In response to ongoing market concerns about customer concentration and capital expenditures, Bernstein’s latest research report conducts an extreme scenario stress test on Oracle: if AI clients like OpenAI completely withdraw, what is the company’s valuation?
According to the Wind Trading Platform, the model constructs a “near-worst-case” downside assumption: all AI contracts signed since Q4 FY2025, including massive orders from OpenAI, fail to generate revenue, with customers neither fulfilling nor renewing their contracts. Meanwhile, core database, SaaS, and non-AI OCI businesses maintain normal growth.
Under this extreme scenario, Bernstein estimates Oracle’s valuation floor at $137 per share, leaving only about 15% downside from the current level of around $160, providing a clear safety margin for investors. In an optimistic scenario with smooth execution, the target price could reach $313.
The conclusion shows a strong asymmetry. Even excluding all AI-related revenue, Oracle’s core enterprise business alone can support a stable valuation. Bernstein believes that the market’s fears over AI customer dependence and capital expenditure pressures are overly priced in, and the current risk-reward profile is attractive.
$248 billion leasing liabilities are not “unbearable”
Bernstein further breaks down market concerns about Oracle’s massive data center leasing contracts. Investors fear that if clients default, Oracle could face up to $248 billion in idle costs. However, the report points out that this risk is significantly overstated.
First, these leasing contracts have durations of 15 to 19 years, with capacity phased in until FY2030. The maximum annual risk exposure is actually only $13 to $16.5 billion, peaking only in 2030. Second, as long as the global AI winter does not occur, data center demand will remain high, and Oracle can fully utilize or sublease idle space.
More importantly, even excluding generative AI, the potential size of the global IaaS/PaaS market has reached $1.2 to $1.4 trillion, and the capacity consumed by traditional enterprise cloud migrations is sufficient to absorb these leasing spaces. In other words, Oracle’s long-term contract structure has ample buffers, and market fears over its capital commitments lack substantive support.
Hardware risks are manageable; won’t be “locked in” by GPU inventory
Addressing market concerns about hardware capital expenditure risks, Bernstein states that the actual exposure is very limited. Server and GPU procurement cycles typically occur 3 to 6 months before capacity goes online. If clients cancel contracts, Oracle can flexibly cancel or delay orders before delivery without significant penalties.
Even if hardware has been delivered, most compute assets such as storage, networking, and memory devices are highly versatile and can be seamlessly repurposed for traditional SaaS and OCI businesses. The report further estimates that every $1 of annual AI revenue corresponds to approximately $2 to $2.5 of compute assets, with an asset lifespan of six years. This means that even in the event of client default, only about 1 to 2 years of compute expenditure is at risk. Overall, the risk of hardware capital expenditure being “locked in” is significantly overstated.
Strong fundamentals of core business
In its latest model, Bernstein further isolates AI business to clearly outline the intrinsic value of Oracle’s core operations. Even if all AI data center-related revenues are eliminated, the company’s traditional database, SaaS, and OCI businesses alone could generate total revenues of $101 billion by FY2030.
More importantly, after accounting for interest costs on debt raised for AI infrastructure (including the recent $25 billion financing), earnings per share could still reach $9.0. Using the peer multiple of 27.3x (as seen with Microsoft, SAP, Salesforce), this earnings outlook implies a valuation of $137 per share. In other words, even if AI business is completely zeroed out, Oracle’s core business alone can support about 85% of its current share price, demonstrating the robustness of its cash-generating capacity.
Bernstein believes that the current Oracle stock price is overly priced in pessimistic expectations, and the risk-reward ratio is entering an extremely attractive zone. With recent debt and equity financing of approximately $25 billion each, the funding bottleneck for data center expansion has been substantially alleviated. The report emphasizes that as long as the global economy does not fall into an “AI winter” scenario, the current valuation provides investors with ample safety margins and asymmetric upside potential.
This insightful content is from Wind Trading Platform.
For more detailed analysis, real-time insights, and frontline research, please join the【**Wind Trading Platform▪Annual Membership**】

Risk Disclaimer and Terms of Disclaimer
Market risks exist; investments should be cautious. This article does not constitute personal investment advice and does not consider individual users' specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their particular circumstances. Investment is at their own risk.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Oracle's most pessimistic assumption: If all AI data center contracts are terminated
In response to ongoing market concerns about customer concentration and capital expenditures, Bernstein’s latest research report conducts an extreme scenario stress test on Oracle: if AI clients like OpenAI completely withdraw, what is the company’s valuation?
According to the Wind Trading Platform, the model constructs a “near-worst-case” downside assumption: all AI contracts signed since Q4 FY2025, including massive orders from OpenAI, fail to generate revenue, with customers neither fulfilling nor renewing their contracts. Meanwhile, core database, SaaS, and non-AI OCI businesses maintain normal growth.
Under this extreme scenario, Bernstein estimates Oracle’s valuation floor at $137 per share, leaving only about 15% downside from the current level of around $160, providing a clear safety margin for investors. In an optimistic scenario with smooth execution, the target price could reach $313.
The conclusion shows a strong asymmetry. Even excluding all AI-related revenue, Oracle’s core enterprise business alone can support a stable valuation. Bernstein believes that the market’s fears over AI customer dependence and capital expenditure pressures are overly priced in, and the current risk-reward profile is attractive.
$248 billion leasing liabilities are not “unbearable”
Bernstein further breaks down market concerns about Oracle’s massive data center leasing contracts. Investors fear that if clients default, Oracle could face up to $248 billion in idle costs. However, the report points out that this risk is significantly overstated.
First, these leasing contracts have durations of 15 to 19 years, with capacity phased in until FY2030. The maximum annual risk exposure is actually only $13 to $16.5 billion, peaking only in 2030. Second, as long as the global AI winter does not occur, data center demand will remain high, and Oracle can fully utilize or sublease idle space.
More importantly, even excluding generative AI, the potential size of the global IaaS/PaaS market has reached $1.2 to $1.4 trillion, and the capacity consumed by traditional enterprise cloud migrations is sufficient to absorb these leasing spaces. In other words, Oracle’s long-term contract structure has ample buffers, and market fears over its capital commitments lack substantive support.
Hardware risks are manageable; won’t be “locked in” by GPU inventory
Addressing market concerns about hardware capital expenditure risks, Bernstein states that the actual exposure is very limited. Server and GPU procurement cycles typically occur 3 to 6 months before capacity goes online. If clients cancel contracts, Oracle can flexibly cancel or delay orders before delivery without significant penalties.
Even if hardware has been delivered, most compute assets such as storage, networking, and memory devices are highly versatile and can be seamlessly repurposed for traditional SaaS and OCI businesses. The report further estimates that every $1 of annual AI revenue corresponds to approximately $2 to $2.5 of compute assets, with an asset lifespan of six years. This means that even in the event of client default, only about 1 to 2 years of compute expenditure is at risk. Overall, the risk of hardware capital expenditure being “locked in” is significantly overstated.
Strong fundamentals of core business
In its latest model, Bernstein further isolates AI business to clearly outline the intrinsic value of Oracle’s core operations. Even if all AI data center-related revenues are eliminated, the company’s traditional database, SaaS, and OCI businesses alone could generate total revenues of $101 billion by FY2030.
More importantly, after accounting for interest costs on debt raised for AI infrastructure (including the recent $25 billion financing), earnings per share could still reach $9.0. Using the peer multiple of 27.3x (as seen with Microsoft, SAP, Salesforce), this earnings outlook implies a valuation of $137 per share. In other words, even if AI business is completely zeroed out, Oracle’s core business alone can support about 85% of its current share price, demonstrating the robustness of its cash-generating capacity.
Bernstein believes that the current Oracle stock price is overly priced in pessimistic expectations, and the risk-reward ratio is entering an extremely attractive zone. With recent debt and equity financing of approximately $25 billion each, the funding bottleneck for data center expansion has been substantially alleviated. The report emphasizes that as long as the global economy does not fall into an “AI winter” scenario, the current valuation provides investors with ample safety margins and asymmetric upside potential.