Premium
Nonlinear Modelling Using the Generalized Exponential Family of Distributions[Note 1. We are grateful to two referees and the Editors ...]
Author(s) -
Creedy John,
Martin Vance L.
Publication year - 1998
Publication title -
bulletin of economic research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.227
H-Index - 29
eISSN - 1467-8586
pISSN - 0307-3378
DOI - 10.1111/1467-8586.00063
Subject(s) - flexibility (engineering) , nonlinear system , variety (cybernetics) , exponential function , econometrics , economics , mathematical economics , variable (mathematics) , distribution (mathematics) , mathematics , mathematical optimization , computer science , statistics , mathematical analysis , physics , quantum mechanics
This paper introduces an approach to nonlinear modelling which is based on the use of the generalized exponential family of distributions. The flexibility of the approach is illustrated using hypothetical data based on an economic model which exhibits multiple equilibria for certain periods of time and a unique equilibrium for other periods. The distributional analogue of multiple equilibria is multimodality. An advantage of this framework is that discrete jumps can be modelled without the need for identifying the timing of jumps ex post . The framework also has the advantage of explaining how smooth changes in market fundamentals can give rise to large and sudden changes in prices. The introduction of economic assumptions into nonlinear models is explained, and it is shown how an explicit form for the distribution of the dependent variable can be derived. It is suggested that the approach has considerable potential in a wide variety of economic contexts.