Testing endogeneity of money supply in selected European countries

This paper examines the character of money supply in selected European countries for the period of 2000-2014. The main purpose of this contribution is to identify whether the supply of money shows some signs of endogeneity, and therefore is created by loans. Testing of causal relationships is based on the mutual interaction between real gross domestic product (GDP) and monetary aggregate M2 in a sample of 12 selected economies divided into two groups: European Union (EU) members (Bulgaria, Czech Republic, Denmark, Croatia, Hungary, Poland, Romania, Sweden, United Kingdom) and non-members (Iceland, Norway, Switzerland). European Union members are further divided into two groups according to their exchange rate regime against the euro. This paper is theoretically based mainly on the research of Nell (1999) or Vera (2001) and its methodology includes the following steps: Whether the time series is stationary or not, the data are verified by the unit root test (Augmented Dickey-Fuller test). Time series with different orders of integration are differentiated and tested using the pairwise Granger causality test. In the case of the same order of integration we apply the Engle-Granger two-step error correction model (ECM). In the first step, time series are tested for the presence of a cointegration vector, and in the second step, we verify the statistical significance of this vector using the ECM. Causal relationships are always tested from one to four quarterly lags, and the determination of the best lag length is based on the lowest values of the Akaike, Schwarz and Hannan-Quinn information criteria. Relevant conclusions are not based solely on the output of Granger causality tests, but also on the qualities of the regression model (parameter significance, economic verification) which is estimated with the chosen lag length.