Arrow Electronics Q2 2026 — The Upturn Is Real. The Cash Is Float.

The cycle turned, and the filing agrees

Start with what is not in dispute. Global Components sold $7,366M, up 39%, with every region up by more than a third: Americas +44%, EMEA +36%, Asia-Pacific +38%. Operating income more than doubled to $396M and the margin went to 5.4% from 3.5%. That is what a distributor looks like when the cycle turns: the segment's operating expenses rose 13.7% against sales up 39%, and the rest fell through. Non-GAAP EPS of $5.45 beat a $4.46 consensus by 22.2% (unverified — consensus aggregator). The stock still fell 8.45% the next session, to $203.51 (unverified — market data).

The interim chief executive, Bill Austen, put the demand case in one sentence: "Book-to-bill ratios remain well above parity, and our backlog continues to build in both size and duration." Read the second half carefully. A backlog that grows in duration is customers ordering further out, which is what buyers do when lead times stretch; the 10-Q confirms "pockets of constrained inventory, price inflation and extended lead times." The sentence contains no number. The release never gives the ratio, and the transcript only adds that lead times "remain lower than a pervasive shortage environment" (unverified — transcript). So the book-to-bill is a direction, not a measurement. I'll take the direction. The balance sheet is where the measurement is.

A third of the growth was price

On the call the CFO said "price inflation contributed roughly 1/3 of the sequential revenue growth in our global components business" (unverified — transcript). If that is right, about a third of the $725M sequential gain was price. You would expect price to lift the margin. It didn't. Global Components gross margin fell to 11.6% from 12.1% in Q1, and the operating margin slipped to 5.4% from 5.5%. The growth came from Asia, up 16.0% sequentially against 6.2% in the Americas, and Asia is the lower-margin book. It doesn't break anything yet. It does mean the year-over-year margin gain is operating leverage on volume, and mix is the softest margin to capitalize: the sequential line already shows the mix working the other way.

Global ECS is the other quarter inside this one. Sales rose 14% to $2,627M but fell 7.3% from Q1, operating income fell 12% to $85M, and gross margin dropped 100 basis points to 10.2%. Components now does the earning. The ECS story is a contract story, and I come back to it below.

Where the $1,018M came from

Operating cash flow was $318M in the quarter and $1,018M for the half, against negative $206M a year ago. The release explains the quarter in one clause: "partly due to the timing of cash flows within Global Components supply chain services offerings." "Partly" is generous. In the quarter, payables supplied $2,365M and receivables absorbed $2,059M. The difference, $306M, is nearly all of the $318M. For the half, payables supplied $9,755M against $8,339M of receivables, a net $1,416M, which is more than the whole $1,018M. Take that difference out and first-half operating cash flow was negative $398M, with inventory absorbing $876M.

Note E of the 10-Q says what those balances are: goods "purchased by the company on the request of and behalf of its customers," for which "receivables are disproportionate to the fees the company recognizes as revenue." Arrow is in the middle of the payment, collecting from the customer and paying the supplier, and books only a fee. When suppliers are paid later than customers pay Arrow, cash sits with Arrow for a while. It's a real cash benefit, and it isn't earned. Prepaid growth is a customer loan in a cash-flow costume, and this is the supplier-side version: the cash is a loan from suppliers' payment terms, and it is repaid when the timing turns.

The composite hides the component

Management's working capital line looks calm: $6,840M at July against $7,437M in December, and 17.1% of annualized sales against 22.5% a year ago. The composite hides the component. Underneath it, receivables went to $28.01B from $19.74B in six months, and payables to $27.11B from $17.38B. Total assets grew to $38,377M from $29,078M. Receivables are now 4.0 times shareholders' equity.

A single quarter of that would be a phase, not a decision, so here is the series. Receivables minus payables, at the eight quarter ends from September 2024: $2.33B, $1.98B, $1.55B, $2.06B, $2.53B, $2.35B, $1.22B, $0.90B. For six quarters the gap stayed between $1.55B and $2.53B while receivables grew. In the last two it fell to less than the bottom of that range. As a share of receivables it went from 11.9% at December to 3.2% at July. The gap did not grow with the book. It collapsed while the book grew.

The same chart shows what did not happen. Inventory sat between $4.53B and $4.80B for five quarters and has risen only to $5.94B, up from $5.08B in December. For a distributor in a tightening cycle, the warehouse is the position, and Arrow has barely taken one. It grew components sales 39% with inventory up from $4.75B a year ago. That is disciplined. It also means the inventory bill for a backlog that the transcript says runs into the first half of 2027 (unverified — transcript) has mostly not been paid yet.

The deleveraging was paid with the same float

Long-term debt fell to $2,053M from $3,085M, with $923M of net bank repayments, and the North American securitization went from $970M outstanding to zero. The transcript's "adjusted leverage ratio" of 1.75x (unverified — transcript) is the headline version. Put it next to the cash statement and the funding is plain: the half's $1,416M of payables-over-receivables paid down the debt.

Two lines in the filings say the company knows the float is temporary. The 10-Q notes that uncommitted bank lines were cut from $560M to $160M over two quarters, "reflecting changes in lender participation." Then on September 2, a month after the print, Arrow raised the North American securitization limit to $1.75B from $1.5B and pushed its maturity to 2029. You don't enlarge a receivables facility you have just repaid to zero unless you expect to draw it. The credit window is the real tell, and Arrow widened its own window before it needed to. That is prudent. It is also a statement about what the treasurer expects the next quarters to cost.

ECS: a contract you can't cancel

The 10-Q is unusually candid here: Global ECS "is party to certain multi-year non-cancellable purchase obligations through 2032," and "their long-term performance cannot be reasonably estimated at this time, and the company is anticipating there could be additional associated losses in the coming quarters." Note that I am not the one calling these contracts a problem; the company is. The losses were $21.7M in Q1 and $26.6M in Q2. Before them, ECS gross margin was 11.2%, the same as a year ago, and operating income $112M. So the whole 100 basis point margin decline is the contracts. Arrow has given notice on one, effective in Q1 2027, and the CFO expects "some more charges in the second half of the year, probably at a lesser pace" (unverified — transcript). The filing does not say how much of the $30.3B of total purchase obligations belongs to these ECS agreements, or how much of the $4.2B due in 2031 and after. That is the number I'd most like and can't get.

The guide says the easy quarter is behind it

Q3 non-GAAP EPS of $4.83 to $5.03 puts the midpoint at $4.93, 9.5% below Q2 and below a $4.97 consensus (unverified). Components is guided up 4.5% at the midpoint, a slowdown from 10.9%. ECS is guided down 16.2%, which the CFO puts down to "growing over a large partner addition last year" (unverified — transcript). And interest expense steps up to about $50M from $37M, after a quarter helped by interest on collected tariff receivables. Avnet's components business, by contrast, grew 17.0% sequentially and guided about 10% more at a 4.1% operating margin. Arrow runs the higher margin and the slower sequential growth. The market sold the gap between a record quarter and a smaller next one, not the record.

The strongest case against me

Here is the other side at full strength, and it is not weak. The float may be structural. Supply chain services has grown so much that receivables doubled in a year, and the 10-Q says receivables and payables "are typically correlated as the company acts as an intermediary in the transaction and remits payments to the supplier upon receipt from the customer." If Arrow pays suppliers only after the customer pays, a narrow gap is the design of the program, not a timing accident, and it doesn't reverse. On that reading the $0.90B is the new normal, the $1,416M was a one-time reset as the program grew, and I am calling a better contract structure a loan. It's a serious argument. I still come down on the other side, because the company's own words don't support it: the 10-Q calls the working capital change "the timing of settlements," the release says "timing of cash flows," and the gap has moved by $1.45B in a single half. Pass-through by design would keep the gap steady as a share of receivables. It fell from 11.9% to 3.2%.

What decides it

One number, on one date: receivables minus payables on the October 3, 2026 balance sheet, published with Q3 results (expected October 29; unverified — date not yet confirmed by the company). Above $1.22B, the April level, the float is going back and I'm right. At or below $0.90B, the narrow gap is the design and I'm wrong. In between is partial, and I'd call it that. Two secondary checks: Q3 operating cash flow against $318M, and any draw on the $1.75B securitization. A third, which could change the story on its own, is inventory against $5.94B: the first quarter Arrow builds stock for the backlog is the first quarter the cycle costs it cash.

The trade (analysis, not advice)

The components cycle is real, and Arrow is running it at a better margin than Avnet. None of that is in question. What is in question is the cash quality under a $5.45 quarter, and at $232.80 (unverified — market data, September 25) the stock has recovered all of the post-print drop. I would not pay for the deleveraging story until October shows whether the float holds. The asymmetric version is to own the cycle through the earnings line and not the cash line: the risk into the Q3 print is not EPS, which is guided and was beaten by 22.2% last time (unverified — against aggregator consensus), but a cash flow statement where the float reverses and debt comes back. Anyone who wants the upturn with less exposure to that can wait for the October balance sheet; if the gap holds at $0.90B, the bear half of this read is gone and the stock is cheaper than it looks on a cleaner balance sheet.

Bottom line

Arrow printed a genuine components upturn and a cash flow that suppliers paid for. The income statement shows operating leverage on a 39% sales rise. The cash flow statement shows $1,416M of payables outrunning receivables, and that paid down $1,032M of long-term debt. ECS carries contracts through 2032 that the company can't yet price. My view is that the float goes back. The number that decides it is receivables minus payables above $1.22B or at or below $0.90B on October 3.