Cass transversality condition and sequential asset bubbles (Q1762759): Difference between revisions

\textit{D.~Cass}, in his seminal papers [``On capital overaccumulation in the aggregative model of economic growth: a complete characterization'', J. Econ. Theory 4, 200--223 (1972), ``Distinguishing inefficient competitive growth paths: a note on capital overaccumulation'', J. Econ. Theory 4, 224--240 (1972)], provided a complete characterization of the efficient growth paths in Solow models in terms of the properties of supporting competitive prices. Such a transversality condition involves the divergence of a certain series. The present study addresses a different issue concerning infinite-horizon stochastic economies, not properly related to the efficiency, but to the existence of valuation bubbles for long-lived assets. It is shown that such an issue turns out to be closely related as well to the behavior of a random series like the one proposed by Cass. This is quite surprising since it is well-known that the connection between inefficiency and the existence of pricing bubbles is fairly loose. The author introduces the short-run pricing equilibria \[ a_tp_t = {\mathbf E}_t [a_{t+1} (p_{t+1} + d_{t+1})], \tag{1} \] where \(p_t\) is the spot prices vector of assets, \(d_t\) is their dividend vector, while \(a_t\) denotes a state-price process. A brief description of asset pricing bubbles according to a given state price processes is discussed. The relation existing between arbitrage-free prices, obeying Eq. (1), and the behavior of the series \[ \sum_{t=0}^\infty\frac{| d_t| }{| p_t| },\tag{2} \] where \(| \cdot| \) is the vector \(l_1\) norm. Several results illustrate the intimate connection between the divergence of (2) and the absence of bubbles. The main one is theorem which establishes the non-existence, unambiguously, of pricing bubbles, provided series (2) goes to infinity uniformly. This simple result can be established thanks to a supermartingale property of a related random process. Weaker versions of this basic property are given. It is also considered the more difficult issue of relating asymptotic behavior of (2) to truly market pricing equilibrium, where some long-lived agent is assumed to exist. While these results are here less sharp, the author proves that the series is forced to diverge or, at least, to behave in a quite unbounded way. Indeed, it is established that the series diverges uniformly under an assumption of impatience of some long-lived agent.