September 9, 2026
Bonus Content: Oracle’s $638B Backlog Hinges on Its Build Timeline
The Copper Window May Still Be Early
The best part of a resource cycle is rarely when everyone is already talking about it.
The more interesting moment is before it gets crowded. Before the big headlines turn into consensus. Before smaller names tied to future supply start getting pulled into the spotlight.
That may be where copper is now.
S&P Global forecasts global copper demand will climb up to 50% over the next 14 years, from 28 million metric tons to 42 million by 2040. AI is adding a new layer to that pressure.
One Microsoft datacenter reportedly used more than 4.8 million pounds of copper. Now there are more than 3,000 data centers planned in the U.S. alone and more than 12,000 worldwide.
Meanwhile, new supply remains difficult to bring online. Declining ore grades, permitting delays, and higher costs are all making future copper harder to develop.
That’s why North American copper stories may matter more now.
This one has more than 62,000 feet of historical drilling, five drill-ready targets, 100% project control, fresh exploration funding, and a proven, seasoned team.
If copper is entering a new cycle, early investors will want to know where the next supply stories are forming.
Oracle’s $638B Backlog Hinges on Its Build Timeline
The partnership announcements look clean on paper. Oracle has deepened multicloud database alliances with AWS, Microsoft, and Google Cloud, placing Oracle AI Database@AWS, Oracle AI Database@Azure, and Oracle AI Database@Google Cloud on the menu for each company’s respective technology partners. A new licensing option lets customers purchase Oracle’s AI database services across all four environments using a single consolidated commitment called Multicloud Universal Credits. Three hyperscaler partnerships, one billing standard, and a database that enterprise IT refuses to abandon. The structure is genuinely elegant.
The problem is what sits behind it.
Remaining Performance Obligations ended Q4 fiscal 2026 at $638 billion, up 363% year-over-year and up $85 billion sequentially from Q3. Oracle said most of the RPO increase in both Q3 and Q4 came from large-scale AI contracts involving customer prepayments and customer-supplied hardware. That backlog is, by any reasonable measure, staggering. It is roughly ten times larger than annual revenue and one of the largest ever disclosed by a public software company. But a signed contract is a promise to build, not proof of delivery.
Oracle reported Q4 revenue of $19.2 billion, up 21% year-over-year, with cloud infrastructure (IaaS) revenue growing 93% to $5.8 billion. Those are hyperscaler growth rates. The underlying demand is real. What the market is currently pricing is whether Oracle can physically build fast enough to honor what it has sold.
The Capital Cost of Ambition
Fiscal 2026 capital expenditures were $55.7 billion, producing negative free cash flow of $23.7 billion despite $32.0 billion of operating cash flow. Oracle raised $43 billion in debt financing and $5 billion in equity financing in fiscal 2026 and said it intends to raise approximately $40 billion more in fiscal 2027. The scale of that debt load is not trivial against any historical comparison for a software company.
Oracle CEO Clay Magouyrk said FY2027 Q1 delivery is approaching one gigawatt, nearly the same capacity Oracle delivered in the previous four quarters combined. That compression of timelines is the whole thesis: construction acceleration translates directly to revenue recognition. CFO Hilary Maxson said FY2027 gross margin is expected to step down due to data center ramp timing and mix, with infrastructure margins expected to improve as data centers reach full contractual revenue contribution.
Bull / Base / Bear
- Bull: Construction timelines hold. Oracle has already sold years of future OCI capacity to the world’s most capital-rich AI companies, and if it can build and light up that capacity on schedule, the revenue and EPS ramp through fiscal 2029 is meaningfully supported by contracted demand. The $90 billion FY2027 revenue target becomes a floor, not a ceiling.
- Base: Execution is uneven. Data center delivery slips by one to two quarters in select regions, margin recovery lags guidance, and ORCL trades sideways at current levels while waiting for free cash flow to inflect positive.
- Bear: The path forward is paved with financial risks. Oracle’s unprecedented capital spending and debt levels mean there is little room for error in its execution. A sustained delay in capacity delivery breaks the RPO-to-revenue conversion story the entire valuation depends on.
What to Watch
The single most important metric going forward is not RPO growth. That number is already enormous. The variable that actually moves the stock is conversion velocity: how many dollars of that $638 billion backlog Oracle can recognize as revenue in FY2027 against its $90 billion guidance commitment. The key metric to watch is the speed at which Oracle’s data centers can turn booked contracts into realized cash flow. Every quarter that gap narrows, the discount the market applies to the backlog shrinks with it.
Bottom Line
Oracle’s multicloud database partnerships have made it structurally indispensable across every major cloud environment. The demand is signed, and much of it is backed by customer commitments that Oracle characterizes as long-duration. The question that determines whether ORCL at current levels is a generational opportunity or a value trap is entirely an operational one: how fast can they pour concrete.
