Alicia Gazely - Academia.edu (original) (raw)

Papers by Alicia Gazely

5th Biennial Conference on Alternative Perspectives on …, 2000

An important element in macroeconomic policy is the relationship between money supply and inflati... more An important element in macroeconomic policy is the relationship between money supply and inflation. The conventional way of measuring the amount of money circulating in an economy is to simply sum the various constituent liquid liabilities of commercial and savings banks. However the belief is widely established that this method of arriving at broad money aggregates is seriously flawed and based on untenable assumptions. From a micro-demand perspective it is hard to justify adding together component assets which have differing yields which vary over time, since the accepted view is that only things that are perfect substitutes can be combined as one ‘commodity’. An important alternative is the Divisia weighted index, in which the component assets are not assumed equal or invariant over time. Components are weighted on the basis of the quantity of monetary services provided by each, expressed in terms of the interest yield. This study utilises a powerful and flexible tool, the Artificial Intelligence technique of neural networks to develop this approach. Can a new empirically weighted measure of money be constructed that more closely captures the monetary services flow provided by the component assets? Sensitivity analyses were derived from neural networks trained with UK data to forecast inflation directly from the components of Divisia, and used to construct a 'new' Divisia style aggregate with superior forecasting performance when compared with the traditional Divisia aggregate. Results to date indicate that the combination of Divisia measures of money with the artificial neural network offers a promising starting point for the development of an optimised Divisia aggregate and hence an improved model of the relationship between money supply and inflation. Our results are reinforced by a growing body of evidence from empirical studies around the world which demonstrate that weighted index number measures may be able to overcome the drawbacks of simpler measures. Ultimately, such evidence could reinstate money as a credible policy tool for controlling inflation.

Forecasting exchange rates with linear and nonlinear models

Spreadsheet designs and the template spreadsheet basics spreadsheet designs advanced techniques.

The Sage Course Companion on Business Information Systems is an accessible introduction to the su... more The Sage Course Companion on Business Information Systems is an accessible introduction to the subject that will help readers to extend their understanding of key concepts and enhance their thinking skills in line with course requirements. Business Information Sytems changes at a rapid pace and the book directs readers to the increasingly wide variety of lighter reading about the field which are available. It provides support on how to revise for exams and prepare for and write assessed pieces.

The subject of enquiry is the effect of high intensities of ambient luminance on task performance... more The subject of enquiry is the effect of high intensities of ambient luminance on task performance by human subjects, and the task employed, the detection of transient light signals in an illuminated display encompassing the whole of the subject's visual field. Experimental conditions were designed to test one particular prediction from previous work, that high intensities of environmental stimulation tend to reduce the range of cues utilised from the environment in performance of a task. ln the case of high intensities of luminance the prediction would be of a reduction in the size of the visual field, or 'tunnel vision'. The data produced does not bear out this prediction. Subjects show a decreased consistency of response when observed under the higher intensities of luminance so that their detection rate for peripheral signals is comparatively lower than for other intensities, but not in a sufficiently clear-cut fashion to be 'tunnel vision' as predicted. [Cont...

IFAC Proceedings Volumes, 2001

Computational Intelligence in Economics and Finance, 2004

Previous work has used neural networks to predict the rate of inflation in Taiwan using four meas... more Previous work has used neural networks to predict the rate of inflation in Taiwan using four measures of 'money' (simple sum and three divisia measures). In this work we develop a new approach that uses an evolutionary strategy as a predictive tool. This approach is simple to implement yet produces results that are favourable with the neural network predictions. Computational results are given.

Applications of Artificial Intelligence in Finance and Economics, 2004

This work applies state-of-the-art artificial intelligence forecasting methods to provide new evi... more This work applies state-of-the-art artificial intelligence forecasting methods to provide new evidence of the comparative performance of statistically weighted Divisia indices vis a vis their simple sum counterparts in a simple inflation forecasting experiment. We develop a new approach that uses coevolution (using neural networks and evolutionary strategies) as a predictive tool. This approach is simple to implement yet produces results that outperform stand-alone neural network predictions. Results suggest that superior tracking of inflation is possible for models that employ a Divisia M2 measure of money that has been adjusted to incorporate a learning mechanism to allow individuals to gradually alter their perceptions of the increased productivity of money. Divisia measures of money outperform their simple sum counterparts as macroeconomic indicators.

Modeling and Control of Economic Systems 2001, 2003

Publisher Summary This chapter discusses the development of an approach that uses an evolutionary... more Publisher Summary This chapter discusses the development of an approach that uses an evolutionary strategy as a predictive tool. This approach is simple to implement yet produces results that compare favorably with the neural network predictions. A preferred inflation-forecasting model is achieved using the networks that employ a Divisia M2 measure of money, adjusted to incorporate a learning mechanism to allow individuals to alter their perceptions of the increased productivity of money. The role of monetary aggregates in the major economies today has largely been relegated to one of a leading indicator of economic activity, along with a range of other macroeconomic variables. The empirical work on Divisia money and, in particular, close monitoring of Divisia constructs that have been adjusted to accommodate financial innovation, may serve to restore confidence in former well established money-inflation links. Neural networks clearly provide more accurate forecasts over larger sample sizes as they can learn the data perfectly. Evolutionary strategies are found to compete very favorably with neural networks over smaller sample sizes and can also learn the data perfectly in extreme cases.

Computing in Economics and …, 2003

Dr. Jane M. Binner and Dr. Alicia Gazely Nottingham Trent University United Kingdom jane.binner@n... more Dr. Jane M. Binner and Dr. Alicia Gazely Nottingham Trent University United Kingdom jane.binner@ntu.ac.uk alicia.gazely@ntu.ac ... economists even believe that this relationship is so strong that the indicator based inflation targeting practiced by many central banks today should ...

The European Journal of Finance, 2002

In this p ap er a weigh ted index measure of money using t he 'Divisia' formulation is constr uct... more In this p ap er a weigh ted index measure of money using t he 'Divisia' formulation is constr ucted for th e Taiwan economy and its in ation forecasting p ot ential is comp ared with th at of it s traditional simple sum counter p ar t. This research ext end s an earlier stud y by Gazely and Binne r b y examinin g th e th eor y th at r apid nancial innova tion, p ar ticularly d uring th e nancial liberalization of th e 1980s , has been resp onsible for th e p oor performance of conve ntional simple sum monetar y aggr egates. Th e Divisia ind ex is ad justed in two ways to allow for the major nancial innovations th at Taiwan has exp erienced sinc e the 1970s. The tech niq ue of neural networks is used t o allow a completely exib le map ping of t he variables and a great er variety of functional form th an is cur rently achieva ble using conventional econometric t ech niq ue s. Results sugge st th at superior tr ackin g of in ation is p ossible for networks th at employ a Divisia M2 measure of money th at has b een ad justed t o incor por ate a learning mechanis m to allow individuals to gr ad ually alter t heir p ercep tions of th e inc reased p rod uctivity of money. Divisia measures of money ap pear to offer adva ntages ove r their simple sum count erp ar ts as macroeconomic indicators.

Central banks continue to publish simple sum measures of the money stock and draw policy inferenc... more Central banks continue to publish simple sum measures of the money stock and draw policy inferences from their behaviour even though it has been demonstrated conclusively that such data violate basic principles of economic and index number theory. (See, for example, Belongia (1996)). Simple sum aggregates are flawed index numbers because aggregating any set of commodities with equal weights implies that each good is a perfect substitute for every other good in the group. If measured money is to internalise pure substitution effects and track the true utility function associated with the monetary services gained from holding a given set of financial assets, monetary aggregates need to be constructed using an index formula from the class of superlative index numbers discussed by Diewert (1976). Barnett (1980) has demonstrated that the monetary index formulation devised by Divisia (1925) is the correct index to encapsulate a monetary aggregate and its concomitant price index. The const...

Global Business and Economics Review, 2009

In this paper we compare the performance of three indices in an inflation forecasting experiment.... more In this paper we compare the performance of three indices in an inflation forecasting experiment. The evidence not only suggests that an evolved neural network is superior to traditionally trained networks in the majority of cases, but also that a risky money index performs at least as well as the Bank of England Divisia index when combined with interest rate information. Notably, the provision of long-term interest rates improves the out-of-sample forecasting performance of the Bank of England Divisia index in all cases examined.

Contemporary Economic Policy, 2004

In this article a Divisia monetary index is constructed for the Taiwan economy, and its inflation... more In this article a Divisia monetary index is constructed for the Taiwan economy, and its inflation forecasting potential is compared with that of its traditional simple sum counterpart. The Divisia index is adjusted in two ways to allow for the financial liberalization that Taiwan has experienced since the 1970s. The powerful artificial intelligence technique of neural networks is used and is found to beat the conventional econometric techniques in a simple inflation forecasting experiment. The preferred inflation forecasting model is achieved using networks that employ a Divisia M2 measure of money that has been adjusted to incorporate a learning mechanism to allow individuals to gradually alter their perceptions of the increased productivity of money. The explanatory power of the two innovation-adjusted Divisia aggregates dominates that of the simple sum counterpart in the majority of cases. (JEL C4, E4, E5) ABBREVIATIONS GDP: Gross Domestic Product VECM: Vector Error-Correction Model *The authors gratefully acknowledge the help of Professor Jim Ford at the University of Birmingham for making the innovation adjusted Divisia data available and three anonymous referees for their valuable comments during the course of this research.

Applied Economics, 2005

Linear models reach their limitations in applications with nonlinearities in the data. In this pa... more Linear models reach their limitations in applications with nonlinearities in the data. In this paper we provide new empirical evidence on the relative Euro inflation forecasting performance of linear and nonlinear models. The well established and widely used univariate ARIMA and multivariate VAR models are used as linear forecasting models whereas NN are used as nonlinear forecasting models. We endeavour to keep the level of subjectivity in the NN building process to a minimum in an attempt to exploit the full potentials of the NN. We also investigate whether the historically poor performance of the theoretically superior measure of the monetary services flow, Divisia, relative to the traditional Simple Sum measure could be attributed to a certain extent to the evaluation of these indices within a linear framework. Results obtained suggest that nonlinear models provide better within-sample and out-of-sample forecasts and linear models are simply a subset of them. The Divisia index also outperforms the Simple Sum index when evaluated in a nonlinear framework.

Applied Economics, 2000

If monetary policy is to be effective in controlling the macroeconomy, accurate measurement of th... more If monetary policy is to be effective in controlling the macroeconomy, accurate measurement of the money supply is essential. The conventional way of measuring the level of the money supply is to simply sum the constituent liquid liabilities of banks. However, a more sophisticated, weighted monetary index has been proposed to take account of the varying degrees of liquidity of the short-term instruments included in money. Inferences about the effects of money on economic activity may depend importantly on the choice of monetary index because simple sum aggregates cannot internalize pure substitution effects. This hypothesis is investigated in the current paper. A Divisia index measure of money is constructed for the USA, UK and Italian economies and its inflation forecasting potential is compared with that of its simple sum counterpart in each of the three countries. The powerful Artificial Intelligence technique of neural networks is used to allow a completely flexible mapping of the variables and a greater variety of functional form than is currently achievable using conventional econometric techniques. The application of neural network methodology to examine the money-inflation link is highly experimental in nature and, hence, the overriding feature of this research is one of simplicity. Superior inflation forecasting models are achieved when a Divisia M2 measure of money is used in the majority of cases. This support for Divisia is entirely consistent with findings based on standard econometric techniques reported from the respective central and Federal Reserve banks of each country. Divisia monetary aggregates appear to offer advantages over their simple sum counterparts as macroeconomic indicators. Further, the combination of Divisia measures of money with the artificial neural network offers a promising starting point for improved models of inflation.

Abstract-The performance of a 'capital certain'Divisia index constructed using the same... more Abstract-The performance of a 'capital certain'Divisia index constructed using the same components included in the Bank of England's MSI plus national savings; a 'risky'Divisia index constructed by adding bonds, shares and unit trusts to the list of assets included in the first index; and a capital certain simple sum index for comparison is compared. The evidence suggests that co-evolutionary strategies are superior to neural networks in the majority of cases. The risky money index performs at least as well as the Bank of England Divisia ...

Global Business and Economics Review, 2008

In this paper, the exchange rate forecasting performance of neural network models are evaluated a... more In this paper, the exchange rate forecasting performance of neural network models are evaluated against the random walk, autoregressive moving average and generalised autoregressive conditional heteroskedasticity models. There are no guidelines available that can be used to choose the parameters of neural network models and therefore, the parameters are chosen according to what the researcher considers to be the best. Such an approach, however, implies that the risk of making bad decisions is extremely high, which could explain why in many studies, neural network models do not consistently perform better than their time series counterparts. In this paper, through extensive experimentation, the level of subjectivity in building neural network models is considerably reduced and therefore giving them a better chance of Forecasting exchange rates with linear and nonlinear models 415 performing well. The results show that in general, neural network models perform better than the traditionally used time series models in forecasting exchange rates.