Optimal dynamic asset allocation using a non-linear discrete time liquidity driven microstructure market model and extended Kalman filtering

Sammanfattning: This paper use a discrete time microstructure model which considers excess demand and market liquidity as two unobservable state variables as determinants whether the market is overvalued or undervalued. The model expresses the variation of conditional variance of price, were the amplitude of the price changes is dependent on the liquidity of the market. It is shown that the filtered process of hidden excess demand and its liquidity is meaningful to apply in an asset allocation strategy and is efficient in terms of producing a residual money gain compared to a passive strategy.