Guggenheim calls Oracle a 'decade stock' despite recent tumble
Oracle (ORCL) is a buy-and-hold, says one Wall Street analyst.
"We continue to believe that ORCL is a decade stock, driven by the massive (and profitable) opportunity ahead for AI training and inferencing," Guggenheim analyst John DiFucci wrote in a note to clients.
Oracle's shares have stumbled in recent weeks — down 9% over the past month as investors digest heavy capex, customer concentration risks, and the financing required for its AI ambitions.
However, its shares are up 75% in the past 6 months and 57% year-to-date.
DiFucci wrote Oracle has an edge, pointing to its "durable technology advantages over hyperscaler competitors ... including clustering technology (RAC) for AI training workloads."
If Oracle wins the "AI training and inferencing" wave, it could meaningfully expand its cloud database and applications business — a segment still under-appreciated by the market, per DiFucci.
Oracle has begun sharing detailed five-year targets for revenue, EPS, and margins. At two recent investor meetings, CFO Doug Kehring discussed the financial mechanics behind Oracle's AI infrastructure push — from major contracts like the one with OpenAI (OPAI.PVT), to how the company plans to fund a build-out that could exceed $500 billion in future obligations.
Kehring highlighted three key features of its new contracts: they're non-cancelable, not tied to milestones or scale-backs, and they lock in GPUs, power, data-center access at signing. In short, DiFucci notes that Oracle says these AI deals are "committed to at signing."
But because many of those AI contracts are still new, Oracle hasn’t locked in how it will finance the massive expansion. Kehring said much of the related spending is 12 to 18 months away, leaving options for debt, vendor financing, or equity.
Still, relying heavily on a major customer like OpenAI introduces concentration risk. The timing uncertainty also raises questions about when cash outflows will hit and how they might affect margins.
DiFucci notes that Oracle's long-term EPS forecast though FY2030 — growing about 28% annually — already factors in financing costs.
Moreover, at its Financial Analyst Day, the company projected its AI data platform could reach about $20 billion in revenue by FY2030 — growing roughly 50% a year for the next five years. If it delivers, that higher-margin business could meaningfully boost earnings.
Yet for investors, heavy spending could slow margin expansion in the near-term, and returns from major AI training deals may take time to materialize.
Content Original Link:
" target="_blank">
