دانلود رایگان مقاله بحران مالی و بایاس برآورد در بازارهای اوراق قرضه بین المللی

Abstract

This paper analyses the impact of estimation bias on various international bond markets during recent financial crises, using a unique empirical design. We estimate the Kalman filter over the period 2004-2014 using weekly data from the US and its main trading partners and construct measures of model forecasts, term premia, and risk premia in the presence of estimation bias, and in its absence. We find that the impact of estimation bias was the strongest for all sampled countries during the Global Financial Crisis of 2007-2010, and the ongoing eurozone sovereign debt crisis.

4. Conclusion

Much recent work has focused on consequences of recent financial crises. In this paper, we have examined how the impact of estimation bias on various measures of model accuracy changes during recent financial crises using data from the US and its main trading partners—Canada, China, the Eurozone, Mexico, and Japan. Our motivation for selecting the US and its main trading partners stems from similarities in the bond market dynamics of these countries. Our objective is to examine the extent to which patterns in the impact of estimation bias on model accuracy are altered in the face of important global economic events or crisis periods. We estimate model forecasts, long maturity risk premia, and long maturity term premia as measures of model accuracy. Motivation for the importance of studying these three measures and their relevance to financial crises were provided in the body of the paper.