The article presents two models to help investors decide how many convertible bonds to convert and keep. These models use Conditional Value-at-Risk (CVaR) to measure risk, which is more reliable than Value-at-Risk (VaR). The researchers used Rockafellar and Uryasev's approach to solve the optimization problems. By applying these models to a convertible bond issued by Hynix, they found a way to balance risk and return effectively.