Bonds Portfolio liquidity risk under stress

Abstract: Abstract The focus of this thesis is to study and model the capacity of the bank to trade on the bonds market under normal and stressed conditions. This capacity is related to the liquidity of bonds market (ie, the ability to trade) and bank’s trading desks capacity to trade. All banks have different models but all are subject to the validation of the bank in a first place and the EU Central Bank afterward. Thereby, this study handles different internal and external recommendations. The study takes also into account market data in order to check out the global liquidity of the bond market. Such a study is important for two reasons. First, it gives a time horizon on the bond’s liquidation process. It provides an estimation of the time that the bank needs to get rid of its bond’s positions. Secondly, it analyses liquidity for different pool of bonds. Thereby, bank can group its bonds by rational liquidity categories and study the liquidity of these groups instead of thousands of products. This information is used by the bank to evaluate the cash that can be raised from bond’s positions in order to get funding. Thus, the cash required for liquidity purpose will decrease if the liquidation process of bonds speed up. The model helps the bank to level out its balance sheet at all time. The research approach is based on several liquidity measures that have been computed on internal data and help to observe trends and periodicity. Then appropriate models are described based on two measures: the daily volume of sells, and the turnover. The main conclusion drawn by the thesis display that bond’s products can be pooled per type (defined in 3.1.2). Indeed, the liquidity between those types is similar and can be compared. The model chosen depends on the stress scenario. Under normal condition the turnover is preferred and under stress the daily volume has been chosen.