Asian Review of Financial Research Vol.29 No.2 pp.149-191
The Index Arbitrage in Expiration Days: Normal Arbitrage Trading or Cross-Market Manipulation?
Key Words : Expiration-Day Effects,KOSPI200,Index Arbitrage,Cross-Market Manipulation,Capital Market Act
The closing price of KOSPI200 on Nov 11, 2010, the expiration day for options, was determined at 247.51 points following a severe fall of 7.11 points, which caused great disorder in the Korean financial market. The Korean Supervisory Service and the Korea Exchange investigated the event and discovered that Deutsche Bank had submitted a huge sales order for 2.44 trillion Won of 199 stocks simultaneously, on condition of holding a large amount of KOSPI200 put options. Deutsche Bank was later indicted for the cross-market manipulation based on the Financial Investment Services and Capital Markets Act (hereafter “Capital Market Act”). In this paper, we investigate the cross-market manipulation using index arbitrage in determining stock closing prices on expiration days. We deal withtwo aspects in relation to this issue. First, in terms of law, we explain the various types of cross-market manipulation regulated by the Capital Markets Act. Specifically, we discuss the concept of “existence of purpose” to satisfy the requirement for cross-market manipulation action. Second, in terms of finance, the hypotheses to verify the “existence of purpose” on the cross-market manipulation are established and tested empirically. This empirical test is also related to Kumar and Seppi's (1992) model, in which the manipulator earns a positive expected profit by holding a futures position and then manipulating the spot price used to compute the cash settlement on the expiration day. In empirical analysis, we identify that expiration effects exist in the Korean market, consistent with previous studies. The volatility and trading volume of the closing price on expiration days are distinctly higher than those on non-expiration days and those of non-KOSPI200 stocks. Then we establish the first hypothesis that if expiration-day effects are caused by index arbitrage activity, the effects are stronger in stocks where the trading volume for index arbitrage is higher. The results show that the portfolio with a greater transaction ratio of index arbitrage has higher volatility and trading volume on expiration days. The differences are statistically significant. The regression analysis using volatility and trading volume on expiration days as the independent variable and index arbitrage as the dependent variable also shows that volatility and trading volume on expiration days are positively correlated with index arbitrage. These results support our first hypothesis that index arbitrage are an important source of expiration-day effects. However, the positive relationship between the expiration-day effect and index arbitrage does not necessarily imply that cross-market manipulation activity exists in the market and affects the closing price on expiration days. Normal index arbitrage can also cause an increase in volatility and trading volume due to its tremendous trading size. To disentangle these two possibilities we use the order submission data of closing call auctions and establish the second hypothesis. Our second hypothesis is that a cross-market manipulator will submit a distinctly low-priced limit sale order or distinctly high-priced limit purchase order, both of which lack economic rationality, to move the closing price as much as possible on expiration days. According to the Kumar and Seppi (1992) model, the cross-market manipulator can earn a positive expected profit by manipulating the spot price used to compute the cash settlement on the expiration day because the profit from the futures position exceeds the loss from the spot position. Analysis of the index arbitrage order data in the closing call auctions on expiration days shows a significant increase in distinctly low-priced limit sale orders and distinctly high-priced limit purchase orders. The limit orders are also concentrated in the last minute of the closing call auction. Finally we test a third hypothesis to confirm the price impact of cross-market manipulation. Our third hypothesis is that orders for index arbitrage have a greater effect on price than other normal orders if there is cross-market manipulation on expiration days, because the manipulators try to maximize the price impact of their orders to maximize their profits in the futures position. The results show that the price impact of index arbitrage orders is larger than that of normal orders. Overall, these results imply that there exist index arbitrage orders issued for the purposes of cross-market manipulation on expiration days in Korea. The findings are also consistent with Kumar and Seppi's (1992) prediction that cross-market manipulators try to manipulate the spot price used in cash settlement on the expiration day.