By Mike Seidenberg, lead manager, the Allianz Technology Trust
Early in my career, I saw the transition from mainframe computing to client-server architecture. Then came the birth of the internet, which fundamentally changed how information and services were distributed. Cloud computing later transformed the infrastructure underpinning the digital economy.
Today, artificial intelligence has the feeling of another one of those large secular shifts.
The challenge for investors is distinguishing between something that is simply niche and something that is genuinely secular or long-term in nature. When a technology becomes secular, it tends to be disruptive and transformative across industries. Our job as investors is to determine whether that inflection point is occurring — and if it is, how best to capitalise on it for our investors.
Early signs of a structural shift
Several factors give me confidence that AI is more than a passing trend. One of the clearest indicators is the sheer level of investment flowing into the ecosystem. Technology companies and hyperscale cloud providers are committing hundreds of billions of dollars to AI infrastructure — particularly to build out the data centres and computing capacity required to run increasingly sophisticated models.
See also: Pictet’s Bastianpillai on AI: Let’s get physical
These are massive projects. When companies allocate capital on that scale, they are doing so because they believe the opportunity is durable. We are also seeing early evidence of productivity gains. Even relatively simple AI applications are already changing the way people work.
In software development, the productivity impact may be even greater. Coding assistants are helping engineers create applications faster and automate tasks that previously required significant manual work. We are seeing similar examples across the workplace.
One or my colleagues recently built a tool using AI that summarises conference calls and identifies the key discussion points. He isn’t a professional developer, yet he was able to create a robust solution in just a few days. That kind of productivity enhancement is meaningful.
Some tasks may be automated, and certain roles may evolve or disappear. But historically, productivity-enhancing technologies have ultimately expanded economic activity. What matters most from an investment perspective is that once a technology becomes embedded in daily workflows, companies rarely go backwards.
Why software valuations have reset
Interestingly, while AI has driven excitement in parts of the technology market, other areas — particularly software have experienced significant valuation pressure. During the pandemic, software companies were trading at extremely high multiples, in some cases around 20 times forward enterprise value to sales. Today, those multiples have fallen closer to five times for many businesses.
Right now there is a broad concern that artificial intelligence could disrupt existing software models. That fear has created a “shoot first” mentality across the sector, where even companies delivering solid results are not being rewarded by the market.
Will AI disrupt some businesses? Undoubtedly. But it would be a mistake to assume that every company will be negatively affected. Many corporate processes rely on structured workflows, whether in cybersecurity, financial systems, or enterprise software. Those systems provide checks and balances that organisations depend on. In many cases, AI will enhance these workflows rather than replace them.
For us, the key is assessing risk versus reward. When evaluating potential investments, we look at both qualitative insights and quantitative analysis. We want to understand whether a company can continue growing, whether that growth can be profitable, and whether the current valuation adequately compensates us for the risks involved.
Managing concentration risk in the AI era
Another important issue for investors is the growing concentration within technology indices. Some of the largest companies now represent very significant weights in benchmarks. While these companies may be exceptional businesses, allocating too heavily to a single stock can introduce meaningful risk.
Our approach is to be benchmark aware, but not benchmark driven. When we identify a powerful secular theme like AI, we try to understand the entire ecosystem surrounding it.
Rather than focusing solely on the most obvious companies — for example, the manufacturers of AI chips — we look at the full “food chain” that supports the technology.
See also: Artificial intelligence: The environment’s ally
If GPUs are being widely deployed in data centres, what other components are required to make those systems work? What companies are supplying the power management systems, the semiconductor equipment, or the specialised connectors inside those data centres?
This is what we often refer to as the “picks and shovels” approach. Instead of trying to predict which AI platform will ultimately dominate, investors can sometimes benefit from owning the companies that supply the essential components used by everyone in the ecosystem.
Artificial intelligence may still be in its early stages. But if history is any guide, the most significant technological shifts often start by quietly improving how people work — before eventually transforming entire industries.
