Editorials

RTX's third-quarter 2026 operating income rose faster than revenue, which is the expected shape of operating leverage. The arithmetic, though, is less simple than that. Revenue barely moved in Q3, so the top line cannot explain the margin gain. The gain came from a cost line, and SG&A worked against it. All figures below are from 10-Q line items for the three FY2026 quarters; derived figures are labeled as such.

Revenue grew about 11.9% from Q1 to Q2, then only about 1.0% from Q2 to Q3. Operating income grew about 10.0% in the first step and about 18.7% in the second. Operating margin slipped slightly in Q2 despite strong top-line growth, then expanded about 200 basis points in Q3.

The incremental numbers make the point. Q3 revenue rose by about $257M over Q2. Operating income rose by about $526M. That is an incremental margin of roughly 200%, which volume alone cannot produce. In Q2, the comparable figure was about 10% ($256M of operating income on $2.63B of added revenue).

A simple bridge from Q2 to Q3 (USD M, derived from the line items above):

• Revenue: +257
• SG&A: −260 (SG&A rose more than revenue)
• R&D: +15 (R&D fell slightly)
• Residual operating costs (revenue less operating income, SG&A and R&D, i.e. mainly cost of sales and other operating items): +514 (residual costs fell from about 79.0% of revenue to about 76.1%)
• Operating income: +526

So the Q3 gain came from the residual cost line, not from growth and not from SG&A. The reported data does not break that residual into cost of sales versus other items, so this is a derived attribution, not a reported one. A reader should treat "cost discipline" here as a statement about cost of sales and other operating costs, not about overhead.

A Buffett-style test asks whether a company is getting more output per dollar of overhead. On that test, Q3 is mixed. SG&A stepped up in both Q2 (+$182M) and Q3 (+$260M), while revenue was nearly flat in Q3. That is the opposite of operating leverage in overhead. R&D, by contrast, fell slightly in Q3, so the company was not spending more on engineering in the quarter when margins expanded.

Whether the SG&A increase is investment, timing, or inflation is not answered by the reported data. The FY2025 MD&A summary cites inflation, tariffs and supply-chain disruption as ongoing margin pressures, but it does not discuss Q3 2026 cost lines, and no FY2026 segment data is available to show where the cost-of-sales improvement sits.

Diluted EPS rose from $1.57 in Q2 to $1.89 in Q3, about 20%. That is faster than operating income growth in the same quarter, but the difference is small relative to the operating story. Weighted diluted shares moved only from about 1,365M to 1,368M, so buybacks did not drive the EPS change. Below operating income, non-operating expense was roughly flat (−$69M in Q2, −$36M in Q3), and the Q3 tax charge was $593M on pre-tax income of $3,301M.

Three questions decide whether Q3 is a lasting shift in economics or a single-quarter event:

1. Is the residual cost decline repeatable? A roughly three-point fall in cost of sales and other operating costs as a share of revenue is large for one quarter. The 10-Q's cost-of-sales detail and any one-time items in MD&A would show whether it is structural (pricing, aftermarket mix, program performance) or timing.
2. Does SG&A stabilize? If overhead keeps growing faster than revenue, the margin gain depends entirely on the cost-of-sales line.
3. Does revenue re-accelerate? Q3 growth of about 1% after about 12% in Q2 is a sharp slowdown. Backlog (FY2025 MD&A cites $268B) is the long-run support for growth, but it does not show up in quarterly revenue on its own.

This piece describes what the numbers show. It does not state a view on RTX's valuation or whether the shares are attractive at any price.