Overexposed: The "end of the dollar" narrative
Headlines
If you only follow the headlines, it seems as if every quarter the world is taking another step towards the end of dollar hegemony. BRICS summits, bilateral oil deals in yuan, central banks buying gold – together they form a persistent narrative: de-dollarisation is underway, the dollar is losing its grip.
Facts
Serious analyses paint a more sober picture. There may certainly be diversification in reserves, but the dominance of the dollar – in FX volumes, trade invoicing and cross-border debt – is still deeply entrenched. IMF and BIS data, among others, show that although the dollar's share of FX reserves has fallen to just under 60%, it is still much larger than that of the euro and yuan combined. De-dollarization is certainly something to keep an eye on, but for the time being there is no good alternative and no substantial evidence that the dollar's dominant status is seriously threatened in the short or medium term.
Weighting
The amount of attention paid to de-dollarization is disproportionate to the actual speed of change. The diversification of reserves is real, but it is happening in millimeters, not meters. The global financial infrastructure—FX markets, trade invoicing, cross-border banking—still runs overwhelmingly on dollars. And nothing in recent data suggests that this will fundamentally shift in the coming years.
Moreover, much commentary fails to recognize the difference between price dynamics and system dynamics. A weak dollar or a decline in the dollar's share of reserves is quickly framed as "proof" that the system is collapsing. But exchange rates respond primarily to interest rate expectations, risk appetite, and capital flows; a weakening of the dollar is far from THE proof of a coordinated exit from the currency.
The most important argument, however, lies in the time horizon. In the short to medium term, de-dollarization is likely to play a much smaller role than acute and more concrete drivers, such as US tariffs and legal uncertainty surrounding trade policy, geopolitical fragmentation, tensions in Asia and the Middle East, structurally higher defense and energy investments, and whether or not the AI bubble will burst. However interesting it may be from a geopolitical perspective, de-dollarization is primarily a long-term story, especially with regard t . That is why this theme does not belong on page 1 of the 2026 risk radar, but in the chapter on "long-term structural trends."
Underexposed: Hybrid warfare in the far north
Headlines
You have to search hard among the analyses of tariffs, Gaza, the Epstein files, and elections, but here and there, security analyses do express concern: the far north as a new front for hybrid warfare. Russian activities along cable routes, drones over the Barents Sea, and Arctic leaders warning of sabotage of undersea infrastructure and energy projects.
Facts
Submarine data cables carry more than 95–99% of international internet and data traffic. A single targeted attack could therefore disrupt cloud services, payment transactions, and trading platforms in one fell swoop. Last year, the ITU and UN were forced to set up a special body to improve the security of these cables after a series of incidents in the Baltic Sea.
Officials in Greenland and the Faroe Islands are already warning that cutting just two cables could shut down their internet for months. Meanwhile, Denmark is spending billions on Arctic security, including new cables and F-35s. At the same time, Russia is trying to further develop its Northern Sea Route (NSR) along Siberia as an alternative to the Suez Canal.
Weighting
The far north is rapidly becoming more important geopolitically, precisely because of a risky mix of climate change, stricter sanctions regimes, protectionism, and pure geographical vulnerability. The Arctic sea routes have few deep-water ports, hardly any infrastructure for emergency assistance, and limited search-and-rescue capacity: if something goes wrong — from a technical defect to a military incident — there are only a handful of ships, airports, and rescue teams that can intervene within a reasonable time. This makes the route extremely vulnerable to accidents, escalation, and strategic pressure.
Although "critical infrastructure" and "gray-zone tactics" now appear in the geopolitical dashboards of banks and think tanks, the level of attention remains relatively low compared to the Middle East or Taiwan. This is strange: in a world where everything – from high-frequency trading to container logistics and AI clouds – runs on fiber optics on the seabed, this is a classic low-probability, high-impact factor (this week's major disruption at Cloudflare showed how vulnerable the internet is due to its heavy dependence on a few choke points).
For companies, this means two things:
- Operational risk – A serious cable failure in the far north (or elsewhere) could disrupt market infrastructure for days or weeks, with implications for trading, clearing, and cross-border payments.
Capital allocation – In the coming years, tens of billions will be spent on cable replacement, redundancy, and monitoring, as well as Arctic defense and satellite alternatives. This is a structural tailwind for certain industrial, defense, and infrastructure niches.
Spot-on: : The battle for power – AI collides with the limits of the physical world
Headlines
AI is constantly presented as the growth engine for the global economy in the coming years, but behind this lies an essential energy and infrastructure story. Not only can chips be a major bottleneck, but so can the power, grid capacity, and hardware needed to manufacture and deliver those chips. In other words, cables, transformers, and permits are just as important to the AI boom as chips and LLM algorithms. For example, the CEO of Hitachi Energy warned last summer that the energy consumption of AI companies is jeopardizing global security of supply.
Facts
According to a recent IEA report, the electricity consumption of data centers—driven by AI—will more than double by 2030 to approximately 945 TWh, comparable to Japan's current consumption. This will make data centers the fastest-growing component of global electricity demand.
The hardware is far from sufficient to meet this demand. The IEA warns that the rollout of high-voltage grids worldwide is lagging behind; delivery times and costs for critical components have risen sharply since 2021. In the US, for example, there is a shortage of large transformers, with delivery times approaching three years. Europe is in the same predicament; in several countries, the power grids are overloaded (including in the Netherlands, Germany, and France) and their expansion is far too slow.
Weighting
Technological ambitions are increasingly coming up against the hard limits of the physical economy. If AI-related demand grows faster than the electricity supply can keep up, this will lead to higher prices, regional inflation shocks, and delays in investment.
In addition, a location battle will ensue: countries with cheap, reliable, and expandable power will attract AI companies. Congested countries will fall behind, which will have consequences for employment, tax revenues, and strategic autonomy.
This has a clear geopolitical dimension: energy security, data centers, critical hardware, digitization, and security are becoming one issue, with the model builders and suppliers of the power, grids, and infrastructure on which those models run becoming increasingly powerful.
Take a look at my latest Global Political Analysis: Will Trump be paralyzed in 2026? (13th November 2025)
