Even if the AI data center business drops to zero, there's no need to worry! Bernstein: Oracle Corporation (ORCL.US) stock is undervalued, maintains $325 price target.
Even if the AI data center business drops to zero, there's no need to worry! Bernstein: Oracle Corporation (ORCL.US) stock is undervalued, maintains $325 price target.
Although the market continues to worry about the debt burden, customer performance risks, and capital expenditure pressure brought by Oracle's heavy investment in AI data centers, Bernstein believes that Oracle's current stock price is already below its theoretical valuation assuming it does not rely on the AI data center business at all.
Despite persistent market concerns over the debt burden, customer performance risks, and capital expenditure pressure brought by Oracle Corporation's (ORCL.US) heavy investment in artificial intelligence (AI) data centers, Bernstein believes that Oracle Corporation's current stock price is already below its theoretical valuation assuming it does not rely on the AI data center business at all.
In a research report released on November 18, Bernstein pointed out that even assuming all of Oracle Corporation's AI data center customers stop honoring their contracts, revenue from the related business drops completely to zero, the approximately US$40 billion in hardware assets already purchased cannot be monetized, and the company still has to bear existing debt and interest expenses, the per-share valuation of its core business could still reach US$133.
Compared with Oracle Corporation's closing price of US$115 on November 17, this valuation under the extremely pessimistic scenario is still about 16% higher. Bernstein therefore reiterated its "Outperform" rating on Oracle Corporation and maintained its US$325 price target, corresponding to potential upside of about 183%.
Oracle Corporation Still Worth US$133 Even If AI Business Earns Not a Single Cent?
Bernstein pointed out that Oracle Corporation has become one of the most controversial stocks in the technology industry. Investor concerns are mainly concentrated in four areas: the economic returns of AI data center investment, massive capital expenditures and financing needs, the company's execution capability, and the long-term sustainability of its existing business.
In particular, the large-scale AI computing power contracts Oracle Corporation signed with technology companies such as OpenAI, although providing support for future revenue growth, have also raised market doubts about whether customers can continue to honor them.
To assess the potential downside risks facing Oracle Corporation, Bernstein constructed an extremely pessimistic valuation scenario. In this model, AI data center customers such as OpenAI, Meta (META.US), NVIDIA Corporation (NVDA.US), and AMD (AMD.US) all stop honoring their contracts, and even though the relevant contracts are in principle non-cancellable, it is assumed that Oracle Corporation cannot obtain compensation for breach of contract.
The approximately US$40 billion in hardware assets Oracle Corporation has purchased, including servers, network equipment, and power and cooling systems, is regarded as worthless, neither usable nor sellable for cash.
More conservatively, Bernstein assumes that Oracle Corporation still has to bear the debt and related interest expenses previously raised to build AI data centers, and incorporates the share dilution caused by equity financing into the valuation.
In other words, this scenario not only completely removes the future revenue and profit of the AI data center business, but also retains the financial burden already generated by the related investments.
Even so, Bernstein estimates that the fair valuation of Oracle Corporation's core business still reaches US$133 per share. This valuation is lower than the US$137 estimated by the institution in February 2026, mainly because Oracle Corporation has since increased debt, purchased more hardware, and expanded the scale of its AI data center business.
However, Bernstein emphasized that this model does not include an "AI winter" scenario in which global generative AI demand collapses completely, but instead assumes that Oracle Corporation's existing AI customers default while other cloud computing and software businesses continue to develop normally.
Traditional Business Still Has Growth Potential; Fiscal 2030 Revenue Expected to Approach US$100 Billion
Bernstein believes that the market is overly focused on the investment risks of Oracle Corporation's AI data centers while ignoring the long-term value of its traditional software and cloud computing businesses. The institution pointed out that Oracle Corporation's database business has high customer stickiness and profit margins, and software services such as enterprise resource planning (ERP) and human capital management (HCM) also have strong business stability.
At the same time, the traditional computing business within Oracle Corporation's cloud infrastructure (OCI) that does not rely on generative AI, as well as sovereign cloud, private cloud, and multicloud services, are still expected to benefit from the ongoing trend of enterprises migrating to the cloud.
According to Bernstein's estimates, even after completely excluding AI data center revenue, Oracle Corporation's core business revenue is still expected to grow from about US$61 billion in fiscal 2026 to US$98 billion in fiscal 2030. Among this, core business revenue is expected to be about US$67 billion in fiscal 2027, increase to US$75 billion in fiscal 2028, and further reach US$85 billion in fiscal 2029.
As revenue grows and operating leverage gradually emerges, Oracle Corporation's non-AI business operating margin is expected to rise from 35% in fiscal 2026 to 39% in fiscal 2030.
While continuing to include the interest costs of AI data center financing, Bernstein expects that Oracle Corporation's core business net profit in fiscal 2030 can still reach US$27.1 billion, with earnings per share of about US$10.8.
In terms of valuation methodology, Bernstein referenced the valuation levels of software companies such as Microsoft Corporation (MSFT.US), SAP (SAP.US), Salesforce, Inc. (CRM.US), and ServiceNow (NOW.US). The average forward price-to-earnings ratio of these companies is about 24.4 times. Bernstein assumed that the compound annual growth rate of Oracle Corporation's core business earnings per share over the next two years would be about 20%, and applied a fair price-to-earnings ratio of 25 times, corresponding to a valuation of about US$133 per share.
Bernstein believes that, considering the competitive advantages of Oracle Corporation's database business and the high stickiness of enterprise software customers, its core business could in theory even enjoy a certain valuation premium, but the institution did not include this factor in its model out of caution.
US$288 Billion Data Center Lease Becomes Market Focus; Bernstein Believes Risks Are Exaggerated
In addition to the performance capability of AI customers, the massive data center lease contracts signed by Oracle Corporation are also an important source of market concern. Bernstein pointed out that Oracle Corporation previously signed data center lease agreements with a total value of about US$288 billion, with lease terms of 15 to 19 years. These lease commitments are closely related to the approximately US$317 billion in non-cancellable contract value the company obtained from customers such as OpenAI.
Investors worry that once major customers such as OpenAI fail to fulfill their computing power procurement commitments, Oracle Corporation may be forced to bear rent for idle data centers for a long time, putting sustained pressure on free cash flow.
However, Bernstein believes that this concern ignores the reuse value of data center assets.
First, Oracle Corporation's data center capacity will not all be put into use at once, but will be delivered gradually over the next several years. Therefore, even if some customers fail to honor their contracts, the company still has time to find other customers or adjust usage. Second, the US$288 billion in lease commitments is spread over contract periods of 15 to 19 years, which does not mean that Oracle Corporation needs to pay the full amount in the short term.
Bernstein estimates that the peak annual expenditure for the related leases is about US$15 billion to US$19 billion, and this level is expected to be reached only around fiscal 2030. In addition, the institution believes that as long as the generative AI industry does not collapse completely, market demand for data centers with sufficient power supply will remain relatively high.
Even if original customers exit, Oracle Corporation may also reduce losses from idle assets by subleasing data centers or using the related facilities for traditional cloud computing business.
Bernstein estimates that the potential annual size of the global infrastructure-as-a-service (IaaS) and platform-as-a-service (PaaS) market is about US$1.2 trillion to US$1.4 trillion, while the current actual market size is less than half of this potential scale.
Therefore, data center demand brought by the ongoing migration of enterprises to the cloud is expected to provide additional utilization space for Oracle Corporation's existing assets.
US$40 Billion in Hardware Assets May Not Be Sunk Costs
Regarding the GPU and server procurement risks that investors worry about, Bernstein also believes that Oracle Corporation has greater flexibility than the market expects. The institution estimates that for every US$1 of annual AI revenue generated by Oracle Corporation, about US$2.5 to US$3 of computing assets need to be invested, and the expected useful life of the related assets is about 6 years.
However, Oracle Corporation usually purchases the relevant hardware only about 3 to 6 months before a data center is put into operation, rather than completing all purchases years in advance. This means that even if AI customers suddenly stop honoring their contracts, Oracle Corporation may not necessarily have already borne the full hardware procurement costs for future years.
Bernstein estimates that Oracle Corporation currently has about US$40 billion in hardware assets in the construction or procurement process, including not only GPU servers but also storage, networking, memory, and power supply and cooling equipment.
Some of these assets can be redirected to traditional cloud computing workloads or recover value through sale or leasing. It is worth noting that in the aforementioned pessimistic valuation model of US$133 per share, Bernstein did not include any recovery gains that these hardware assets might generate. The institution believes this further demonstrates the conservatism of its valuation assumptions.
AI Investment Returns Expected to Gradually Materialize; Fiscal 2030 Free Cash Flow May Reach US$80 Billion
Under the normal operating scenario, Bernstein remains optimistic about Oracle Corporation's growth prospects over the next several years. The institution expects that as AI data center contracts gradually convert into actual revenue, Oracle Corporation's total revenue in fiscal 2030 is expected to reach about US$234 billion, higher than the company's previously stated target of US$225 billion. Among this, AI cloud infrastructure business revenue is expected to reach about US$136 billion, becoming an important driver of the company's overall growth.
However, the gross margin of the AI data center business is relatively low, and as its revenue share increases, Oracle Corporation's overall gross margin may decline from about 66% in fiscal 2026 to 47% in fiscal 2030.
Nevertheless, Bernstein expects that as data centers are gradually put into operation and capital expenditure growth slows, Oracle Corporation's free cash flow position will improve significantly. According to financing and cash flow plans previously disclosed by management, after completing about US$20 billion in new financing, the company is expected to need no further fundraising and is expected to achieve positive free cash flow in the second half of fiscal 2028.
Bernstein expects Oracle Corporation's free cash flow to reach about US$30 billion in fiscal 2029 and further increase to about US$80 billion in fiscal 2030. Over the longer term, the institution's model shows that Oracle Corporation's revenue in fiscal 2035 is expected to approach US$396 billion, operating profit of about US$160 billion, and free cash flow may reach US$233 billion.
However, these long-term forecasts are highly dependent on continued growth in AI data center demand, existing customers honoring their contracts, and the company smoothly executing its expansion plans, and are not operating results that are certain to be achieved.
Stock Price Has Fallen Sharply from Its Year-High; Bernstein Optimistic About Risk-Reward Ratio
As of the close on November 17, Oracle Corporation's stock price was US$115, down about 47.7% over the past 12 months and down about 41% year to date.
Bernstein believes that Oracle Corporation's current stock price not only fails to fully reflect the value that the AI data center business may create in the future, but is even lower than its core business valuation under the extremely pessimistic assumption. The institution pointed out that the market may be assigning too low a valuation to Oracle Corporation's traditional software business, or may have already priced in some of the risk of a collapse in AI industry demand in advance.
However, Bernstein believes that Oracle Corporation's core database and enterprise software businesses remain highly competitive, and once AI data center investments begin to generate the expected cash flow, they will bring significant additional value to the company.
The institution expects that as more data centers are put into operation in the coming quarters, customer contract execution is verified, and investor confidence in OpenAI's long-term business prospects improves, Oracle Corporation's stock price is expected to gradually recover.
Bernstein said that a shift in market sentiment could take 18 months, or it could happen sooner. Based on the support provided by the core business valuation and the potential long-term growth opportunities of the AI data center business, Bernstein reiterated its "Outperform" rating on Oracle Corporation and maintained its US$325 price target.
Related Articles

CITIC SEC: Cost reduction and profit growth; the economic inflection point for domestic standalone energy storage is approaching.

FIN STREET SEC (01476): Chairman Zhu Yanhui is under investigation and subject to retention measures

China Securities Co.,Ltd.: AI advanced packaging drives upgrading of glass substrate materials; equipment takes the lead in the early stage of industrialization.
CITIC SEC: Cost reduction and profit growth; the economic inflection point for domestic standalone energy storage is approaching.

FIN STREET SEC (01476): Chairman Zhu Yanhui is under investigation and subject to retention measures

China Securities Co.,Ltd.: AI advanced packaging drives upgrading of glass substrate materials; equipment takes the lead in the early stage of industrialization.






