For Hewlett Packard Enterprise, Monday was a bad day. After Evercore ISI lowered its rating and identified what analysts called a “tougher setup” with fewer near-term catalysts, the stock dropped nearly 11%, from about $62 to close just above $55. A single-day move like that usually attracts attention, and it did. However, it’s important to consider the true implications of the selloff before making any judgments.
because it is difficult to ignore the underlying numbers. With adjusted earnings per share of $1.11, significantly higher than the $0.91 Wall Street had predicted, HPE reported revenue of $12.2 billion in the most recent quarter, up 34% year over year. The total AI backlog is getting close to $7 billion, with orders for AI servers totaling $2.4 billion for the quarter alone. In addition to raising its fiscal 2026 EPS guidance, management projected 13% to 17% revenue growth in 2027. That doesn’t sound like a business that is losing steam.
Monday’s decline also has a different context given the stock’s overall performance this year. Prior to the decline, HPE had increased by about 159% in 2026, rising from a 52-week low of about $19.84 to a high of $64.25. After a triple-digit run, a 10% correction is not out of the ordinary. In actuality, it’s pretty standard. Combining that kind of gain with an AI backlog that is still growing and analyst price targets that are comfortably above the current stock price is less common.
As this develops, there’s a sense that the Evercore downgrade may have overstated the worry about near-term catalysts, but it did capture something genuine—the stock had priced in a lot of optimism very quickly. Unlike Nvidia, HPE is not a pure AI chip designer. It manufactures and markets the networking infrastructure and servers that data centers genuinely require to run AI workloads. Although it’s not as glamorous, that job is long-lasting. Physical infrastructure is necessary for both hyperscalers and enterprises, and HPE’s backlog and increased orders indicate that demand isn’t declining.
There’s something appropriate about the company’s headquarters being in Spring, Texas rather than Silicon Valley or the Bay Area. HPE has always kept a low profile. Under CEO Antonio Neri, the company, which was founded in 1939 and separated from HP Inc. in 2015, has spent the last ten years stealthily regaining its significance as a hardware and hybrid cloud enterprise. The AI wave confirmed the direction Neri had already been leading the company, rather than significantly altering HPE’s character.
Right now, the valuation picture is really intriguing. The stock may be trading about 30% below its modeled value, according to a discounted cash flow estimate that places intrinsic value at about $90 per share. The management has set a goal of at least $5 billion for fiscal 2027, with free cash flow over the previous twelve months coming in at about $3.8 billion. These projections are not conjectural. They are based on an order pipeline that has been growing and an existing backlog. The gap itself is evident, but whether the market eventually closes it is another matter entirely. It’s still unclear whether the larger cycle of AI infrastructure spending will continue at its current intensity.
The Evercore downgrade might not have been a fundamental reevaluation, but rather just tactically sound timing. The stock had moved quickly. Valuations were stretched. Taking profits was probably long overdue. However, the consensus price target is approximately $67, and 63% of the 27 analysts covering HPE still consider it a buy. The analyst community has not given up on that stock.
Investors will likely need to determine whether the AI infrastructure buildout has an additional year or two of runway and whether HPE is well-positioned to continue turning that demand into profits given its server and networking exposure. Yes, at least for the time being, according to the backlog. Be patient, the downgrade advises. It’s possible for both to be true simultaneously, and this tension is likely what makes HPE interesting to watch rather than simple to classify.
