Abstract: Pledged inventory is one of the typical financing modes of supply chain finance (SCF), and the fluctuation of pledge value is the main risk faced by SCF. Therefore, how to measure the risk of pledge price fluctuation is the focus of academic and industry circles. As a risk measurement tool mainly promoted by the Basel accord, Value at Risk (VaR) is widely used in academia and industry for risk measurement. However, existing methods of VaR measurement suffer from 1) error in distribution assumptions, 2) inaccurate characterization of nonlinear relationships and 3) potential challenges for the utilization of mixed-frequency data. Thus, this paper proposes a new method to measure the VaR of SCF pledge: MIDAS-SVQR. This model uses kernel functions to deal with nonlinear relationships and directly outputs quantiles without distribution assumptions; meanwhile, it uses MIDAS to process the mixed frequency data to increase the ability of the model to extract the information of the mixed data. In addition, this paper presents the solution process of MIDAS-SVQR in detail based on the standard quadratic programming technique. This paper selects six typical pledges such as copper and aluminum as the research objects. Five different GARCH models and three quantile regression models are selected as the benchmark models. Three back-tests such as the Kupiec test are used to evaluate the accuracy of the model. The empirical results show that MIDAS -SVQR has the highest average P value of all three back-tests in all samples, and it is found that the quantile regression models generally perform significantly better than the GARCH models. Finally, the MIDAS-SVQR proposed in this paper can effectively measure the VaR of SCF pledge and provide new technology support for SCF’s risk management.

Key words: supply chain finance, pledged property, Value at Risk, MIDAS-SVQR, mixed-frequency data, support vector quantile regression