Abstract
<jats:p>The natural rate of interest (r*) plays a central role in monetary policy analysis, but its estimation is inherently model dependent. This paper studies how alternative econometric specifications affect the estimates of r* in the euro-area. We propose a Hybrid Common Trend Vector Autoregression that incorporates a large information set, stochastic volatility, and fat-tailed innovations. Using euro-area data spanning the last 25 years, we find that these features materially improve inference, reducing the width of 95\% credible intervals for r* by as much as 100 basis points relative to conventional specifications. Our results point to a persistent decline in the natural rate between 2000 and 2020, followed by a moderate post-pandemic recovery. The decomposition of r* identifies the trend of labor productivity growth as the dominant driver of its long-run movements. The findings suggest that uncertainty surrounding r* is partly econometric in nature and can be reduced through richer empirical specifications.</jats:p>