Piper Sandler: Oracle Cloud FY27 Revenue May Exceed Expectations
Miles Bennett
Piper Sandler analyst Billy Fitzsimmons estimates Oracle's cloud infrastructure could generate roughly $23 billion in new FY27 revenue — $2.2 billion above his current forecast; the key test is whether capex converts to live capacity on schedule.
How did he arrive at "above expectations"?
Fitzsimmons built a per-megawatt cost-and-revenue model using public disclosures from Crusoe and CoreWeave as benchmarks.
Key assumptions: ~$46 million per MW in build cost, ~$13.5 million per MW in IaaS revenue — IaaS being the most basic "rent computing power" layer of cloud services.
In plain terms = he calculated how much it costs to build one megawatt of data-center capacity and how much revenue it can generate, then scaled that up to Oracle's planned total.
Where does the extra $2.2 billion come from?
Based on Oracle's expected FY27 capex range, Fitzsimmons projects ~2,400 MW of new usable capacity, translating to roughly $23 billion in incremental OCI revenue.
That is $2.2 billion above his current estimate of $20.8 billion — implying roughly 12% upside to FY27 OCI revenue growth.
This means → if all planned capex lands, total FY27 OCI revenue could reach $41.1 billion, materially above current consensus.
How reliable is this estimate — and where is the risk?
Fitzsimmons maintains an Overweight rating on Oracle with a $225 price target — he stands behind the model.
The critical assumption: capex must convert to live, sellable capacity on schedule — building the data centers is only half the equation.
In plain terms = money spent does not equal revenue earned. The ramp from "construction starts" to "customers running workloads" takes time, and that ramp period is the biggest source of uncertainty.
Content is for reference only, not financial advice.