Loan Supply Shocks in Macedonia: A Bayesian SVAR Approach with Sign Restrictions

Abstract

This paper analyses the effects of loan supply, as well as aggregate demand, aggregate supply and monetary policy shocks between 1998 and 2014 in Macedonia using a structural Vector Auto Regression with sign restrictions and Bayesian estimation. The main results indicate that loan supply shocks have no significant effect on loan volumes and lending rates, as well as on economic activity and prices. The effects of monetary policy on lending activity are fairly limited, although there is some evidence that it affects lending rates more than loan volumes. Monetary policy shocks have strong effects on inflation, while the central bank reacts strongly to adverse shocks hitting the economy. Baseline results are fairly robust to several extensions and robustness checks. According to historical decomposition, the lending activity was supporting economic growth before and during the crisis, but its contribution became negative during the recovery and it was a drag on growth until the end of the period. Pre-crisis GDP growth is mostly explained by the supportive interest rate of the main monetary policy instrument. However, the restrictive policy during the crisis for the purposes of maintaining monetary policy goals was associated with a fall in GDP, while the policy became supportive again during the early stages of the recovery. Policy rates in the recent years mostly reflect subdued lending activity and aggregate supply factors, which the central bank tries to counteract with a more accommodative policy.