Title: Multi-factor model for contingent convertible bonds
Authors: Renata Latocha - ODDO BHF Asset Management GmbH (Germany) [presenting]
Abstract: The aim is to identify the cross sectional multi-factor structural model that provides insight into the drivers of prices of contingent convertible (CoCo) bonds. The OLS regression performance with cross-sectional data show the effectiveness of this method when dealing with CoCos market which is too immature to draw any far reaching conclusions on modeling and calibration from the market data. The OLS estimator minimizes the squared distance between the plane with multiple factor loadings in the multidimensional matrix and the regressand vector of CoCos prices. The identified model retains predictive power for the portfolio of 9705 global corporate bonds including CoCos. The values of estimated parameters of the bonds prices functions indicate that the risk-free interest rate, maturity of the bond, credit spread, coupon level, current economic sytuation and expected one, discount factor and distance-to-default factor influence in prices of CoCos significantly. The CoCos specific factors like trigger spread or conversion type have moderate affect the prices of CoCos. The results suggest that the constructed model is a pragmatic approach to analyze and explain CoCos prices.