Morgan Stanley's Deep Dive with Dell COO on "This Time Is Different": AI Agents Are Not Another Refresh Cycle but TAM Expansion; Probability of $756 Bull Case Rises

Zhitong Finance APP learned that Morgan Stanley released a research report stating that after a one-on-one meeting with Chief Operating Officer Jeff Clarke at Dell's headquarters, the firm gained a deeper understanding of the bullish logic for Dell but maintained its "Equal-weight" rating and $511 price target. The core debate in the report is whether AI agents represent just another peak in the hardware cycle or the beginning of a multi-year structural boom in local infrastructure. Morgan Stanley explicitly stated that the meeting led it to reassess its previously bearish judgment and believes the probability of achieving the $756 bull case scenario within the next 12 months has increased.

This Time Is Different: From Refresh Cycle to TAM Expansion

Clarke emphasized to Morgan Stanley that the hardware market has experienced eight boom-and-bust cycles over the past 40-plus years. However, these cycles were primarily driven by fluctuations in refresh rates resulting from changes in per-capita device penetration, rather than a structural expansion of the total addressable market (TAM). For example, the PC refresh cycle accelerated during the pandemic as the ratio of PCs to people increased, but this ratio normalized after society reopened, meaning the PC TAM was not larger than it was before the pandemic.

The difference with AI agents is that they decouple cognitive output from headcount for the first time, allowing enterprises to complete more work without adding staff, with productivity gains reaching 10x or even 100x. Dell believes this return on investment is clear, and various enterprises will proceed along this path. By 2030, inference tokens are expected to grow 87-fold, adding 200 gigawatts of computing power, with ZettaFLOPS growing to 830. Since agents require 1 billion more tokens per unit of work than basic chatbots, token growth will force data centers to restructure towards accelerated computing and higher CPU-to-GPU ratios, catalyzing a structural rise in computing and storage demand.

Dell also believes that AI agents will adopt a hybrid deployment model. Enterprises will dynamically allocate inference workloads between public clouds and on-premises facilities based on security and cost trade-offs. Dell itself will place content-related workloads on the public cloud but will never move proprietary source code or telemetry data out of Dell facilities. Given that inference tokens are projected to grow 87-fold by 2030, "even taking a small slice of the large pie will result in a very large number." The proliferation of open-source weight models will also drive enterprises to increase investment in local infrastructure to optimize costs.

Servers and Storage: Density Increases Coexist with Supply Bottlenecks

Traditional server shipments are still declining year-over-year, but Dell argues this is not due to weak demand, but because servers are becoming more powerful, denser, and more efficient. The 17th/18th generation Dell servers can replace up to 13 units of 14th generation traditional servers. In the short term, server unit counts are falling while average selling prices (ASP) are rising; as data centers rebuild around accelerated computing and the adoption of agents boosts demand for CPU servers, denser server units will eventually resume growth. Clarke even pointed out that if inference tokens truly grow 87-fold over five years as expected, traditional server unit growth could also become exponential. Storage benefits similarly, as behaviors like KV Cache and agent artifact retention will create greater demand for data storage and knowledge preservation.

Supply is the biggest challenge facing Dell. Clarke stated that historical commodity cycles typically last two to four quarters, driven by mature hardware replacement cycles or supply decision booms/busts; however, if one believes in the growth algorithms for gigawatts and tokens, shortages of memory, HDDs, etc., are not short-term cycles, and shortages could persist for another five years. When asked what keeps him awake at night, Clarke's answer was not macroeconomics, geopolitical conflicts, data center overbuilding, or power shortages, but supply. He described the current environment as being in "Neverland," with Dell assuming long-term shortages of key components.

Margin Debate: Not Simply "Profit Stacking"

Morgan Stanley previously noted in an industry report that infrastructure gross margins for Dell and Hewlett Packard Enterprise (HPE.US) had expanded unprecedentedly, attributing this mainly to "profit stacking," i.e., leveraging rising memory costs to secure higher profits amid rigid demand. Clarke acknowledged that year-over-year server and storage margins are expanding but offered a different perspective: in a world with limited bit supply, companies will allocate limited parts to the highest-margin products; meanwhile, the structural growth of the infrastructure TAM reduces the willingness to engage in competitive pricing for incremental customers. When the TAM is growing, rather than selling servers at narrow margins to acquire new customers, it is better to maintain stricter pricing discipline and raise the price floor. Clarke believes that the resulting improvement in server and storage margins is structural.

Regarding operating expenses, Dell's internal AI modernization is in full swing. Clarke emphasized that Dell is leveraging AI to enhance productivity and efficiency in supply chain optimization, customer service, coding, sales, and R&D. These investments will not slow down in the short term because Dell is seeing "huge benefits" from them, so operating expenses are unlikely to decline year-over-year. However, as revenue grows faster than investment, Dell will still release operating expense leverage.

Valuation Divergence: Bulls and Bears Clash on Durability

Morgan Stanley pointed out that investor communications show that both bulls and bears agree that Dell's revenue and earnings growth over the next 18 months will be extremely strong. Buy-side expectations for fiscal 2028 earnings per share (EPS) have risen to the $40–$50 range, while Morgan Stanley forecasts $36.47, and the market consensus is $30.44.

The real divergence lies in the durability of growth: Bulls agree with Clarke's worldview and are willing to assign a 15x P/E multiple to fiscal 2028 EPS, corresponding to a stock price of $600–$750; Bears believe this is another boom/bust cycle, with more AI value accruing to the cloud, and that Dell's earnings could halve when rigid demand turns into a demand collapse, thus only willing to assign an 8–10x P/E multiple to peak fiscal 2028 EPS, corresponding to a stock price of $320–$500.

Morgan Stanley admitted that it was previously closer to the bearish view, but Clarke's meeting prompted the firm to reassess, focusing not on token growth itself, but on whether this growth will truly bring long-term prosperity to local infrastructure. The firm currently maintains its "Equal-weight" rating and $511 price target but believes the probability of the $756 bull case scenario being realized in the next 12 months is higher. Its bull case is $756, based on 18x the bull-case fiscal 2028 EPS of $42; the base case is $511, based on 14x the fiscal 2028 EPS of $36.47; and the bear case is $306, based on 12x the bear-case fiscal 2028 EPS of $25.53.

In terms of financial modeling, Morgan Stanley expects Dell's fiscal 2027 revenue to grow by 93.9%, fiscal 2028 by 33.1%, and fiscal 2029 to decline by 4.7%; non-GAAP gross margin to drop from 20.4% in fiscal 2026 to 16.0% in fiscal 2028, and 16.2% in fiscal 2029; operating margin to rise from 8.8% to 11.1% in fiscal 2027, then fall back to 10.0% and 10.2%; and free cash flow to rise from $8.552 billion to $21.174 billion and $28.876 billion, reaching $27.910 billion in fiscal 2029. The firm's forecast for fiscal 2028 revenue is $293 billion, higher than the market consensus of $230.4 billion; EPS forecast is $36.47, higher than the consensus of $30.02; and net profit forecast is $22.813 billion, higher than the consensus of $19.269 billion.