| This thesis mainly studies future-spot arbitrage and calendar-spread arbitrage with CSI 300 stock index future. After necessary introduction to CSI 300 index and CSI 300 index future, both future-spot and calendar-spread arbitrage model in imperfect market are given with detailed analysis of model parameters such as dividend rate, impact cost and margin etc. Using 2007 high-frequency mock trading data from China Financial Futures Exchange in combination with spot market data, the two arbitrage models are analyzed and various risk factors are discussed. During future-spot arbitrage analysis, two different types of index replication method including industry stratified sampling and ETF portfolio replication are investigated. I find that a lot of arbitrage opportunities exist in mock trading and the size of arbitrage profit is considerable. In terms of future-spot arbitrage, 88% sample points of front-month continuous contract lie beyond no-arbitrage interval. While the positive miss pricing error(MPE)decreases with time to maturity (TTM), the negative MPE does not show any significant relationship with TTM. Average return of positive and negative arbitrage strategy are 4. 13% and 3% respectively. In terms of calendar-spread arbitrage, actual spread between back-month and front-month contract is unreasonable, which lead to unstable equilibrium spread. As the spread trend has to be subjectively judged, calendar-spread arbitrage may involve high rollover risk. Arbitrage trading is a mainstream in international capital market. Not only can it pull future price to reasonable level thus maintaining market efficiency, but also can gain risk-free profit. Through the analysis of CSI 300 stock index future arbitrage, this paper is in hope of providing some scientific operational guidance for the policymakers as well as index future investors. |
PDF Full Text Request