ABSTRACTThis paper suggests an optimal execution strategy to minimize expectedcost of a large size order within a fixed time period. Based on ... Show moreABSTRACTThis paper suggests an optimal execution strategy to minimize expectedcost of a large size order within a fixed time period. Based on [42]’s price impactmodel, I include time varying bid-ask spread, a measure of market width as aparameter into the problem, and let not only width, but also depth (order booksize) and resiliency time dependent in a trading day. In addition, I utilize meanreversion regression models to estimate mean resiliency ratio as a parameter inthe execution strategy, with S&P 500 stock data in year 2012. U-shaped intradaypatterns of resiliency are presented when measured by bid-ask spreads, whileCotangent-shaped patterns are shown measured by market depths. Resiliencymovement is then predicted using machine learning techniques. In the end, Iconduct empirical experiments with all three time dependent liquidity parametersand obtain same conclusions with numeric examples. I find out higher expectednet cost savings comparing to costs from model with constant liquidity parameters.Market depth is the primary parameter to the strategy while width and resiliencyare not ignorable. When resiliency is low, cost saving is substantial. Show less
