Did the turmoil affect money-market segmentation in the Euro area?

Abstract

Yes. For the preturmoil period, exogeneity tests from vector autoregression (VAR) models suggest the presence of a transmission channel from longer maturities to the overnight. This disappears in the subsample starting in 9 August, 2007. The results of the semiparametric tests of Cappiello et al. (2005a) report evidence of an increase in volatility contagion within the longer end of the money-market curve, which takes place in the lower tail of the empirical distributions.

Notes

¹The lag length is chosen by minimizing the BIC criterion. This choice helps dealing with short sample available after 8 August 2007. For the preturmoil model, four lags are used, whereas the model estimated on the turmoil sample includes one lag.

²The reader should bear in mind that the analysis carried out in this article disregards the effects of the communication policy of the ECB on the money market. Trichet (2008) stresses that, during the ongoing market correction, the ECB has put efforts in justifying the decisions on liquidity policy in a prompt way. Rosa and Verga (2008) use data on the term structure of Euribor futures to show that the unexpected part of the explanation of policy decisions explains most of the variability of rates around decision dates.

³The decomposition is based on the standard Choleski decomposition.

⁴To investigate the specification of the CAViaR model, I compute the DQ test of Engle and Manganelli (2004). This null of the DQ tests the hypothesis of no autocorrelation in the exceedances of the quantiles. The specification with unconditional quantiles is rejected over the entire domain. The test statistics are not reported for brevity.