Research On The Liquidity Measurement,Transmission And Risk In The Chinese Agricultural Commodity Futures Market

Liquidity is the vitality of modern financial market system and the lubrication of financial resources allocation.Since the beginning of the 21st century,the Chinese commodity futures market has deepened increasingly the liquidity pool,with an increase of futures contracts in varieties and trading volumes.Last decade has witnessed enhanced financial attribute and growing liquidity volatility in agricultural commodity futures,which has been a potential threat to agricultural market stability and food security in China.Strategic portfolio allocation and intersection of orders not only help to achieve cross-market benign interaction,but also build the bridge of risk contagion among markets.It leads naturally to some scientific issues that how does liquidity performance?how has it evolved individually and transmitted across markets?and to what extent does individual liquidity risk contribute to the systemic risk?Therefore,there is a need to investigate liquidity level,cross-market transmission and liquidity risk in the Chinese agricultural commodity futures market.Based on that,it attempts to identify liquidity"risk sources" to provide a reference point for risk monitoring and preventing,which is of significance to maintain the agricultural security and economic and social development in China.Using a unique dataset containing both high-frequency and low-frequency trading information,this paper bases on high-dimensional measurements and effect analyses,uses liquidity proxy to investigate individual evolution and cross-market transmission of market liquidity,and eventually constructs liquidity risk assessment and transmission network in Chinses agricultural commodity markets.With a combination of theoretical and empirical analyses,this paper develops an analytical framework to study three aspects of liquidity issues:liquidity measurement,liquidity transmission,liquidity risk.The main contents and findings in this dissertation are as follow:(1)High-dimensional measurements on market liquidity and in-depth analyses of liquidity effects in Chinses agricultural commodity markets.Taking soybean,soybean meal,soybean oil,palm oil,rubber,corn,cotton and sugar for example,this section is two-fold:the first part is to measure market spread,depth,immediacy and resiliency in high resolution based on real-time trading database of 2016-2018;the second part is to analyze the role of liquidity in shaping trade activity and futures price using regression model and asset pricing model with consideration of periodic effects and based on the "market microstructure" theory to explain intraday(inverse)U-shaped and intraweek(inverse)V-shaped liquidity distributionResults show that:Firstly,the trading size is a crucial factor in shaping comprehensively liquidity evaluation,of which large-sized trading usually deepens market but widens spread,reduces immediacy and lower resilience.Secondly,multi-dimensional and comprehensive measurements of liquidity illustrate that sugar and soybean markets where small-sized trading prevails are highly liquid,while corn and rubber markets with large-sized trading are low-liquid.Thirdly,informed trading is prevailing in Chinese agricultural futures market with opening and closing effects,and there appears a "liquidity premium" and an "economics of scale" concurrently on Monday.(2)Historical evolution of agricultural commodity futures market liquidity based on an optimal liquidity proxyThis section firstly introduces liquidity proxies widely-accepted in developed markets(i.e.,Roll,Gibbs,Effective Tick,Zeros,FHT,High-Low Spread,Amihud,Amivest and others)and identifies a more suitable proxy for liquidity in Chinese commodity markets by goodness-of-fit test with liquidity benchmarks and robustness test to different data frequency and market types.Furthermore,based on the optimal proxy,daily trading data of futures contracts since listing are used to capture historical patterns of liquidity evolution in agricultural commodity markets,in order to well observe how market liquidity changes with various historical background and eventsResults show that:Firstly,FHT spread is most competitive in fitting liquidity benchmarks,hence identified as the optimal proxy for historical evolution of agricultural commodity futures market liquidity.Secondly,FHT spread captures cross-market similarity in historical patterns of liquidity development,of which 2008 global financial crisis made import-oriented markets(e.g.,rubber,palm oil and soybean-related markets)encounter a greater problem of liquidity dry-ups and 2016 nationwide financial crisis gave corn and cotton futures with a heavier liquidity shock under the market-pricing scenario.Thirdly,agricultural commodity futures experience a high similarity in liquidity movement with industrial and metal futures,despite a lower volatility than them(3)Transmission effects of market liquidity within agricultural commodity markets and across agricultural,industrial and metal commodity marketsThis section firstly applies rolling window approach to co-integration and error correction analyses to capture time-varying transmission effects of market liquidity across markets and investigates and detects multi-breakpoints in these linear transmission effects based on Bai-Perron test.Secondly,to tackle issues of nonlinearity,heavy tail and asymmetry etc.,time-varying Normal and SJC Copulas are further used to study liquidity dependences in general and extreme scenarios,with a focus on asymmetrical tail dependence.Results show that:Firstly,soybean oil and palm oil play a dominant role in affecting the systemic and others liquidity level,followed by soybean meal,soybean and rubber Secondly,industrial commodity affects more the liquidity of agricultural commodity than what metal commodity does,especially after financial crisis in 2008 when tail dependences have gotten stronger.Thirdly,extreme events make cross-market dependences departure from normality,of which joint falling of liquidity under negative shocks is a more normal state across markets(4)Measurement of liquidity risk and its transmission effect within agricultural commodity markets and across agricultural,industrial and metal commodity marketsThis section firstly attempts to build up a real-time VaR model which highlights upward co-movements in spreads based on the definition of liquidity risk,to capture dynamic evolution of individual and sector-wide liquidity risk in agricultural commodity futures markets.Furthermore,delta CoVaR models by Copula-GARCH and GJR-GARCH-DCC estimations are applied to measure transmission effects of liquidity risk within agricultural commodity markets and across agricultural,industrial and metal commodity markets,identify respectively inner and outer "risk source" and draw a pass-through network of liquidity risks in the agricultural commodity futures marketResults show that:Firstly,low-liquidity varieties are prone to confronting with liquidity risks,and CoVaR in investment portfolio is systemically higher than individual VaR.Secondly,there is no single liquidity "risk source" in the agricultural commodity market where oil-fat futures serve as the framework in a risk pass-through network Thirdly,sector-wide liquidity risk of agricultural commodity futures is higher than industrial and metal futures,of which industrial futures transmits more risk to agricultural futures and hence are cited as a primary outer "risk source".There are four-fold contributions of this dissertation.Firstly,it completes comprehensive consideration and rating for liquidity level in agricultural commodity futures markets based on multi-index and high-dimensional measurements,and forms procedure-oriented selection of optimal proxy by constructing a new acceptance standard from high-frequency benchmark to low-frequency proxy.Secondly,it disentangles empirically influence mechanism of liquidity on other economic variables,e.g.,investor behaving and futures pricing,based on the information asymmetry and liquidity premium theories.Thirdly,it defines liquidity risk in commodity markets based on "spread"attribute as transaction cost,identifies multi-"risk source" and creates an intra-system network of risk transmission.Fourthly,it constructs a unique set of continuous and active trading dataset,which not only provides sufficient data support for high-dimensional measurements of market liquidity,but also effectively smooths the jump of variables in contract nodes.