Thomas C. Baker sells EUR 1.4m after CNXN's Q2 surge

PC Connection does not sell a story. It sells hardware, software, and services, and the business lives or dies on how much enterprise and public-sector demand it can capture, how well it mixes higher-margin product into the order book, and how much of the quarter is pulled forward by vendor cycles or delayed by them. That is why the 2026 backdrop matters here. IT distribution has had a better year because AI infrastructure spending has kept data-center systems busy, lead times have been less sloppy than they were in the post-pandemic hangover, and the broader technology budget picture has not rolled over the way a lot of people feared.

That backdrop is not abstract for a name like CNXN. When distributors report improving book-to-bill ratios and accelerating AI-related orders, the market stops treating them like pure plumbing and starts treating them like a levered read on enterprise refresh demand. PC Connection sits in that lane with peers such as CDW and TD Synnex, both of which live off the same basic machine, though each has its own mix of enterprise, public-sector, and vendor exposure. The point is simple. If the cycle is firming, distributors can show it quickly. If it is not, they show that too.

PC Connection's second quarter gave the bulls something real to work with. The company reported record net sales of $854.0 million for the quarter ended June 30, 2026, up 12.4% year over year. Gross billings reached $1.2 billion, up 14.0%. Gross profit rose 14.3% to $157.5 million, and gross margin expanded by 30 basis points to 18.4%. Net income increased 33.8% to $33.2 million, or $1.31 per diluted share versus $0.97 a year earlier. That is a clean quarter for a distributor. Not perfect, because distributors rarely are, but clean enough that the market had a reason to lean in.

The filing came after the quarter, not before it

The insider piece is straightforward. Baker, the senior vice president, CFO, and treasurer, sold 20,000 shares of common stock on July 31, 2026, in two open-market transactions at weighted-average prices of $83.17 and $83.89 per share. The filing landed on August 3, after the company had already released second-quarter results on July 29. CNXN closed at $83.72 on July 31, which puts the sale right on top of the market, not at some dramatic discount or premium.

That matters because timing changes the read. A sale before earnings can look like a hedge against a bad print. A sale after a strong quarter is more often a liquidity event, a portfolio trim, or a routine reduction after a run. You do not get to assign motive from the filing alone. You do get to notice that the CFO sold after the company had just posted record sales and a sharp jump in profit. That is a different posture than buying into weakness.

The euro-normalised filing value was about EUR 1,400,406. InsiderTrades data tags the transaction as a cluster sale from a high-weight role, and the role matters here. A CFO is not a random director with a small side bet. The office sits closest to the numbers, the capital structure, and the cadence of the quarter. Our scoring also leans on the fact that the sale represented about 0.08% of the company's market value, which is enough to register without pretending it is a balance-sheet event.

What the quarter says about the business underneath the filing

The quarter matters more than the filing because it tells you what Baker was selling against. PC Connection's revenue mix is still tied to the old distributor logic, but the current cycle has given that logic a better backdrop. AI infrastructure spending has been a real tailwind for the broader hardware stack, and the company is operating in a market where technology budgets have not been cut to the bone. Deloitte's 2026 outlook points to worldwide IT spending surpassing $6 trillion, with data center systems boosted by AI infrastructure among the leading growth segments. That is not a CNXN-specific forecast. It is the weather. CNXN still has to catch the rain.

The company also sits in a sector where the market is rewarding evidence of demand resilience. Peer commentary around CDW and TD Synnex has focused on how much of the AI hardware cycle is flowing through distribution versus going straight to hyperscalers and large direct channels. That split matters. If enterprise and public-sector customers keep refreshing fleets, distributors can keep earning. If the AI spend stays concentrated in a few giant data-center builds, the benefit is less evenly spread. PC Connection's quarter suggests it is getting enough of the broad demand to matter.

There is a reason the stock can trade like this after a strong print. The market is not paying only for the quarter that just happened. It is paying for the possibility that the quarter is a sign of a steadier run rate. That is where the insider sale becomes useful, because it forces you to separate business momentum from personal positioning. The company can be doing well while an executive trims. Both can be true. The filing does not erase the quarter, and the quarter does not erase the filing.

InsiderTrades data, read in context

Our data gives this filing a display score of 51. That is not a verdict, and it is not a promise. It is a compact way of saying the trade sits in a role bucket we care about, it came with cluster context, and it was sized at a level that is large enough to notice without being absurd for a CFO after a strong run.

The historical cohort behind the role and size bucket is the part that deserves discipline. For CFO buys at mid-cap names, InsiderTrades data shows a 48.3% 90-day win rate and a 3.99% average return over 90 days, with a 61.55% average return over 365 days. That is historical cohort data, not a forecast for CNXN and not a promise that this filing will behave the same way. It is a useful reminder that role-specific patterns can be noisy in the short run and much less tidy than people want them to be.

The strategy frame is there for readers who want the broader system view, but it should stay in the background. Our live out-of-sample headline sits at 0.81, with 26.4 and 51.5 on the same restricted EU venue universe, and those figures survive only in that narrow regime. They are a screen, not a prophecy. The point is to keep the filing in a tested context without pretending the context can tell you what CNXN does next.

Why the sale is easier to read after a strong quarter

The cleanest way to think about Baker's sale is not as a verdict on the company, but as a transaction that arrived after the market had already been given a good reason to re-rate the name. CNXN had just reported record sales, stronger gross profit, and a wider margin. The stock was sitting at $83.72 on the day of the sale. In that setting, a CFO trimming 20,000 shares is not the same thing as a CFO heading for the exits.

Still, you should not flatten the filing into a harmless footnote. A cluster of sales from four distinct insiders, with 12 recent declarations in the cluster picture and several of them landing on August 3 and July 24, tells you this was not a one-off administrative form. InsiderTrades data marks the cluster as real. That does not mean the group knows something the market does not. It does mean the selling was broad enough to deserve a glance rather than a shrug.

The market often overreacts to insider sales when the underlying business is weak and underreacts when the business is strong. Here, the business is the anchor. CNXN is not a speculative software name with no earnings and a story stock multiple. It is a distributor with a quarter that showed real operating leverage. If you want to argue the sale matters, you have to argue against that operating backdrop, not in a vacuum.

CDW, TD Synnex, and the shape of the 2026 channel

The peer set is useful because it keeps the story grounded in the channel, not in a single ticker. CDW and TD Synnex operate in the same broad IT distribution and solutions space, and the 2026 conversation around the group has been about how much AI-related hardware demand is flowing through the channel, how much enterprise refresh is still alive, and which distributors have the best exposure to resilient end markets. That is the real comparison set. Not whether one company has a prettier investor deck.

CNXN's quarter suggests it is participating in the same favorable current. Record net sales, higher gross billings, and a margin that moved in the right direction are not the numbers of a business fighting for oxygen. They are the numbers of a distributor with enough demand and enough mix to make the quarter work. The market can still decide that the multiple is too rich, especially after a run. But the operating evidence is there.

That is why the insider sale is interesting rather than decisive. If the company had just printed a weak quarter and the CFO sold, the filing would read one way. If the company had printed a strong quarter and the CFO bought, it would read another. Here you have a strong quarter and a sale into it. That combination is common enough to avoid melodrama, but specific enough to keep on the page. The stock closed almost exactly where the CFO sold. That is the kind of detail that keeps the story honest.

What to watch when the next filing lands

The next useful question is not whether Baker sold. He did. The question is whether the company can keep turning the same demand backdrop into the same kind of operating leverage. If the AI infrastructure cycle stays firm, if enterprise refresh demand holds, and if gross margin keeps inching higher, the market will have to decide whether the post-earnings strength was a one-quarter pop or the start of a more durable rerating.

Watch the next quarter for the same things that mattered in the last one, because those are the levers that move a distributor. Net sales growth. Gross billings. Gross margin. Net income. If those stay constructive, the insider sale will look more like a trim after a good run than a warning shot. If they soften, the filing will get more attention in hindsight. That is how these things work. The filing is a datapoint, not a thesis.

The dividend also keeps CNXN in a slightly different lane from the pure growth crowd. The company has maintained a dividend, with the most recent $0.20 per share declaration carrying an ex-date of August 11, 2026. That does not change the insider read, but it does remind you that this is a cash-generative distributor, not a venture-style story. The market can still reprice it, and insiders can still sell into strength, but the business underneath is more grounded than the headline would suggest.

The filing, the quarter, and the part that still matters

Baker sold after a record quarter, not before it. The sale was large enough to matter, about EUR 1.4m on a euro-normalised basis, and it came from a CFO in a cluster of insider selling. That is enough to keep the filing on the desk. It is not enough to overpower the quarter, which showed a business with real revenue growth, better margins, and stronger earnings in a sector that has finally found a tailwind.

If you are reading CNXN properly, you are not choosing between the quarter and the filing. You are putting them in the same frame. The quarter tells you the business is participating in the 2026 IT distribution upswing. The filing tells you one senior insider chose to monetize some of that strength. Both are true. The market will decide which one matters more when the next set of numbers arrives, and the next Form 4 will tell you whether this was a one-off trim or part of a longer pattern.

Dig deeper: PC Connection INC's full insider filing history and Baker Thomas C's filing track record.

This is not investment advice.