Money demand estimation: An application of the bootstrap.
Item
-
Title
-
Money demand estimation: An application of the bootstrap.
-
Identifier
-
AAI9807975
-
identifier
-
9807975
-
Creator
-
Narine, Kenneth.
-
Contributor
-
Adviser: Temisan Agbeyegbe
-
Date
-
1997
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Economics, General | Economics, Theory
-
Abstract
-
Money demand studies attempt to quantify the linkages between money, income, prices, and some measure of the opportunity cost of holding money. The standard econometric models used in these studies are reduced form single equation models which produce estimates of the short run parameters. The long run parameters are deduced from the short run estimates.;This study was designed to: (1) Produce a reduced form single equation model of the long run demand for money which explicitly shows the long run coefficients, and, using parametric techniques, retrieve the coefficient estimates together with the relevant statistics for hypothesis testing. (2) Apply a non parametric procedure, the bootstrap, to estimate the same model, and to derive the relevant confidence intervals. (3) Evaluate the results of the two procedures and compare the estimates with those produced by other researchers. (4) Assess the non-parametric estimates with regards to stability, convergence, and sample size.;The literature is summarized. Conventional demand functions, time series models, and bootstrap techniques and models are analysed. The long run model is developed and applied.;The results show: (1) Parametric and bootstrap estimates of the model are of comparable magnitudes with those derived by other researchers in the field, and identify periods of instability. (2) Confidence intervals produced by the bootstrap accept the regression values of the income elasticity parameters. These intervals are stable and converge rapidly. (3) Strong conclusions cannot be made with regards the bootstrap and sample size.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
