Research On Financial Market Microstructure Based On Continuous Double Auction Trading Mechanism

Continuous Double Auction trading mechanism has been widely used in globalfinancial markets.Based on the systematic review of the literatures relevant to thefinancial market microstructure,we find that the existing literatures in the financialmarket microstructure study under continuous double auction trading mechanism hascarried out some useful results,but has not achieved substantial headway in this area yetand the entire theoretical framework has been awaiting the improvement.Thus,according to the uniqueness of the institutional constraints and investor structure inChina security market,we synthetically adopt the market microstructure theory,stochastic process and simulation,and optimization theory to extend and improve theexisting literatures.Our methodology mainly relies on the theoretical methods,butcombines the empirical methods as well.Taking the liquidity supply and demand andtheir dynamics as the central topics,we follow the technical routes that are from generalto specific short-term price behavior,analyzing the market dynamics from puremicrostructure and trading mechanism effects to their interactions with the asymmetricinformation,from the situation of without to with short-sale constraints,to further studythe financial market microstructure based on continuous double auction tradingmechanism.Firstly,we examine the dynamics and statistical properties of the generalshort-term price and trading volume from the viewpoint of the interaction between orderflows and limit order book,and explicitly get the mathematical expressions of executionprobability,best bid and ask,bid-ask spread,transaction price,the mean and variance ofbuy-side and sell-side trading volumes and total trading volume and also theirinfluencing factors.We further analyze the equilibrium properties of continuous doubleauction including the competitive equilibrium price it converges to and correspondingelapsed time and do comparative statics with respect to their influencing factors.Theresults demonstrate that the execution probability of the new incoming limit orders isaffected by current bid-ask spread,tick size and its aggressiveness.Three tradingvolumes aforementioned are jointly affected by order arrival rates from liquidity demanders,average order size,proportional structure between buy and sell orders,execution probability and the time interval we focus on.The uncertainties of threetrading volumes come from that of order flows.Transaction price evolves with themid-point of best bid and ask,which is regarded as the best estimation of the true assetvalue in common,and is the outcome of the adjustments to the mid-point according tothe order-type-weighted execution probability and current bid-ask spread.Moreover,theresults regarding the equilibrium analysis indicate that the transaction price whenbid-ask spread reaching zero is the competitive equilibrium price,and if the depth oflimit order book is deeper or/and the tick size is smaller,then the competition uponliquidity providing is stronger,as a result,price converges to competitive equilibriumfaster thus the price discovery efficiency is higher.Secondly,we extend Sand(?)s(2001) by introducing information process and tradingprocess and then focus on the impact of short-sale constraints and comparatively studythe optimal quote price and depth relationship and strategic behavior of liquidityproviding and their influencing factors between with and without short-sale constraints.The results imply that for liquidity provider on limit order book,if the probability ofinformation event occurs is higher or/and the proportion of the informed trader isgreater or/and the accumulated depth of limit order book is deeper or/and the order sizeof liquidity demander is larger,then the adverse selection risk and cost they face ishigher,therefore their pricing strategy is less aggressive,which means the buy side willlower their bids and the sell side will raise their asks,ceteris puribus.Furthermore,it isseemingly surprising but actually reasonable that the impact of short-sale constraintsupon liquidity providers on the buy side of limit order book merely relates to therelative proportions not absolute magnitude of unconstrained informed traders andnon-discretionary uninformed traders.Meanwhile,if this relative proportion is greater,the adverse selection risks the liquidity providers on the buy side face are higher,hencetheir pricing strategy is less aggressive and bid prices are lowered.Thirdly,based on improving the prevailing empirical methods,we furtherempirically study the magnet effect and liquidity dynamics it accompanies byemploying the high-frequency transaction data from January 4,2002 to December 31,2002 on Shanghai stock market.We find significant evidence of magnet effect.Byexploring the dynamics of bid-ask spread,trading volume and order size prior to limit-hits,we find asymmetric liquidity pattern between ceiling and floor magnet effects,that is,compared to ceiling magnet effect,liquidity in floor magnet effect is worse.Theunique short-sale constraints may be one of the most important reasons for this liquidityasymmetry.The post-limit-hit analysis of stock price performance based on event studymethod shows significant price continuation after ceiling hits and price reversal afterfloor hits,which reflects that it is the irrational and panic selling psychology ofindividual investors that promote the floor magnet effect and the liquidity asymmetry.This supports the results by Wong et al.(2009) that the magnet effect is mainly causedby the individual investors.Finally,we develop a theoretical mierostructure model based on the noisy rationalequilibrium expectation framework to explore the effects of short-sale constraints uponmarket quality including market liquidity,information revelation and volatility,andfurther give a formal theoretical explanation to the liquidity asymmetry that the magneteffect accompanies aforementioned.The results indicate that compared to the case thatshort-sale constraints are non-binding,the total private information and signal-to-noiseratio are less in the case that the short-sale constraints are binding,which leads to a lessinformative price,a larger price impact from the uncertainty of random supply and aweaker reponse from price to private information thus a worse liquidity,a lowerinformation revelation and a higher volatility.With the increase of the informed tradersthat are not short-sale constrained,the market quality would turn better.