JPMorgan's $24 Billion IREN Forecast Implies 34x Growth
JPMorgan's 2030 revenue projection for IREN sits about 34 times above the $707.00 million the company last reported for a full fiscal year. Here is what that number assumes.
JPMorgan projects that IREN Ltd. (NASDAQ: IREN) will generate $24 billion in annual revenue by 2030, roughly 34 times the $707.00 million the company reported for its most recent full fiscal year.
JPMorgan Chase & Co. (NYSE: JPM) has put a number on IREN Ltd. (NASDAQ: IREN) that dwarfs anything the company has actually reported: $24 billion in annual revenue by 2030. Against the $707.00 million IREN booked in its most recent full fiscal year, that is an increase of roughly 34 times over the span of the forecast — an illustrative ratio derived from the two figures, not a growth rate JPMorgan has published in that form.
The distinction matters more than usual here. This is a sell-side projection. It is not a reported result, and it is not company guidance. Nothing about it obliges IREN to do anything, and nothing about it has been audited. It is one bank's model of what a data-center operator's revenue base could look like at the end of the decade if a very specific set of assumptions holds.
What a 34x revenue assumption actually requires
Revenue multiples of that size do not come from selling more of the same thing. They come from a change in what the business sells. For IREN, that change is the pivot from bitcoin mining — where revenue is a function of hash rate, network difficulty and the coin price — toward leasing high-density compute capacity to customers training and running artificial-intelligence models.
Those two businesses have very different revenue characteristics. Mining revenue is spot-priced, volatile and capped by the block reward. AI compute is typically sold on multi-year contracts at a price per GPU-hour, and a single large customer commitment can add more annualized revenue than a mining fleet generates in total. That is the mechanism by which a bank can get from hundreds of millions to tens of billions inside a five-year model without assuming anything implausible about any single quarter.
What such a model does have to assume is that the power is there, the buildings are built, the chips are delivered, and the customers sign. Each of those is a separate execution risk, and each one compounds. Miss on energization timelines and the contracted revenue slides into a later year. Miss on chip allocation and it does not arrive at all. The 24/7 Wall St report frames the figure as a projection rather than a commitment, which is the right frame for anything with a 2030 date on it.
The shares went the other way into the weekend
Price action did not obviously reflect a fresh burst of optimism. IREN last traded at $44.13, down 4.39% on the session from a previous close of $46.15, with a day range of $43.41 to $46.49, as of the close on Fri, 25 Sep 2026, 20:00 GMT.
That was a decline against a broadly firm tape. The S&P 500 tracker closed at $771.35, up 0.54%. The Nasdaq 100 proxy finished at $744.50, up 0.46%. The Dow 30 vehicle ended at $517.49, up 0.94%. So the stock underperformed all three major benchmarks on the day — a reminder that names levered to AI infrastructure capital spending trade on their own news cycle, and on sentiment about financing and power contracts, rather than on the index.
JPMorgan itself closed at $343.06, up 1.33% from a prior close of $338.56, inside a day range of $338.64 to $343.31.
Where the $707 million baseline sits in the story
The reported figure is the anchor investors should keep in view. $707.00 million of full-year revenue is a real, historical, disclosed number. The $24 billion is a modeled endpoint. Everything between the two is assumption, and the gap between them is where the entire debate about this stock lives.
A few things follow from that.
- Valuation is being set by the forecast, not the base. Any investor paying today's price is implicitly underwriting some meaningful fraction of the growth path, not the current revenue run-rate.
- Contract disclosure is the scoreboard. The practical way to track whether the model is on course is announced customer agreements, their term lengths and their contracted capacity — not quarterly mining output.
- Capital intensity is the constraint. Building gigawatt-scale compute requires financing well ahead of revenue. Equity issuance, converts and project debt all have to be weighed against the revenue they are meant to unlock.
- A projection is a hypothesis. Analyst targets for 2030 have a long history of being revised — upward and downward — as each year's actuals land.
What would confirm or break the thesis
Everything between the two is assumption, and the gap between them is where the entire debate about this stock lives.
The next several quarters will do more to validate or undermine JPMorgan's number than any further research note. Watch for energization milestones at announced sites, the pace at which contracted megawatts convert into billed revenue, gross margin on AI cloud versus mining revenue, and the terms on which new capital is raised. Watch too for whether the revenue mix disclosed in filings actually shifts toward contracted compute, or whether bitcoin remains the dominant line.
There is also a sector-level question that no single company controls. The entire AI data-center buildout rests on sustained hyperscaler and model-developer spending. If that spending plateaus, or if the market concludes that compute supply has run ahead of demand, contract pricing softens for everyone — and long-dated revenue models get marked down across the group regardless of individual execution.
For now, the honest summary is this: a major bank has published a large number with a distant date attached, the company's last audited full-year revenue was $707.00 million, and the shares closed the week lower at $44.13. The arithmetic gap between those first two figures is the whole investment case, and it will be settled by delivery, not by forecasts.
Frequently asked questions
What exactly did JPMorgan project for IREN?
JPMorgan projected that IREN will generate $24 billion in annual revenue by 2030. It is an analyst projection produced by the bank's research process. It is not a reported financial result and it is not guidance issued by IREN itself, so the company has made no commitment to reach that figure by that date.
How does $24 billion compare with IREN's actual revenue?
IREN reported $707.00 million in revenue for its most recent full fiscal year. Dividing the projection by that base gives a ratio of roughly 34 times, an illustrative calculation using only those two figures. The gap between the reported base and the modeled endpoint is the core of the debate over the stock.
Where did IREN shares close most recently?
IREN last traded at $44.13, down 4.39% from a previous close of $46.15, with a session range of $43.41 to $46.49, as of the close on Friday, 25 September 2026 at 20:00 GMT. The market was closed at the time of writing, so this is the most recent traded price rather than a live quote.
Did IREN move with the broader market that day?
No. The S&P 500 tracker closed at $771.35, up 0.54%; the Nasdaq 100 proxy at $744.50, up 0.46%; and the Dow 30 vehicle at $517.49, up 0.94%. IREN fell 4.39% on the same session, underperforming all three benchmarks and trading on company- and sector-specific factors rather than index direction.
What would drive revenue growth of that magnitude?
A shift in the revenue mix from bitcoin mining toward contracted AI compute capacity. Mining revenue is spot-priced and volatile; AI compute is typically sold on multi-year contracts, where a single large customer commitment can add substantial annualized revenue. That change in business model, not incremental mining growth, is what a large long-range forecast implies.
What should investors track to test the forecast?
Announced customer contracts and their term lengths, site energization milestones, how quickly contracted capacity converts into billed revenue, margins on AI cloud revenue versus mining, and the terms of any new financing. Because the buildout requires capital ahead of revenue, funding costs and dilution are as relevant as the revenue targets themselves.
