Despite the extensive advancement of knowledge in the field of empirical asset pricing, little is known about how this literature applies to asset classes beyond common stocks and bonds. In this paper we apply recent developments in financial economics, which posit an important role for the leverage of financial intermediaries and limited stock market participation, in understanding real estate returns. Consistent with these theories, we find that luxury consumption, funding liquidity and the capital share of income have significant explanatory power for the cross-section of equity REITs. However, this relationship is the opposite of what we expected, and the results point to a more complex set of findings that are difficult to reconcile with risk-based explanations. Our results suggest systematic mispricing of real estate assets that is heavily influenced by investor sentiment. stocks, directly or indirectly. Guo ( 2004) incorporates these market frictions, i.e., idiosyncratic labor income shocks, borrowing constraints, and limited stock market participation, in an otherwise standard consumption-based CAPM. Guo shows that the modified heterogeneous-agent consumption-based CAPM provides a coherent explanation for several well-known stock market stylized facts such as the equity premium puzzle, stock market return predictability, and excess volatility puzzle. 1 In Guo's model, assets are priced by shareholders' consumption. Vissing-Jorgensen (2002) and Aï t-Sahalia, Parker, and Yogo (2004) find that shareholders' consumption and luxury-goods consumption, respectively, provide a better explanation for the cross-section of stock returns than aggregate consumption. While these empirical findings are encouraging, it is a challenging task to test the limited stock market participation model empirically because it is difficult to measure the consumption of marginal shareholders who are likely to be very wealthy and underrepresented in the Consumer Expenditure Surveys used by Vissing-Jorgensen (2002).The intermediary asset pricing model is built on the premise that a financial intermediary is the marginal investor whose "consumption" sets asset prices. For example, in He and Krishnamurthy's (2013) model, only sophisticated investors, i.e., financial intermediaries, can trade risky assets. Unsophisticated investors, i.e., households, can invest in risky assets only through a financial intermediary. While this assumption is clearly unrealistic, it might hold approximately, especially for complex assets, e.g., mortgage-backed securities and credit default swaps, and perhaps commercial real estate, which are traded mainly by sophisticated investors such as large financial intermediaries. That is, the intermediary asset pricing model is a variant of 1 The representative-agent models by Campbell and Cochrane (1999) and Bansal and Yaron (2004) can also explain these stylized facts. However, unlike Guo (2004), these models cannot explain the unstable relation between stock market volatility and the dividend yield documented by ...