The object of the thesis is to investigate, measure and analyse the impact of liquidity on portfolio value, risk and execution. We consider the formalism of [Acerbi and Scandolo, 2008] to value portfolios in markets exposed to illiquidity through the use of Marginal Supply Demand Curves. We show that future portfolio returns become fat-tailed when liquidity risk is introduced. Further, we investigate the market impact model of [Almgren et al., 2005], who estimates supply curves on equity instruments by considering a large database of executed orders. Since such data are highly confidential, we propose to use transaction data to estimate the same supply curves. This may enable more market participants to assess their liquidity risks and costs. Transaction data does not contain the same information as order data. To bridge the information gap between the two data sets, we introduce a 'strategy identifier'. By using regression and filtering techniques we show that using transaction data together with the strategy identifier give results comparable to using order data. Finally, we combine the formalism of [Acerbi and Scandolo, 2008] with the supply curves of [Almgren et al., 2005] and expand the notion of Marginal Supply Demand Curves to a stochastic object to model future liquidation prices. We find that a portfolio owner required to liquidate a large position will be faced with a trade off between liquidating fast to a high liquidity cost, versus liquidating over a longer time span but with higher market and liquidity risk.