Title: Analyzing and testing the forward bias puzzle
Authors: Julia Reynolds - Universita della Svizzera italiana (Switzerland) [presenting]
Leopold Soegner - Institute for Advanced Studies (Austria)
Martin Wagner - University of Klagenfurt (Austria)
Abstract: New econometric tools are applied to explore the so-called ``forward bias puzzle'', or the finding that the forward unbiasedness hypothesis, which states that forward exchange rates should predict expected future spot rates, fails to hold in real-world datasets. Modelling the forward unbiasedness hypothesis as a cointegrating regression allows us to apply previous monitoring tools, which monitor the stability of a cointegrating regression over time. The results show that deviations from the forward unbiasedness hypothesis are much more likely at longer maturities, implying a potential liquidity effect in the forward bias puzzle.