The Swiss National Bank’s latest working paper, authored by Simon Beyeler, analyses why the post‐COVID inflation surge in Switzerland surprised forecasters. The study notes that shocks initially deemed transitory proved to be persistent, leading to sizable forecast errors.
The research replicates a U.S. exercise that applied Unobserved‐Components Stochastic Volatility (UC‐SV) models to real‐time data. It adapts the approach for Switzerland by examining (i) total Swiss consumer‐price index (CPI) inflation and (ii) sectoral inflation across 22 expenditure groups, adding an explicit seasonal component because the Swiss CPI is not seasonally adjusted.
Both model specifications attribute the bulk of the 2021‐2023 inflation surge to a permanent trend rather than to temporary noise. Importantly, the models reach this conclusion using only data that would have been available at the time, suggesting that the persistent nature of the shock could have been recognised earlier.
In the sector‐specific analysis, the authors find that individual sector trends explain most of the variance in inflation, while the common (economy‐wide) inflation trend plays only a minor role. This highlights the importance of sectoral dynamics in understanding overall price movements.
The paper underscores the potential of UC‐SV modelling for real‐time inflation monitoring and suggests that incorporating sectoral trend information could improve forecast accuracy, a point of relevance for monetary‐policy analysts monitoring price stability.