PTC Inc. PTC

Key data

The business

PTC sells the software manufacturers use to design and manage physical products. Two products do the work. Creo is the 3D computer-aided-design tool an engineer opens to draw a part. Windchill is the system of record that stores every version, bill of materials and change order behind that part, so a company with thousands of engineers and suppliers works from one truth. Codebeamer extends the same idea to software-defined products, the code and requirements inside a modern car or medical device. The customers are the industrial base: automakers, aerospace, industrial equipment, medtech. Once Windchill holds a company's entire product history, moving off it means re-platforming the engineering department, so almost nobody does, and that retention is the moat. Nearly all revenue is recurring subscription, gross margin runs 84%, and in fiscal 2026 PTC sold ThingWorx and Kepware, its internet-of-things and factory-connectivity units, to concentrate on this core.

The numbers

The reported figures now hide as much as they show, because two businesses left mid-year.

The quarterly line deceives twice. Q2 FY2026's $590.7M of net income and $4.98 of EPS carry a one-time gain of roughly $3.50 a share from selling ThingWorx and Kepware, and stripped of it the quarter earned about what the others did. Q3 FY2026 then fell the other way, sales down 6.8% year on year to $600.0M, not because customers left but because the divested units no longer count. The honest gauge is ARR, which the company does disclose: PLM annual recurring revenue excluding the sold units grew 8%, and 10% in constant currency, led by Windchill and Codebeamer, with Europe up 16% constant currency.

FY2025 revenue of $2.7B and $6.08 of GAAP EPS are the last clean full year, and the 9-month FY2026 EPS of $7.43 is inflated by the same gain. What compounds underneath is cash.

Free cash flow grew from $415.8M in FY2022 to $856.7M in FY2025, a 27% annual clip, on capital spending near zero. Stock compensation runs 8% of revenue, so the owner keeps a free-cash margin closer to 24% than the 32% headline, and even that funds a share count down about 2.5% year on year on roughly $300M of annual buybacks. On non-GAAP earnings PTC beat expectations every quarter this year. The variant is plain: the market prices reported revenue, which the divestiture pushed to roughly flat for FY2026, and pays 5.4 times sales near a twelve-year low, while the memo's view is that an 8% to 10% organic ARR base is worth more than flat optics. The next two ARR prints settle it.

Management

Neil Barua took the CEO chair in 2024 and has run one playbook: prune the portfolio, concentrate on PLM, let ARR and cash compound. The ThingWorx and Kepware sales fit that, and the roughly $300M of FY2025 buybacks were spent below today's price against a 2.5% lower share count. The other side of the record is insider selling: no open-market buys in twelve months, $11.5M sold across sixteen transactions, the largest CFO Kristian Talvitie's $8.7M in December 2025 and general counsel Aaron von Staats's $0.6M the day before. Plan status is not disclosed in the filings, so read them as discretionary until a Form 4 footnote says otherwise. Long-term debt rose to $1.40B funding the buyback and cash build, but at 0.75 times EBITDA and 22 times interest cover it is not a constraint.

How it fails or surprises you

Organic ARR slips below high single digits (downside). PLM ARR excluding the divested units grew 8%, 10% constant currency, this quarter, and the whole 17x forward multiple rests on that staying near double digits. Americas PLM ARR including divested units actually fell 4%. If the ex-divestiture line drifts toward 5%, the stock is expensive, not cheap. Watch the next two ARR disclosures.

The trailing multiple is a trap (downside). The 13.6x trailing P/E and "cheapest in years" optic exist only because Q2's $590M divestiture gain sits in the numerator. Clean trailing earnings are near 20x, forward 17x. If organic growth disappoints there is no hidden value in the trailing number to catch you, and the first ARR quarter without divestiture noise is what reveals it.

PLM reaccelerates into the model-based enterprise (right tail). Europe PLM ARR grew 16% constant currency excluding divestitures, faster than the 10% group. If Windchill's role as the system of record for AI-driven and software-defined product design pulls the whole base to sustained double digits, free cash flow compounds mid-teens per share and 5.4 times sales, near its twelve-year floor, re-rates toward its 8x norm. The market pays for none of this today.

Closing thoughts

The evidence is an 84%-gross-margin subscription base throwing off a growing $850M-plus of cash, wrapped in a reported income statement the divestiture made unreadable for a year. The fatter tail is modestly to the upside: the downside is a slow de-rate if organic ARR fades to mid-single digits, a real but gradual loss, while the upside is a base re-rating off a twelve-year-low sales multiple if PLM holds double digits. What is genuinely at risk is the growth premium, not the business, because the cash and the switching costs are not in doubt.

The bet is still that engineers keep working inside Windchill and Creo and keep paying more each year to stay. It breaks if organic PLM ARR, the 8% to 10% constant-currency line, slides toward mid-single digits while reported revenue stays flat, because then the 17x forward multiple has nothing to grow into. One pair of numbers tells you first: PLM ARR growth excluding the divested units, set against the reported revenue line, over the next two quarters. If those diverge further, the market was right to look through the reported flatness. If they converge upward, it was not.