Research On Trading Strategies Based On Implicit Trading Costs On The Futures Markets

The fast development of the futures markets makes it necessary to enhance underlying theory research. Along with internationalized finance, algorithmic trading will play a more and more important role in market competition as it can accurately compute the trading costs, fund efficiency, optimal trading time, and so on. Occasionally happened market manipulation and squeeze make all market participants realize the importance of insights into market risk.Based on these backgrounds, this thesis focuses on the research on trading strategies based on implicit trading costs on the futures markets and aims at providing market traders with decision-making support. Market traders include hedgers, arbitragers and speculators. The approaches proposed in this paper are mainly for hedger and arbitrager, with little contribution to speculators. Firstly, this thesis researches on the measurement of implicit trading costs such as reserve margin and impact costs, and then puts forward some strategies for fund management and block trading execution. Secondly, a no-arbitrage pricing model is built considering the implicit trading costs, and then the trading timing decision strategies are puts forward. Lastly, based on the analysis of squeeze risk, some market risk responding strategies are set up. This thesis mainly consists of the following aspects:(1) Measure the requirement of the reserve margin in futures markets, derive safety fund leverage and safety fund utilization rate, complete empirical studies on China copper and aluminum futures markets, and finally put forward the fund management strategies for different traders. This research result helps market participants optimize their funds allocation and control default risk.(2) Measure impact costs, analyze their influence factors, finish empirical studies on China copper and stock index futures markets, and finally suggest some block trading executive strategies. These strategies are helpful for institution investors in reducing trading cost, designing trading strategies, ordering structure, and so on.(3) Design the no-arbitrage pricing boundary model considering reserve margin and impact costs on futures markets, prove the relationship between market state and arbitrage, conduct empirical studies on China copper futures markets, and finally put forward the trading timing decision strategies which help all market participants make trade decisions.(4) Design a squeeze risk index (SRI) and put forward an alarming method combining quartile thresholds and constant thresholds. After finishing the empirical studies on China rubber futures markets, this thesis puts forward some market risk responding strategies. These results are helpful for all market participants in market risk identity and risk management.The thesis draws the following conclusions:(1) The reserve margin lookback option model is suitable for market participants to manage fund. The empirical results show that the investors need more reserve margin than initial margin. The investors will need more reserve margin under short positions than under long positions, more reserve margin for distant contracts than for nearby contracts. The back testing results show that the model can control price risk very well. The reserve margin model is best for hedgers, less useful for futures-spot arbitragers and long term speculators, and no use for spread arbitragers and short term speculators. The paper then presents the fund management strategies for different traders in details.(2) Impact cost can be described through the measure formulas of permanent impact, temporary impact, realized impact and implied impact. The linear regression equation for impact costs is very helpful for trader to forecast impact costs. The empirical results show that there's significant price impact to copper futures and HS300 stock index futures. The impact costs are both less than the equity markets'impact costs. There is no significant asystematry between ask and bid in the realized impact. The level of the realized impact under different delivery periods is consistent with contracts liquidity. Different from the literature, it implies that the greater information asymmetry in futures markets than in equity markets. The analysis the influence factors of impact costs suggests that the explanatory power of the model improves significantly comparing with previous results. Most of the explanatory power comes from the trader-identification dummy variables, daily price variables and daily settle price variables. The block trading executive strategies are helpful for hedgers and arbitragers, but little use for speculators. Some suggestions are also put forward.(3) Taking into account of various explicit and implicit trading costs, the paper presents more scientific no-arbitrage pricing bound model, and then proves there is no cash-and-carry arbitrage opportunity in the normal market. The empirical results on China copper futures markets show that arbitrage profit decreases approaching to maturity. In the normal market, there are few arbitrage opportunities with low profit. In the inverted market, there are many reverse cash-and-carry arbitrage opportunities with high profit. The paper proves the inevitability of the results according to the theories and realities. The trading timing decision strategies based on the no-arbitrage pricing model are most helpful for arbitragers, less helpful for hedgers, and no use for speculators. The paper also puts forward some concrete trading timing decision strategies for different traders.(4) The squeeze risk index and the united alarming method with the quartile thresholds and constant thresholds constructed in this paper are suitable for identifying squeeze risk. The empirical study to China rubber futures markets shows that it is better with the combination of quartile threshold and constant threshold under subsection sample setting by China's margin adjustment system. And the paper finds that there is significant squeeze risk on the ru0306 and ru0407 contracts. It is consistent with the results in practice. The market risk responding strategies based on squeeze risk are most helpful for arbitragers, less helpful for hedgers, and no use for speculators. The paper puts forward concrete responding strategies for different traders.The major contributions of this thesis are:(1) Appling the lookback option model, the paper develops the reserve margin demand under holding the short or long positions. And the paper analyzes reserve margin level on China copper and aluminum futures markets.The setting of initial margin has been intensively studied by the literature on futures markets. However, the setting of reserve margin has not been studied yet. The paper is the first to analyze reserve margin in futures markets. Reserve margin describes the extreme price change in life-span of the holding period, different from initial margin that describes the extreme price change intraday. This path dependence is the same with lookback option. So, the paper develops the reserve margin model according the lookback option model of Gonze and Viswanathan's (1991). Furthermore, the paper introduces the formulas of safety fund leverage and safety fund utilization rate under holding short or long positions. This paper is also the first empirical analysis on the reserve margin level on China copper and aluminum futures markets and use back testing to examine efficiency.(2) Putting forward the measuring formulas of impact costs and the linear regression model of its influence factors of the block trades, and finishing empirical studies on China copper and stock index futures markets.Previous research has made many investigations on market-impact costs. However most of the literature focuses on equity markets, but little has been done on futures markets. Specially, the research on China futures markets is absent. The paper re-defines impact costs and classes from the prospective of elasticity and persistency. Based on impact costs formulas of Frino and Oetomo's (2005), the papers puts forward the measuring formulas of the permanent impact, temporary impact, realized impact and implied impact, then gets the conditional inequalities and impact identities. Otherwise, the linear regression equation is built with six independent variables for analyzing the influence factors of the impact costs. Comparing with the previous research, the paper firstly consider the range of intraday price variable, daily settle price variable, daily volume variable and daily open interest variable leading to improve the whole explanatory power. The paper also firstly analysis the influence of the different delivery periods to impact costs.(3) Taking into account of various emplicit and implicit trading costs, the paper presents the more scientific no-arbitrage pricing boundary model, proves the relationship between market state and arbitrage, and is followed by an empirical study on China copper futures markets.No-arbitrage equilibrium analysis is the core technology in finance field. More and more friction factors are taken into futures no-arbitrage pricing account. But the empirical study on commodity futures is little. The paper indicates that implicit trading costs are important for futures no-arbitrage pricing. So the paper firstly adds reserve margin and impact cost into no-arbitrage pricing model, proves the relationship between market state and arbitrage, and then accomplishes an empirical study on China copper futures markets. The paper distinguishes normal markets and inverted markets in empirical study. The paper also indicates it's theoretically and empirically inevitable that there are many reverse cash-and-carry arbitrage opportunities with high profit. (4) The paper designs squeeze risk index on the futures markets. And the united alarming method with the quartile thresholds and constant thresholds are put forward. There is firstly an empirical study on China rubber futures markets.About squeeze risk identification, the previous research focuses on building index system. But there is not a synthesis index to describe squeeze risk up to date. The out-of-sample efficiency might be a problem for nonlinear mapping and KLR signal approaches. And it maybe exist reliability problems of ahead alarming to SV and GARCH approaches. So the paper puts forward a real-time united alarming method with the quartile thresholds and constant thresholds. Through comparing the whole sample with subsection sample windows in empirical study, the paper finds that the united alarming method under subsection sample windows is better.