Earnings Season Tracker

Mike Dastoor ran this call like a man who knows his numbers are real. No hedging on the headline: $44.5B in FY2027 revenue, up 24% on a base that already grew 21%, with AI-related revenue jumping from $14.4B to $22.1B — that is not a continuation, that is acceleration. Greg Hebard was equally clean: Q4 revenue came in $1B above the June midpoint, free cash flow beat the initial full-year outlook by more than $200M, and the balance sheet sits at 1.3x debt-to-EBITDA. The detail that the Street should be sitting with is the customer concentration shift inside Intelligent Infrastructure — Jabil went from one hyperscaler above $1B in revenue two years ago to five in FY2027, with the second hyperscaler alone expected to cross 10% of total company revenue. That is a structural change in the customer mix, not a good quarter. The ripple outward: Hanley Energy's acquisition is being positioned as a margin-accretive services wedge — deployment and maintenance revenue on top of manufacturing — which starts to look like a data center services play that puts Jabil in partial competition with infrastructure services firms over a 3-5 year horizon. Meanwhile, Frank McKay's prepared remarks on memory tightening across non-AI end markets was the most important supply chain signal on the call — HBM and DRAM being reallocated toward hyperscale AI means companies in healthcare, automotive, and industrial automation are fighting for allocation. Jabil's 120 strategic supplier relationships are a genuine moat here, not a talking point.

Manish Bhatia ran this call with the measured confidence of someone who has already seen the order book and likes what is in it. The headline from Q4 2026 is not the quarter itself but the forward posture: Micron now says supply and demand will remain tight through 2028 with no line of sight to when balance arrives. That is a stronger statement than anything the company offered three months ago, and Bhatia backed it with specifics. More than 75% of fiscal 2027 shipments are already committed. Ten new supply commitment agreements were signed this quarter, bringing the total to 36, and newer contracts are priced to current tight-market conditions rather than the softer Q2 pricing that underpinned earlier deals. The NVHBM co-design announcement with NVIDIA is the call's most consequential detail and the one most likely to be underappreciated by the Street. Scott DeBoer described it as the first major custom HBM product to market, developed over more than a year on a foundry process. This is not an incremental product refresh. A co-designed, NVIDIA-specific HBM4E built to a bespoke specification gives Micron a product that SK Hynix and Samsung cannot exactly replicate without their own equivalent co-design relationship. The second-order read here is that TSMC and Samsung Foundry are now embedded in the HBM supply chain as logic-layer partners, which reshuffles competitive dynamics in advanced packaging. Mark Murphy closed the call by volunteering guidance details that no analyst had yet asked for, which is what a CFO does when he wants the market to have the numbers and is comfortable with where they land.

Sanoke Viswanathan ran this call with genuine energy, not the rehearsed optimism you hear when a CEO is papering over cracks. The headline number is historic: $86 million of organic ASV in a single quarter, the largest in FactSet's (NASDAQ:FDS) history, and full-year organic ASV growth of 7% that beat the top of guidance by a meaningful margin. Josh Warren was more careful with language, particularly around the consumption-revenue divergence and the Q1 free cash flow warning, but neither man sounded defensive. The more consequential disclosure was buried in the MCP adoption figures: API call volume grew seven times in Q4 over Q3, and Q3 itself had grown thirteen times over Q2. That is not a product launch curve, that is something closer to a network effect igniting. The chain reaction worth tracking here runs directly through the large-language-model providers. Every time an enterprise deploys a frontier model and plugs FactSet's MCP connector in, FactSet captures a new data consumption relationship that didn't exist before, often without displacing its own workstation seat. That dynamic is validating a two-sided growth engine: MCP pulls in new logos that then convert upward into broader enterprise contracts. The implication for Bloomberg's terminal business is uncomfortable: FactSet is now winning mandates at bulge-bracket banks and the five largest Canadian wealth managers simultaneously, and the weapon it is using is open-architecture data delivery that Bloomberg's proprietary model structurally cannot match.