OVHcloud Sacrificed Short-Term Gains to Secure Its Future
Some companies publish results to reassure their shareholders; OVHcloud published theirs to declare a war.
On April 9, 2026, the French cloud computing company confirmed an adjusted EBITDA margin of 40.9% for the first half of its fiscal year 2026, the highest level since its IPO. Revenues reached €555.3 million, with organic growth of 5.5%. Public Cloud grew 15.1% on a comparable basis. These are numbers any management team would frame on the wall.
However, in the same report, leveraged free cash flow deteriorated to negative €14.2 million, capital investment reached 42.9% of revenues, and net debt rose to €1.13 billion. This is not a balance sheet celebrating success; it’s the balance of a company that made a painful and conscious decision: sacrificing liquidity today to avoid a loss of position tomorrow.
From my perspective as a strategist, that deserves more than just a headline.
Record Margin Conceals a High-Risk Bet
The superficial narrative would suggest that OVHcloud is thriving. The EBITDA margin increased from 40.0% in the same prior period. Revenues surpass half a billion for the semester. The CEO, Octave Klaba, explicitly confirmed it: "We achieved an adjusted EBITDA margin of 40.9%, a record level since our IPO."
The more uncomfortable, and therefore more useful, narrative is different.
OVHcloud decided to front-load massive hardware purchases in response to inflationary pressures on components. This bumped capex to 42.9% of revenues in the first half, up from 36.0% in the prior period, resulting in a 12.2% drop in unleveraged free cash flow to €32.3 million. Net income fell to €5.9 million from €7.2 million. The net operating margin decreased to €35.4 million from €42.4 million.
This dynamic has a precise name in strategic analysis: cost concentration to secure a competitive position before the market allows it. OVHcloud is not mismanaging its cash. It is purchasing infrastructure capacity at prices it deems reasonable compared to what is to come. If this market read is correct, capex will decrease to 30-32% of revenues projected for the full year, and leveraged free cash flow will turn positive. If incorrect, the €1.13 billion debt at 2.6 times leverage will begin to weigh heavier.
What I find strategically honest is that the company is not pretending this reality doesn’t exist; it names it.
The Sacrifice No One Reports
Behind the growth numbers lies a far more revealing decision than any profitability metric: OVHcloud is choosing whom to grow with and, therefore, whom not to.
The acquisition of Dragon LLM to create a vertical service-oriented AI lab is not a generic move toward "AI." It is a specific bet on building proprietary capabilities rather than relying on the ecosystem of large American providers. This has an enormous opportunity cost: every euro invested in its own development is a euro not going into capturing medium-volume clients, where AWS, Azure, and Google Cloud compete with aggressive pricing and infinite catalogs.
Establishing a dedicated defense unit aimed at European ministries is even more selective. There is no tougher market in terms of data sovereignty, long sales cycles, and specific technical requirements. OVHcloud is betting on clients that very few providers can serve, precisely because that segment demands what American hyperscalers cannot offer for regulatory and political reasons.
These two initiatives are not additional business lines launched to capture opportunities. They are the operational declaration of whom this company serves and consequently whom it chooses not to serve. A company attempting to also capture the high-volume startup market, the mass consumer segment, and large multinational corporations without geographical restrictions with the same capital and structure would simply be less effective at everything. OVHcloud is doing the opposite: narrowing its focus to deepen its advantage where the terrain favors it.
That sacrifice comes at a cost, and it is the most strategic aspect of this report.
Debt Is the Clock No One Can Ignore
There’s one element that the enthusiasm for the record margin tends to push to the bottom of the page: €1.13 billion in net debt with a leverage of 2.6 times. The company has €242 million in available liquidity with coverage until 2030. There is no imminent debt crisis, but a real constraint limits how long the transformation plan has to show results.
Return on capital employed grew 2.9 percentage points from 2021 to 2025, indicating that the capital structure is being used with greater discipline. Net income improved from a loss of €10.3 million in the 2024 fiscal year to €0.4 million positive in 2025. The trajectory is correct, but speed matters as much as direction when the debt clock ticks in parallel.
The projection for the full fiscal year 2026 includes organic growth of 5 to 7%, EBITDA margin above 40.4% from the previous year, capex in the range of 30-32%, and positive leveraged free cash flow. If those four variables are to be met simultaneously, the transformation narrative has mathematical backing. If capex does not decrease as projected, or if organic growth remains at the lower limit of the range, the margin for maneuver tightens significantly.
There is a difference between a company managing a calculated bet and a company racing against its own balance sheet. For now, data points to the former. The upcoming second-quarter results for fiscal 2026, expected between May and June, will reveal whether execution aligns with the design.
Focus Is the Only Advantage No One Can Buy
Amazon, Microsoft, and Google can match any price OVHcloud sets on standard infrastructure. They can replicate technical capabilities faster and with more capital. What they cannot buy is the sovereign credibility of a European company in front of a German defense ministry or a French regulatory institution that needs guarantees that no contract with an American provider can structurally offer.
That advantage does not stem from technology; it comes from opting out of competition where the big players are unbeatable to build a position where the big players are ineligible. The in-house AI lab supports that same logic: if the agentic product they offer contains third-party components controlled from Silicon Valley, the promise of sovereignty dissipates. Building capacity internally, even if it's slower and more expensive, is the only way to uphold the value proposition in the chosen segment.
OVHcloud has a clear guiding policy: to be the benchmark provider for European entities that cannot or do not want to depend on infrastructure controlled outside Europe. Every action reported this semester, including those that pressure cash flow, points in the same direction. This is strategic coherence, and coherence in an industry where capital dispersion is the norm is scarcer than it appears.
The C-Level that succumbs to the temptation to "also capture" the standard corporate market, to compete on price with AWS in segments where sovereignty is not a decision criterion, or to build a service portfolio so broad that the message dilutes, will destroy in months what this architecture took years to build. The only way to sustain a 40.9% margin with €1.13 billion in debt is for each action to reinforce the same bet. The discipline of not growing where there is no structural advantage is not an abstract virtue; it is the condition for the survival of the plan.
