Leveraging from a long expertise of our teams in thematic investment, five years ago DPAM decided to include a diversified approach across themes at global level. After five years of the multi-thematic expertise at DPAM, our portfolio managers have had quite a unique experience investing across the whole value chain of sectors oriented to a future society.
For this occasion, Dries Dury and Tom Demaecker, co-managers of this investment strategy, explain in this interview the evolution of key disruptive themes and sectors over the years, and the most thrilling innovation trends for the future.
We are celebrating five years of multi-thematic investing at DPAM. How has it been?
It’s been 5 years of fear and excitement. We’re just coming out of the greatest pandemic in a generation, and after years of unlimited monetary stimulus and zero interest rates, we are now dealing with this inflationary burst. And what strange things have we seen in the market? Bitcoin at 65,000 USD, meme stocks, ARK Innovation exploding and imploding.
What is the value of thematic investments in a broader portfolio?
Interest in thematic investments does not seem to be stopping. This is mainly about two things: long-term performance and diversification. When you invest thematically, you are investing in companies that have the wind at their back. However, from time to time, some themes might be “hyped”, overvalued, or overinvested. A multi-thematic approach gives the flexibility to adapt the portfolio to the theme offering the best risk-reward profile. Multi-thematic strategies have a high active share, meaning they deviate from the benchmark substantially. That means you can truly diversify a broader portfolio.
What is really driving the Artificial Intelligence revolution and which companies profit?
We are at the tipping point, where Artificial Intelligence (AI) is entering our everyday lives, although you don’t always notice it is there.
ChatGPT is one of the most recent developments in Artificial Intelligence (AI) and has quickly become hyped. But there are already many use cases, where AI will make people more productive. Today it already acts as a kind of personal assistant for developers or creative designers, where they can do a part of the work in no time, increasing the productivity of people where there are real shortages and labour is very costly. Making code developers 20% more productive means huge savings.
For a good AI model, you need 1/ model, 2/ data, and 3/ a lot of hardware. What is hard to know is who has the best AI model, or who will have the best data? That remains still undefined. Will AI create value for society? Absolutely, but the question is who will capture that value? Will it be a good investment?
If you take a step back, what’s really driving the AI revolution is the cloud, basically on-demand ultra-cheap and ultra-fast computing power and storage enabled by superfast data communication infrastructure and big data centers that have been built in the last two decades. The CEO of Nvidia, Jensen Huang, said in their latest earnings call that they accelerated AI processing by a million times over the last decade, and they expect they will accelerate AI by another million times.
Cloud is probably the single biggest investment theme to look at today. You can only do AI in the cloud, and so companies willing to adopt AI, need to go to the cloud. Doing that means also their costs are lower, more flexible and allows for shorter development cycles. Today, only 27% of all workloads happens in the cloud.
Therefore, Cloud & Semiconductors are the best way to benefit from AI today. Amazon and Microsoft will be beneficiaries. But it goes much beyond that. Software companies that have state-of-the-art AI capabilities are very interesting in this theme. Semiconductor companies may be the best way to profit from this trend. They can be considered the picks and shovels of AI. In the gold rush, you need to sell shovels. Today, to do AI, you need huge computing power, which means that you need the best chips, and you need a lot of them.
These AI models tend to become bigger and bigger, and require more computations, from 175 billion parameters now, to multi-trillion parameters in the future. That drives the need for cloud infrastructure and semiconductors. GPUs in the cloud data centres are behind this work. There are two steps to AI: Making the model, called training, and usage of the model, called inference. Microsoft and Amazon Web Services are training these workloads on Nvidia chips. Inference is done on AMD chips. These are cutting-edge chips, designed with software from Synopsys and made with ASML, ASMI, and BESI equipment.
Why do you need to look beyond the usual suspects when searching for companies with an ecological impact?
Going to ecology—where valuations are high—stock picking can make the difference.
The Ecology theme is not only about EVs. In Europe, there are some very interesting mid-cap companies providing ecological solutions—that’s the advantage of being able to invest globally and across market capitalizations. We have been able to select key winners early on in this theme, such as Nibe in heat pumps, Tomra in recycling, and Spirax-Sarco in steam energy. Today, stocks related to Ecology are quite expensive, like wind energy or even solar. This is driven by large inflows in Clean Energy ETFs, which are price-agnostic investors.
When valuations are too demanding, the first step is to look for less obvious winners within Ecology, e.g., Epiroc, a company that enables underground mining for copper or nickel, much-needed materials to decarbonize our world. Epiroc is a more mature company which fits well with our barbell approach, but they are not less innovative. They provide electrified underground vehicles, which reduce emissions from mining operations. They also try to automate the equipment, so fewer accidents happen in the mines. Innovative, growing, and sustainable.
If good alternatives don’t exist in the market, it is interesting to keep a minimum exposure, which makes sense as it would surprise you how well many of these clean energy names have performed, despite being expensive for the last couple of years.
How to profit from innovations in healthcare and biotech with fewer risks?
Healthcare is a defensive industry, since people rely on healthcare in any macroeconomic environment. This type of theme counteracts the more cyclical parts of a multi-thematic portfolio. But it is also a growing industry with a lot of innovation. The attractiveness is that innovations automatically get protected through regulation or patents, and healthcare professionals are very sticky to a solution. The barriers to entry are really high. Therefore, growth is typically somewhat slower than hypergrowth tech, but trends are long, visible, and durable.
Two of the promising trends we see in this sector are personalized healthcare, thanks to innovative biological drugs, and a trend towards minimally invasive surgery, which improves patient outcomes and overall healthcare costs.
Investing in drugs comes with risks; biotech companies’ drugs may fail in trials, they may lose patents, they may have a very narrow pipeline, or competition may be coming in. How do you navigate that?
We think that innovation in this industry will not stop; more and better drugs will be discovered and manufactured in the future. Pharmaceutical companies are forced to do so, to protect their earnings after patents expire. Today’s medicines are much more complex to manufacture. Innovation is increasingly coming from smaller biotech companies that need to outsource the manufacturing. In the biopharmaceutical value chain, companies that provide services to the biotech and pharma companies are interesting to us.
Be they outsourced manufacturing companies like Catalent, or manufacturing equipment providers like Danaher, or clinical trials facilitators like IQVIA and Tigermed, we think more targeted biological drugs will be discovered and used, as they provide better outcomes for patients. As long as this remains the case, these companies will thrive, and we don’t need to worry about failing drugs, patent cliffs, or binary outcomes.
What can we expect for the next five years?
A multi-thematic setup allows you to shift weights according to the opportunity. That has been very valuable during the period of COVID, where a lot of these trends were accelerated. Examples were ecommerce and gaming themes.
We are seeing no slowdown in the mid- to long-term outlook of the themes we have selected, and some are even accelerating, such as the adoption of AI. Valuations of high-growth companies have reset due to rising interest rates, but company execution remains strong.
