I initiated a ‘Buy’ rating on Hewlett Packard Enterprise (NYSE:HPE) in July 2024, highlighting their strong growth potential from cloud core, edge computing and AI servers. HPE announced their Q3 results on September 4^th after the bell, reporting a strong revenue growth. There are market concerns regarding the gross margin of HPE’s AI server, I trust their gross margin will gradually improve in the future. Given the attractive valuation, I am upgrading to ‘Strong Buy’ with a one-year target price of $28 per share.

Low Gross Margin for AI Servers

In Q2, the adjusted gross margin declined by 410bps year-over-year and 130bps sequentially, primarily due to lower margin from AI server businesses. Due to the rapid growth of AI workloads, HPE’s Server business grew by 35.1% year-over-year. Due to the higher revenue mix towards AI Servers, HPE’s gross margin declined during the quarter.

I think the low gross margin for AI server is temporary and HPE’s overall margin will improve in the future. Key reasons are:

• As presented in the slide below, AI Systems include both products and services. Currently, services only contribute a tiny portion to HPE’s total AI revenue. Based on cumulative orders, services represent around 10% of total. Looking at other hardware companies, such as Cisco (CSCO), I estimate services could potentially reach 20% of total AI system revenue in the future. As HPE has experienced 35.1% growth in AI servers, hardware will continue to account for the majority of total revenue. When hardware business growth starts to moderate, the revenue mix from services is expected to increase over time. As services carry much higher margin for HPE, an increased revenue mix towards services will enhance HPE’s gross margin over time, in my view.

• In a recent conference, Dell’s (DELL) management indicated that currently, the tier 2 cloud service providers (CSP) currently account for a significant portion of total demands for AI servers. As these tier 2 CSPs typically make large-volume purchases, the gross margin from CSPs is lower than enterprise customers. In other words, CSPs have stronger price negotiation power. With the rapid growth in AI training and inference, I anticipate that enterprise customers will gradually become the major end-users for AI servers. When AI enters into the inference stage, there will be more edge computing demands from enterprise customers.
• Lastly, HPE has been investing in their AI solutions, including HPE Private Cloud AI in partnership with Nvidia (NVDA). HPE needs to offer diversified generative AI use cases for various industries. These initiatives require up-front investments, which may temporarily negatively impact the company’s overall margin.

Other than the decline in gross margin, HPE delivered a strong financial result in Q3, with 10.1% growth in revenue, as shown in the chart below:

Update For H3C Technologies

On September 4^th, HPE received proceeds of approximately $2.1 billion from the partial sale of its equity position in H3C Technologies Co., Limited. As discussed in my initiation report, HPE sold 30% of its stake in China-based H3C Technologies. The proceeds could potentially be used for shares repurchase in the future. Over the past nine months, HPE repurchased $100 million of own stocks, compared to $366 million during the same period last year. HPE exited the quarter with only $500 million in debt and $3.4 billion in cash. As such, I think HPE has sufficient capital for shares repurchase, given the stock price is only trading at around 12x forward free cash flow.

Outlook and DCF Update

The company anticipate 1%-3% constant revenue growth and 0%-2% adjusted operating profit growth for FY24, as shown in the slide below. The modest profit growth reflects the margin pressure from the increased revenue mix towards AI servers.

As HPE has already delivered three quarters of earnings, I don’t anticipate their full-year result will deviate significantly from their guidance. I am considering the following factors for their future growth:

• Server: The business segment currently represents more than 55% of total revenue. Due to the rapid growth in AI servers, I anticipate the Server segment will increase to 70% of total revenue in the near future. I think HPE is likely to experience strong server growth in AI workloads, offset by decline server growth in traditional on-premises networks. As such, I estimate their Server business will grow by 10%, contributing around 8% to the overall topline growth.
• Hybrid Cloud and Intelligent Edge: I project both segments will deliver 5% organic revenue growth, driven by growing demands for storage, GreenLake Flex Solutions, Private Cloud and switch businesses. The growth rate is in line with historical market growth.
• Financial Services: The segment is a relatively small business, representing only 11% of total revenue. I estimate it will grow by 2% annually.

Overall, I calculate HPE will grow their revenue by 7% on a normalized basis. I have adjusted margin assumptions from my previous DCF model. As AI Servers are more likely to be a headwind for gross margin expansion in the next two years, I model HPE’s margin will decline by 10bps in FY24 and FY25. However, as I analyzed previously, with AI entering into inference stage and edge computing, more AI server demands will come from enterprise customers, and HPE will be likely to generate higher gross margin from enterprise customers. As such, I forecast their operating margin will be expanded by 10bps annually from FY26 onwards.

With these parameters, the new DCF can be summarized as follows:

I calculate the total free cash flow from equity (FCFE) as follows:

The cost of equity is calculated to be 11.4% assuming: risk free rate 3.7%; beta 1.1%; equity risk premium 7%. Discounting all the FCFE, the one-year target price is calculated to be $28 per share.

Key Risks

HPE has been increasing their inventory levels during the first nine months of FY24, resulting in almost $3.2 billion in cash outflow for inventories. The company ended the quarter with $7.67 billion in inventories, up from $4.6 billion last year. The inventory increase is primarily because of the rapid demands for AI servers and strong backlogs. While I am not concerned about the rising inventory, it will impact the company’s working capital and cash flow from operations in FY24.

End Notes

I am not concerned about the current low gross margin for AI server business, and I trust the overall margin will improve gradually when AI moves into edge computing and inference stage with increasing demands from enterprise customers. I am upgrading to ‘Strong Buy’ with a one-year target price of $28 per share.