XBP Global HoldingsXBP
Recorded

XBP Global Holdings 2026 Q3 Earnings Call

Review the key takeaways and the transcript of this earnings call.

PeriodQ3 2026Duration34 minParticipants6

Transcript

Preview the first fifteen paragraphs, organized by speaker.

Operator

today, and thank you for standing by. Welcome to the XBP Global Third Quarter 2026 Financial Results Conference Call. At this time, all participants are in a listen-only mode. After the speaker's presentation, there will be a question-and-answer session. To ask a question during the session, you will need to press star one one on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star one one again. Please be advised that today's conference is being recorded. I would now like to turn the conference over to your speaker for today, David Shamis, Head of Investor Relations.

David ShamisVP of Investor Relations and Corporate Finance

Please go ahead. Thank you and good afternoon, everyone.

David ShamisVP of Investor Relations and Corporate Finance

Welcome to XBP Global's Third Quarter 2026 Earnings Call. Joining me are Chief Executive Officer Andrej Jonovic and Chief Financial Officer Dejan Avramovic. Before we begin, please note that today's remarks may contain forward-looking statements, including statements regarding our future performance, outlook, and strategy. These statements are based on current expectations and assumptions and are subject to risks and uncertainties that can cause actual results to differ materially from those described. For a detailed discussion of these risks and uncertainties, please refer to our filings with the Securities and Exchange Commission, including our most recent annual report on Form 10-K and our proxy statement and other filings with the SEC, copies of which are available on our investor relations website at investors.xbpglobal.com. During this call, we will also reference certain pro forma and non-GAAP financial measures.

David ShamisVP of Investor Relations and Corporate Finance

Reconciliations of these non-GAAP measures to the most directly comparable GAAP measures are included in our earnings release and the appendix to our investor presentation, which are available on our investor relations website. With that, I will turn the call over to Andrej.

Andrej JonovicCEO

Good afternoon, everyone, and thank you for joining us today. Before I walk through the quarter, I want to frame my results around three themes. Number one, where revenue stands as last year's client exits finish running off. Two, the margin expansion our AI-first operating model is delivering. And three, how our pipeline is evolving. First theme, the revenue. Third quarter revenue was $182.8 million, down 12.6% year-over-year as reported, and 17.1% on a pro forma basis, and down 4.5% from the second quarter. The sequential decline came from two places: the tail end of legacy client exits and volume reductions that were mostly known to us last year, and lower platform revenue in the Americas in our Technology segment. A majority of the year-to-date decline was expected and reflects exits we already knew about rather than new client losses. That runoff is still in its tail.

Andrej JonovicCEO

As we said in August, we expect revenue to inflect in the second half, and we expect that to show in the fourth quarter, with revenue stepping up from the third quarter. The second theme, our margin expansion. Our margins are expanding because of how we're using AI. I want to start with what AI means for our business. Across our industry, the key question is, what happens to a services business when AI does the work people used to do? Clients are moving away from paying for headcount and toward paying for transactions and outcomes. Contracts are being repriced at renewal, and some clients are automating work themselves. The lesson across the industry is simple. The service provider that owns the automation keeps the client. We would rather lead the transition with our clients than have it done to us.

Andrej JonovicCEO

Let me explain how we do that and then where it is showing up. Turning to slide five of our investor deck, which shows the intelligence pipeline many of our workflows run through. We're actively migrating legacy rules-based workflows into agentic AI pipelines. At the start of a client deployment, we typically auto-resolve 40%-60% of transactions in the first pass. This rises to 85% or more as models mature. The more the system resolves on its own, the less labor each transaction needs. We've built the technology first and wrapped the service around it, not the other way around. We track the AI-first impact in our enterprise across three key areas. Number one, through our internal functions. Two, our delivery teams. And three, the solutions we build for clients. I'll take each in turn.

Andrej JonovicCEO

In our internal functions, a great example is software delivery, where AI agents now handle much of the work from requirement to deployment, with our people approving the work at defined checkpoints. Delivery cycles have gone from days to hours, and headcount across our Technology and IT functions is down about 30% this year. For our delivery example, I would like to talk about Healthcare Flex, which is our AI pipeline for healthcare workflows, where we typically process healthcare insurance claims for our healthcare payer clients. In the past, this work relied on large teams of trained staff who had to learn every form and rule, which took considerable time and investment. Today, AI reviews each claim first and processes it automatically when it is confident. When it is uncertain, a staff member reviews only the part that needs attention.

Andrej JonovicCEO

As a result, we can handle the same volume with fewer staff and much less training and supervision. Accuracy remains high because staff still review anything the AI is unsure about, and the AI keeps learning from the staff's corrections. We are now faster than we were. Our straight-through processing rate is higher than it has been, with further improvements as the models continue to be sharpened. Close to half of the cost benefit from this migration comes from support functions, such as training, quality, and production support, which a trained staff model needs and Healthcare Flex largely doesn't. People still review every point where the system is unsure, so accuracy holds while costs continue to fall. The third area we track is client solutions, where clients are now buying the same model. A year ago, many enterprise conversations were about AI experimentation.

Andrej JonovicCEO

Today, clients want production-grade AI running inside their own environment, under their own controls, but with human oversight. That is particularly true in heavily regulated sectors like healthcare, financial services, and the public sector, where precision, auditability, and data sovereignty are non-negotiable. Clients are discounting broad AI promises and rewarding partners who can show real outcomes in production. A good recent example of a client solution is British Airways, a deal we announced in Q3. It also illustrates the four controls on slide six of our investor deck. That slide focuses on how we're able to deploy AI to suit our clients' needs, including where the AI runs, which models are used, how the work gets done, and ensuring humans retain decision authority. We've managed British Airways' engineering records archive for more than 20 years.

Andrej JonovicCEO

It documents every maintenance action on every part across a fleet of more than 250 aircraft. To give you a sense of scale, the documentation for a single Airbus A380 aircraft runs to about 400,000 pages. The information in that large archive has always been complete. The difficulty was retrieving it. The archive was indexed on a small number of fields, so locating a record depended on engineers who knew both the aircraft and the archive, and this took a lot of time. We're deploying our Plexus AI solution to address this gap. Our agentic AI reads the existing records, including handwriting and inspection stamps, and vectorizes them into a private database without rescanning the physical archive. On infrastructure, the solution runs inside British Airways' private cloud, so the airline's data stays within its own environment.

Andrej JonovicCEO

On models, we use an open-weight model selected to meet the airline's business and regulatory requirements, which also avoids the usage-based token charges of a third-party model. On workflow, AI agents manage access to the data within defined permissions, and the search structure is configured by British Airways' own engineering organization. Finally, on decisions, the AI finds the evidence, and the airline's engineers make the call. The practical benefit is straightforward. Engineers can search in their own terminology and receive every relevant record. When an aircraft or a major component is sold or comes off lease, the buyer or new operator requires its complete maintenance history. We expect to reduce the time needed to assemble and verify that documentation from weeks to hours, which speeds up the airline's own transactions. It also allows British Airways to consolidate some of its legacy software.

Andrej JonovicCEO

We won this contract on the strength of a longstanding relationship, domain knowledge, and technology we could demonstrate on the client's own critical records. This all translates into profitability. Normalized EBITDA for the quarter was $22.6 million, up from $21.9 million in the second quarter and $15.6 million in the first, our second consecutive quarter of sequential growth, even as revenue declined. Our normalized EBITDA margin expanded to 12.4%, up 90 basis points sequentially and 120 basis points year-over-year, a record high for the combined company. Returning now to our three themes and the last one, our pipeline. Bookings were solid in the quarter, with total contract value of $147.5 million and new ACV of $37.5 million, both well ahead of a year ago.

Andrej JonovicCEO

That said, bookings can be lumpy from quarter to quarter, but it is encouraging that the new TCV we've booked in the first nine months of the year already exceeds our total for all of 2025. Regarding the pipeline, what matters to us is the kind of work we're winning and focusing on. AI-led engagements, often with clients who already trust us with their data. There is a lag between signing and revenue for these engagements, which require implementation and integration into the client's environment before they ramp. Skipping ahead to slide eight of our investor deck, which summarizes the quarter. To what I've already covered, I would add just one item. Under our AI-enabled operational efficiency program, we're also raising our targeted annualized operational efficiencies to the $70 million-$75 million range, up from $60 million-$75 million, as we continue to identify additional opportunities across the organization.

Andrej JonovicCEO

In short, the work we've done to rebuild the business is now showing up in bookings, pipeline quality, and margins. Our job now is converting that momentum into revenue. With that, I'll turn the call over to Dejan Avramovic, our CFO.

FULL TRANSCRIPT

Continue the full translated transcript in StockNow.

Log in to unlock every statement, the English original, and speaker-by-speaker history.

Log in for the full transcript

Other earnings calls that day

View calls from this day