IBM will host its conference call to discuss fourth-quarter 2025 results on Wednesday, January 28, 2026 at 5:00 p.m. ET. The live webcast will be available on IBM’s Investor Relations site, with a replay, charts and prepared remarks posted shortly after the call. This event follows a year in which IBM’s Cloud & Cognitive Software segment grew revenue by 8% year-over-year, contributing to total company revenue of $60.1 billion in the first nine months of 2025. Investors will be watching for guidance on full-year 2026 earnings per share, capital expenditure plans—which reached $4.3 billion in 2025—and any updates to the company’s targeted free cash flow margin of 15% to 18%.

IBM has expanded its collaboration with Datavault AI to deliver watsonx-powered, zero-trust micro data centers that enable secure, ultra-low-latency AI processing at the network edge. The companies plan to deploy more than 500 micro data centers by mid-2026 across manufacturing, utilities and retail clients in North America and Europe. Each unit integrates IBM’s Power10 processors and Datavault’s hardened hardware, reducing data-transit delays by up to 90% compared with cloud-only architectures. IBM projects that edge AI solutions will contribute $2 billion in incremental revenue by 2027 and support an operating margin expansion of 50 basis points in its Infrastructure segment.

Although IBM shares have rallied approximately 18% over the past 12 months—outperforming the S&P 500’s 12% gain—the company still trades at a forward price-to-earnings multiple near 14 times consensus estimates, below the 26 times multiple for larger AI-focused peers. Analysts project IBM’s revenue to grow at a 6% compound annual rate from 2025 to 2028, supported by hybrid cloud and AI services, while operating margins are expected to expand by 200 basis points over the same period. The relatively attractive valuation and strong free cash flow yield of 5% may appeal to income-oriented investors, even as higher-growth competitors capture market share in generative AI workloads.