Development of the ADCC model in Europe and the United States : analysis of the coefficient during crises, its drivers and an application to risk management

The recent developments regarding a possible Grexit have raised concerns about the resistance of the Euro area to negative shocks. One way to assess financial stability of an entity consists in observing whether bad news impacting an asset market are later transmitted to other assets market. The study of stock-bond correlation is obviously central to this approach. Other uses of stock-bond return comovements include applications to risk management and asset allocation. The aim of this thesis is, on one hand, to explain the level of stock-bond return relation in periods of market stress in order to find whether there has been transmission of negative shocks in Europe and the United States. By further analysing the drivers of the correlation coefficient, we intend to deeply understand how it can be influenced and hence to obtain accurate forecasts. In the end, the emphasis is given to a way of handling stock-bond comovements in order to mitigate the risk exposure of a portfolio composed of equities and government bonds. Through our study of stock-bond correlation, we find the occurrence of various Flights-To-Quality and of a likely Flight-To-Liquidity, justifying the low levels it has been at lately. The analysis of the drivers indicates that the correlation coefficient is mainly influenced by investors’ sentiment. A set of explaining factors having an explicit relation with it is obtained and then applied to manage to make projections of the value in the future. Finally, forecasts of the covariance matrix are applied to efficiently reduce the Value-at-Risk of a portfolio.