The primary challenges in systemic risk measurement involve determining an overall reserve level of risk capital and allocating it to different components based on their systemic relevance. In this paper, we introduce a multivariate loss ratio measure (MLRM), which is the minimum amount of capital to be injected into a financial system such that the ratio of the multivariate shortfall risk over the total capital to be injected falls within a specified degree of tolerance. The degree of the tolerance controls the balance between the expected systemic risk to be reduced and the amount of the capital to be injected. The MLRM recovers the well-known multivariate shortfall risk measure when the degree of tolerance is zero. Under some moderate conditions, we demonstrate that the MLRM is monotonic, continuous and convex, and that the optimal capital allocation based on the MLRM is unique. Moreover, we show that the risk capital allocation based on the MLRM optimizes the overall systemic performance (measured by a combination of the systemic multivariate shortfall risk and the cost of risk capital) with a minimum capital requirement. Furthermore, we analyze the sensitivity of the MLRM and the associated risk allocations with respect to variation of the degree of tolerance as well as underlying random data. Finally, we report numerical test results which highlight the efficiency and stability of the risk capital allocation under the MLRM.