Browsing by Author "Sephton, Peter, S."

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    A Note on the Probability of Recessions: Can Statistics Cananada's Leading Index Predict as Well as MARS
    Sephton, Peter, S.
    In this note we estimate the probability of recession using the revised leading index published by Statistics Canada. The results are compared to probabilities derived from a new non-parametric regression routine. While the indexes provide similar information on the probability that the economy is currently in recession, the non-parametric approach appears to offer more reliable information. Two out-of-sample forecasting exercises demonstrate the potential benefits to the use of the multivariate adaptive regression spline model.
  • Thumbnail Image
    Item
    A Note on the Probability of Recessions: Can Statistics Cananada's Leading Index Predict as Well as MARS
    Sephton, Peter, S.
    In this note we estimate the probability of recession using the revised leading index published by Statistics Canada. The results are compared to probabilities derived from a new non-parametric regression routine. While the indexes provide similar information on the probability that the economy is currently in recession, the non-parametric approach appears to offer more reliable information. Two out-of-sample forecasting exercises demonstrate the potential benefits to the use of the multivariate adaptive regression spline model.
  • Thumbnail Image
    Item
    Are Real Exchange Rates Stationary?: More Evidence from Panel Unit Root Tests
    Sephton, Peter, S.
    A number of recent studies using the panel unit root test of Levin and Lin (1992) suggest that real exchange rates are mean-reverting. These findings are at odds with traditional univariate unit root/stationarity test results. The purpose of this note is to extend the analysis to indexes of real effective exchange rates, and to demonstrate the sensitivity of the results to the lag lengths adopted in the panel testing procedure and the variables included in the panel. The empirical findings suggest the evidence in favour of a Cassellian view of purchasing power parity is less conclusive than it may appear.
  • Thumbnail Image
    Item
    Are Real Exchange Rates Stationary?: More Evidence from Panel Unit Root Tests
    Sephton, Peter, S.
    A number of recent studies using the panel unit root test of Levin and Lin (1992) suggest that real exchange rates are mean-reverting. These findings are at odds with traditional univariate unit root/stationarity test results. The purpose of this note is to extend the analysis to indexes of real effective exchange rates, and to demonstrate the sensitivity of the results to the lag lengths adopted in the panel testing procedure and the variables included in the panel. The empirical findings suggest the evidence in favour of a Cassellian view of purchasing power parity is less conclusive than it may appear.
  • Thumbnail Image
    Item
    Extended Critical Values for a Simple Test for Cointegration
    Sephton, Peter, S.
    Leybourne and McCabe (1993) have extended the Kwiatkowski et al (1992) stationarity test to examine the null hypothesis of cointegration. The purpose of this note is to provide an extended set of critical values for use in applied research.
  • Thumbnail Image
    Item
    Extended Critical Values for a Simple Test for Cointegration
    Sephton, Peter, S.
    Leybourne and McCabe (1993) have extended the Kwiatkowski et al (1992) stationarity test to examine the null hypothesis of cointegration. The purpose of this note is to provide an extended set of critical values for use in applied research.