Analysing the main drivers of forward bias using a VARX model and how they were impacted by the global financial crisis

Abstract

The primary aim of this dissertation is to analyse the main drivers of forward bias in terms of causation and behaviour from period 2001 to 2018. These drivers are extracted from literature as several authors attribute forward bias to risk premiums, failure of rational expectations, central bank intervention, monetary policy, liquidity effects, volatility effects, anomalies and more. This study will contribute to the literature by identifying which fundamentals cause forward bias in terms of magnitude and direction (i.e. sign). These drivers are modelled in a Vector Auto Regression (VAR) with exogenous variables which is labelled as VARX. A VARX model is applied in order to capture three dynamics which causes forward bias; the cross-correlation effects of currencies, momentum/reversal effects and exogenous drivers. Currencies applied are free floating exchange rates from four developed countries namely the USD/EUR, USD/GBP and USD/JPY. Supporting the primary aim, the secondary objective of this research is to build on the VARX model by taking the identified drivers and see how these fundamentals might have been impacted when there is an economic regime change such as the recent financial crisis of 2007 (Global Financial Crisis). None of the authors analysed as to what happens to these variables when a crisis occurs. Do their impact on forward bias change? Or remain the same? This will further assist my research in uncovering the behaviour of these drivers.