Executive Summary / Key Takeaways

• Pension Derisking Is Accelerating and Material: Unisys has transferred over $500 million in U.S. pension liabilities through annuity purchases in 2024-2025, with management targeting $600 million by end-2026. This removes the single largest source of financial volatility that has historically obscured the company's operational performance and repelled institutional investors.

• Margin Expansion in Ex-L&S Business Is Structural, Not Cyclical: Despite revenue headwinds, Unisys has improved Ex-L&S gross margins by nearly 600 basis points over three years through AI-driven automation, workforce optimization, and disciplined contract selection. This demonstrates a fundamental shift from labor-augmented to technology-led delivery that could sustain profitability even if top-line growth remains elusive.

• License & Support Revenue Provides a Defensive Moat: The ClearPath Forward ecosystem is generating approximately $430 million in high-margin (70%+) revenue for 2025, with expectations of $400 million annually through 2028. This stable, annuity-like cash flow funds pension obligations and transformation investments while competitors face pressure in discretionary services.

• Scale Disadvantage Versus Peers Is Real but Niche Positioning Is Defensible: At $2 billion in annual revenue, Unisys is a fraction the size of DXC ($12B), IBM ($60B), and Leidos ($18B). However, its proprietary Stealth security micro-segmentation and ClearPath Forward platform create switching costs in government and financial services that pure-play services competitors cannot replicate.

• Revenue Growth Remains the Critical Uncertainty: While margins expand, all three segments face revenue pressure from delayed client spending, PC cycle headwinds, and government funding uncertainty. The investment thesis hinges on whether AI-driven solutions like the Service Experience Accelerator can catalyze new business signings and offset legacy revenue attrition.

Setting the Scene: A 150-Year-Old Company Reinventing Itself

Founded in 1873 and headquartered in Blue Bell, Pennsylvania, Unisys Corporation has survived multiple technology epochs by transforming from a hardware manufacturer into a services-led enterprise computing provider. The company generates revenue through three distinct business models: high-margin software license and support (L&S) contracts anchored by its ClearPath Forward platform; managed services in digital workplace and cloud infrastructure; and specialized business process solutions for government and financial services clients.

Unisys operates in a bifurcated market structure. On one side, it competes with scale-driven IT services giants like DXC Technology and IBM , who leverage global delivery footprints and hyperscaler partnerships to win cost-sensitive commercial contracts. On the other, it vies with specialized government contractors like Leidos and CGI for mission-critical public sector work. The company's unique position lies at the intersection of proprietary enterprise computing software and services—a combination that creates both opportunity and complexity.

The industry is experiencing three concurrent transitions: enterprise AI adoption shifting from pilot to production, hybrid cloud migration accelerating in regulated sectors, and IT budgets facing pressure from macroeconomic uncertainty. These forces create both headwinds—clients delaying discretionary projects—and tailwinds, as organizations seek to modernize legacy systems to support AI workloads. Unisys's challenge is capturing this demand while overcoming a scale disadvantage that limits its bargaining power with technology partners and constrains its ability to compete on price.

Technology, Products, and Strategic Differentiation: AI as a Delivery Model, Not a Feature

Unisys's competitive moat rests on two proprietary technologies that are increasingly valuable in an AI-driven environment. ClearPath Forward is a software-defined operating environment for high-intensity transaction processing that delivers mainframe-grade reliability with cloud-compatible deployment. This platform generated $241.7 million in L&S revenue through Q3 2025, with full-year expectations raised to $430 million—$40 million above original guidance. The "why" matters: ClearPath's "fairly fixed cost base" means incremental revenue flows directly to operating profit, creating 70%+ gross margins that fund the entire transformation effort.

Unisyst Stealth provides micro-segmented , identity-based security that enables zero-trust architectures without the performance degradation common in encrypted environments. This is not a generic cybersecurity overlay; it's embedded in the network fabric, making it particularly valuable for government agencies and financial institutions handling sensitive data. The technology underpins the company's ability to cross-sell CA&I solutions into its ECS client base, where workflow knowledge combined with security clearances creates switching costs that pure-play cloud providers cannot match.

The Service Experience Accelerator (SEA) exemplifies Unisys's AI-driven delivery transformation. Early client adopters are seeing 40% deflection from human to automated support, 28% higher user engagement, and 24% lower abandonment rates. This isn't incremental improvement—it's a fundamental re-architecting of service delivery from labor-intensive to technology-led. Management has invested in Salesforce's (CRM) Agentic AI to automate field service scheduling and rescheduling, continuously learning to optimize dispatch efficiency. The "so what" is profound: these capabilities shrink the size advantage historically held by larger competitors, enabling Unisys to compete for mid-market commercial clients ($1B-$5B annual revenue) where personalized service and less organizational complexity become differentiators.

The January 2025 organizational restructuring centralizes application development capabilities within CA&I and integrates business processing solutions across both ECS and CA&I. This creates synergies that management is already harvesting: CA&I maintained 19.6% gross margins despite a 4.8% revenue decline, benefiting from automation and workforce optimization. The application factory is yielding a growing pipeline, particularly in public sector platform modernization for criminal justice information and identity access management—areas where Unisys's domain expertise creates barriers to entry.

Financial Performance: Margin Expansion as Evidence of Strategic Shift

Unisys's Q3 2025 results appear alarming at first glance: a $308.9 million net loss versus $61.9 million in the prior year, driven by a $227.7 million pension settlement loss and $55 million goodwill impairment in DWS. However, these GAAP figures obscure operational progress that directly supports the investment thesis. The pension settlement loss is a one-time accounting charge from transferring $316 million in liabilities—strategically desirable derisking. The goodwill impairment reflects DWS's slower client signings but also management's discipline in not overpaying for growth.

The critical metric is Ex-L&S gross margin of 18.6% in Q3 2025, up 70 basis points year-over-year. This marks nearly 600 basis points of improvement over three years, achieved while revenue declined. Why does this matter? It proves the technology-led delivery model is working. As CEO Michael Thomson noted, "What we're adding is a higher margin addition" while lower-margin hardware and accreted-off contracts exit the mix. The company is actively shedding unprofitable revenue and replacing it with AI-automated services that generate superior unit economics.

Segment performance reveals the strategy in action. DWS revenue fell 4.3% to $125.3 million, but gross margin held steady at 16.2% despite PC cycle headwinds and Microsoft's Windows 10 extension dampening refresh activity. The company became an authorized Apple reseller, adding MacBooks and iPads to its device subscription service—expanding addressable market in the mid-market commercial sector where hardware choice drives decision-making. CA&I's 4.8% revenue decline to $180.4 million was offset by margin stability at 19.6%, achieved through workforce optimization and centralized application capabilities. ECS's 13.5% revenue drop to $133.2 million reflects timing of software license renewals, not demand destruction; specialized services and next-generation compute solutions grew 1.7% in Q3, and L&S revenue expectations for 2025 were raised to $430 million.

The balance sheet shows strategic financial engineering. The June 2025 issuance of $700 million in 10.62% Senior Secured Notes due 2031, combined with a $250 million discretionary pension contribution, helped reduce the U.S. pension deficit while extending debt maturity to 2031. Concurrently, shifting pension asset allocation to fixed income securities removes market and interest rate volatility from future contributions. CFO Debra McCann's guidance that contributions through 2029 "are not expected to fluctuate more than 3% in aggregate per annum" provides certainty that has been absent for decades.

Outlook and Guidance: Execution Risk on Revenue Recovery

Management's guidance frames a clear path to margin expansion but leaves revenue growth as the critical uncertainty. The company remains on track to meet or exceed the midpoint of its 8-9% non-GAAP operating profit margin guidance for 2025, implying a Q4 margin in the mid-teens due to L&S revenue concentration. Pre-pension free cash flow is forecast at $110 million for the full year, though McCann cautions that "the exact timing of some larger L&S collections could shift collections between the fourth quarter and first quarter of 2026."

The L&S revenue outlook provides a defensive floor. Full-year 2025 expectations of $430 million represent a $40 million increase from original guidance, driven by AI-related consumption and longer-duration contracts. Management projects approximately $400 million annually for 2026-2028, creating a $1.2 billion revenue base over three years at 70%+ margins. This annuity-like stream covers a significant portion of corporate overhead and pension obligations, de-risking the transformation effort.

However, revenue headwinds persist across all segments. DWS faces continued PC cycle pressure and Microsoft's Windows 10 extension, which "dampened a pickup in PC refresh activity." CA&I's public sector exposure suffers from "lingering effects of the government shutdown on project work" that management expects to persist "for a couple of quarters." The broader market dynamic shows clients "pausing or delaying project initiation" due to IT budget constraints, while competitive pricing pressure intensifies as rivals undercut on "aggressive assumptions for future AI-related efficiencies."

The key execution variable is whether AI-driven solutions can catalyze new business signings. Year-to-date new business TCV is "slightly positive relative to 2024," but the pace remains insufficient to offset legacy attrition. The Service Experience Accelerator's early success—40% deflection rates—must translate into scaled deployments. The application factory's pipeline in public sector modernization shows promise, but conversion to signed contracts has been slower than hoped.

Risks and Asymmetries: Where the Thesis Can Break

The most material risk is revenue growth failure. If AI-driven solutions like SEA and the application factory cannot generate sufficient new business to offset legacy contract attrition, Unisys will become a shrinking company with improving margins but declining enterprise value. The mechanism is clear: DWS's mid-market commercial opportunity remains "untapped," but penetration requires sales execution that has historically been uneven. CA&I's public sector pipeline is growing, but government funding uncertainty creates timing risk that could delay revenue recognition for multiple quarters.

Competitive pricing pressure threatens margin expansion. Thomson acknowledges that "clients want to share in AI cost savings" and competitors are "undercutting on price based on aggressive assumptions." If Unisys's disciplined approach—"willing to accept certain attrition, especially at clients prioritizing cost over value"—results in excessive market share loss, the margin improvement story collapses. The risk is asymmetric: larger competitors like DXC and IBM can absorb price cuts through scale, while Unisys cannot.

Pension obligation volatility, while diminishing, remains a tail risk. Any "material deterioration in the value of global defined benefit pension plan assets" or changes in legislation could require cash contributions beyond the $360 million forecast through 2029. The company's negative book value of -$4.21 per share reflects accumulated pension liabilities; while not operationally relevant, it limits financial flexibility and could trigger debt covenant issues if performance deteriorates.

Technology execution risk in AI transformation is underappreciated. The shift from labor-augmented to technology-led delivery requires sustained R&D investment and cultural change. If SEA's 40% deflection rates cannot be replicated at scale, or if Agentic AI implementations face integration challenges, the margin expansion thesis reverses. Competitors like IBM (with Watson) and ServiceNow (NOW) (with workflow automation) are investing heavily in similar capabilities, potentially leapfrogging Unisys's early progress.

Competitive Context: Niche Moats Versus Scale Disadvantages

Unisys's competitive positioning reveals a stark trade-off between specialized technology and operational scale. At $2 billion in annual revenue, it is one-sixth the size of DXC ($12 billion), one-thirtieth of IBM ($60 billion), and one-ninth of Leidos ($18 billion). This scale disadvantage manifests in higher customer acquisition costs, limited bargaining power with suppliers like Microsoft (MSFT) and Dell (DELL), and constrained ability to compete for mega-deals that drive growth at larger peers.

However, Unisys's proprietary technology creates defensible moats in specific niches. Stealth's micro-segmentation provides security efficiency that DXC's generalized cybersecurity suites cannot match, enabling faster deployment in zero-trust environments—a critical advantage for government agencies facing advanced persistent threats. ClearPath Forward's software-defined architecture delivers mainframe-grade reliability with cloud-native flexibility, appealing to financial institutions that cannot tolerate downtime during digital transformation. These capabilities drive the 70%+ gross margins in L&S and support premium pricing in services.

The competitive dynamics are most acute in DWS, where Unisys faces DXC and IBM in digital workplace services. DXC's scale enables materially lower delivery costs, while IBM's AI-infused platforms accelerate innovation cycles. Unisys counters with SEA's demonstrated 40% deflection rates and its Apple (AAPL) reseller partnership, which differentiates its device subscription service. The mid-market commercial sector ($1B-$5B revenue) represents a "significant untapped opportunity" where Unisys's personalized service and less complex organizational structure can win against larger competitors' bureaucracy.

In CA&I, CGI and Leidos dominate public sector cloud modernization. CGI's 9.7% revenue growth and Leidos's 7% expansion contrast sharply with Unisys's 4.8% decline. However, Unisys's application factory and multi-cloud expertise enabled a 7-year extension with a U.S. state government and a cyber vault solution introduction—demonstrating that domain expertise and security clearances can win sticky contracts despite scale disadvantages.

The key competitive asymmetry is pricing discipline. Thomson's explicit commitment to "stay disciplined in the contracts we're signing" means Unisys is willing to lose low-margin business to protect profitability. This contrasts with competitors chasing top-line growth through aggressive AI efficiency assumptions that may prove unrealistic. If Unisys's approach proves sustainable, it could capture a smaller but more profitable market share, while competitors face margin compression when promised AI savings fail to materialize.

Valuation Context: Distressed Pricing with Operational Improvement

At $2.93 per share, Unisys trades at a market capitalization of $209 million and an enterprise value of $656 million—approximately 0.34x trailing revenue and 5.06x EBITDA. These multiples place it at a significant discount to services peers: DXC trades at 0.43x revenue, IBM at 5.21x, CGI at 1.98x, and Leidos (LDOS) at 1.65x. The discount reflects two decades of pension overhang, inconsistent execution, and revenue decline.

The negative book value of -$4.21 per share and price-to-book ratio of -0.70 are artifacts of accumulated pension liabilities and goodwill impairments, not operational insolvency. More relevant is the pre-pension free cash flow guidance of $110 million for 2025, implying a 16.77% free cash flow yield on enterprise value—a figure that suggests either undervaluation or unsustainable cash generation. The truth lies in between: the free cash flow is real but must be allocated to pension contributions ($342 million in 2025), leaving limited discretionary capital.

The L&S business provides a valuation anchor. With $430 million in 2025 revenue at 70%+ margins, this segment alone could justify a significant portion of the enterprise value if valued as a software annuity. Assuming a 12x EBITDA multiple on $300 million of L&S EBITDA implies $3.6 billion of value—far exceeding the current EV. The discount reflects investor skepticism that services revenue can stabilize and that pension liabilities are truly contained.

Peer comparisons highlight the opportunity and risk. DXC's (DXC) 2.93% profit margin and 8.09% operating margin show that scale doesn't guarantee profitability in IT services. IBM's (IBM) 57.81% gross margin and 17.17% operating margin demonstrate what a software-led model can achieve—margins Unisys is approaching in its L&S business but not in services. CGI's (CGI) 10.42% profit margin and 16.76% operating margin represent a realistic target if Unisys can execute its AI-driven transformation.

The critical valuation variable is revenue stabilization. If Unisys can return to even low-single-digit growth while maintaining margin expansion, the current valuation would appear severely depressed. If revenue continues declining at 3-5% annually, the margin improvement becomes a value trap—profitable but shrinking.

Conclusion: A Show-Me Story with Asymmetric Upside

Unisys Corporation stands at an inflection point where financial engineering and operational transformation are converging. The pension derisking strategy—having transferred over $500 million in liabilities and extended debt maturities to 2031—removes the primary obstacle that has obscured the company's true earnings power for decades. Simultaneously, AI-driven automation has delivered nearly 600 basis points of Ex-L&S margin expansion, proving that technology-led delivery can overcome scale disadvantages.

The investment thesis hinges on a simple but critical question: can AI-driven solutions like the Service Experience Accelerator and application factory generate sufficient new business to offset legacy revenue attrition? The L&S business provides a defensive floor—$1.2 billion of high-margin revenue through 2028—that de-risks the transformation effort, but services revenue must stabilize for the equity to re-rate.

The asymmetry lies in the valuation. At 0.34x revenue and a 16.77% free cash flow yield, the market prices Unisys as a melting ice cube. If management executes on its 8-9% operating margin target while achieving even modest revenue growth, the multiple expansion potential is substantial. If revenue continues declining, the margin improvement becomes a value trap. For investors, the critical variables to monitor are new business signings growth in DWS and CA&I, public sector funding normalization, and competitive pricing discipline. The pension albatross is being slain; now the company must prove it can grow.