Oracle Corp. (ORCL, Financial) is currently spending heavily on its artificial intelligence infrastructure, which is driving higher fundamental growth forecasts by analysts. Despite this positive, the stock's valuation is likely too high right now and I believe that it is one of the least likely companies to develop AGI first.

The result of this analysis is that a better entry point may be worth considering for Oracle despite the formidable niche moat it is carving in special use cases for AI-assisted SaaS.

Key developments

The most important operational developments for Oracle are currently threefold, in my opinion.

First, Oracle has announced a new collaboration with Palantir (PLTR, Financial), integrating the company's foundry and AI platforms with Oracle Cloud Infrastructure ("OCI"). This is a significant partnership for strengthening Western defense capabilities, and it is one of the symbols that the U.S. is potentially beginning to focus more on utilizing commercial sector capabilities for international security. The integration should help the company meet regulatory, performance and security needs, enhancing the quality and utility of large-scale AI deployments.

Further, Oracle has reportedly canceled its potential deal with Elon Musk's xAI for expanded cloud computing to train large-language models. There was a potential $10 billion agreement on the table, but xAI has decided to build its own data center using Nvidia (NVDA, Financial) chips instead. Despite this loss, which contributed to a 3% price drop for the stock, it is worth noting that Oracle still has major cloud AI contracts, including with OpenAI.

Finally, and perhaps most importantly, Oracle has recently expanded its AI capabilities by introducing new Nvidia GPU options on OCI. This will help a full range of businesses of different sizes leverage AI capabilities through cloud access. The new offerings include Nvidia L40S GPUs for virtual machine options, Nvidia H100 Tensor Core GPUs to support a wide range of AI computations and up to 65,000 Nvidia GPUs supported by OCI Supercluster, which enables large-scale AI model training with hundreds of billions of parameters.

Oracle benefits from the fact it focuses on sovereign AI solutions, meaning it ensures its data remains within national borders. This is particularly attractive to government clients, so I think the company has the potential to become more involved in AI defense capabilities as its cloud infrastructure scales.

In addition, not only is Oracle's approach to AI accretive, it is also necessary. A failure to compete in the AI arms race would likely pose a medium-term to long-term existential threat to its business. The nature of information and database work is rapidly changing as a result of machine learning capabilities, particularly in regard to automation and efficiency of data access. As a result, Oracle is doing the right thing by investing heavily to stay at the cutting edge of advanced SaaS capabilities.

Moderately overvalued despite expanding fundamental growth

The GF Value Line currently shows Oracle to be modestly overvalued. I think this is a fair analysis as it presents some caution to investors. However, the valuation expansion, which includes a price-earnings ratio of 37, a forward price-earnings ratio of 22 and a price-sales ratio of 7.30, is due to the growing consensus that the company is likely to experience significantly higher growth than in recent years. For comparison, Oracle's earnings multiple has a 10-year median of 20. Therefore, the expansion in multiples already largely prices in the higher fundamental growth forecasts analysts have reported.

The question is whether Oracle can achieve higher growth than what is currently anticipated on Wall Street. I think it is possible, but estimates are already relatively aggressive.

In the fiscal fourth quarter of 2024, Oracle signed over 30 contracts totaling more than $12.50 billion, including significant deals with companies like OpenAI to train ChatGPT on OCI. This demand is expected to continue despite the lower demand forecasted for Nvidia GPUs as the first wave of AI infrastructure build-out culminates.

In my estimation, it is likely big tech companies will reach a threshold where more GPUs do not equate to more computational capability, especially as the majority of vast data sets get tapped and adequate computational power is provided. Nvidia's growth may momentarily slow, but companies like Amazon (AMZN, Financial) and Oracle should continue to generate large growth over the long term from the third-party use cases of their data center capabilities powered by Nvidia GPUs. As a result of this analysis, I am very bullish on Oracle.

Despite the price-earnings ratio being quite high, it is not as high as Amazon's, but it is higher than Meta's (META, Financial); both competitors have higher fundamental growth forecasts than Oracle. Meta has a future three to five-year earnings per share without nonrecurring items growth rate of 21%, while Amazon has a future growth rate of 37%. Oracle's rate is 14%.

Meta seems like the better play here because its future growth is higher, but its valuation is lower than Oracle's. Furthermore, I think the long-term narrative and operational direction currently being constructed by Meta is very strong. Based on my analysis of its open-weight approach to frontier model AI, I think Meta could be the first to develop AGI.

I believe that Meta AI will begin to be tapped by a range of commercial use cases, so its long-term growth trajectory with regard to AI is vaster than Oracle's as a result of its higher capital expenditures on computation. Meta's capex for 2024 is projected to be between $37 billion and $40 billion. In contrast, Oracle has spent about $7.50 billion.

AI spending concerns, competitive pressure and integration risks

There is currently growing uncertainty about the return on investment of substantial AI infrastructure investments. If the AI and cloud services do not generate the expected revenue growth, it could have a straining effect on Oracle's earnings. Much of the success of these investments depends on the market's adoption of its AI and cloud services, and there may be pricing power that larger moats like those being developed by Meta, Amazon and Microsoft (MSFT, Financial) command that could undercut Oracle. This is why it is important for Oracle to continue to focus on industry-specific data management with enhanced AI capabilities because this will allow the company to retain a strong position in its dominant fields of operation.

I think it is quite likely that Oracle falls further behind competitors; not just in infrastructure moat, but also in AI innovation. The simple reason for this is that Oracle has a significantly lower market cap than the other major tech companies in my analysis.

The result of this is lower capex, lower research and development and lower levels of innovation. The AI arms race is absolutely a game of monetary and computational power; it is highly unlikely that smaller players will have the moat necessary to command the power that deployed AGI will present.

There are also integration challenges that could arise, as well as difficulty with customer adoption. There is a likelihood of some industries being more receptive to AI capabilities than others, and management needs to ensure the transition is seamless for its customers. I have had first-hand experience using Oracle services in professional settings before, and the experience is certainly less smooth than using Apple (AAPL, Financial) interfaces. As much of the attraction of AI is automation and ease of use, Oracle must ensure it tries to engender this ethos on its updated data management software; a failure to do so could mean it loses funds to larger, more streamlined competitors.

Conclusion

Oracle is a player in the AI arms race, but it is both unlikely to win and unlikely to develop AGI anytime soon. This does not mean its AI investments are unworthy; they simply need to be added to the context of the business profile when considering an investment. For example, with a current higher valuation but lower future fundamental growth rates than Meta, I think Meta is certainly the better investment at this time, as I think sentiment could grow for it more readily than Oracle in the medium term.

That being said, both companies appear overvalued and I think it is worth refraining from too heavy an investment in AI at this time when market sentiment is already incredibly high. It is arguably better to wait for a more appealing valuation, which I think could present itself once Nvidia begins to show contracting fundamental growth rates, potentially reducing its stock price and the overall sentiment on AI stocks.