Texas Instruments’ miss is not the story; the margin trough and capex reset are

The print says less about end-market demand disappointment than about where the investment debate has migrated: from whether revenue is recovering to whether Texas Instruments can turn that recovery into cash returns while carrying new fab depreciation. What was priced in was a roughly in-line quarter, with the Street at $4,437.0 million and $1.29, and management’s own language matched that broad expectation when Haviv Ilan said, “Revenue came in about as expected at $4.4 billion, a decrease of 7% sequentially and an increase of 10% from the same quarter a year ago.” What actually surprised was the small negative spread to consensus, with actual revenue of $4,423.0 million versus estimate $4,437.0 million for a -0.3% surprise, and EPS of $1.27 versus estimate $1.29 for a -1.6% surprise. The market’s likely mistake is to treat the miss and 55.9% gross margin as proof that the analog cycle is losing steam. The numbers point to a different conclusion: the company is already growing year on year at +10.4% in Q4 FY2025, the March quarter in the history shows $4,825.0 million, +9.1% sequentially and +18.6% year on year, and the gross margin line has already moved to 58.0%. That combination makes the Q4 disappointment a cost-absorption and seasonality event, not a demand rollover.

That distinction matters because the quarterly trajectory does not support a bearish demand read. Revenue rose from $4,069.0 million in Q1 FY2025 to $4,448.0 million in Q2 FY2025 and $4,742.0 million in Q3 FY2025 before falling to $4,423.0 million in Q4 FY2025, a sequence of +1.5%, +9.3%, +6.6%, and -6.7% sequential movement. The year-on-year line improved from +11.1% in Q1 FY2025 to +16.4% in Q2 FY2025, +14.2% in Q3 FY2025, and +10.4% in Q4 FY2025. A company losing cyclical traction does not normally follow a Q4 revenue base of $4,423.0 million with Q1 FY2026 revenue of $4,825.0 million and +18.6% year-on-year growth. The top-line miss versus Street was real, but at -0.3% it is not the basis for a thesis change. The more important read is that the analog cycle is broad enough to absorb a seasonal Q4 drop while keeping the next quarter above every quarter in the provided history, including $4,742.0 million in Q3 FY2025.

The margin debate is where the print looks worse at first glance, and also where the market may be extrapolating the wrong thing. Gross margin fell to 55.9% in Q4 FY2025 from 57.4% in Q3 FY2025 and 57.9% in Q2 FY2025, while EPS fell to $1.27 from $1.48 in Q3 FY2025 and $1.41 in Q2 FY2025. That is the bear case in one sentence: revenue slipped -6.7% sequentially and gross margin compressed to the lowest point in the FY2025 sequence. But the broader table cuts the other way. Gross margin was 56.8% in Q1 FY2025, 57.9% in Q2 FY2025, 57.4% in Q3 FY2025, 55.9% in Q4 FY2025, and 58.0% in Q1 FY2026. The 55.9% number matters, but the recovery to 58.0% on $4,825.0 million suggests the Q4 margin trough was not the new run rate. The market may be missing that the revenue recovery is already showing up with operating leverage, even as depreciation from capacity additions becomes more visible.

The capacity story explains why the EPS reaction can be harsher than the revenue reaction. Rafael R. Lizardi put a hard boundary around the next leg of the investment cycle when he said, “We continue to expect CapEx for 2026 between $2 and $3 billion.” That phrase is not a generic spending comment; it reduces the risk that 2025’s $4.6 billion of capital expenditures becomes the normalized forward burden. In 2025, cash flow from operations was $7.2 billion and capital expenditures were $4.6 billion, while free cash flow was $2.9 billion or 17% of revenue, an increase of 96% from 2024. In Q4 alone, cash flow from operations was $2.3 billion and capital expenditures were $925 million. The variant perception is that the Street is still debating the old capex cycle while the company is already pointing to a lower capex envelope for 2026. Against that, depreciation does not vanish: management now expects $2.2 to $2.4 billion on depreciation in 2026. That is the conflict to underwrite. Capex moves from $4.6 billion in 2025 to a stated 2026 range of $2 and $3 billion, while depreciation rises into a $2.2 to $2.4 billion range. The equity argument depends on free cash flow improvement outrunning depreciation pressure.

That cash-flow framing also changes the interpretation of capital returns. The company paid $1.3 billion in dividends and repurchased $403 million of stock in the quarter, while increasing the dividend per share by 4% to $1.42 per share. Total debt outstanding was $14 billion with a weighted average coupon of 4%, and inventory ended the quarter at $4.8 billion, down $25 million from the prior quarter. None of those figures points to a balance sheet crisis, but neither do they allow a lazy “shareholder return” victory lap. The point for PMs is tighter: if capex falls into the $2 and $3 billion range while quarterly cash flow from operations remains near the Q4 figure of $2.3 billion, the stock’s debate shifts from “fab overbuild” to “how much cash returns can grow without stretching the balance sheet.” If revenue fades back toward Q4’s $4,423.0 million while depreciation reaches $2.2 to $2.4 billion in 2026, the 55.9% gross margin print becomes a warning instead of a trough.

The demand mix gives more support to the constructive version of the thesis than the headline miss does. Management’s own end-market framing showed Industrial at $5.8 billion, up 12% year on year and 33% of revenue; Automotive at $5.8 billion, up 6% year on year and 33% of revenue; Data center at $1.5 billion, up 64% year on year and 9% of revenue; Personal electronics at $3.7 billion, up 7% year on year and 21% of revenue; and Communications equipment at about $500 million, up about 20% year on year and 3% of revenue. The striking point is not that data center is the biggest business, because at 9% of revenue it is not. The point is that the fastest-growing bucket, up 64% year on year, is additive to two large analog anchors that are still growing, with Industrial up 12% and Automotive up 6%. That mix reduces the probability that a single consumer or server pocket is carrying the recovery. It also limits the read-through to AI-adjacent enthusiasm: data center can help the multiple, but at 9% of revenue it cannot fully offset weakness in Industrial and Automotive if those two 33% categories roll over.

That customer and supplier read-through is specific enough to matter. For Arrow Electronics, the distributor read is that Texas Instruments’ Q4 inventory was $4.8 billion, down $25 million from the prior quarter, while revenue was $4,423.0 million and year-on-year growth was +10.4%; that combination suggests sell-in was not being supported by a visible inventory build at the manufacturer level. For National Instruments, the test and measurement link is tied less to Q4 revenue than to the capital plan: Texas Instruments spent $925 million in Q4 capital expenditures, $4.6 billion in 2025 capital expenditures, and still expects 2026 CapEx between $2 and $3 billion, leaving a smaller but still material equipment envelope. For UMC, the analog and specialty wafer implication is positive but capped by Texas Instruments’ internal capacity strategy: Q1 FY2026 revenue of $4,825.0 million and +18.6% year-on-year growth support specialty analog demand, while 2026 depreciation of $2.2 to $2.4 billion signals that more internal fab capacity is moving through the P&L. For JCET, the IC assembly and test read is volume-linked: Q4’s -6.7% sequential revenue decline was followed in the history by Q1 FY2026 revenue of $4,825.0 million and +9.1% sequential growth, so the near-term read-through is a rebound in assembly/test activity rather than a continued Q4 downdraft.

The peer comparison reinforces the same split between revenue recovery and margin quality. In the latest reported quarter table, TXN shows $4,825.0 million of revenue, 58.0% gross margin, and +18.6% revenue YoY. That gross margin is higher than NXPI at 56.2%, INTC at 39.4%, IFNNY at 38.7%, STM at 33.8%, 6724.T at 35.0%, RNECY at 51.2%, and 6758.T at 30.8%. The relative revenue growth picture is not best in group, with RNECY at +25.4% and STM at +22.8%, but TXN’s +18.6% is above NXPI at +12.2%, INTC at +7.2%, IFNNY at +7.9%, 6724.T at +8.9%, and 6758.T at +15.4%. That matters because the bear debate on Texas Instruments often treats the company as choosing margin durability over growth. The latest peer set shows both 58.0% gross margin and +18.6% revenue YoY. The market does not need to pay for a hypergrowth analog story; it only needs to stop penalizing the shares as if the capacity build permanently impaired through-cycle gross margin.

The call delivery was better on surface sentiment but not clean enough to ignore. The tone history shows Q4 FY2025 sentiment at 0.13, up from 0.10 in Q3 FY2025, with guidance_tone at 0.34 versus -0.21 in Q3 FY2025; the Q1 FY2026 call then moved sentiment to 0.19, while guidance_tone fell to 0.13. The call-over-call delta from Q1 FY2026 versus Q4 FY2025 was sentiment +0.07, guidance_tone -0.22, tone_confidence -0.01, prepared_sentiment -0.05, qa_sentiment -0.01, ai_optimism -0.08, uncertainty -3.8, and qa_evasiveness +10.6. That combination is mixed in a very specific way: management sounded more positive overall and uncertainty fell, but guidance language became less positive and Q&A evasiveness rose. The market should not treat the tone data as a green light. It says the prepared narrative improved around recovery and capex, while investor pushback on the moving pieces, depreciation, offsets, and cash returns became harder to answer directly.

The guidance language is the clearest place where management chose a bounded rather than aggressive posture. Lizardi said, “Turning to our outlook for the first quarter, we expect Texas Instruments' revenue in the range of $4.32 billion to $4.68 billion and earnings per share to be in the range of $1.22 to $1.48.” The commitment matters because it brackets a Q1 outcome around Q4’s $4,423.0 million base rather than promising a straight-line rebound in the company’s own outlook language. That is why the historical Q1 FY2026 line of $4,825.0 million, 58.0% gross margin, and $1.69 EPS is so important to the thesis if investors are using the table as the latest reported quarter. It shows what confirmation looks like when the recovery clears the guided posture: revenue above the $4.68 billion high end, EPS above the $1.48 high end, and gross margin back to 58.0%. It also explains why the -1.6% EPS miss in Q4 should not dominate the debate if the next data point is $1.69.

The risk to this view is not the tiny revenue miss; it is that free cash flow improvement gets absorbed by depreciation before investors see it in per-share economics. Q4 operating expenses were $967 million, up 3% from a year ago, and trailing twelve-month operating expenses were $3.9 billion or 22% of revenue. Operating profit was $1.5 billion in the quarter or 33% of revenue and was up 7% from the year-ago quarter. Those are not broken operating metrics, but they leave less room for disappointment if gross margin stalls near 55.9% instead of recovering toward 58.0%. The company’s own accounts show gross profit in the quarter was $2.5 billion or 56% of revenue, net income was $1.2 billion or $1.27 per share, and cash flow from operations was $2.3 billion. The equity case works if the next few quarters show the revenue base absorbing fixed costs and depreciation. It breaks if sequential revenue turns negative again while depreciation moves toward $2.2 to $2.4 billion in 2026.

What to watch next is therefore concrete. First, revenue must stay above Q4’s $4,423.0 million and ideally validate the Q1 FY2026 history at $4,825.0 million; a return toward the guided low of $4.32 billion would break the demand part of the thesis. Second, gross margin needs to hold near the Q1 FY2026 level of 58.0%, because another print near 55.9% would suggest the Q4 trough was not transitory. Third, EPS must show that the $1.27 Q4 result was the seasonal low, with $1.69 in Q1 FY2026 as the confirmation level and the guided $1.22 to $1.48 range as the management-commitment baseline. Fourth, capital expenditures should track toward the stated 2026 range of $2 and $3 billion after $4.6 billion in 2025, while depreciation should be monitored against the $2.2 to $2.4 billion expectation. Finally, on the next call, sentiment above 0.19 with guidance_tone above 0.13 and qa_evasiveness below 21.1 would confirm that management can explain the cash-flow inflection without sounding more defensive; if qa_evasiveness rises again from 21.1 while gross margin slips below 58.0%, the market will be right to keep underwriting the capex overhang rather than the recovery.