Should you buy Argan (AGX) stock in 2026?

2026-08-03 · By Lubin Danilo, founder of Lubin Investment

AGX: see the full analysis on Lubin Investment

Argan validates all 10 of my quality criteria, a perfect score, driven by explosive electricity demand from AI data centers. But my model flags a severe 52% overvaluation: the narrative is real, the price has already priced in much of it. Here is how I separate the story from the price.

The quiet contractor behind the AI power rush

Argan is not a household brand, but its subsidiary Gemma Power Systems sits at the heart of one of 2026's hottest infrastructure narratives: the race to build enough electricity generation to power AI data centers. Gemma designs, builds and commissions large natural-gas power plants for developers and grid operators, several of which are explicitly justified as firming capacity for new, power-hungry data center clusters. The company is currently building roughly 5.5 gigawatts of thermal power generation across the US and Ireland, including four major US natural gas projects representing more than 4.1 gigawatts.

The immediate trigger is the first-quarter fiscal 2027 report (quarter ended June 30, 2026): earnings per share of $3.24, well above the $2.33 analysts expected, a beat of nearly 39%, described by the company as a record quarter, with further AI-linked backlog expansion. The stock was also added to the Russell 2000 index. For fiscal year 2026, ended in late January, the consolidated backlog stood at $2.9 billion, with the power segment alone growing from $1.3 billion to $2.7 billion, driven by $2.5 billion of new contract value signed.

An engineering-contractor model that ties up almost no capital

To make sense of Argan's extreme numbers, you need to understand its business: it does not own the power plants it builds. Argan is an engineering, procurement and construction (EPC) contractor: developers or utilities hire it to design, procure equipment for, and build a plant, and Argan then hands it over turnkey without ever carrying it on its own balance sheet. That explains a genuinely extreme cash return on capital employed: 313%. With very little of its own capital tied up in the business (mostly working capital for ongoing projects, not owned power plants), every dollar of capital employed generates more than three dollars of cash each year, a level far beyond reach for a traditional industrial company that owns its heavy assets.

A second number confirms this mechanic: Argan's net cash conversion cycle is negative, at minus 9 days. In practice, the company collects payment from its customers, on average, before it even has to pay its own suppliers and subcontractors, a float model similar to what Amazon or Costco practice, where the company effectively funds its operations with other people's money rather than its own. Combined with a 38.7% free cash flow margin (an extremely high level for a construction-type business) and a 292% cash conversion rate (nearly triple the accounting profit turns into real cash), the picture that emerges is a genuinely elite cash-generation machine, not just a good growth story.

Fast growth, funded without any debt at all

Argan passes all 10 of my financial criteria. Its sales have grown 19.8% a year on average over five years, its free cash flow per share has climbed 85.7% a year over the same period, it buys back its own shares (the share count has fallen 1.83% a year), and above all, it holds a net cash position: zero net debt. A company growing fast, self-funding, and buying back shares, all while carrying no net debt, is a rare combination.

The real risk to watch is not in these numbers, it is in the nature of the business itself: an EPC contractor's revenue is booked project by project, as work progresses, which can make quarterly revenue uneven. A lost or delayed contract, or a major project winding down without an immediate replacement, can swing a given quarter significantly, unlike a subscription business whose revenue is smooth from one quarter to the next.

The price: cheap versus its sector, expensive versus its own history

Argan currently trades at 20.8 times its free cash flow (P/FCF), a level I would call reasonable in absolute terms. Compared with its engineering and construction sector peers (sector median of 28.1 times), that multiple places it at the 41st percentile: cheaper than the median of its direct competitors.

But set against its own five-year history, that same 20.8-times P/FCF sits at the 88th percentile: Argan has rarely been this expensive relative to itself. The AI power narrative has pushed the stock up faster in recent quarters than this specific company's own valuation history would typically suggest, even though it still looks cheap next to other engineering contractors less exposed to that narrative.

Third step, the verdict: my model puts the fair buy price at $282.15, against a current price of $592.66, a 52.4% overvaluation. Even a genuinely elite, debt-free, 10-out-of-10 business can remain a bad buy if the price paid is wrong: the AI power narrative is likely real and durable, but my model shows the market has already priced in a large share of that future growth today.

The real debate: will the AI power supercycle last long enough?

The whole Argan thesis rests on this tension. On one side, data center power demand is a documented, concrete narrative: Gemma is actively building or bidding on gas plants explicitly justified by data-center-cluster demand, the backlog is growing fast, and zero debt leaves enormous flexibility to keep bidding on new projects. On the other, construction remains cyclical and concentrated in a limited number of large contracts: a slowdown in data center capital spending, or tighter permitting for new gas plants, would hit the backlog directly. Argan's own valuation history suggests today's price already assumes this supercycle continues without a hitch. If you believe AI power demand is only getting started, the perfect score justifies keeping Argan on your watchlist. If you think the market has already paid upfront for several years of this growth, my $282 buy price remains the benchmark to watch for.

How I settle it

Argan is not a mediocre business riding a fad: it is a genuinely elite, debt-free cash machine (10/10) whose price has simply run ahead of its own valuation history. Both judgments coexist: perfect quality, a price that is no less demanding. I am not betting against Argan, I am marking a price and waiting for it to come to me. You can follow these numbers live on Argan's analysis page, and understand in detail how I calculate this fair buy price in my full methodology.

Argan validates all 10 of my quality criteria (a perfect score) through its subsidiary Gemma Power Systems, which builds gas-fired power plants to meet explosive AI data center electricity demand (about 5.5 GW under construction).
An engineering-contractor model that ties up almost no capital: a 313% cash return on capital employed, a negative cash conversion cycle (-9 days, it collects before paying suppliers, a model similar to Amazon or Costco).
Fast growth (sales +19.8%/year, cash per share +85.7%/year over 5 years) funded without any net debt, with the power-segment backlog growing from $1.3 billion to $2.7 billion during the fiscal year.
P/FCF of 20.8 times: cheap versus its sector (41st percentile, median 28.1 times) but expensive versus its own history (88th percentile), a sign the AI narrative has already pushed the stock up.
My model flags a severe 52.4% overvaluation (fair buy price of $282.15 against a $592.66 price): perfect quality, but a price that has already priced in much of what comes next.

FAQ

What does Argan actually do?

Argan is an engineering, procurement and construction (EPC) contractor, through its subsidiary Gemma Power Systems. It designs and builds large power plants, mainly natural-gas fired, for developers, without ever owning the plants itself. A large share of its recent business responds to AI data center electricity demand.

Why is Argan's return on capital so extreme (313%)?

Because Argan does not own the plants it builds: it ties up very little of its own capital in the business, which mechanically inflates the return on every dollar of capital employed, a level out of reach for an industrial company carrying heavy assets on its balance sheet.

What is a negative cash conversion cycle?

It means Argan collects payment from its customers, on average, before it has to pay its own suppliers, a float model similar to Amazon or Costco, where the company funds its operations with other people's money rather than its own.

Isn't a 20.8x P/FCF reasonable for Argan?

It is versus its sector (41st percentile, cheaper than the median). But versus its own five-year history, that same level sits at the 88th percentile: Argan has rarely been this expensive relative to itself, a sign the AI narrative has already pushed the stock up.

Should you buy Argan stock now?

My quality screen is perfect (10 of 10 criteria), but my price model shows a severe 52.4% overvaluation that leaves no margin of safety at the current price. This is not personalized investment advice, do your own research before any decision.

About the author

Written by Lubin Danilo, founder of Lubin Investment. A self-taught individual investor, I find fundamental analysis fascinating, and it has delivered excellent results. For three years now, my performance has beaten the S&P 500. But analyzing every stock took too much time: sites with incomplete data, calculation methods and criteria never aligned with mine. And spotting the best stocks was just as time-consuming, even with my own well-defined checklist. So I put my software development background to work to build this software, base my investment strategy on its results, and share it with people who share the same passion as me. It judges a company's quality and its price separately, using criteria drawn from the financial literature (Warren Buffett, Michael Mauboussin, Aswath Damodaran).