AI: fad or margin effect ?
” Maintaining net margins at high levels appears largely subject to the rapid and continued spread of productivity gains associated with AI.”
A disappointing earnings season (for now)
For the first time in 10 years, European companies are reporting results below consensus. The negative surprises are limited to just -2%, suggesting there’s no immediate cause for alarm. However, it would make sense to examine the sustainability of margins at this point in the cycle. In the United States, the operating margin of S&P 500 companies has risen from 13% in 2019 to over 17% expected for 2025. In the meantime, this same margin had fallen to 10% in 2020 amid the Covid crisis.
Net margins remain high
What is most surprising, though, is the trend in net margins. The rise in interest rates and the resulting increase in financing costs should have eroded this margin. Instead, the net margin rose slightly. While government support during the Covid crisis, fiscal policies backed by central banks, and positive consumer sentiment have played a role in the pick-up in net margins, the duration mismatch between assets and liabilities is one of the main reasons for this trend. Since debt has a longer maturity than assets (essentially cash or cash proxies), companies have seen the return on their assets more than offset the rise in the cost of debt. As a result, the inversion of the yield curve – i.e. where short-term rates surpass long-term rates – continues to generate positive cash flows.
What are the key parameters for margin growth?
1. Economic growth
2. Rise in the prices of goods and services
3. Unit labour costs
4. Productivity gains
5. Pricing power
6. Taxes and regulation
Are current margin levels sustainable?
Our central scenario is based on a moderate slowdown in global growth (‘soft landing’) and quasi-stagnation in Europe. Against this backdrop, margins are unlikely to rise overall. However, major sectoral differences need to be taken into account, as a soft-landing scenario is not necessarily synonymous with falling margins. For example, quarterly earnings expectations in the technology sector remain markedly different from those in the chemicals sector, which is still suffering from the sharp rise in energy costs.
An in-depth fundamental analysis is required to identify those companies that will benefit more from the current cycle. In terms of the cycle, we are still in a period of normalisation of inflation following the successive shocks of recent years. After peaking at almost 10%, consumer price inflation in the United States has fallen back to 3.10% year-on-year.
Disinflation is a driver of margin reduction, as costs are now tending to fall less than inflation. Also in the United States, producer prices, which had been falling rapidly since the spring of 2022, now appear to have stabilised at around 0% year-on-year since the summer of 2023.
Let’s try to understand the dynamics between labour costs, selling prices and margins. Using a simple regression model, a 1% rise in unit labour costs reduces the margin by 0.5%, while a 1% rise in selling prices increases the margin by 0.3%. The net effect is a 0.2% reduction in the overall margin. Therefore, if unit costs and selling prices rise at the same rate, the margin will automatically fall. But that’s without taking into account any productivity gains, pricing power that exceeds the market, or lower taxes.
- Regarding productivity gains, the potential offered by generative Artificial Intelligence (AI) is substantial but will take time to spread. More stringent environmental regulation and re- onshoring are negative factors (from a purely financial point of view).
- Regarding pricing power, after the multi-sector phenomenon of “greedflation”, when all companies across all sectors were able to pass on price rises that exceeded inflation, we are seeing a harsh return to reality for sectors where barriers to entry are lower.
- Regarding taxes, while the share of taxes in price rises has been below its historic rate since mid- 2022, the opposite trend seems to be underway since last autumn… The trajectory of budget deficits and debt/GDP ratios does not bode well in this regard.
In summary, maintaining net margins at high levels (without even considering a potential increase) appears largely subject to the rapid and continued spread of productivity gains associated with AI.
How should investors position themselves?
We favour sectors that will benefit from the exponential development of Artificial Intelligence. AI as an investment theme has accelerated sharply in 2023 thanks to generative AI, but this secular trend is only in its infancy. The theme is spreading from semiconductors to software. Data-intensive sectors such as healthcare and automotive will follow in the next few years. With this in mind, we are maintaining
our preference for software over semiconductors, for AI enablers over AI adopters, for the United States over Europe and the emerging markets, and for large caps over small caps. Finally, we are starting to shift our weighting slightly back towards the healthcare sector (Life Science).
We reiterate our recommendation to invest in companies with the ability to pass on price rises regardless of the cycle. In that respect, the Luxury Goods sector makes sense, but at the top end of the market. Brands such as Burberry or Kering do not have the same barriers to entry as Richemont or LVMH. Not to mention Ferrari or Hermès, which benefit from a low volume effect leading to multi-year waiting lists. The purchase of ultra-luxury items caters to the desire for exclusive group membership, where the notion of price becomes secondary. This is the sub-sector we are favouring.
Finally, we draw attention to the impact of regulation on margins. The recent farmers’ crisis in France has made us aware of the micro-economic impact of certain government decisions. Our aim is not to judge whether a particular decision is rational, but to analyse its consequences. More regulation has never helped companies. Our advice is simple: avoid segments or sectors that are sensitive to government action. To name just a few: telecoms, utilities and even banking.
To conclude, if we had to choose 2 criteria for observing the evolution of margins, those would be productivity gains and pricing power. They have the advantage of taking you partly out of the economic cycle, and even out of geopolitical uncertainty. But above all, looking at investments from this angle is akin to adopting a resolutely optimistic stance about future growth and value creation. To those who will argue that valuations are too high or that companies are more sensitive to interest rates than envisaged in this editorial, I would say that, as long as the growth in the quarterly earnings of these companies is at least double nominal global growth, 2-digit multiples are justified. At the very least, it’s a fine way to emerge from secular stagnation.
