The Magyar Nemzeti Bank (the central bank of Hungary) introduced a “funding for lending” type loan program aimed at small and medium sized enterprises (SMEs) in mid-2013. We combine firms’ balance sheet data with two loan data sets to study the program’s impact on firm level investment in 2013. We start from a simple difference-in-differences (DID) estimator, but argue that the parallel trend assumption that underlies the method is likely violated. Therefore, we propose a correction based on the idea that the selection process involved in securing a market loan in a pre-program year is similar to the selection process into the program. Our results indicate that the program succeeded in generating extra investment in the SME sector that would not have taken place otherwise; specifically, we attribute to the program about 30% of the total investment undertaken by participating firms. Nevertheless, the effect is markedly heterogeneous with respect to firm size, being proportionally larger for smaller firms.
JEL: D04, G38, E58.
Keywords: funding for lending, program evaluation, difference-in-differences estimation, unconventional monetary policy.