The objective of this article is to criticize neoclassical models of asset price bubbles and to argue that a general theory of asset price cycles demands the substitution of rational expectations hypothesis for bounded rationality assumption. In order to do that we will initially present the two neoclassical approaches for the problem of asset price bubbles. The first one, based on models of multiple equilibria with rational expectations, take financial markets as competitive and investors's behavior as based on perfect and complete information. In this setting, asset bubbles are a logically possible but unprobable phenomenon since their ocurrence will be associated with problems of dynamic ineficience which are not a relevant problem for most of capitalist economies. The second approach, initially developed by Krugman, take as a starting point the idea that financial market are far from perfect. In fact, these markets have a great number of imperfections as, for example, moral hazard. In this approach, asset price bubbles are the result of trading in assets with low supply-price elasticity as, for example, equities and land. Although this second approach is more realistic than the first, it is not capable to explain in a unified framework the appearance, propagation and burst of the speculative bubble; i.e the phenomenon of asset price cycles. This second approach is only capable to show the conditions for the existence of an asset bubble; but it is not capable to explain the dynamic evolution of the bubble. This question is better adressed by heterodox literature based on the hypothesis of bounded rationality.The objective of this article is to criticize neoclassical models of asset price bubbles and to argue that a general theory of asset price cycles demands the substitution of rational expectations hypothesis for bounded rationality assumption. In order to do that we will initially present the two neoclassical approaches for the problem of asset price bubbles. The first one, based on models of multiple equilibria with rational expectations, take financial markets as competitive and investors's behavior as based on perfect and complete information. In this setting, asset bubbles are a logically possible but unprobable phenomenon since their ocurrence will be associated with problems of dynamic ineficience which are not a relevant problem for most of capitalist economies. The second approach, initially developed by Krugman, take as a starting point the idea that financial market are far from perfect. In fact, these markets have a great number of imperfections as, for example, moral hazard. In this approach, asset price bubbles are the result of trading in assets with low supply-price elasticity as, for example, equities and land. Although this second approach is more realistic than the first, it is not capable to explain in a unified framework the appearance, propagation and burst of the speculative bubble; i.e the phenomenon of asset price cycles. This second approach is only capable to show the conditions for the existence of an asset bubble; but it is not capable to explain the dynamic evolution of the bubble. This question is better adressed by heterodox literature based on the hypothesis of bounded rationality.