A risk analysis proposition for fixed income securities

Managing financial risk continues to be an integral part of assessing financial instrument performance. It is important to note that many empirical studies have looked on factors such as total risk and diversifiable risk or even beta, standard deviation and variance as risk representation. Due to the urgent need for a single risk measure, Value-at-Risk (VaR) has attained more demand in replacing standard deviation or volatility as the most widely used risk measure. However, VaR has so far not been exploited extensively in explaining fixed income financial risk within specific parameters, assumptions and data characteristics. In addition, most literature with regards to the usage of VaR has associated the measure with the assumption of normal distribution. Maintaining a normality assumption and failure to account for any financial time series imperfection will undoubtedly lead to underestimating or overestimating VaR and should the risk of heavy-tailed events fail to be quantified, the financial distress implications of trader’s actions will not be captured accurately. This paper highlights the proposition to fill the gap in the knowledge of financial risk measures by adding a new parameter dimension to the quantification of VaR for fixed income securities. This will be done by extending the measure through the inclusion of several volatility models under a different assumption of return distribution. The new dimension includes associating VaR estimation based on Bayesian distribution. Within this new parameter dimension, the financial risk modelling for the fixed income securities should be able to portray the actual traits of the return thus providing more accurate financial risk estimation