The valuation of equities remains a challenge both in practice and in theory. The value of an asset must reflect the present value of the future payments that accrue to the holder of the asset. Hence, the key question in equity valuation is how to determine the value today of a sequence of uncertain future cash flows or earnings by accounting for both the time value of money and the uncertainty (i.e., the risk) associated with these payments. Since the time value of money is readily observable, the challenge remains in how to account for risk. Not only is risk a central issue but also taxes, inflation and growth affect the present value of an asset.The standard approach to equity valuation is to use the CAPM or the Fama-French three-factor model to determine a risk-adjusted cost of capital used to discount expected future cash flows or earnings. Unfortunately, the valuations are quite imprecise and very sensitive to estimates of betas and market risk premiums. This way of accounting for risk is justified if all shocks to future cash flows or earnings are fully persistent on a percentage basis, i.e., if cash flows or earnings follow a geometric Brownian motion (Bach and Christensen, 2016). However, the competitive advantage of a company is often only temporary and the force of competition in the product markets implies that the cash flows or earnings exhibit mean-reversion, i.e., contains both persistent and transitory shocks. This mean-reversion of future cash flows or earnings for firms and industries has been documented by Nissim and Penman (2001). Moreover, the standard valuation models are based on single-period asset pricing models, which do not easily carry over to multiperiod settings. Theoretically, there is a call for a valuation model, which provides a better link between the time-series properties of a firm’s fundamentals and the associated risk-adjustments.Multi-period asset pricing models suggest that the risk-adjustments should be done in the numerator (Rubinstein, 1976). These risk-adjustments are determined by how the firm’s fundamentals covary with a so-called valuation index (a normalized eventprice deflator). This is in stark contrast to the standard valuation models where the risk-adjustments are done in the denominator. A consumption-based capital asset pricing model (CCAPM) is presented and compared to the standard valuation model. This model is based on modern asset pricing theory where the prices of financial assets are linked to (the marginal utilities of) consumption. Since financial markets allow investors to shift consumption opportunities across time and states, equilibrium asset prices should reflect investors’ desire to make these shifts, which is closely related to their marginal utilities of consumption.The CCAPM-based valuation model developed by Christensen and Feltham (2009) is theoretically superior to the standard CAPM-based valuation model. However, the basic consumption-based asset pricing model gives rise to several asset pricing puzzles as the model is unable to explain some stylized empirical facts. The most famous puzzle is the equity premium puzzle, which in equity valuation results in very conservative riskadjustments as the stochastic variation in real aggregate consumption is too smooth to fit observed risk premia in asset prices for reasonable relative risk aversion parameters. This has motivated us to extend the equity valuation-framework with recursive utility, offering a possible resolution to the equity premium puzzle in the model. The purpose of this thesis is to develop an Epstein-Zin asset pricing model, which will then be extended 6 into an equity valuation model. This serves as our contribution to the equity valuation research.In short, this thesis will study three equity valuation models, namely (i) the standard CAPM-based valuation model, (ii) the CCAPM-based valuation model and (iii) the Epstein-Zin-based valuation model. We begin by reviewing basic asset pricing theory from which all three valuation models are derived. Having set the theoretical framework, we will link the asset pricing theory to equity valuation by presenting a set of accounting relations used to re-express the dividends in terms of accounting numbers. In order to make the two consumption-based valuation models empirically implementable, we derive equilibrium solutions for the risk-adjustments in the models. We compare the performance of the valuation models in an empirical study along two dimensions: (a) the efficient market perspective and (b) the fundamental valuation perspective
