Sanov meets Cramér again

We introduce a new method for proving large-deviation principles (LDPs). This method amounts to "mixing" a collection of LDPs with a sequence of probability measures that obeys itself an LDP. As an illustration, we construct from empirical measures a sequence of capacity-valued maps that can be considered as an indexed collection of LDPs. The index 1 coincides with Sanov's theorem. By taking a Poisson mixture we establish a new connection with Cramér's theorem.     

Computing equilibria of GEI by relocalization on a Grassmann manifold

An algorithm is described to compute equilibria of the general economic model with incomplete asset markets, that is, of GEI. The algorithm is based on the existence of a route of zeros of a homotopy whose domain includes the price simplex and a Grassmann Manifold. This route is followed, in effect, by localizing and following diffeomorphic pieces in Euclidean space, and by relocalizing as is necessary.     

Buildings for the Japanese information economy : Neighbourhood and resort offices

A new physical infrastructure to support activities based on communications and the transmission of information is being developed. Part of this infrastructure includes the construction of ‘intelligent’ buildings which are seen to play as important a role in improving efficiency of office workers as automation has played on the shopfloor of manufacturing industry. The location of these buildings is important. This article focuses on two types of experimental office development—neighbourhood offices and resort offices.     

Reporting Discretion and Private Information Communication through Earnings

We model a two-period pure exchange economy where a risk averse manager, who has private information regarding future earnings, is required to issue an earnings report to investors at the end of each period. While the manager is prohibited from directly disclosing her private information, she is allowed to bias reported earnings in the first period, subject to GAAP rules that require that a specified proportion of the bias be reversed subsequently. We show there is a minimum threshold of reversal, such that, when the proportion of required reversal is above this threshold, the manager smooths income and communicates her private information through reported earnings. Consequently, the market attaches greater weight to reported earnings than under a regime that allows no discretion. When the required reversal is below the minimum threshold, the manager increases reported earnings without limit and the equilibrium degenerates. When the manager is not endowed with any private information, the market unravels the "true" earnings and price is unaffected by earnings management. Our results underscore the importance of both allowing and restricting reporting discretion through formal mechanisms.