The oil shock has pushed inflation risk and interest rate uncertainty back to the forefront. For Matthias Ramser, however, the stakes go well beyond the Middle East. Market concentration in equities and massive investment in artificial intelligence call for a rethink of portfolio diversification, without losing sight of long-term growth prospects.
As oil has reached its highest level since May, at $115 a barrel, what are the main economic consequences to fear from a prolonged price shock?
Higher oil prices are already putting pressure on the economy. Consumers have less money to spend, while companies face rising costs. At current levels, however, we do not expect oil prices alone to push the economy into recession. In our view, oil prices would probably have to remain well above $140 per barrel for an extended period to push the economy into recession. The duration of the shock is therefore particularly important.
With the US midterm elections approaching, political uncertainty adds another dimension. We do not know whether Trump will decide to launch another attack on Iran after the elections. What is clear is that he has a strong interest in keeping energy prices under control ahead of the vote. And he has reason to be concerned. Voters are holding the president responsible for high fuel prices, and Trump risks losing control of both chambers of Congress. Interestingly, history suggests that this would not necessarily be bad news for investors. Since 1950, the S&P 500 has risen in the twelve months following every midterm election, regardless of the outcome.
Our base case remains that any renewed escalation would be relatively short-lived. A prolonged energy crisis is neither in the interest of the major economies nor of Iran, which itself depends on functioning oil exports.
Even if the conflict proves short-lived, we expect oil prices to remain above their pre-war levels for some time. Depleted inventories need to be rebuilt, while energy security is likely to become an increasingly important priority for governments in a more fragmented geopolitical environment. Both factors suggest that a geopolitical risk premium will remain embedded in oil prices.
How should a sustained rise in energy prices influence central banks' policies?
Rising energy prices present central banks with a dilemma. On the one hand, they push inflation higher; on the other, they erode purchasing power and slow economic growth. The key question is whether the increase in inflation remains temporary or becomes more persistent through second-round effects, such as higher wage demands.
The situation is particularly challenging for the ECB. Europe is heavily dependent on energy imports and therefore especially vulnerable to this type of supply shock. At the same time, economic growth is already weak, making the trade-off between controlling inflation and supporting economic activity even more difficult.
In the US, second-round effects have so far remained limited, and longer-term inflation expectations are still relatively well anchored. This argues against a prolonged monetary tightening cycle. We expect one or two further Fed rate hikes, followed by a pause. Equity markets should be able to absorb this. However, if inflation expectations start to move away from the Fed's target and force the central bank to tighten policy further, we would regard this as one of the biggest risks for global equity markets.
Which adjustments to asset allocation could currently offer the most effective protection against a renewed escalation of the conflict in the Middle East?
Since the escalation of the conflict in the Middle East, we have recommended maintaining our exposure to energy companies. These stocks would probably come under pressure in the event of a clear de-escalation. However, we expect oil prices to remain above their pre-conflict levels even after tensions ease. The need to rebuild inventories and the growing importance of energy security both support a lasting risk premium in oil prices. This should also provide a favourable backdrop for energy companies.
Alongside energy stocks, we continue to favour gold as a portfolio diversifier against geopolitical risks. Following the recent correction, we see attractive entry opportunities at prices around $4,000 per ounce.
CalPERS has reduced its direct exposure to the Magnificent Seven by 26%. Do you believe concentration risk in technology stocks has become excessive?
Concentration in US equities is undoubtedly very high. But there are good reasons for this. The largest technology companies have delivered above-average earnings growth, so their market dominance is at least partly justified by fundamentals rather than simply driven by excessive expectations.
Where I see a particular risk is at the portfolio level. Many investors are more concentrated in a handful of companies than they realise. They hold the same large technology stocks through index funds, actively managed funds and thematic investments. Even outstanding companies, however, can become overvalued.
Against this background, I can understand CalPERS' decision to reduce its exposure. I see it less as a negative view on technology and more as a deliberate diversification measure. We also remain convinced of the long-term opportunities offered by artificial intelligence. The key is to capture these opportunities within a balanced portfolio rather than becoming overly dependent on the performance of just a few companies.
How would you reallocate the 26% reduction made by CalPERS?
That depends very much on the composition of the rest of the portfolio. We would certainly retain meaningful exposure to the Magnificent Seven and to high-quality AI infrastructure providers. At the same time, we increasingly see opportunities in companies that benefit from the broader adoption of artificial intelligence.
Banks are a good example. AI offers significant potential for efficiency gains, particularly in administrative processes and customer service. I also find biotechnology especially interesting, as AI could substantially accelerate drug discovery and development. Imagine a major medical breakthrough, for example in cancer treatment, made possible by AI. Such an achievement could fundamentally change how investors perceive the economic potential of this technology.
We also favour established healthcare companies. Many have solid fundamentals and attractive valuations. They offer useful diversification, particularly if the high expectations surrounding AI are not fully realised.
The idea is therefore not to turn away from AI, but to capture its investment opportunities more broadly. While infrastructure providers have so far been among the main beneficiaries, companies applying the technology could increasingly move into the spotlight. This is reminiscent of the internet revolution some 25 years ago. The greatest long-term economic benefits did not come from building the infrastructure alone, but from its widespread adoption across almost every industry.
At what point could the massive capital expenditure on AI infrastructure become a concern for equity investors?
History shows that major technological revolutions are often accompanied by overinvestment. We saw this with the expansion of railways, electrification and the internet. It would be surprising if artificial intelligence did not follow this pattern, at least to some extent.
I remain convinced of the technology's potential. The remarkable advances in mathematics, coding and complex reasoning give us good reasons to expect substantial productivity gains. The key question, however, is how quickly and to what extent these advances translate into corporate revenues and earnings. There is a real risk that investment in AI infrastructure could temporarily run too far ahead of actual demand.
On the other hand, AI could spread much faster than previous technologies. Unlike during the early days of the internet, companies do not need to build an entirely new infrastructure from scratch. They can leverage existing digital systems, which should accelerate the commercial adoption of AI.
For investors, the critical question is whether AI-related revenues and cash flows can keep pace with the enormous investments being made. Here, we need to distinguish between two risks. If AI as a technology fundamentally fails to meet expectations, broad corrections across equity markets would be almost inevitable. If the technology delivers on its promise but individual business models or infrastructure investments fail, the losses are more likely to be concentrated among the companies and investors directly involved.
A slowdown in AI investment growth would therefore not necessarily be bad news. It could also reflect greater capital discipline, with companies placing more emphasis on efficiency and profitability.
Ultimately, what matters to me is not simply how much money is being invested in AI, but whether these investments generate attractive returns. The risks of overinvestment are real. However, given the technological progress and the fundamentals we see today, I see no reason to fundamentally question the long-term investment opportunities offered by artificial intelligence.