This paper examines whether a tipping point exists for real GDP growth in Italy above which the ratio of non-performing loans (NPLs) to total loans falls significantly. Estimating a heterogeneous dynamic panel-threshold model with data on 17 Italian regions over the period 1997-2014, we provide evidence for the presence of growththreshold effects on the NPL ratio in Italy. More specifically, we find that real GDP growth above 1.2 percent, if sustained for a number of years, is associated with a significant decline in the NPLs ratio. Achieving such growth rates requires decisively tackling long standing structural rigidities and improving the quality of fiscal policy. Given the modest potential growth outlook, however, under which banks are likely to struggle to grow out of their NPL overhang, further policy measures are needed to put the NPL ratio on a firm downward path over the medium term.
AbstractThis paper examines whether a tipping point exists for real GDP growth in Italy above which the ratio of non-performing loans (NPLs) to total loans falls significantly. Estimating a heterogeneous dynamic panel-threshold model with data on 17 Italian regions over the period 1997-2014, we provide evidence for the presence of growththreshold effects on the NPL ratio in Italy. More specifically, we find that real GDP growth above 1.2 percent, if sustained for a number of years, is associated with a significant decline in the NPLs ratio. Achieving such growth rates requires decisively tackling long-standing structural rigidities and improving the quality of fiscal policy. Given the modest potential growth outlook, however, under which banks are likely to struggle to grow out of their NPL overhang, further policy measures are needed to put the NPL ratio on a firm downward path over the medium term.JEL Classifications: C23, E44, G33.