CGNX — Q2 2026 Financial Report Analysis

Cognex, which makes the cameras and software that factories and warehouses use to read barcodes, inspect parts and guide automated equipment ("machine vision"), posted record quarterly revenue of $291.3 million in Q2 2026, up 17% from a year earlier. The bigger story is profit: operating income doubled to $85.5 million because almost every extra dollar of sales fell straight through to profit. Gross margin rose to 70.6% from 67.4% on a better mix of end markets and higher volume, while operating expenses actually fell 3%.

At a glance

• Revenue $291.3M, +17% (+16% in constant currency): growth came from most end markets at once, led by Greater China (+47%) and the Americas (+27%), not from a single big order.
• Operating margin 29.4% vs 17.4%: revenue rose $42.2M and operating income rose $42.1M. Costs barely moved, so the extra sales became profit almost dollar for dollar.
• Diluted EPS $0.43 vs $0.24 (+79%): this is operating improvement, not financial engineering. The share count rose slightly and the tax rate went up, both working against EPS.

Key figures

Source: Form 10-Q for the quarter ended July 5, 2026; adjusted figures and quarterly cash flow from the Q2 earnings release (Form 8-K, August 5, 2026). Cognex closes its quarters on a Sunday, which is why the quarter ends July 5 rather than June 30.

Where the growth came from

Cognex reports revenue by where the customer is based, and the regions moved in very different directions:

"Constant currency" means the growth rate after stripping out exchange-rate moves, so it shows how much more the company actually sold.

Two things in the regional numbers need context:

• Americas and Europe partly offset each other. The 10-Q says the Americas benefited from "consumer electronics customer procurement changes as purchases shifted from entities based in Europe to the Americas." In other words, a large consumer electronics customer started buying through its US entity instead of its European one. That inflates the Americas growth rate and depresses Europe's. Europe's 15% constant-currency decline also reflects "ongoing softness in the automotive end market," which is a real weakness, not just a booking shift.
• Greater China was the strongest region on genuine demand, up 42% in constant currency on "higher revenue from customers in the consumer electronics and semiconductor end markets."
• Other Asia was held back by the April 1, 2026 sale of a Japan-focused trading business (acquired with Moritex in 2023), which removed some packaging revenue. Excluding currency, the region still grew 14% on semiconductor demand.

Why margins jumped

Three things combined:

1. Gross margin (the share of revenue left after the cost of making the products) rose 3.2 points to 70.6%. Management attributes this to "more favorable end-market mix and higher sales volume, which improved fixed-cost absorption." Put simply, factory overhead was spread over more units, and more sales came from higher-margin customers. The earnings release adds that "tariff refunds were not a material contributor."
2. Operating expenses fell 3% to $120.3M while revenue rose 17%. Research and engineering spending fell $0.7M and selling, general and administrative costs fell $3.5M. The company credits "savings from continued cost management activities and lower stock-based compensation expense." On a constant-currency adjusted basis, operating expenses fell 5%.
3. Operating leverage did the rest. Spending stayed flat while sales rose, so expenses fell from 50% of revenue to 41%.

What the headline numbers hide

• Cash conversion is good but working capital is absorbing cash. Year-to-date operating cash flow was $114.2M against $124.5M of net income. Accounts receivable (money customers owe) rose to $216.2M from $146.7M at the end of 2025. The year-to-date cash-flow statement shows a $68.5M receivables build, against $38.1M in the first half of 2025, while first-half revenue rose 20%. Part of this is a late-quarter sales skew, but it is worth watching that receivables keep pace with revenue rather than outrunning it. The offset is positive: deferred revenue and customer deposits rose to $49.0M from $21.1M, so customers are paying in advance for more future orders.
• The GAAP vs adjusted gap is small. Adjusted EPS of $0.45 vs GAAP $0.43 excludes only acquisition-intangible amortization ($0.01), reorganization charges of $1.3M ($0.01, mostly severance and consulting) and their tax effect. The year-to-date reorganization charge is larger at $6.4M, plus a $1.5M pre-tax loss on the Japan divestiture.
• The prior-year quarter included a one-off. Q2 2025 other income included a $2.1M Employee Retention Credit refund from the IRS. Without it, the comparison would be slightly more favorable to 2026.
• None of the EPS growth came from buybacks or tax. Cognex bought back $105.2M of stock in the first half, but employees exercised a large number of stock options (net proceeds of $132.7M), so diluted shares rose 0.8% to 170.0M. The effective tax rate rose to 19% from 16% as more profit was earned in higher-tax countries. Both worked against EPS, so the 79% EPS growth came entirely from the business.
• Guidance was beaten, then raised. In May, Cognex guided Q2 revenue of $280-300M, adjusted EBITDA margin of 28-31% and adjusted EPS of $0.40-0.44. It delivered $291M, 32.2% and $0.45, beating the top of the range on both profit measures.

Takeaway: Revenue grew 17% but operating income doubled, because costs stayed flat. The incremental margin was close to 100%. That is the clearest evidence yet that Cognex's cost program is lasting. The risk is the other side of that leverage: if a consumer electronics or semiconductor order cycle turns, profit would fall nearly as fast as it rose.

Outlook

Management's guidance from the August 5 release, in its own adjusted (non-GAAP) terms:

Q3 2025 included a one-time benefit from a commercial partnership with a medical lab automation channel partner. Excluding it, Q3 2025 revenue was $264M, so the Q3 midpoint implies about 17% underlying growth vs 12% reported.

Our read. The first half ($559.7M) plus the Q3 midpoint ($310M) leaves about $270M for Q4 at the full-year midpoint. That is roughly 7% above Q4 2025's implied ~$252M (derived from the reported $994M full year less the first three quarters), a clear slowdown from the 17-20% pace so far. Either management is building in caution after a strong consumer electronics and semiconductor stretch, or it expects those order cycles to cool. Q4 is where that question gets answered.

RealSense acquisition (announced after the quarter). On September 22, 2026, Cognex agreed to buy RealSense, the 3D depth-camera maker spun out of Intel in 2025, for about $500 million in cash. On top of that come a three-year cash retention program of $56.5 million at target and about $50 million of restricted stock for RealSense employees. Cognex says RealSense expects $80-90 million of 2026 revenue, more than 50% growth. Closing is expected in Q4 2026. The price is about two-thirds of the $755M of cash and investments Cognex held at quarter-end (it has no debt). The August guidance predates the deal and does not include it. The deal brings Cognex a faster-growing robotics business, but also integration and amortization charges that will widen the gap between GAAP and adjusted earnings from Q4 onward.

What to watch in Q3: whether revenue lands in the $300-320M range without another big shift in where a consumer electronics customer books its purchases; whether receivables growth slows to match revenue; and whether Europe's automotive weakness spreads to logistics or packaging.