December 10, 2019

SVARs, the central bank balance sheet and the effects of unconventional monetary policy in the euro area

This discussion paper presents further evidence that the most important published estimates of the effects of unconventional monetary policy are not reliable. It is a further elaboration of the ideas in the CPB discussion paper "Do zero and sign restricted SVARs identify unconventional monetary policy shocks in the euro area?". Previous empirical studies seem to show that the unconventional monetary policy of the ECB, also known as balance sheet policy, has a positive effect on growth and inflation. However, this conclusion is unfounded, because institutional features of monetary policy in the euro area make it impossible to identify unexpectedly exogenous variation in monetary policy.

Read CPB Discussion Paper 391 "Do zero and sign restricted SVARs identify unconventional monetary policy shocks in the euro area?".

VAR modeling shows the effects of unexpected exogenous variation in monetary policy, also known as policy shocks. This discussion paper presents a number of reasons why the existing literature is unable to isolate unexpected variation in monetary policy. 

First, almost all of the major unconventional monetary policies have been announced in advance. The consequence of this is that the observable movements in the balance sheet of the ECB are predictable, which leads to the well-known ‘foresight problem’ and, therefore, unreliable estimates. Secondly, the most important unconventional monetary policies up to 2015 have been implemented as "Fixed Rate Tenders with Full Allotment" (FRTFA). FRTFA policy means that the ECB offers as much liquidity to counterparties as they want if they have good quality collateral. Therefore, the amount of liquidity is determined by the demand of the counterparties, not by the supply of monetary policy. Third, many different policies have been implemented, which means that the parameters of the empirical models used are not constant throughout the sample. Fourth, there are relatively few observations in the period of unconventional monetary policy, which also leads to unreliable estimates.

This study replicates the most important estimates from the existing literature that investigates the empirical effects of unconventional monetary policy. In this study I show that the apparently exogenous variation on which these studies are based has no relation to what should be expected from these shocks. The estimated shocks in one of the most influential empirical studies even have the wrong sign in 11 of the 13 most important periods since unconventional monetary policy has been implemented.


Foto Adam Elbourne
Adam Elbourne +31 6 11627514 Read more