Arrow’s beat was not a cyclical inventory snapback; it was a mix-and-working-capital trade the Street is still underpricing

The thesis from this print is that Arrow is being misread as a simple components-cycle beneficiary when the numbers point to a more specific setup: revenue is recovering faster than gross margin, but the earnings surprise came from mix, operating-cost absorption, and interest expense rather than a clean margin inflection. That matters because a components distributor can screen optically cheap into an upcycle while cash conversion and working capital quietly take the next leg of earnings hostage. What was priced in was a recovery in distribution revenue after the 2024 trough, visible in the sequential revenue lift from $6,814.0 million in Q1 FY2025 to $7,579.9 million in Q2 FY2025 and $7,712.5 million in Q3 FY2025. What was not priced in was the magnitude of the Q4 upside: street-comparison revenue landed at $8,746.4 million versus $8,206.5 million, a +6.6% surprise, and EPS landed at $4.39 versus $3.55, a +23.7% surprise. The variant view is that the EPS beat should not be capitalized as a straight-line earnings run-rate, because the company’s own next-quarter EPS guide of $2.70 to $2.90 is far below the Q4 EPS beat level even as Q1 sales guidance of $7.95 billion to $8.55 billion still implies year-over-year growth at the midpoint.

That distinction between what surprised and what was already expected is essential because the headline growth was not a hidden variable by February. The recovery had already moved through the income statement: revenue was down -20.7% YoY in Q1 FY2024, down -19.0% YoY in Q2 FY2024, down -14.8% YoY in Q3 FY2024, down -7.2% YoY in Q4 FY2024, then improved to -1.6% YoY in Q1 FY2025, +10.0% YoY in Q2 FY2025, +13.0% YoY in Q3 FY2025, and +20.1% YoY in Q4 FY2025. The surprise was not that Arrow had turned; the surprise was that Q4 revenue reached $8,746.4 million, roughly back to the $8,736.4 million level from Q1 FY2023, while gross margin at 11.5% remained below the 12.7% posted in Q1 FY2023. In other words, the top line has nearly recovered to prior-cycle scale, but the gross-margin structure has not. That is not bearish by itself, because EPS still beat materially, but it says the upside came from a cost, mix, and below-operating-line package that is more fragile than a broad-based gross-margin reset.

The revenue and gross-margin path therefore argues against treating Q4 as a clean re-rating event. Gross margin recovered from 10.8% in Q3 FY2025 to 11.5% in Q4 FY2025, but it was 11.0% in Q4 FY2024 and 12.6% in Q4 FY2023, so the latest quarter sits between trough and prior-cycle economics rather than confirming a full margin recovery. The sequential revenue acceleration of +13.4% in Q4 FY2025 was the largest step in the provided history, but gross margin only returned to 11.5%, the same level recorded in Q3 FY2024 when revenue was $6,823.3 million. That is the core tension in the print: the company is moving more revenue through the model, but the gross-margin dollars are not yet supported by the same percentage economics investors saw in 2023. The earnings upside, then, should be decomposed. Per Rajesh Agrawal, Q4 non-GAAP diluted EPS “increased 48% to $4.39,” and the attribution matters because he tied it to “favorable sales results, a higher mix of our value-added services and lower interest expense.” The phrasing is valuable because it does not claim pricing power across distribution; it identifies a combination of volume, mix, and financing costs.

The operating-cost bridge supports that interpretation rather than contradicting it. Q4 non-GAAP operating expenses increased $53 million sequentially to $669 million, yet OpEx as a percent of gross profit declined 700 basis points sequentially and 100 basis points year-over-year to 67%. That means the beat benefited from scale absorption even as absolute operating expense rose. Non-GAAP operating income was $336 million, 3.8% of sales, and interest and other expense was $44 million in Q4. Lower average debt levels helped the quarter, but the next guide assumes interest expense of approximately $60 million, so one piece of Q4 EPS leverage becomes a headwind in Q1. The same point shows up in the revenue guide. The company expects Q1 sales of $7.95 billion to $8.55 billion and non-GAAP diluted EPS of $2.70 to $2.90. That combination is not a demand-collapse message, since Rajesh Agrawal said the sales range represents “an increase of 21% year-over-year at the midpoint of the range.” It is a margin-and-mix reset message, particularly because Q4 EPS on the street-comparison basis was $4.39 and the Q1 EPS guide tops out at $2.90.

The segment detail explains why investors should separate components recovery from ECS seasonality and services mix rather than put one multiple on all of Arrow’s growth. Global components sales increased $1.1 billion year-over-year and $326 million sequentially to $5.9 billion in Q4, above guidance and up 6% versus the prior quarter. Global components non-GAAP operating income increased approximately $20 million sequentially to $219 million, up 10% from the prior quarter, which says the components recovery did generate incremental operating profit. But the next-quarter components guide is only $5.75 billion to $6.15 billion, representing sequential growth of 1% at the midpoint, so the company is not guiding another Q4-style acceleration in its largest segment. ECS had different optics: Q4 ECS sales increased approximately $400 million year-over-year to $2.9 billion, above the midpoint of guidance and up 16% versus the prior year, or up 11% on a constant currency basis. Total ECS billings were $7.1 billion, up 16% year-over-year. Yet the Q1 ECS guide is $2.2 billion to $2.4 billion, up approximately 13% at the midpoint year-over-year, which embeds a step down from Q4 dollars. That is why Q4’s EPS beat and Q1’s EPS guide can both be true.

The balance sheet is where the bullish and cautious readings collide most directly. Net working capital grew sequentially in Q4 by approximately $180 million, ending the quarter at $7.4 billion to support growth, while cash flow from operating activities was $200 million in Q4 and $64 million for the full year. Return on invested capital increased 190 basis points year-over-year to 11.1%, so the working-capital build has not yet destroyed returns, but cash conversion is not giving investors the same message as EPS. Gross balance sheet debt declined sequentially by $44 million to $3.1 billion, and the company repurchased $50 million in shares in Q4 and $150 million in 2025. William Austen’s capital-allocation framing was intentionally long-dated: “And since 2020, we have returned approximately $3.6 billion to shareholders through share repurchases.” The important part for PMs is the gap between that historical buyback scale and the current-year figure of $150 million in 2025. Arrow still has repurchase capacity as a capital-allocation lever, but the latest year’s deployment was constrained relative to the multi-year history while working capital was $7.4 billion.

That cash-versus-EPS split is the reason the post-print debate should not be reduced to whether revenue beat by +6.6%. Full-year 2025 consolidated revenue was $30.9 billion, up 10% versus the prior year and up 9% on a constant currency basis, with global components revenue up 8% and ECS revenue up 18%, or 7% and 15% on a constant currency basis. Full-year non-GAAP diluted EPS increased 4% to $11.02. Those full-year numbers say the recovery is already in the P&L, but the earnings growth rate lagged revenue growth because the model is still absorbing mix, margin, operating-cost, and working-capital frictions. Q4 broke that pattern temporarily with $4.39 of EPS on the company’s non-GAAP basis and a +23.7% street surprise, but management’s own Q1 non-GAAP EPS guide of $2.70 to $2.90 reintroduces the question: how much of the Q4 earnings profile repeats when ECS normalizes and interest expense rises to approximately $60 million? The answer is not “none,” because components are still guided to sequential growth of 1% at the midpoint. The answer is also not “all,” because Q1 EPS guidance is explicitly far below Q4.

The read-through to the semiconductor supply chain is more positive for sell-in than for margin capture. Arrow’s Q4 global components sales of $5.9 billion, up $1.1 billion year-over-year and up $326 million sequentially, points to better distribution logistics demand for Infineon, NXP, and Texas Instruments. The magnitude matters: Arrow guided Q1 global component sales to $5.75 billion to $6.15 billion, with sequential growth of 1% at the midpoint, so the supplier signal is continued normalization rather than another inventory restocking surge. For Texas Instruments and NXP, both named suppliers to Arrow in semiconductor distribution and logistics, that is a cleaner signal for channel movement than for end-demand acceleration, since Arrow’s gross margin remains 11.5% in Q4 FY2025 compared with 12.6% in Q4 FY2023. For Infineon, the same read-through applies, but without a linkable ticker in this data pack: the channel is moving more product, yet Arrow’s own margin data does not prove tight supply.

Relative to distribution peers, Arrow’s latest reported quarter shows the recovery is not merely company-specific, but Arrow is also not the highest-margin way to own it. In the peer table, ARW revenue was $9,473.5 million with revenue YoY of +39.0% and gross margin of 11.1%, while AVT revenue was $7,119.8 million with revenue YoY of +33.9% and gross margin of 10.4%. That comparison favors Arrow on growth and margin versus AVT in the same subsector. But 8070.TW posted gross margin of 18.2% on revenue YoY of +20.4%, and 5434.TW posted gross margin of 13.6% on revenue YoY of +17.6%, so Arrow’s scale-led acceleration does not translate into peer-leading gross margin. The market may therefore be right to pay for Arrow’s recovery versus slower or lower-margin distributors, but wrong if it prices the Q4 EPS beat as evidence that Arrow has escaped the distribution margin ceiling.

The call delivery reinforces the same mixed but investable message: management sounded more controlled than exuberant, and the model’s tone supports taking the guide seriously rather than assuming sandbagging. The tone history shows Q4 FY2025 sentiment at 0.55 versus 0.51 in Q3 FY2025, guidance_tone at 0.64 versus 0.69, tone_confidence at 0.46 versus 0.43, prepared_sentiment at 0.70 versus 0.61, and qa_sentiment at 0.19 versus 0.33. The call-over-call delta from Q4 FY2025 to Q1 FY2026 then moved sentiment down -0.14, guidance_tone down -0.13, prepared_sentiment down -0.10, and qa_sentiment down -0.02, while tone_confidence rose +0.12 and ai_optimism rose +0.17. That combination is not a management team trying to sell a breakout story; it is a team giving a more confident but less positively worded forward frame. Uncertainty fell from 64.5 in Q3 FY2025 to 48.1 in Q4 FY2025, then increased to 50.4 in Q1 FY2026, while qa_evasiveness moved from -6.3 to -5.7 and then to -7.0. The numbers line up with the financial setup: better visibility than the downcycle, but not a clean admission that Q4 earnings are the new base.

That tone matters because the Q1 guide is the fulcrum for the stock after a large EPS beat. If investors focus only on Q4, they see revenue of $8,746.4 million versus $8,206.5 million and EPS of $4.39 versus $3.55. If they focus only on Q1, they see sales guidance of $7.95 billion to $8.55 billion and EPS guidance of $2.70 to $2.90. The right interpretation is between those extremes: the demand recovery is durable enough that Q1 sales at the midpoint would be up 21% year-over-year, but the earnings power implied by Q4 is not durable unless value-added services mix, OpEx leverage, and interest expense all cooperate again. The most defensible positioning takeaway is to own the recovery but not pay for a 2023-style gross-margin reset until gross margin moves materially above 11.5% and cash flow follows EPS. Q4 proved Arrow can beat when revenue, mix, and financing costs align. It did not prove that the new normalized model earns Q4 EPS every quarter.

What to watch next quarter is therefore specific. The thesis is confirmed if Q1 sales land inside or above the $7.95 billion to $8.55 billion range while global component sales meet or exceed the $5.75 billion to $6.15 billion guide and EPS holds within or above $2.70 to $2.90 despite interest expense of approximately $60 million. It strengthens further if gross margin moves above 11.5%, because that would show Q4 was not just scale and mix, and if operating cash flow improves against the full-year 2025 base of $64 million while net working capital does not expand materially beyond $7.4 billion. It breaks if Q1 revenue growth fails to support the stated 21% year-over-year midpoint framework, if components fall below the 1% sequential growth midpoint implied by guidance, or if EPS misses the $2.70 to $2.90 range while interest expense is near approximately $60 million. The next date that matters is the Q1 FY2026 report for the period ending 2026-04-04, because that is where the market learns whether Q4 was the first clean quarter of a recoverable earnings cycle or a mix-aided spike inside a still cash-hungry distribution model.