Dynamic VWAP Strategies In China’s Stock Market

Along with the rise and development of algorithmic trading, the study ofvarious algorithms has become one of the main issues in the field of modernfinancial research. As a basic algorithmic trading strategy, VWAP(Volume-Weighted-Average-Price) strategy has already been one of the mostused strategies among institutional investors. VWAP strategies could reducethe market impact and then reduce the execution cost by splitting the mainorder to several sub-orders according to the distribution of market volume. It isquite important for the institutional investors who may incur heavy marketimpact during trading executions. The core of VWAP strategies is to estimatethe distribution of intra-day trading volume, and its accuracy will play a vitalrole in the execution of strategies. That’s why we are aimed to better estimatethe intra-day volume distribution to improve the performance of VWAPsplitting strategies.However, most of previous literatures focused on the static and historicalVWAP strategy using low frequency data. In this paper, our objective is tocreate a dynamic VWAP framework that could incorporate the observable, intra-day and stock-specific information. We will start with historical VWAPmethod and gradually adjust it with instantaneous high frequency data dividedby5-minutes bins. We will calculate the tracking errors between the volumefrom VWAP strategies and the market volume to evaluate the performance. Asa result, dynamic VWAP could outperformance historical VWAP with anaverage probability of56.7%. As to the robustness test, dynamic VWAPstrategy stays the same among stocks with different intra-day characteristics.Thus, the dynamic VWAP strategy has a wide applicability in China’s stockmarket, while it seems to have smaller tracking errors for stocks with lowturnover and high daily volume.After we have modeled and confirmed that dynamic VWAP strategy couldoutperformance the historical strategy, we will next focus on the orderingmethods after the splitting the orders by making a research on the optimalliquidation strategies based on limit orders. In this paper, an exponentialprobability function for stock deals will be given and further deduction willrefer to the optimal solution of limit orders to a liquidation problem. Analyticalsimulations, as well as Monte-Carlo methods will help make an averagedtrading curve. This method also took the liquidity risk into consideration. If theorders couldn’t be absorbed by market volume, then the limit orders willchange to market orders automatically. In short, this paper studied the combination strategy of both the dynamicVWAP splitting strategy and the ordering strategies by best ask-price for limitorders. Based on both modeling and solid demonstration, the optimizedexecution strategy will help China’s institutional investors participate in stockmarket with low execution cost, which will naturally reduce the market impactand improve the liquidity of the whole china’s stock market.