We present an order-driven asset price model where liquidity determines both price impact and the intensity of subsequent orderflow. This allows for a natural form of market resilience and also provides a direct connection between market liquidity and asset price volatility. We then investigate how a trader will process dynamically generated trading signals to maximize her expected utility from her terminal position after liquidation, taking into account that her trades affect the market in the same way as exogenously generated order flow. Mathematically, this is accomplished by using the theory of Meyer sigma-fields and by deriving and solving numerically the corresponding HJG equation. (This is joint work in progress with Alvaro Cartea, Oxford, and Laura Koerber, TU Berlin.)