Title: Semi-structural credit gap estimation
Authors: Peter Welz - European Central Bank (Germany) [presenting]
Jan Hannes Lang - European Central Bank (Germany)
Abstract: A theory-based approach is proposed to identifying excessive household credit developments. An equilibrium-relationship for the level of household credit is derived using an overlapping generations model that takes into account the demographic age structure of the economy. In the scope of this framework it is shown that potential GDP, a measure of the equilibrium real interest rate, information about the debt distribution by cohorts and the level of institutional quality are important determinants of the trend household credit stock. Semi-structural household credit gaps are obtained as deviations of the real household credit stock from this fundamental trend level. The resulting model is estimated as an unobserved components system for 12 EU countries using quarterly data for the period 1980 - 2015. Estimates of these credit gaps yield credit cycles that last between 15 to 25 years with amplitudes of around 20\%. The estimated credit gaps possess superior early warning properties for financial crises relative to a number of established statistical credit gaps, notably the commonly used Basel total-credit-to-GDP gap and its household credit-to-GDP gap variant. The proposed semi-structural household credit gaps could therefore provide useful information for the formulation of countercyclical macroprudential policy.