Oracle carries a $638 billion AI backlog but funds expansion through debt, unlike cash-rich rivals Microsoft, Amazon, and Alphabet.

More than half of Oracle's backlog ties to OpenAI alone, creating dangerous customer concentration that justifies its discount to peers.

Down 57% from its 52-week high, Oracle trades at 13.6x forward earnings. That looks cheap if its massive backlog converts to recurring revenue.

Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks — and Oracle didn't make the cut. Grab the names FREE today.

Artificial intelligence has created an unusual investing environment. Companies willing to spend hundreds of billions of dollars building data centers are being rewarded with enormous growth expectations, while those sitting on the sidelines risk falling behind. The challenge is that AI infrastructure is expensive, and not every company has the balance sheet of Microsoft (NASDAQ:MSFT), Alphabet (NASDAQ:GOOG), Amazon (NASDAQ:AMZN), or Meta Platforms (NASDAQ:META). 

Oracle (NYSE: ORCL) is trying to join that elite club by borrowing aggressively to finance its cloud expansion. After its worst one-week stock performance in roughly 25 years, investors are beginning to ask whether the market is finally pricing in the risks as much as the opportunity.

Oracle's cloud infrastructure business (OCI) has become one of the fastest-growing AI platforms, driven by demand for GPU clusters and large language model training. According to Oracle's latest earnings release, the company now has an AI-related backlog of approximately $638 billion, one of the largest in the cloud industry.

Revenue estimates illustrate why investors have been excited.

Fiscal Year

Revenue Estimate

Growth

2026

$89.9 billion

33%

2027

$128.6 billion

43%

2028

$184.7 billion

44%

2029

$206.2 billion

12%

2030

$230.5 billion

11%

Earnings are expected to follow a similar trajectory.

Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks — and Oracle didn't make the cut. Grab the names FREE today.

Fiscal Year

EPS Estimate

Growth

2026

$8.09

5%

2027

$11.01

36%

2028

$15.57

42%

2029

$19.71

27%

2030

$22.27

13%

Those numbers explain why Oracle has been willing to take on substantial debt to expand capacity. Management is effectively betting today's borrowing costs against years of future AI demand.

The problem is that this isn't the same business model employed by hyperscalers. Microsoft, Amazon, Alphabet, and Meta generate tens of billions of dollars annually in free cash flow that can help fund expansion internally. Oracle must rely much more heavily on debt markets.

Borrowing itself isn't necessarily dangerous if the assets produce predictable cash flow. Utilities have operated that way for decades. Oracle's challenge is concentration.

More than half of its AI backlog is tied to OpenAI. That makes Oracle's investment case dependent not simply on AI demand remaining strong, but on one customer continuing to honor commitments over many years.

Granted, OpenAI remains one of the fastest-growing AI companies in the world. But customer concentration always deserves a discount because investors lose diversification. If OpenAI's infrastructure needs change, develops more internal capacity, or shifts workloads elsewhere, Oracle's return on those massive data center investments becomes less certain.

That's the risk investors appear to be repricing today.

With Oracle stock down 57% from its 52-week high -- and nearly 24% year-to-date -- the sell-off has compressed the stock to roughly 14 times forward earnings and less than 15 times projected 2028 EPS. Those valuation multiples look inexpensive for a company expected to grow revenue more than 40% annually through fiscal 2028.

Here's how Oracle stacks up against the competition:

Company

Primary AI Driver

Balance Sheet Advantage

Forward P/E

Microsoft

Azure

Massive free cash flow

19.2x

Alphabet

Google Cloud

Net cash position

22.8x

Amazon

AWS

Strong operating cash flow

23.1x

Meta Platforms

Llama

Strong liquidity

15.7x

Oracle

OCI

Debt-funded expansion

13.6x

The discount exists for a reason. Oracle is financing growth differently than its larger competitors, and investors are demanding compensation for that added risk.

In short, Oracle no longer looks expensive. At roughly 14 times forward earnings, much of the financing risk appears reflected in the share price. If Oracle converts even a large portion of its $638 billion backlog into recurring cloud revenue, today's valuation could prove unusually attractive.

That said, this is no longer a straightforward AI infrastructure story. It has become a wager that OpenAI continues expanding aggressively and fulfills the commitments underpinning much of Oracle's future growth. Until Oracle broadens that customer base, the stock probably deserves to trade at a discount to its hyperscale peers.

For long-term investors comfortable with customer concentration risk, today's valuation offers an appealing entry point. For more conservative investors, waiting for evidence that Oracle can diversify its backlog beyond OpenAI may be the more prudent path.

Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks — and Oracle didn't make the cut. Grab the names FREE today.