Resume af teksten:
Store teknologileverandører forventes at se positiv vækst med stigende indtægter og EBITDA som følge af øget adoption af AI-baserede tjenester. Trods dette viser teknologiske aktier stor volatilitet, med Microsofts aktiekurs faldet 20% og Oracles 27% i første halvdel af 2026, mens Alphabets aktiekurs steg 14%. Investeringer i AI-datainfrastruktur stiger, påvirker indtjening pr. aktie og aktievurderinger, dog kan mest dække investeringerne gennem driftskontanter. Uden store aktietilbagekøb, der hidtil har hjulpet EPS, kan investorernes villighed til at betale høje værdiansættelser mindskes. Oracle står over for specifikke udfordringer grundet ambitiøse investeringsplaner, hvilket kan påvirke deres kreditvurdering. Nvidia’s høje driftsmargener kan også blive udfordret af kundernes udvikling af egne chips. Markedet oplever usikkerhed omkring investeringsafkast fra AI-projekter og fremtidig EPS vækst, hvilket bidrager til prisvolatilitet.
Fra ING:
We see a positive future for large technology providers, as the adoption of AI-based services continues to increase. Expected revenues are set to climb, as is EBITDA, as noted here . Yet investors remain nervous. Technology stocks have recently shown large intraday declines. During the first half of 2026, Microsoft’s share price declined by 20%, while Oracle fell by 27%. Alphabet, on the other hand, rose by 14%.
In this note, we discuss the factors behind this share price volatility. Although we think the overall technology market can sustain heavy investment in new AI data centre infrastructure, earnings per share growth is likely to slow, influencing equity valuations. Broadly speaking, most credit profiles continue to look fine. Nevertheless, some companies such as OpenAI, Anthropic and Oracle are investing at a pace that exceeds incoming cash flows.
Despite some occasional doom-mongering by regulators such as the Bank for International Settlements, we think current investments in AI infrastructure are made based on sound economic rationale. For example, Microsoft invested US$65bn in Cloud and AI infrastructure in F2025, and the company reports annualised run-rate revenues from AI-based services of US$37bn. This shows there’s demand for its investments. Also, technology companies have stated that they cannot deploy AI capacity fast enough to meet customer demand. It means demand for AI computing exceeds supply.
Importantly, most companies can fund these investments through operating cash flow. As a result, industry leaders have little need to take on additional debt, and their balance sheets generally remain strong. The ability to cover investments from existing cash flow can be seen in the chart above (where EBITDA is used as a proxy for operating cash flows). As widely documented, Oracle is a bit of an outlier, with a more aggressive investment programme. This will be discussed below.
As investment spending rises, though, free cash flow (EBITDA less capex) is likely to be lower than in the past, as the benefits to revenue growth may take time to materialise. This reduces the scope for large shareholder returns in the form of share buybacks. As can be seen from the chart below, share buybacks have been material for Microsoft, Alphabet and Meta. Alphabet has reduced its share count by 9.5% through buybacks from YE20 to YE25. This provided a positive tailwind for its earnings per share (EPS) metric; profits are now divided by fewer shares. Growth in this metric is important when evaluating equity valuations based on EPS. This is because the multiple can fall over time as EPS grows. Although investments can be funded with cash flows, equity investors may need to evaluate future EPS growth and a company’s ability to buy back shares.
On a trailing 12-month basis, the Nasdaq trades on a price-to-earnings ratio of around 35x. This is above the 24x level seen at the end of 2022 but below the 40x peak reached in 2021 and 2024. Given expectations for strong earnings growth, however, forward valuations are arguably more relevant. On a two-year forward basis, the Nasdaq trades at around 20x earnings, close to the lower end of its historical range.
Without the support of share buybacks, investors may be less willing to pay 40x earnings multiples in future. The trajectory of EPS growth will depend not only on the revenues generated by AI investments, but also on whether rising costs, including depreciation charges, offset some of those gains.
As the first chart shows, investment levels are higher than a few years ago. This matters because these investments must ultimately be depreciated. If revenues from AI-based services take time to materialise, rising depreciation charges could put pressure on operating margins. This risk is particularly relevant for AI infrastructure. Server hardware typically has a shorter replacement cycle than traditional data-centre assets, resulting in higher depreciation rates.
Case in point: Alphabet. Based on consensus estimates, its FY26 capex-to-sales ratio is expected to reach around 44%, while depreciation is projected to be 14% of sales. If depreciation charges rise over time to reflect this higher level of investment, profit margins could come under pressure, creating a headwind for earnings per share. For Microsoft, the equivalent figures are 35% and 14%, while Amazon is expected to post ratios of 24% and 13%, respectively.
During periods of heavy investment, capex typically exceeds depreciation. The AI sector is currently in such a phase. The key questions for investors are whether these investments generate returns above the cost of capital and whether the anticipated revenue growth ultimately materialises.
As discussed above, we believe most of the uncertainty, and therefore much of the volatility, stems from questions around project returns and fair valuations. However, investors must also contend with several company-specific risks associated with the current AI investment boom. This is particularly the case with Oracle, Nvidia, OpenAI and Anthropic.
Oracle has announced ambitious plans to support Project Stargate, a large-scale US AI infrastructure programme. However, its internal cash generation is lower than that of Microsoft and Alphabet. As a result, Oracle’s projected EBITDA less capex is expected to turn negative (based on Refinitiv’s consensus forecast), despite the company already operating with relatively low investment-grade credit ratings. While Oracle’s balance sheet remains sound, its access to funding is more constrained than larger peers. To preserve its investment-grade rating, the company may ultimately need to raise additional equity capital. So, volatility in Oracle’s share price is driven by factors different from those affecting most large technology companies.
Nvidia’s operating margin could exceed 60% in FY26, according to consensus estimates. This makes sense, as it designs state-of-the-art semiconductors that are in high demand. However, major customers such as Microsoft, Alphabet and Amazon are developing their own custom chips to help manage AI infrastructure costs (capex efficiency). As these capabilities are expanded, Nvidia’s pricing power may face greater competition than it has in recent years. This could make today’s exceptionally high margins harder to sustain over the long term, even as the company expands into new business lines.
Companies such as OpenAI and Anthropic are developing state-of-the-art models for generative AI. They’re investing huge sums to become leaders in the field and to develop models specific to customer workflows. OpenAI appears willing to tolerate significant cash burn to establish a dominant position, much as Netflix did in streaming and Amazon did in e-commerce and cloud computing. Yet the long‑term economics of AI remain less proven than those earlier industries at similar stages of their development.
When evaluating the developments in the market, we don’t see evidence of unsustainable investment. How, then, should we explain the volatility in share prices and the steady stream of news articles highlighting concerned investors and regulators? In our view, the market is undergoing a process of price discovery, assessing how to value the future returns from AI‑related investments. Key sources of uncertainty include the revenue contribution of AI infrastructure spending, the associated profit margins, and the possibility of slower EPS growth.
Kilde: ING, https://think.ing.com/articles/why-tech-investors-are-revaluating-ai-investments/
Hurtige nyheder er stadig i beta-fasen, og fejl kan derfor forekomme.
