Edward J. O'Brien

Investigating Nonlinearity: A Note on the Estimation of Hamilton's Random Field Regression Model

In this paper we give an account of the approach to nonlinear econometric modelling proposed by Hamilton (2001) and briefly describe some of the methods of nonlinear optimization that may be used in the Gauss computer program provided by Hamilton for the implementation of his methodology. The performance of this program is investigated using data relating to Hamiltons example concerning the US Phillips curve, two versions of the Gauss software, and a range of alternative numerical optimization options and values for the Gauss parameter _oprteps. The impact of changes in initial parameter estimates and the use of pairs of optimization algorithms are also briefly examined. Finally, the effects of changes in the sample data on the results produced by Hamiltons procedure are explored. The results presented suggest some clear conclusions, which will be of value to those contemplating working with Hamiltons new method.

Exploring nonlinearity with random field regression

Random field regression models provide an extremely flexible way to investigate nonlinearity in economic data. This article introduces a new approach to interpreting such models, which may allow for improved inference about the possible parametric specification of nonlinearity.

A Note on Spurious Nonlinear Regression

This paper explores the power of two tests for nonlinearity against spurious nonlinear regression. Results show that while the BDS test is susceptible to spuriousness, an approach introduced by Pena and Rodriguez [Pena, D. and Rodriguez, J., 2005, Detecting nonlinearity in time series by model selection criteria, International Journal of Forecasting 21, 731-748.] is powerful, regardless of sample size.

Investigating Nonlinearity: A Note on the Implementation of Hamilton's Methodology

In this paper we give an account of the new approach to nonlinear econometric modelling proposed by Hamilton (2001) and briefly describe some of the methods of nonlinear optimization that may be used in the Gauss computer program provided by Hamilton for the implementation of his methodology. The performance of this program is investigated using data relating to Hamiltons example concerning the US Phillips curve, two versions of the Gauss software and a range of alternative numerical optimization options and values for the important Gauss parameter _oprteps. Finally, the effects of changes in the sample data on the results produced by Hamiltons procedure are explored. The results presented suggest some clear conclusions, which will be of value to those contemplating working with Hamiltons new method.

Weak instruments in estimating business cycle effects on banks' interest income

This article explores the link between the real business cycle and core bank earnings. Using bank-level data and an estimation technique which corrects for weak instruments, evidence confirms that pre-provision Net Interest Income (NII) is determined by the term structure of interest rates rather than output fluctuations. Output growth is only found to be significant when Loan-Loss Provisions (LLP) are taken into account.

Nonlinearity and structural breaks in Irish PPP relationships: an application of random field regression

Using nominal and real exchange rates for Ireland relative to Germany and the UK from 1975 to 2003, this article explores likely sources of nonlinearity in Purchasing Power Parity (PPP) relationships and difficulties in employing an I(1)/I(0) econometric framework. Tests for fractional integration and nonlinearity, including random field regression-based procedures, are applied. Results reveal shortcomings in the standard cointegration and smooth transition autoregression approaches to modelling, and point to multiple structural changes models. Such a model for the case of Ireland and Germany suggests that PPP holds not only in the long-run but also in the medium to short term.

Nonlinearity as an explanation of the forward exchange rate anomaly

This paper shows that nonlinearity can provide an explanation for the forward exchange rate anomaly (Fama, 1984). Using sterling-Canadian dollar data, and modelling nonlinearity of unspecified form by means of a random field, we find strong evidence of time-wise nonlinearity and, significantly, obtain parameter estimates that conform with theory to a high degree of precision: the anomaly disappears.

Economic Base Multipliers Revisited

This paper takes a fresh look at the estimation of economic base multipliers. It uses recent developments in both nonstationary and nonlinear inference to consider issues surrounding the derivation of such multipliers for Northern Ireland. It highlights the problem of distinguishing between nonstationarity and nonlinearity in empirical work. The results of standard unit root and cointegration analysis call into question the adequacy of that framework for estimating employment multipliers. There is strong evidence of nonlinearity; and modelling using random field regression supports the findings of Harrison and Bond (1992) that there is substantial parameter instability or nonlinearity in the data.