The Relationship Between U.S. Corporate Bond and CDS Spreads: Evidence from 2004-2017
Hacklin, Joonas (2018)
2018
Kuvaus
Opinnäytetyö kokotekstinä PDF-muodossa.
Tiivistelmä
The purpose of this thesis is to study long run association and price discovery process between two similar credit risk markets for US financial corporate bonds and credit default swaps (CDS). These two products were playing a big role in the latest financial distress. The regulation of CDS market has increased significantly in the post crisis period, which created a great opportunity to do more recent research of the subject topic.
Before testing long run association, it is essential to apply Augmented Dickey- Fuller test to see if the data is non-stationary at levels and stationary at first- difference. After ADF tests, this study applies Johansen cointegration trace test to see long run equilibrium. Furthermore, to investigate price discovery process between bond and CDS spreads, the study uses Granger Causality test.
Thesis employs time series dataset, which consist 3,652 trading day observations of nine different S&P-500 listed financial companies. The data period starts from the beginning of 2004 and goes all the way to the end of 2017. The data period is divided into three sub-periods, pre crisis (1/2004–6/2007), crisis (7/2007– 2/2009) and after crisis (3/2009–12/2017) to see effects of financial crisis. Additionally, the study examines how bond and CDS markets vary in different economic situations.
The results show that the long run association is becoming stronger after the crisis period compared to financial turmoil and prior time. Furthermore, the findings suggest that bonds are leading the price discovery process prior and during the financial distress but CDS market has a slight lead in the post crisis period.
Before testing long run association, it is essential to apply Augmented Dickey- Fuller test to see if the data is non-stationary at levels and stationary at first- difference. After ADF tests, this study applies Johansen cointegration trace test to see long run equilibrium. Furthermore, to investigate price discovery process between bond and CDS spreads, the study uses Granger Causality test.
Thesis employs time series dataset, which consist 3,652 trading day observations of nine different S&P-500 listed financial companies. The data period starts from the beginning of 2004 and goes all the way to the end of 2017. The data period is divided into three sub-periods, pre crisis (1/2004–6/2007), crisis (7/2007– 2/2009) and after crisis (3/2009–12/2017) to see effects of financial crisis. Additionally, the study examines how bond and CDS markets vary in different economic situations.
The results show that the long run association is becoming stronger after the crisis period compared to financial turmoil and prior time. Furthermore, the findings suggest that bonds are leading the price discovery process prior and during the financial distress but CDS market has a slight lead in the post crisis period.