Proper risk measurement can be useful in determining investment decision choices in the financial asset market. There are several models for measuring risk, including Value-at-Risk (VaR) and Modified Value-at-Risk (MVaR). Which of the two risk measures are more appropriately applied to analyze multiple stock prices, depending on the behavior pattern of each return of the financial assets analyzed. This paper aims to choose a more appropriate risk measurement model, if the return of financial assets has a certain behavior pattern. It is assumed that the return of financial assets follows the asset-liability model. Similarly, it is assumed that the return data on financial assets follows the time series pattern. In this paper, we formulate VaR and MVaR risk measurement models under liability assets. Furthermore, an analysis of the comparison between Value-at-Risk (VaR) and Modified Value-at-Risk (MVaR) is carried out under the asset-liability model, using a time series approach. The evaluation of the performance of a risk measure has been done using the Lopez II approach. The results of the analysis show that some stocks are more suitable to be measured by Value-at-Risk (VaR), and others are more suitable to be measured based on Modified Value-at-Risk (MVaR). So, to choose a more appropriate risk measure, it depends on the behavior pattern of stock returns.