IBM tops expectations as it struggles to bounce back from Q2 mainframe slump

IBM's stock surged over 4% in regular trading after the company reported better-than-expected third-quarter financial results, only for it to be erased in late trading. The company posted adjusted earnings of $2.95 per share, surpassing Wall Street's forecast of $2.86 per share. Revenue for the period increased by 5% from the same quarter last year, reaching $17.1 billion, just above the Street's target of $16.82 billion.

IBM's net income grew to $2.19 billion, up from $1.74 billion the previous year. CEO Arvind Krishna highlighted how clients are deploying 'Super Intelligence', transforming entire operating models, rather than just addressing narrow or novel problems. Software revenue rose 8% to $7.8 billion, while infrastructure revenue grew by 4% to $3.71 billion.

Consulting revenue remained relatively flat at $5.34 billion. However, AI's gross margin narrowed to 27.4% during the quarter, down from 28.9%, and adjusted gross margin stood at 58.5%, below the estimated 59.1%. CFO Jim Kavanaugh noted that AI now accounts for around 30% of consulting revenue and 35% of the overall order backlog.

The company's quarterly revenue growth, though modest, is still below the 9.5% growth seen in the first quarter. IBM reaffirmed its full-year revenue growth forecast of between 4% and 5%, but did raise its full-year free cash flow forecast by over $1 billion, up slightly from the previous July forecast. IBM's quantum computing business is a key focus, with the company recently securing a significant vote of confidence from the U.S. government for its progress towards building a useful, error-corrected quantum computer by 2033.

IBM also acquired quantum processor design startup HRL Laboratories LLC, which is developing a new quantum computing system slated to launch in three years. This system is expected to deliver 20,000 times more calculations than current quantum computers, potentially offering the 'quantum advantage' needed to solve complex, real-world problems that classical supercomputers cannot.