تداوم عادت، موانع تعدیل عامل تولید، و بازده دارایی ها در مدل تعادل عمومی با انتظارات خودبرآورد

I examine asset returns in the context of real dynamic stochastic general equilibrium economies with multiple equilibria (indeterminacy) that allow for aggregate fluctuations due to non-fundamental belief shocks. The two models include habit formation in preferences. Model 1 combines restrictions on factor mobility and adjustment costs in a one-sector economy. Model 2 uses restrictions on factor mobility in a two-sector economy. Results demonstrate that Model 1 fails to match the stylized financial facts. Model 2 replicates the low risk-free rate and the standard deviation of the return on the risk-free asset, but underestimates the equity premium and standard deviation of the return on equity.

In this study I investigate the asset pricing implications of real dynamic stochastic general equilibrium (DSGE) economies with multiple sunspot equilibrium solutions. Multiple equilibria or indeterminacy arise when agents act based on expectations, not grounded in fundamentals of the economy. By their behavior these beliefs may become self-fulfilling. Models with self-fulfilling expectations enjoy a measure of success in explaining the stylized facts of the business cycle, and in this regard they are difficult to distinguish from the neo-classical real business cycle (RBC) models with unique equilibriums. However, neo-classical (determinate) and indeterminate economies differ greatly with respect to the shock propagation mechanism, and, consequently, in their policy implications. [See Benhabib and Farmer (1999) for a detailed survey of literature on indeterminacy.] Financial implications of models with multiple equilibria have not been extensively studied. Indeterminacy of equilibrium in a model economy allows for the introduction of belief or “sunspot” shocks. Sunspot equilibria may involve large fluctuations in macroeconomic variables. In the context of DSGE models, in which asset prices and macroeconomic variables are endogenously connected, sunspots influence asset returns. This connection is a potentially attractive feature of the models with indeterminacy, since financial markets are driven to a large extent by investors' beliefs. My first objective in this research is to evaluate the extent to which self-fulfilling expectations change the asset pricing implications of DSGE models. I focus on matching the equity premium, average risk-free rate, and the volatilities of asset returns observed in US data. My second objective is to contrast the financial implications of indeterminate models with asset pricing results obtained in their counterpart models with determinate (unique) equilibrium. The two models, I analyze, share both habit formation in investors' preferences and factor market inflexibilities in production. However, they differ in their production technologies. The first specification combines capital adjustment costs with increasing returns to scale in a one-sector model with predetermined labor. The second is a two-sector model with increasing returns in the investment producing sector only, and restricted capital and labor mobility between sectors. The logic underlying this particular choice of features is that habit formation in preferences of representative investors breaks the link between the model's stochastic discount factor (pricing kernel) and the rate of consumption growth. Without habit, the standard deviation of the pricing kernel, itself proportional to consumption growth, must be low in order to replicate the standard deviation of the consumption growth rate of about 1% estimated in the data. Alternatively, the Hansen and Jagannathan (1991) study implies that accounting for the high equity premium in the US data in the context of the consumption-based DSGE framework requires the standard deviation of the stochastic discount factor to be at least 50% annually. Habit formation dramatically increases investors' desire to smooth consumption while factor market inflexibilities, such as adjustment costs, predetermined labor, and restricted factor mobility between production sectors, prevent instantaneous costless factor adjustments. Researchers often use a combination of the two elements to improve asset pricing implications in general equilibrium in the determinate (no sunspots) setting. Mehra and Prescott (2003) and Campbell (1999) present extensive reviews of the asset pricing literature on the determinate DSGE framework. To explore the effectiveness of these features in explaining the stylized financial facts in a multiple equilibria setting, I first establish that they are compatible with indeterminacy. The next step is to quantitatively evaluate the asset pricing implications of the two models. I find that the one-sector model with self-fulfilling expectations is not consistent with the high equity premium and the volatility of the return on equity. In contrast, the determinate one-sector model matches these features of asset returns well, although it produces an extremely volatile risk-free rate. The two-sector model with sunspots is able to replicate the average risk-free rate and the standard deviation of the return on the risk-free asset observed in the data. This is an improvement over the corresponding determinate model developed in Boldrin, Christiano and Fisher (2001), which results in extremely volatile return on the risk-free asset. On the other hand, the indeterminate two-sector model does not reproduce the high equity premium observed in the US data. The corresponding determinate model matches the equity premium. The paper proceeds as follows: Section 2 outlines the two models with self-fulfilling expectations and presents the solution method that I use throughout this paper. In Section 3 I describe my choice of parameters and analyze the implication of habit formation and restrictions on factor mobility for indeterminacy. In Section 4 I discuss the quantitative implications of two indeterminate models and contrast them with quantitative implications of the corresponding determinate models. Section 5 concludes. 2. Two models with self-fulfilling expectations In this section I outline two decentralized production economies and explicitly define the prices of financial securities.

In this paper I focus on the ability of DSGE models with self-fulfilling expectations to match the equity premium, the average risk-free rate, and the standard deviations of asset returns observed in the US data while maintaining their ability to replicate stylized business cycle facts. To improve their asset pricing implications, the models include habit formation in preferences and restrictions on instantaneous factor adjustments. My results show that the indeterminate one-sector model produces a near-zero equity premium and excessively smooth asset returns. In contrast, the corresponding determinate model replicates the equity premium and high standard deviation of the return on equity found in the data, but it substantially exaggerates the standard deviation of the risk-free rate. I attribute the difference between two sets of results to the size of capital adjustment costs. To preserve multiple equilibria, adjustment costs have to be much smaller in the indeterminate model than in its determinate counterpart. With mild capital adjustment costs, agents can adjust their consumption plans and respond to shocks by trading in equity security; therefore, stocks do not earn a high premium. In the indeterminate two-sector model, the magnitude of the equity premium has the right order, but it is still 3.5 times smaller than in the data. The standard deviation of the return on equity is lower than its empirical value. The determinate two-sector model replicates the equity premium and the standard deviation of the return on equity, but produces an unrealistically volatile risk-free rate. My conclusion is that the task of explaining the stylized asset pricing facts in the context of local indeterminacy is even more challenging than in the models with unique equilibriums. The elements of DSGE models that facilitate self-fulfilling expectations, such as increasing returns and the costless capital adjustment mechanism, work against the ability of these models to produce high equity premium and substantial volatilities of financial returns.