Practical development of a CCP risk management system for institutional investor settlement in the Korean stock market

Abstract

The authors have redesigned the logic for setting the net debit cap and provided the architecture for a credit and market value at risk (VaR) measuring system for risk mitigation in institutional investor settlement in the Korean stock market, where the Korea Securities Depository (KSD) as a central counterparty (CCP) assumes the payment obligation of participants and settles by DVP2 on settlement date T + 2. In total, 6,345 data points of net settlement funds from 16th January to 27th August, 2012, were used for 89 members to redesign the logic of the net debit cap and 11,541 data points of net settlement funds from 16th January to 27th August, 2012, were used for the 125 account groups (a set of securities settlement accounts designated for participants’ convenience, such as management of margin surplus value and payment and delivery of settlement fund) to calculate risk exposure. For setting the net debit cap, all data were divided into four different groups by quartile and four different net debit caps on quartile 3 were set, with 75 per cent points for each group. According to the simulation result, the total amount of the guaranteed net settlement fund could be reduced by 52.22 per cent while maintaining operational efficiency of members. In addition, 62 artificial default rates for credit VaR were generated by using the bond default rates of the Korea Financial Investment Association (KOFIA) and the global and Asia-Pacific corporate default rates of Standard & Poor’s (S&P). For market VaR, the International Monetary Fund (IMF) crisis, 9/11, sub-prime crisis and Lehman’s bankruptcy were chosen to build the ‘stress-test’ environment. Four risk exposures — ‘high risk’, ‘risk’, ‘semi risk’ and ‘low risk’ — of the net settlement fund were devised for measuring the degree of risk exposure of the net settlement fund. For further study, measurement of a net debit cap to be calculated with an average of three highest values may not be sufficient, so, alternative methods to measure the optimal number of settlement funds to represent the whole settlement fund may be needed.