This week, Hewlett Packard Enterprise (NYSE: HPE) took the stage at the Goldman Sachs Communacopia + Technology Conference, one of the most important investor conferences in tech. AI dominated the conversation, as it does everywhere these days. I wanted to see what management had to say about where things stand. What I heard backed up the thesis I laid out in July, and I still find the valuation attractive.

A stock that is up around 130% since the start of the year looks expensive at first glance, like a candidate for taking profits. I see it the opposite way. That rally, to me, simply reflects a massive undervaluation working itself out, and I don't think that process is finished. The price-to-cash-flow ratio currently sits at 12.5, based on market estimates for the current fiscal year 2025/26 (ending in October). That implies a cash flow yield of 8%, above my sweet spot for a long-term investment of 7%. The valuation gets even more attractive looking ahead: based on estimates for fiscal 2026/27 and 2027/28, the ratio drops to 9.2 and 8.4. Against a compound annual growth rate of 21.6% for the next 2 years, I still find that compelling, even if growth could slow in 2027/28.

HPE reported strong earnings on September 2, with record cash flow and orders in the traditional server business up 75% year over year. What interested me more was why management sounded so confident at the Goldman Sachs conference. Two things stood out.

A few weeks ago, HPE announced a major deal with Oracle for a new data center. Now we know why it won: this was a genuine competitive bid against other vendors. The deciding factor was a new switch, fully liquid cooled instead of air cooled, which matters because AI data centers consume enormous amounts of power and run extremely hot. By HPE's own account, it was roughly six months ahead of the competition on this technology. That's a real, verifiable edge, not a lucky break. It came from combining Juniper's networking technology with HPE's long-standing cooling expertise from its server business, which to me underscores the strategic logic of the Juniper acquisition. Powerful networking is exactly what's needed to actually put growing AI computing power to use.

HPE's supply chain also looks considerably more secure. The company more than quintupled its purchase commitments for memory chips and networking components within a single quarter, from $6 billion to $30 billion. That shows HPE locked in long-term supply agreements early, in a market where many competitors are currently struggling with shortages instead. That matters because it shrinks one of the biggest risks to growth: having strong demand but not being able to deliver.

The risks from my original write-up on July 10, 2026, remain in place: intense competition, lower margins in the AI server business, and dependence on overall enterprise appetite for AI spending. What's new is that management itself acknowledged it built the Oracle contribution into the raised fiscal 2026/27 networking guidance conservatively. To me, that's a sign there's more upside than the numbers already capture, but also that part of the forecast still hinges on how this one customer develops.

I still see HPE well positioned to benefit as more and more companies scramble to get ready for the AI era. The infrastructure behind the scenes is what makes the growing wave of everyday AI applications actually possible. That confirms my view, and I'm holding on to my HPE position.

Is HPE a buy? At current levels, I still consider HPE a buy despite the stock's 130% run this year. It trades at a price-to-cash-flow ratio of 12.5 for fiscal 2025/26, implying a cash flow yield of 8%, and that multiple drops further as earnings grow. The biggest risk is that part of the raised fiscal 2026/27 outlook depends on how one customer relationship, the Oracle networking deal, continues to develop.

Disclaimer: This newsletter is for informational purposes only and does not constitute investment advice. I am not a financial advisor. Always do your own research before making any investment decision.

Disclosure: I may hold direct or indirect positions (including options) in any securities mentioned in this newsletter. My opinions are my own and always honest.