Big trends win out in the end. New realities can take time to emerge, but when they do, things can break quickly. The US outlook is a case in point.
When we commented back in January that ‘Higher interest rates and a seemingly overvalued exchange rate. One can’t help thinking that reality will bite the US exceptionalism narrative sooner or later.’, we did not expect some of that reckoning to come so soon. The shift in sentiment – and its speed – illustrate two key features of a world shaped by evolving trends.
First, big trends win out in the end. The world changes, and anything that is not compatible with the evolving trend will eventually need to adjust to be in line with it.
Second, when the current reality rests in large part on people’s beliefs and behaviour, then the timing of a shift to be compatible with the new reality will be hard to predict. But when it happens, it can happen quickly. The collapse of communism in a so-called ‘preference falsification cascade’ is the canonical example. The shifting US narrative is a double-bump version of the same mechanism. First came the ‘vibe shift’ break from the prevailing sociopolitical norms following the election, and then the break of the US exceptionalism narrative in recent weeks. We don’t see that narrative coming back.
Moore’s law galore
One of the elements of the US exceptionalism narrative had been a bullish view of US dominance in AI and crypto, spurred on by actual and expected policy announcements. The presumption had been that the US would have an unassailable first-mover advantage. Its deep capital markets and depth of engineering talent complemented the access to data advantages of a handful of US-headquartered firms. DeepSeek’s release of its AI chatbot – particularly the decision to release the technology as open source – punctured that bubble. Yet an understanding of the economics of technological progress would tell you not to have believed in an unassailable advantage in the first place.
It is in the nature of technologies in the early, rapid phase of development that costs and prices decline quickly as technology advances. Compute-intensive technologies are especially vulnerable to this dynamic because of Moore’s Law, the empirical regularity that chips’ computing power doubles roughly every two years.
It is also in the nature of compute-intensive technologies that advances can sometimes occur in leaps. When your product is purely algorithmic, you are vulnerable to someone else coming along with a better algorithm, cutting costs and prices by an order of magnitude in the process.
Electricity shock treatment
This combination of rapid cost decline and occasional big leaps is not confined to obviously computer-oriented technologies. Genomics is another example: recall that the Human Genome Project cost about $US2.7 billion back in 2000 to sequence a single human genome. Nowadays, that cost is about $US1000, and people in the field talk about the $100 genome being not far off. Part of that cost decline lines up with Moore’s Law, but there have clearly also been some big leaps along the way. While this might seem like an esoteric example, the potential implications for health outcomes are profound.
An example with more obviously pervasive implications comes from the energy sector. The relative cost of renewable generation technologies continues to decline. And while some of the costs of different energy sources are a policy choice – think approval processes and safety standards – the underlying driver of this relative cost shift is technological.
Relatedly, it is easy to forget that technological progress also improves the efficiency of energy consumption. If you do, you can end up overcooking your forecasts of future energy demand (see Figure 5.1 in the Finkel Report, for example). New appliances are typically more energy efficient: for example, a new washing machine typically consumes about half the electricity used by a 15-year-old model. Improvements in insulation, logistics planning and other activities also matter. Here, too, there are the occasional leaps, such as the switch from halogen to LED, lowering energy consumption for lighting by an order of magnitude.
The exponential trends that emerge from these technological drivers are among the underpinnings of the adage that people overestimate the change that will occur over the next couple of years but underestimate the change that will occur over the next ten years.
If, suddenly, there is an alternative
The dynamics of technological change should have led to more scepticism about the technological elements of the US exceptionalism narrative. The financial market context has a similar ‘this can’t last’ flavour.
The US dollar is overvalued. That is the big trend that will win out. History would suggest that periods of exchange rate overvaluation eventually correct via depreciation, though it can take a few years. It is essentially impossible to say exactly when the next leg down will occur, especially as fiscal trends imply US rate differentials are likely to at least partly compensate for the growth risks. This overvaluation could therefore persist for a while, just not forever.
Another reason why the US dollar can stay overvalued for a while is that investors have a bit of a TINA problem: There Is No Alternative to US Treasuries. Investors feeling nervous about the prospect of a significant depreciation of USD-denominated assets over the next few years would need to invest in something else. So far, though, no other low-risk asset class has the size and depth of the US Treasury market that would allow it to be that alternative.
This is why some of the recent shifts in policy in Europe could have profound implications. European sovereign bonds are currently fragmented with different risks and circumstances. If that fragmentation were to diminish, European sovereign bonds would all be of comparable credit standing. Indeed, a structural shift in European debt dynamics could be underway already. Consider the growing issuance of pan-European bonds by the European Commission, the collective move to raise and co-ordinate defence spending across the continent, and success in narrowing fiscal divergences between member nations. If investors gain confidence in this longer-term evolution, we could see attitudes shift towards treating European bonds more like a single, deeper market – and thus a true alternative to Treasuries.
As recent weeks have shown, when human collective beliefs and decisions are involved, things can unravel much sooner and faster than you think.
