A semi-structural credit gap for Malta: A multivariate filter approach

This paper presents a credit gap for Malta derived from a semi-structural multivariate filter. This modelling approach has several advantages over univariate approaches typically used, for example to construct the Basel gap. The multivariate filtering of observed data into trends and cycles is informed by economic relationships, making estimates of gaps more sensible and robust, and the framework is flexible, allowing for further model development with relative ease. The estimated credit gap is cyclical with an average duration of 13 years, it was positive between the years 2006-2013, and has turned positive again since 2019. The model also provides estimates of other economic concepts like potential output, trend inflation and the house price gap. The semi-structural credit gap estimated in this paper correlates with other existing measures of cyclical risks and is shown to have early warning properties. Moreover, the gap can be decomposed into contributions from the household and firm sectors, yielding a better picture of the drivers of the financial cycle, thereby guiding policy on the appropriate macroprudential policy tool to deploy.