Iris Energy Is Sitting on $7.6 Billion in Cash and Building One of the Largest AI Data Centers in the World. Is IREN Stock a Buy?

Iris Energy Is Sitting on $7.6 Billion in Cash and Building One of the Largest AI Data Centers in the World. Is IREN Stock a Buy?

David Beren

Mon, September 21, 2026 at 7:43 PM GMT+3 5 min read

• IREN +2.79%
• BTC-USD +1.93%

Key Stats for IREN Limited

• 52-Week Range: $28.93 to $76.87

• Street Mean Target: $80.21

• 5-Yr Beta: 4.28

• LTM Gross Margin: 68.9%

• Market Cap: ~$18.4 billion

From Bitcoin Miner to AI Infrastructure Company

Iris Energy (IREN) came out of the gate as a Bitcoin miner with a specific angle: run the operation entirely on renewable energy, starting with cheap hydroelectric power in British Columbia before expanding into Texas.

The clean power focus kept costs down and gave the company a story to tell at a moment when institutional investors were scrutinizing how much electricity Bitcoin mining was consuming. It worked well enough as a positioning strategy, but the company has moved well beyond it.

The business IREN is building now centers on Childress, Texas, where a 510-megawatt data center campus is under development to serve AI and high-performance computing customers.

The company has signed customer agreements for initial capacity blocks and is still actively expanding its Bitcoin mining hash rate while building out the AI cloud infrastructure in parallel.

CEO Daniel Roberts has described the Childress asset as one of the most significant power-contracted AI data center developments in the United States, targeting hyperscale customers who need large, reliable blocks of computing capacity and cannot easily find it elsewhere.

The operating income chart tells you what it looks like financially to build something at this scale. Operating income turned positive in FY2025 at around $21 million, then swung hard to negative $379 million in FY2026. The swing is not a business deterioration signal.

It reflects large non-cash depreciation and impairment charges tied to the Childress expansion and mining fleet upgrades, which flow through the income statement even though no cash leaves the door. The underlying mining business kept generating revenue, and AI cloud services are ramping.

Q4 FY2026 revenue reached $71.5 million, up significantly year over year, with energized hash rate rising to 31 EH/s.

The Capital Raise Changes the Runway Picture Entirely

Cash position is the metric that matters most for an early-stage infrastructure company. It decides how long the business can build without needing to go back to the market, how much leverage it has in customer and supplier negotiations, and whether management can afford to be patient rather than desperate on deal terms.

Cash jumped from $565 million at the end of FY2025 to $7.6 billion at the end of FY2026. That is not operating cash generation, it is a major equity capital raise executed specifically to fund the Childress buildout.

The effect is that IREN can now complete the 510-megawatt facility without being forced into dilutive financing at whatever terms the market offers during construction.

The LTM net debt-to-EBITDA ratio of 50.69x looks alarming if you read it cold, but it is a distorted number for a company that is pre-revenue on its largest business line. The real question is whether $7.6 billion is enough to build and fill Childress, and at current construction costs and contract rates, management says it is.

Forward revenue estimates of around 227% annual growth over the next two years reflect the expected ramp as signed customer deployments come online.

Whether IREN hits those numbers comes down almost entirely to whether Childress builds on schedule and customers activate their contracted capacity when the time comes.

What the TIKR Model Says, and How to Read It

The TIKR valuation model mid case puts a near-term price target of around $58 on IREN over the next 4.8 years, implying roughly 5% annualized returns from current levels. Revenue growth assumptions of around 82% annually in the mid case are aggressive in absolute terms but actually conservative relative to what a fully ramped Childress would deliver.

Net income margins are projected near zero at around 0.7%, because infrastructure depreciation and operating costs absorb most of the near-term revenue growth.

The near-term model is honest but incomplete for a company at this stage of development. Extend the horizon to 2035, and the mid-case forecasted return reaches around 284% at roughly a 17% IRR, which reflects the compounding effect of data center revenue scaling over a longer period once the facility is built and occupied.

The Street's mean target of around $80 implies about 70% upside from current levels, and analyst consensus has stayed broadly constructive on the Childress thesis.

The gap between 5% near-term annualized and 17% longer-horizon IRR is the whole trade in one number: the business needs time and execution to deliver, and patience is not optional.

Should You Buy IREN Stock?

The bull case works for investors who can underwrite Childress execution. The facility has $7.6 billion in cash behind it, signed customer agreements, renewable power infrastructure already in place, and a management team that has been deliberately building toward this specific moment.

If the data center ramps as planned and AI computing demand keeps growing, the forward revenue trajectory is achievable, and the longer-horizon valuation case becomes genuinely compelling.

The bear case is execution risk and time. A 510-megawatt data center does not get built without complications, and any delays in construction, customer deployment, or power availability push the revenue ramp out in ways that the near-term model does not absorb gracefully.

A beta of 4.28 means large moves on any piece of news, the operating income chart shows how easy it is to misread the financial statements without context, and the near-term model implies modest returns even in a reasonable scenario.

IREN is a high-conviction infrastructure bet on AI computing demand, and getting the timing wrong means sitting through volatility and potential dilution while waiting for a thesis that may simply take longer than expected to prove out.