US productivity may fuel interest rate divergence and dollar strength – Fed study

Research from the Federal Reserve suggests productivity gains from high-tech firms and artificial intelligence (AI) may be contributing to the strength of the dollar and could widen the gap between the federal funds rate and interest rates in other advanced economies.

“Improved prospects for US productivity could fuel some divergence in the natural rate of interest between the US and G-6 and an appreciation of the dollar,” say the authors, Danilo Cascaldi-Garcia and Hyunseung Oh. “The lack of productivity spillovers may sound counterintuitive as technology and ideas should have no borders.”

Their paper, published on July 19, says US business dynamism and financing conditions may explain the lack of spillover.

“All told, history suggests that brighter prospects for US productivity could amplify the growth divergence between the US and other advanced economies and lead to the further strengthening of the dollar,” they write.

Some productivity gains can be anticipated because of the long lag between when investments in new technologies are made and when they come to fruition. “Some future productivity gains are anticipated well ahead of time and are reflected in firms’ investment and stock prices,” say the researchers.

The authors’ model looks at productivity, investment and stock prices, and includes 13 quarterly variables stretching back to 1973. To examine the international spillover effects of US productivity, they compared the country’s GDP, labour productivity, consumer price inflation, nominal exchange rate and short-term interest rate figures with those for Canada, France, Germany, Italy, Japan and the UK.

The authors point to higher private investment in AI in the US than in any other country. China and the UK were in second and third place, respectively, by distant margins.

They add that new business registrations have surged in the US since the pandemic, “unlike in the euro area where new business registrations remained flat”.

The high-tech industry in the US, they say, is disproportionately driving this trend, and past productivity fluctuations have largely been driven by high-tech businesses.

Despite a lack of concrete evidence that AI is going to drive productivity gains, the authors find heightened exuberance: “The US might be on the verge of higher productivity in the future, even though the empirical evidence up to 2023 is not yet close to the 1990s boom.”