Wim Zwanenburg: The HALO strategy in the age of AI disruption

August 21, 2026

In an era in which AI is increasingly disrupting business models and geopolitical uncertainty is on the rise, HALO companies are increasingly being seen as potential safe havens for investors. The rise of AI is forcing a reassessment of competitive advantages. Whereas expertise, scale and software used to be the key differentiators, an AI model can now perform tasks in minutes that previously took days. As a result, differentiation is diminishing and the risk of software products becoming commoditised is growing. Traditional SaaS companies, in particular, are under pressure from generative AI and AI agents. The recent share price falls at companies such as Salesforce, ServiceNow and even Microsoft, as well as data providers such as RELX and Wolters Kluwer, illustrate the fear of a broader ‘SaaS apocalypse’.

Against this backdrop, investors’ focus is shifting. Analysts are noting a growing preference for companies with physical, hard-to-replace assets. After all, AI cannot lay power grids, build pipelines or set up production chains. It is precisely sectors such as energy, industry, materials and logistics that form the backbone of the real economy and are, moreover, essential for the functioning of AI itself.

A crucial bottleneck in the ongoing AI revolution is the rapidly increasing demand for energy. Data centres and AI applications require enormous amounts of electricity. US energy demand is expected to rise by 20% to 25% by 2030, partly due to the explosive growth in data processing. This brings utility companies and providers of energy infrastructure – such as energy producers, network operators and other service providers in this sector – back into the spotlight. Electricity companies combine high capital intensity with a long economic lifespan: precisely the characteristics of HALO.

Previously, software companies were regarded as exponential organisations. Whether Microsoft produces 100,000 packages or 100 million, marginal production costs do not increase and distribution costs are virtually zero due to online distribution via the internet. With the AI revolution, the question is whether this will remain the case. To develop AI models, the so-called hyperscalers must make enormous capital investments in chips and data centres.

For investors, this may signal a structural shift. Whilst software companies have long benefited from economies of scale and minimal or even falling marginal costs, the playing field is changing. The development of AI models requires enormous investments in chips, data centres and energy supply. The era of ‘asset-light’ business models therefore appears to be partly over, particularly as semiconductor prices continue to rise due to strong demand, a shortage of supply, shortages of memory chips and rising production costs.

That does not, however, mean that software giants will lose their position immediately. Their cash flows are still robust enough to invest heavily in AI and thus maintain their competitive advantage. Nevertheless, the narrative is shifting: from purely digital scale to a combination of digital intelligence and physical infrastructure. Software companies may be growing less exponentially and are therefore less exceptional.

The key question for investors is therefore whether the current market movement marks a structural shift or a temporary overreaction. In any case, it is clear that in a world where AI is rewriting the rules of the game, companies that form the physical foundation of that very same AI economy are gaining strategic value. HALO is therefore more than just a trend. It is a revaluation of the fundamentals of economic growth.

Read the column in the digital edition of Financial Investigator magazine