Europe is on the cusp of a new era of higher growth and interest rates. Our analysis suggests that over the coming decade, its economic potential will increase, closing the gap with the US. The investment implications could be significant: with Europe’s asset markets on a sounder economic footing, the single currency and euro denominated bonds have the potential to become strategic allocations within global fixed income portfolios.
We expect US economic growth to steady over the next 10 years, converging towards its long-term potential of 2.3% – slightly above the 2.0% rate we forecast for 2026 but below the 2.5% level it has averaged over the past 10 years. The steady performance reflects a likely slowdown in investment and depletion of savings: the years of fiscal largesse and outsized public investment, which saw its public debt balloon to 123% of GDP from around 85% over the past 16 years, are now behind us.
For the euro zone, however, our models suggest the opposite. They point to an increase in the potential GDP growth to 1.8% from 1.4% thanks to an extensive investment programme, focused on defence, infrastructure, AI and the green transition and funded by a much more expansive fiscal policy. In contrast to the US, there is capacity for largesse: euro zone spending has until now been much more modest, with debt increasing to 89% of GDP from 85%.
AI potential
Europe's investment spree should help it catch up with the US on tech adoption, enabling it to fully benefit from the potential AI-driven productivity boost – which we expect to add 0.25 percentage points per year to euro zone GDP.
While the US may still dominate in developing the technology, the biggest economic impact will come from companies and businesses adopting these advances. In this area, Europe should be able to hold its own.
Some 77% of European firms have embraced advanced digital technologies, and 37% are using generative AI, according to research by the European Investment Bank – compared to 78% and 36% in the US, respectively.https://www.eib.org/en/stories/eu-firms-ai-big-data
Higher rates
Europe’s improving growth prospects, in turn, mean that it can tolerate a higher neutral level of interest rates.
The Taylor Rule – one of the most established interest rate models – states that nominal base rates should be close to the rate of nominal GDP growth plus or minus a coefficient that measures how far inflation is from the central bank's target, and how growth is faring relative to its long-term trend. We have adjusted the formula to account for exchange rates and to focus on expected future fundamentals rather than historical ones. This approach, we believe, more accurately reflects central banks’ approaches to monetary policy.
Assuming the US Federal Reserve retains its credibility and inflation returns to target, our model points to US rates at 3.70% by mid-2035 – broadly in line with current levels. This empirical evidence thus also shows that the current Fed funds rate is at its neutral level at present and that even in the current macro environment monetary conditions are fully appropriate in the US. Euro zone rates, meanwhile, are projected at 3.68%, up from the current 2.25%. The yield gap will thus narrow from 150 basis points to just 2 basis points.
Actual and estimated policy rate for euro zone and US based on the nonlinear Taylor Rule, %
Source: Bloomberg, Pictet Asset Management. Data covering period 01.01.2002-01.01.2026.
Adjusted for inflation, the model suggests long-run equilibrium real rates of 1.7% for both US and euro zone.
As Europe’s economic footing improves thanks to public investment, the adoption of AI and higher productivity, the region’s bonds – both sovereign and corporate – will benefit from more solid fundamentals.
The model’s findings tally with our Secular Outlook which anticipates a narrowing of the yield gap between the US and Europe, making European bonds relatively more attractive for yield-seeking investors. Add in the much higher fiscal risk for the US and the expected steady depreciation of the dollar, and Europe looks even more attractive. With the US premium fading, Europe’s strengthening fundamentals put it in a strong position to benefit from the repricing of risk and a rotation of global capital.
Pictet's modified Taylor Rule model
-
Our modified Taylor Rule takes a nonlinear, semi parametric approach, factoring in currency fluctuations and allowing for the two-way interaction between exchange rates and inflation. Based on our projections of inflation, real GDP growth and real effective exchange rates 10 years from now, our model aims to provide a good approximation of equilibrium short-term interest rates.
Unlike the traditional model, we base our calculations on expected rather than realised inflation, output, growth and exchange rates – a formula which we believe more closely reflects how central banks themselves approach the setting of monetary policy.
Over time, we would expect growth rates to trend towards their long-term potential while inflation returns to target (assuming a credible central bank). Thus, the 10-year forecasts in our Secular Outlook serve as proxies for potential growth and target inflation in the absence of persistent macroeconomic shocks. By plugging these forecasts into the modified Taylor Rule formula, we can estimate the equilibrium interest rate.
We don’t specify an interest rate smoothing term in the equation, as our goal is to determine the equilibrium policy interest rate (the target rate).
As the responsiveness of the central bank to macroeconomic variables can change depending on the state of the economy (recession, expansion, recovery) and over time, we have also added a nonlinear functional form of the policy reaction function. This allows for the central bank setting a different policy interest rate for the same level of inflation, output growth and exchange rate depending on the state of the economy and their priorities at any given point.
For the CPI inflation projections, we have assumed that inflation will revert to central banks’ targets (2% in most countries) over the next 10 years. For the GDP projections, we have estimated potential growth using an average of non-accelerating inflation growth of output (NAIGO), non-accelerating unemployment growth of output (NAUGO) and the Hodrick-Prescott filter, adjusting for the impact of AI on productivity and for demographic changes.
