour-cost percentage of revenue. I do not insert an estimated labour share as if it were reported fact. More than 590,000 employees create exposure to statutory minimum wages, collective bargaining, immigration rules, overtime and local labour scarcity. In an inflation shock, wage inflation can be easier to forecast than food commodities but harder to reverse.

The 2022–2024 inflation episode provides some evidence that pass-through works. Compass restored and then expanded underlying margins while carrying significant food and wage inflation, and management subsequently moved more of the book toward dynamic pricing or indexation. The remaining risk lies in timing and tender competition: a client may ultimately reimburse cost inflation yet reject an aggressive renewal price, turning a margin problem into a retention problem. That is why margin and retention should be monitored together.

The sensitivity is unusually high because the starting margin is low.

The calculation uses Compass's June consensus revenue of $50.7bn and management's indicated 25.5% tax rate. It is a static sensitivity, not a forecast.

The International margin gap creates optionality. At H1 rates, a 50-basis-point margin gain on roughly $16–17bn of annualised International revenue would add around $80m of operating profit. A full convergence toward North America is unrealistic because service mix, wage structures, contract maturity and acquisitions differ. Even partial convergence can contribute meaningfully to group earnings. The mirror-image risk is North America: with roughly two-thirds of revenue there, a 25-basis-point North American squeeze can overwhelm substantial International improvement.

Real moats are scale purchasing, sectorised sales and operating know-how, client retention, and financial capacity to mobilise large contracts. Brand awareness among end consumers is secondary. The relevant "brand" is the client's belief that Compass can recruit thousands of people, handle food safety, procure reliably, mobilise a complex location and deliver across multiple sites. A 96% retention rate and repeated $4bn-plus annualised gross wins are stronger evidence of that moat than consumer recognition.

Sectorisation reinforces that moat because hospital catering, university dining, stadium concessions and remote energy-site food are operationally different despite all involving meals. Dedicated brands and sales teams accumulate reference clients, menu expertise, regulation knowledge and labour pools within each niche. The sectorisation deep dive shows International retention improving as this model has been rolled out, while Healthcare management sees more than half of its target market still in-house.

Switching costs are moderate, not absolute. A client can retender, and Compass loses roughly 4% of business annually at current retention. The moat shows up in probabilities: replacing a caterer across a university, hospital network or corporate estate is operationally disruptive, so an incumbent that delivers reliably begins renewal with an advantage. Compass then uses new wins to replace the small percentage that leaves. This moat would weaken quickly if retention fell to the low-90s because gross wins would be consumed replacing losses.

Capital strength matters in a way it does not for a neighbourhood caterer. Large new sites can require meaningful upfront contract fulfilment assets, prepayments, kitchens and technology. Compass spent $689m in FY2025 on contract-fulfilment assets and contract prepayments alone. Financial capacity therefore helps win larger contracts, although those mobilisations also raise the capital intensity of growth.

The industry structure remains unusually fragmented for Compass's size. Management's $360bn outsourced-food estimate implies Compass's sub-15% share leaves enormous room around the largest global players. Nearly three-quarters of the opportunity is described as self-operated or served by regional suppliers. The fundamental industry growth driver is penetration, not population or restaurant visits: hospitals, companies, campuses and venues deciding that food service is complex enough to outsource.

The profit pool belongs to operators that combine local execution with purchasing and corporate infrastructure without building too much overhead. A pure regional caterer can be nimble but lacks purchasing scale and balance-sheet capacity. A giant integrated facilities manager can cross-sell more services but may treat food as one activity among many. Compass's decision to remain overwhelmingly food-led, 86% of FY2025 revenue, has produced a clearer operating model than many facilities-management peers.

The cycle classification is "defensive growth with cyclical pockets." Hospital and school meals do not behave like discretionary restaurants; workplace attendance, sports calendars and corporate construction do. Inflation can initially help nominal organic revenue through pricing but hurts margins if pass-through lags. Recessions can reduce B&I volume and investment while increasing incentives to outsource. Interest rates matter primarily through valuation, acquisition funding and client capital budgets rather than through direct end-consumer financing.

Regulation enters mainly through labour law, food safety, public procurement, healthcare requirements and environmental standards. The regulatory intensity is operational rather than licence-scarcity based. A tightening wage regime can move billions of dollars of cost through the P&L, which is manageable only if contracts permit repricing. Geopolitical shocks feed through food, energy and transport inflation, while Defence and Remote work can actually grow with higher government and infrastructure spending. Compass has less geopolitical concentration risk than a manufacturer with a single supply chain, but substantial inflation transmission risk.

Governance is conventional UK public-company governance. The 2025 annual report shows KPMG LLP as external auditor. The 2024 audit was selected for Financial Reporting Council inspection, and the audit committee subsequently focused on goodwill impairment, uncertain tax positions and acquisition accounting, including 4Service and Dupont Restauration. KPMG issued an unmodified opinion. Total auditor fees were $12.4m in FY2025, of which non-audit fees were $0.8m.

Remuneration is reasonably aligned with the variables that matter economically. For FY2026, 65% of the annual bonus is tied to PBIT growth, 25% to cash conversion and 10% to safety. Long-term incentives use ROCE for 40%, adjusted free cash flow for 40% and relative TSR for 20%. The chief executive's maximum annual bonus is 250% of salary and maximum LTIP award 500%; one-third of the bonus is deferred into shares for three years, and the chief executive has a five-times-salary shareholding requirement. Those metrics at least force management to balance growth with capital returns and cash.

Capital allocation is where governance is being tested now. Compass states that it first invests in organic growth and bolt-on acquisitions, seeks to retain a strong investment-grade rating and 1.0–1.5-times leverage, pays roughly half of underlying earnings as ordinary dividends, then returns surplus capital. That framework produced $2.3bn of buybacks from 2022 through the completion of the latest programme in December 2024. The present acquisition wave has removed "surplus" balance-sheet capacity. I interpret the lack of a new buyback as prudent, provided management actually deleverages instead of simply resetting its acceptable leverage upward.

The shareholder register is institutionally dispersed. At the FY2025 reporting date, disclosed significant holdings included BlackRock at 9.99%, FMR at 5.09%, Artisan Partners at 4.96%, Invesco at 4.95% and Massachusetts Financial Services at 4.60%. These are disclosure snapshots, not necessarily exact economic holdings on 18 September 2026. I found no reliable evidence yet that the April dollar trading line has materially changed US institutional ownership. The currency change is only five and a half months old, so claiming such an effect would be premature.

Horizontal Analysis and Current Fundamentals

Compass belongs to a competitive set that needs to be separated into two groups. Sodexo and Aramark compete directly for many food-service accounts. ISS is more facilities-management-heavy and provides a useful benchmark for large outsourced-contract economics. Elior is a smaller catering-focused European operator whose recent strategy emphasises contract profitability over growth. Food distributors such as Sysco are suppliers and cost benchmarks, not direct peers; Cintas and Rentokil are useful analogues for retention and route/service economics but sell very different services.

Compass has become the food-services pure-play with the strongest mix of North American scale, contract growth and margin. Sodexo remains formidable in global food and facilities but entered fiscal 2026 with serious execution problems. Its official H1 FY2026 results reported only 1.7% organic growth and a 3.7% underlying operating margin, down 140 basis points at constant currencies, causing full-year guidance to be cut to 0.5–1.0% organic growth and a 3.2–3.4% margin. Later Q3 performance improved enough for guidance to move to roughly 1.2–1.5% organic growth, but the central contrast remained: Compass was running at about 7% organic growth and expanding margin while Sodexo was repairing execution.

Sodexo's customer proposition remains broad. It offers food and facilities services across many of the same corporate, healthcare and educational clients. That breadth can help integrated bids; it can also produce greater organisational complexity. The 2026 weakness in North America, including education and business accounts, shows why Compass's stronger local sector execution commands investor attention. What Compass needs is a more consistent selling, mobilisation and retention process, not a fundamentally superior sandwich.

Aramark is the closest US operating comparator. It serves educational institutions, Fortune 500 companies, sports teams, healthcare providers, municipalities and destinations, and it operates in 16 countries. Its strongest overlap with Compass is North American education, workplace, sports and facilities services. Aramark also has deep sports and entertainment relationships and a recognised US concessions franchise. I could retrieve its current IR filing index and operating description, but not a clean, same-date fiscal-2026 financial table from the primary filing, so I do not populate a pseudo-comparable table with secondary numbers.

ISS is a different animal. It is primarily an integrated facilities-services company, making food only one element in much larger workplace relationships. H1 2026 organic growth was 8.2%, Q2 organic growth 8.9%, and H1 operating margin 4.6%. ISS guided for more than 6% organic growth, an operating margin around 5.25% and underlying free cash flow above DKK3.1bn for 2026. Its mid-term target calls for average organic growth above 5% during 2026–2028 and margin of 5.5–6.0% in 2028. The relevance for Compass is contract economics: ISS shows that large, complex, multi-service outsourcing can also compound through retention and net wins, though its lower margins reflect a different service mix.

Elior shows the other end of the spectrum. In its fiscal 2025 first half, organic growth slowed to 1.5%, but adjusted EBITA rose sharply as management exited unattractive contracts and concentrated on profitability; it raised margin expectations while cutting revenue-growth guidance. Elior is therefore the clearest reminder that revenue growth in contract catering is not automatically valuable. Losing an underpriced contract can create value; winning one at inadequate labour or food assumptions can destroy it.

The cleanest current primary-data comparison is operating momentum, not a same-date market-multiple spreadsheet:

The accounting bases and period ends differ, so the margin numbers are directional comparisons, not a league table. Compass's margin advantage partly reflects a food-heavy portfolio and North American mix; ISS has much more facilities content, and Sodexo's FY2026 margin is depressed by restructuring and execution issues. The quality signal is that Compass combines strong growth, high retention and expanding margin at the same time.

I do not publish a same-date peer P/E table because I could not establish a uniformly primary, same-period earnings set and 17 September market values for all four peers. Mixing statutory IFRS earnings at one company, adjusted earnings at another and vendor forward estimates at a third would create false precision. The qualitative valuation conclusion is still clear: Compass deserves a premium to a struggling Sodexo or turnaround-stage Elior because its organic growth and contract economics are stronger. The case for an unlimited premium is weaker. Contract catering remains a labour-intensive, competitive business with client retenders and little proprietary technology.

The ecological niche is attractive. Compass is the global scale leader that has remained sufficiently food-focused to keep operational complexity below that of a broad facilities conglomerate. It takes profit pools from self-operated institutional kitchens and smaller regional caterers. The threat is more likely to come from competitors matching Compass's procurement, digital ordering, unattended retail and sector-specific selling while bidding more aggressively for renewals than from a restaurant chain or food distributor suddenly replacing the model.

The last four reported checkpoints support the same picture. FY2025 organic growth was 8.7% with 7.2% operating margin. Q1 FY2026 delivered 7.3% organic growth, ahead of the 7.1% market expectation cited by Reuters. H1 held organic growth at 7.2%, increased margin to 7.4% and led management to raise full-year constant-currency operating-profit-growth guidance above 11%. Q3 delivered 7.1%, restored net new business to the 4–5% target and left full-year guidance intact.

North America is currently doing more of the work. H1 organic growth was 7.2%, with Business & Industry in double digits and 97% retention. International also grew 7.1% organically with 95% retention, but its like-for-like momentum moderated in Q3 as inflation fell and Sports & Leisure calendar comparisons became less favourable. That pattern is healthy: growth is not dependent on a single event or region, although North American B&I remains disproportionately important.

The World Cup gave North American volume only a modest Q3 benefit. This is useful because it means 7.5% North American Q3 organic growth reflected more than an event spike. Conversely, investors should expect International like-for-like growth to look less spectacular when inflation continues to normalise. As pricing falls from 2–3% towards something closer to long-run inflation, net-new business must remain near 4–5% for the mid-to-high-single-digit organic algorithm to hold.

The market is currently trading three narratives simultaneously. First is the established "quality compounder" story: retention around 96%, 4–5% net new business and margin progression. Second is the AI workplace debate. February's selloff proved that investors can look through a quarterly beat when they fear structural office headcount erosion. Third is European M&A: investors must decide whether Vermaat, Pro Care, Dupont and 4Service are the early stages of a North-America-style acquisition playbook or evidence that organic opportunities alone no longer absorb Compass's capital.

The AI debate is more balanced than the share-price reaction implied. Reuters reported that technology, professional and financial-services clients account for about 20% of Compass revenue. Management said the white-collar office exposure it considers directly vulnerable is smaller, and entry-level jobs viewed as most susceptible to AI represent only a fraction of it. At the same time, Compass is winning data-centre food and support contracts from the same technology ecosystem. A severe white-collar employment shock would still reduce cafeteria transactions; data-centre revenue is a hedge, not an exact substitute.

The bull case rests on four pieces of evidence. Retention has remained around 96% for years, gross wins continue to grow double digits, half of wins are still first-time outsourcing, and operating margin has continued to rise as inflation normalises. Those are observable rather than narrative claims. The bear case rests on equally tangible facts: pricing contribution is already moderating, leverage has risen above target, M&A exclusions are recurring, and the current equity valuation assumes that the 4–5% net-new regime lasts.

Valuation, Cash-Flow Passthrough, Risks, Catalysts and Tracking

The correct starting point is cash flow because accounting presentation materially affects Compass's valuation. FY2025 statutory attributable profit was $1.868bn and cash generated from operations was $3.366bn, giving a latest-year operating-cash-flow/net-income ratio of approximately 1.80 times. Depreciation, amortisation, working-capital timing and non-cash acquisition accounting explain much of the excess. A fully consistent five-year aggregate OCF/net-income ratio cannot be reproduced from the latest recast disclosure without combining older reporting-currency and classification regimes; instead of inventing a precise five-year number, I use the company's multi-year free-cash-flow record and disclose this as a research limitation.

FY2025 underlying free cash flow was $1.975bn and statutory free cash flow $1.865bn. The $110m difference largely reflects items Compass excludes, including strategic/pension and acquisition-related transaction costs. Because acquisition transaction costs recur in a serial-acquisition strategy, my valuation starts with statutory FCF rather than automatically adding them back.

Maintenance capex requires an explicit assumption. Compass spent $1.514bn in FY2025, including $492m of contract-fulfilment assets and $197m of contract prepayments. Those two categories are strongly associated with winning, retaining and mobilising customer sites. PPE and intangibles include a mixture of maintenance and growth. I assume economic maintenance capex equals 45% of total reported capex, approximately $681m, with the remaining $833m treated as growth/contract-acquisition capital. A reasonable sensitivity band is 40–50% maintenance.

On that basis:

Growth capex is still a real cash cost. The point is that a shareholder valuing the existing earning machine should distinguish capital required merely to maintain current operations from capital consciously deployed to win future revenue. Future growth projections then have to earn an adequate return on that growth capital. The 5.1% yield is therefore more informative than raw FCF yield, while M&A expenditure remains separately chargeable to capital allocation.

The headline underlying P/E gives a less favourable picture. FY2025 underlying EPS was $1.319, so the current price is roughly 23.6 times trailing underlying earnings. Using FY2026 consensus underlying operating profit of $3.743bn, management's roughly $350m finance-cost indication and 25.5% tax rate produces an indicative underlying net profit around $2.53bn, or approximately $1.49 per share, implying around 20.9 times forward underlying earnings. The owner-earnings and underlying-P/E approaches differ by less than 30%, so the divergence threshold at which underlying P/E should be abandoned in favour of owner earnings alone is not reached; I nevertheless weight owner earnings heavily because mobilisation capex and acquisition accounting are economically central here.

The statutory/underlying reconciliation is important enough to show directly:

Compass's current comparative disclosure gives a clean FY2024–H1 FY2026 bridge. I could not independently reconstruct FY2023 under precisely the same USD/recast classification basis from the primary documents I could access, so I do not insert an estimated FY2023 number. The economic conclusion does not depend on it: acquisition-related adjustments are recurring enough that transaction and integration cash costs belong in normalized economics even when acquired-intangible amortisation is treated separately.

Historical valuation needs similar caution because the trading currency changed in April. A statistically precise historical percentile would require a fully currency-adjusted daily price and point-in-time earnings series. I did not obtain one clean enough to publish. I do not claim a false "63rd percentile" style number. The more defensible statement is that the stock continues to carry a quality-growth valuation: about 21 times indicative FY2026 underlying earnings and around 20 times FY2025 estimated owner earnings despite leverage above target.

The peer premium is economically justified to a degree. Compass is growing far faster and earning a much higher margin than Sodexo's depressed FY2026 level; it also has a clearer food-services identity than ISS. A premium becomes dangerous once it implicitly requires 4–5% net new business, annual margin expansion and successful European M&A all at once. Absolute valuation matters more than simply observing that weaker peers trade cheaper.

My three-scenario framework values normalized owner earnings, cross-checked against underlying P/E and cash yield. Near-term fair value is intended as a 12–18-month research value, not a trading target.

This is valuation-scenario analysis, not investment advice.

The conservative scenario assumes no crisis, only that pricing normalises, net new business drifts towards 3–3.5%, margin ceases expanding and the market assigns a more ordinary outsourcing multiple. That alone produces value in the high-$20s. The current price is above that range, which tells us there is no conservative-case discount.

The base case assumes the core management algorithm largely works: 4–4.5% net new business, approximately 2% pricing over time, modest volume, incremental International margin improvement, and leverage moving back towards 1.5 times. I do not require full North American/International margin convergence. That supports low-$30s fair value and a 3–5-year total return mainly from earnings growth and dividends, not immediate multiple expansion.

The optimistic case needs more than ordinary execution. It assumes the 4–5% net-new regime proves structural even through a softer economy, European platforms raise International margin, data-centre and unattended-retail opportunities offset office headwinds, and acquisition returns eventually support group ROCE rather than dilute it. Under those conditions, $38–40 is defendable. Above the low-$40s the valuation begins to pre-spend too much of that success.

The expectation gap at the next earnings report therefore centres on four figures: FY2026 net-new business, exit margin, year-end leverage and initial FY2027 guidance. A 7% organic number alone is unlikely to surprise investors. Evidence that net new remains above 4% with less pricing would be more valuable because it tests the structural thesis. Likewise, an operating-margin beat accompanied by weak cash conversion would be less impressive than a smaller margin beat paired with deleveraging. Full-year results are scheduled for 24 November 2026.

The margin-of-safety test is deliberately harsher than the fair-value exercise. Current price of $31.10 is about 7–15% above the conservative $27–29 valuation, so the discount to conservative value is negative. Margin of safety at the present quotation is therefore zero.

The most fragile base-case assumption is that net-new business remains near 4% or better after the post-pandemic outsourcing cycle matures. Cutting the base organic-growth contribution attributable to sustained share gains to roughly 70% of my assumption reduces normalized growth towards 5% and lowers my base valuation to about $29–30 per share. That would place the current quote above rather than near fair value.

A flat-earnings test is harsher still. With earnings unchanged for three years and the valuation multiple unchanged, expected return is essentially the dividend yield, approximately 2–2.5% annually at today's price. The US Treasury's published yield-curve material has shown ten-year yields in the mid-4% area during 2026; the specific Treasury page I retrieved displays 4.44%, although its embedded table date is earlier than the report date, so I do not present 4.44% as a 17 September close. The comparison holds under modest rate changes: flat Compass earnings plus its present dividend would underperform a roughly 4% risk-free yield. There is no margin of safety at this buy price.

Margin-of-safety verdict: none.

The permanent-loss risks are concentrated enough to describe individually.

The first is a structural fall in net-new business. I assign medium probability and high impact. If retention falls below roughly 95% while gross wins slow, the 4–5% net-new engine can fall towards 2–3%. Organic revenue would then lose one to two growth points before considering lower pricing, turning a 7% growth asset into perhaps a 4–5% one. The immediate transmission is lower revenue; the larger loss could come from the market reducing the earnings multiple because Compass would no longer deserve a share-gain compounder premium. The observable indicators are gross wins, first-time-outsourcing share, retention and net-new mobilisation.

The second is margin/pass-through failure. I assign medium probability and high impact because food and labour inflation are volatile even though contract protection has improved. A 50-basis-point group margin shortfall on consensus FY2026 revenue equals about $254m of operating profit, roughly 7% of expected underlying operating profit. If the cause is underpriced multi-year contracts, the effect can persist until renewal; if Compass reprices too aggressively, retention can then weaken. The key indicators are pricing versus external food/wage inflation, North American margin and retention.

The third is acquisition-return dilution. Probability is medium and impact high. Compass has put more than $3bn of acquisition capital to work across FY2025 and FY2026, with leverage now above target. If Vermaat and the European platforms cannot raise profit fast enough, ROCE can fall while acquisition amortisation and transaction costs continue. The market could then stop capitalising acquisition-related amortisation as harmless and begin treating M&A as an ongoing maintenance requirement for growth. Monitor ROCE, International margin, leverage, acquisition-related cash costs and whether management resumes buybacks only after debt returns inside the target range.

The fourth is AI-driven workplace contraction. I assign medium probability but uncertain magnitude. The clearest risk is concentrated in technology, financial and professional-service clients. Falling office headcount or attendance lowers transactions at existing B&I sites and can reduce new-building opportunities. Canteen's unattended formats and data-centre work reduce the net exposure, but they do not eliminate it. The early warning signal is North American B&I like-for-like volume becoming persistently negative while data-centre wins fail to offset conventional office weakness.

The fifth risk is valuation compression independent of business failure. Probability is medium and impact medium-to-high. A company at around 20 times normalized earnings does not need an earnings recession to generate poor shareholder returns; a rerating to 15–17 times on unchanged earnings would be enough. The February 2026 share-price response shows how quickly the market can reprice duration risk even after a revenue beat.

Positive catalysts over the next year are unusually measurable: FY2026 net-new business closing within 4–5%; FY2027 guidance retaining mid-to-high-single-digit organic growth; International margin moving convincingly above the low-6% range; free-cash-flow conversion returning towards the high-80s; and net debt/EBITDA heading back to 1.5 times. A return to shareholder buybacks after those conditions would be evidence of genuine balance-sheet capacity, not a catalyst I currently assume.

Negative catalysts are the reverse: FY2027 organic guidance below 6%, retention below 95%, a margin stall caused by food or labour costs, leverage remaining around 1.7–1.8 times despite a full year of acquisition earnings, another large platform acquisition before deleveraging, or evidence that B&I office volumes are becoming structurally negative.

The tracking dashboard should stay compact:

The first four indicators test the growth algorithm. Margin, FCF and capex show whether growth is being bought at increasingly poor economics. Leverage and ROCE test the acquisition strategy. All should be read together: 8% organic growth with worsening retention and 4.5% capex is lower-quality growth than 6.5% organic growth with 96% retention, normal capex and falling leverage.

Cross-Synthesis, Final Research Conclusion, Key Data, Uncertainties and Sources

Vertically, Compass has proven one capability more convincingly than any other: it can turn a mundane, labour-heavy local service into a scalable global compounding model without centralising every local operating decision. The company's history repeatedly returns to the same architecture. Local teams run kitchens and client relationships. Sector brands specialise the offer. Group purchasing and technology reduce unit cost. The balance sheet funds mobilisation. Corporate overhead remains relatively small. Scale is then recycled into higher win rates and enough cash to fund both dividends and additional growth.

Past success was partly helped by an era tailwind. Institutional outsourcing has expanded for decades because food service has become more complicated: allergen rules, labour shortages, menu expectations, digital ordering, sustainability requirements, data analytics and procurement inflation all make self-operation harder. Post-pandemic disruption probably accelerated that trend. Compass was fortunate to be positioned for it.

Yet luck does not explain five years of net-new performance. A favourable outsourcing market is available to Sodexo, Aramark, ISS and regional operators too. Compass has consistently converted that market into 95–97% retention and unusually high gross wins, while some competitors have struggled. That points to management systems, sectorisation and scale as genuine capabilities.

Horizontally, Compass's advantage is best described as focused scale. Sodexo has comparable global reach but more complex execution and substantially weaker FY2026 operating momentum. Aramark is a formidable US competitor with deep sports and education franchises but less global scale. ISS is excellent at integrated facilities contracts but is a different service mix. Elior is improving profitability after retrenching from unattractive contracts. Compass occupies the useful middle: large enough to enjoy procurement, technology and capital advantages, focused enough that food remains the operating centre of gravity.

Its weakness is also visible. Compass earns about seven cents of operating profit on each dollar of revenue, so seemingly small execution errors matter. It employs more than 590,000 people, buys food at enormous scale, has to mobilise thousands of sites and renews contracts continuously. There is no patent protecting that margin. The business depends on repeating mundane actions very well.

The most important evidence for the 3–5-year view is whether net-new business remains above its old pre-pandemic norm when pricing and macro conditions normalise. Pricing cannot stay around 3% forever without equivalent inflation. Like-for-like volume has contributed less than one point recently. If net new stays around 4–5%, Compass can still produce mid-to-high-single-digit organic growth in a normal inflation environment. If it falls towards 3%, revenue growth will look much more ordinary.

That is the metric the market is most likely misjudging in either direction. The bullish error would be treating the current win rate as a transient reopening phenomenon when sectorisation and first-time outsourcing have actually created a structurally higher rate. The bearish error would be extrapolating five unusually strong years indefinitely and ignoring the possibility that the first-time-outsourcing backlog was accelerated by pandemic-era labour and inflation shocks.

The first-time-outsourcing share gives bulls meaningful evidence. Half of $4.3bn of trailing wins were customers entering outsourcing for the first time. That is healthier than merely exchanging accounts with Sodexo or Aramark. Healthcare remains underpenetrated, Sports & Leisure International is still generating first-time business, and Canteen widens the market to locations too small for staffed kitchens. Data centres create a new B&I niche.

The recession test has not yet been fully passed under the current model. A serious downturn could hit workplace attendance, client construction projects and sports hospitality at the same time. Cost pressure might increase outsourcing demand, but the sequence matters: Compass can experience weak existing-site volume before new outsourcing contracts are signed and mobilised. That is why I would not value 4–5% net new as a contractual certainty.

Margin has more visible protection than five years ago. Two-thirds dynamic pricing and indexed residual contracts make a repeat of uncontrolled inflation easier to manage. Procurement scale and group-purchasing organisations add purchasing leverage. The next gains are harder, however. North America is already above 8%; expecting years of expansion there asks a lot from a labour-heavy service. International has more space, but extracting it requires integration and better local execution.

The European M&A strategy is therefore the decisive new variable. Compass spent decades building North American scale through acquisitions; management is trying to reproduce part of that process in Europe. Vermaat adds a premium platform and €700m-scale sales base. Pro Care expands purchasing power. Dupont and 4Service deepen France and Norway. This is strategically coherent. The financial question is whether the platform prices allow sufficient returns.

At approximately 2.1 times Vermaat's indicated sales, Compass paid a meaningful price for a catering business. Strong margins and growth can justify it, but Compass has not disclosed enough standalone EBITDA, expected cost synergy or medium-term return data to let an outside investor prove the deal clears an 18% group ROCE hurdle. The return burden should remain on management instead of being assumed.

Leverage disciplines that process. At 1.7 times EBITDA Compass is nowhere near financial distress; liquidity and covenants are comfortable. The importance of the 1–1.5-times target is behavioural. Returning below it would show that management still treats the range as a binding capital-allocation rule. A new large acquisition before deleveraging would imply that the range has become aspirational and would reduce my confidence in capital discipline.

The accounting debate reaches the same conclusion. Underlying profit is useful for understanding operations because acquired-intangible amortisation is non-cash. Yet Compass now incurs acquisition transaction and integration costs repeatedly. Adding all of them back while giving full credit for acquisition-driven earnings is asymmetric. My owner-earnings approach keeps those recurring cash costs in the economics, while adding back estimated growth capex that is deliberately spent to create future contracts.

This produces a business that is cheaper than its headline 23.6-times trailing underlying P/E but still not cheap. My estimated FY2025 owner-earnings multiple is about 19.6 times, with a 5.1% owner yield. That valuation can compound acceptably if earnings grow around 7–9% and the multiple is reasonably stable. It offers weak protection if growth settles around 4–5%.

The 12-month setup differs from the 3–5-year setup. Over twelve months, Compass has to clear an already high operating bar. The market expects around 7% organic revenue growth and more than 11% constant-currency operating-profit growth for FY2026. A merely "good" FY2027 outlook may produce limited rerating. The largest short-term upside would come from margin and leverage beating expectations simultaneously. The largest downside would come from net-new business rolling over just as pricing normalises.

Over three to five years, the opportunity is larger because the relevant variables are outsourcing penetration, sustained win rates and International returns. A company with a sub-15% share of a management-estimated $360bn pool does not need the market itself to grow explosively. Retaining clients and taking one or two points of share over time can produce substantial revenue. That is what makes a stable 4% net-new rate so valuable.

Currency is secondary to operating economics but must be handled correctly. Compass reports in dollars, earns roughly two-thirds of revenue in North America and now trades in dollars in London. Q3 management estimated that 15 July spot rates, if sustained, would add about $438m to FY2026 reported revenue and $23m to operating profit. This is mainly translation, not evidence of stronger underlying demand. Euro debt associated with European acquisitions also introduces translation and financing effects. UK shareholders still receive dividends in sterling by default unless they elect dollars.

The share-price currency switch itself removes the awkwardness of reporting in dollars while trading in pence and may make analysis simpler for international investors, without making Compass more valuable. Compass explicitly says it does not alter the listing or FTSE membership. There is not yet enough evidence to claim it has materially changed liquidity or the shareholder base.

For the next year, I would focus on net-new business, margin and deleveraging. For three years, I would focus on International margin and acquisition ROCE. For five years, the defining question is whether Compass remains a 4–5% share-gain machine once first-time outsourcing has moved through a full economic cycle.

The conditions for a materially better investment are straightforward. The company can improve without the stock becoming a better investment if valuation expands at the same time. I want either a lower price or stronger evidence that the current growth algorithm deserves a higher multiple. A return towards the low-$20s with retention near 96%, net new above 4% and leverage moving below 1.5 times would create an unusually attractive combination. At a price above $40, the business would need to outperform my already demanding optimistic scenario.

The judgment should be overturned on the positive side if International margin rises faster than expected, recent acquisitions sustain group-level returns and net-new business remains around 5% through a recession. It should be overturned on the negative side if retention slips below 95%, net-new business stays below 3.5%, ROCE falls towards the mid-teens or management makes another major acquisition before restoring leverage.

Bull reasons

• Compass has delivered 4–5% net-new business for four completed years and expects a fifth, while group retention remains 96%, providing a growth engine largely independent of restaurant-style consumer traffic.
• Half of $4.3bn of trailing new wins are first-time outsourcing, giving Compass access to a profit pool larger than competitor share-shifting alone.
• H1 FY2026 margin reached 7.4% while organic growth remained 7.2%, showing that inflation normalisation has not yet broken operating leverage.
• The company has below-15% share of its own estimated $360bn outsourced-food market, while much of Healthcare, Sports & Leisure and institutional catering remains self-operated or regional.
• Canteen, data centres and European platform acquisitions extend the addressable market beyond conventional staffed workplace cafeterias.

Bear reasons

• The current $31.10 price is above my conservative $27–29 value and implies roughly 20 times normalized owner earnings, leaving no conservative-case margin of safety.
• Leverage rose to 1.7 times after $2.3bn of first-half M&A, above Compass's 1.0–1.5-times target, while ROCE had already fallen from 19.0% to 18.2%.
• Acquisition-related exclusions reached $371m in FY2025 and another $234m in H1 FY2026, making a portion of "non-underlying" cost economically recurring.
• At a 7%-plus margin, a 50-basis-point cost/pass-through miss is worth roughly $254m of operating profit on FY2026 consensus revenue.
• B&I exposure to technology, financial and professional-services offices creates a real downside pathway if AI permanently reduces white-collar employment or site attendance.

Pre-mortem. The first plausible 50%-loss script begins in FY2027. AI-driven restructuring cuts employment and attendance across technology and professional-services campuses while a broader slowdown delays new corporate construction. North American B&I like-for-like volume turns negative, gross wins slow and retention falls to 94–95%. Net-new business drops from 4–5% to 2%. Pricing normalises below 2%, so organic growth falls near 3%. With wage inflation still elevated, group operating margin drops from roughly 7.4% to 6.7%. Underlying EPS could fall rather than grow, and investors could reset the multiple from roughly 21 times forward earnings to 13–15 times. A share price around $15–18 would be plausible even without balance-sheet distress.

The second script starts in Europe. Vermaat, Dupont, 4Service and Pro Care integrate more slowly than expected through 2027–2028; labour and client-contract differences prevent rapid margin convergence. Compass keeps reporting acquisition adjustments, International margin remains around 6%, ROCE falls below 16%, and net debt remains above 1.5 times because another platform deal consumes cash that had been expected to deleverage. Investors then stop treating M&A amortisation and integration expense as harmless. A lower statutory earnings base combined with a 14–16-times multiple could again produce 40–50% equity downside from the current quote.

The present company is financially stronger than either pre-mortem implies, so these are stress tests, not base forecasts. They show how permanent loss would occur: operating deterioration and multiple compression would reinforce one another.

Final research conclusion. Compass is a high-quality operator in a structurally attractive outsourcing niche, with better evidence of durable share gains than most labour-intensive service companies. Its 96% retention, repeated 4–5% net-new business, purchasing scale and sectorisation constitute a real moat. I think the evidence favours the view that part of the post-pandemic improvement is structural. I do not assume all of it is permanent.

The valuation is the constraint. At $31.10, the stock sits close to my $31–34 base value and above the $27–29 conservative value. The owner-earnings yield is around 5%, leverage is temporarily above target and the European platform strategy still lacks enough disclosed return data to prove that recent acquisitions will earn group-like returns. Existing holders are being paid to own a good compounding business; new buyers are not being offered a meaningful cushion against an ordinary growth slowdown.

That leads to a different stance over the two horizons. Over twelve months I see modest upside unless FY2027 guidance, International margin or deleveraging exceeds expectations. Over three to five years, the business can plausibly compound total returns around the high single digits under the base case. A price materially below current levels would make that long-duration quality much more attractive.

【Company-profile scores】

• Fundamental quality: high
• Growth: high
• Moat: strong
• Financial soundness: strong
• Management credibility: high
• Valuation attractiveness: medium
• Risk level: medium
• Suitable investor type: long-term growth

【Investment rating】

• Rating: Hold
• One-line thesis: Retention and net-new business support durable compounding, but a roughly 5% owner-earnings yield and acquisition-heavy leverage leave little margin of safety.

【Ideal Buy Price】21–22 USD

Basis: this sits at least 20% below the midpoint of my $27–29 conservative valuation, while requiring retention to remain roughly 95–96%, net new business to remain around 4%, and leverage to demonstrate a path back towards 1.5 times.

• Acceptable hold price: $29–35, a band around the $31–34 base-case fair value.
• Clearly overvalued price: $44–47, at least roughly 10% above the $38–40 optimistic valuation.
• Current-price classification: acceptable hold.
• Whether to wait for a better price: yes for new capital. My preferred trigger is $22 or below while retention remains at least 95%, net new remains near 4% and leverage is clearly falling. The opportunity cost is foregoing roughly a 2–2.5% dividend yield and potential earnings compounding if the price never revisits that level.
• Target holding horizon: 3–5 years.
• Expected annualized return: conservative roughly 1–4%; base roughly 8–10%; optimistic roughly 12–15%, including dividends and assuming the respective operational paths hold.
• Max-loss risk: about 50% in the pre-mortem case, with a potential $15–18 share price if net-new business falls near 2%, margin drops into the mid-6% range and the earnings multiple compresses into the low-to-mid teens.
• Reassessment-trigger signals: retention below 95% for two reporting periods; net-new business below 3.5%; underlying operating margin below 7.2% without a clearly temporary cause; leverage above 1.8 times or failure to move towards 1.5 times; group ROCE below 16%.

【Valuation Range】

• current: 31.10 USD (close as of 2026-09-17)
• bear (conservative · ideal buy zone): [21, 22]
• base (fair · acceptable hold zone): [29, 35]
• bull (optimistic · above the clearly-overvalued line): [44, 47]

A few data-quality points matter. The market capitalisation of approximately $52.9bn is the product of the $31.10 close and about 1.700bn voting shares. The 30 June 2026 total-voting-rights figure is 1,700,442,428. I did not retrieve a later primary voting-rights RNS; MarketScreener's $52.88bn market-cap figure independently reconciles almost exactly to that share count, but I treat the June share count as a residual timing uncertainty rather than claiming a later date.

The other principal uncertainty is acquisition disclosure. Vermaat's purchase price and revenue scale are available, but a clean standalone operating-profit/EBITDA baseline and quantified synergy bridge were not sufficiently disclosed for a reliable purchase multiple or IRR. The maintenance-capex assumption is likewise mine because Compass does not label capex cleanly as maintenance versus growth. Those two inputs are the largest modelling blind spots.

A third limitation is historical valuation. Because the LSE trading currency changed from GBp to USD on 1 April 2026, vendor price histories need explicit currency conversion before calculating decade-long P/E percentiles. I did not obtain a fully verified point-in-time dollar-normalised series, so the report deliberately avoids a spurious historical percentile.

A fourth limitation concerns FY2023 comparability. The most recent disclosures provide clean FY2024, FY2025 and H1 FY2026 acquisition-adjustment data, but I could not reconstruct FY2023 on an exactly consistent current USD/recast basis from the primary materials retrieved. I therefore use the later reconciliations to judge the recurring nature of acquisition exclusions instead of publishing an estimated FY2023 bridge.

The principal research sources are Compass Group's FY2025 Annual Report, H1 FY2026 results, Q3 FY2026 trading update, Q3 and half-year management transcripts, sectorisation materials, capital-allocation and remuneration disclosures, trading-currency announcement and company-compiled analyst consensus.

External cross-checks include Reuters reporting on the Vermaat transaction and the February 2026 AI-related selloff, MarketScreener for the verified 17 September dollar close, Sodexo and ISS primary FY2026 disclosures for horizontal analysis, Aramark's investor materials for business scope, Elior reporting for turnaround context and US Treasury data for the risk-free-rate check.

Other tickers mentioned

• SW.PA: Sodexo is Compass's closest global food-services and facilities-management competitor, with materially weaker FY2026 operating momentum.
• ARMK.US: Aramark competes directly in North American education, workplace, healthcare and sports food services.
• ISS.CO: ISS is an integrated facilities-services benchmark for large outsourced-contract retention, growth and cash economics.
• ELIOR.PA: Elior provides a smaller European contract-catering comparison focused on restoring contract profitability.
• SYY.US: Sysco is a food-distribution supply-chain benchmark rather than a direct catering competitor.
• USFD.US: US Foods is relevant as a North American food-procurement and inflation benchmark.
• PFGC.US: Performance Food Group is another distribution-side comparator for food-cost conditions.
• CTAS.US: Cintas provides a useful route-based service benchmark for recurring contracts, retention and capital discipline.
• RTO.LSE: Rentokil Initial is a contract-service comparison for recurring revenue and execution risk rather than a direct food-services peer.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.