The number jumps off the page: $523 billion in contracted backlog against $66.5 billion in trailing revenue. That’s a 7.9x ratio—nearly eight years of current revenue already committed on paper.
Compare that to the hyperscalers: Microsoft at 1.4x, Alphabet at 0.5x, Amazon at 0.3x. Oracle isn’t playing the same game. It’s playing a different game entirely.
This is either visionary positioning or concentration risk dressed up as strategy.
The backlog surge is almost certainly anchored by a single counterparty: OpenAI. When one customer can move your backlog-to-revenue ratio from peer-comparable to industry-outlier, you haven’t diversified your future—you’ve outsourced it to someone else’s success.
The capex tell is even more revealing. Oracle’s FY25 capital expenditure jumped to $50 billion, up from $35 billion—a 43% increase in a single year. That’s not incremental capacity building. That’s a company racing to construct infrastructure for demand it has contractually locked in but not yet delivered.
The -32% stock decline over three months suggests the market sees the risk embedded in the opportunity. Eight years of pre-committed revenue sounds like stability until you consider: What happens if OpenAI’s trajectory changes? What happens if the AI infrastructure market commoditizes faster than Oracle can convert backlog to margin?
The strategic paradox: Oracle has potentially secured a decade of relevance in the AI era through a single relationship. But that security comes with a dependency that makes its fate inseparable from its largest customer’s.
In enterprise software, backlog is supposed to reduce risk. At 7.9x revenue, it might be the risk.
