|Full time||1 year||September 2017|
|Part time||2 years||September 2017|
This course is ideal if you would like to pursue a career in which the application of econometric methods plays a major role – such as market research, evidence-based planning and policy advice, or financial modelling and forecasting – or if you would like to enter a PhD programme. It focuses on the application potential of a broad range of econometric techniques and the effective communication of empirical results, rather than the statistical theory upon which econometric methods depend.
Find out more about this course:
You will study core econometric concepts and methods alongside modules that focus on different application contexts, such as time-series analysis, microeconometrics, financial econometrics, macroeconometric modelling and business forecasting. You will also develop your computing and communication skills. You will then write a dissertation to show evidence of your knowledge, skills and abilities to undertake complex, self-managed tasks with tight deadlines.
To include: essays, small research reports, practical exercises, formal examinations, and final dissertation.
Please note that this is an indicative list of modules and is not intended as a definitive list.
This module introduces and develops core econometric topics. It provides fundamental knowledge on statistics, multivariate linear regression and other selected advanced econometrics topics with applications to cross-sectional data. On successful completion of the module students will be able to demonstrate econometric skills for assessment of economic theories with matching data, which will be helpful for various career opportunities and further doctoral study of economics.
Your dissertation is produced in consultation with a supervisor on a topic chosen by yourself and written over a 12-week period.
If you are working, you can use your own professional circumstances as research material, in consultation with your employers. Graduates often use the dissertation to demonstrate their professional development to future employers.
This module considers the techniques and methods used to construct forecasting models for national economies. Students learn about appropriate econometric methods and practice applying these and interpreting and reporting their results.
The module focuses primarily on the econometric techniques relevant to dynamic models estimated on macroeconomic time-series data and the production of quantitative forecasts from estimated models. We look at both the Cowles tradition of simultaneous structural equation models and also the more recent approach of Vector Autoregression (VAR) models, including their SVAR and VECM variants.
The module also pays attention to associated topics, for example one or more of: the elements of macroeconomic theory that underpin macroeconometric models; solution methods for non-linear models; stochastic simulation; evaluating forecast accuracy; integrating model-based forecasts with expert judgement; the difficulties posed by the quality; and publication schedules of official data; other approaches to macroeconometric modelling and forecasting.
This module considers time series methods and models used primarily to produce forecasts of business, economic and financial data. Students learn about fundamental time series methods and how to apply them using dedicated econometric software. They also learn how to interpret and report results from the application of these methods.
The module focuses primarily on the main fundamental univariate time-series techniques relevant to time-series data commonly encountered in business, economic and finance environments and the production of quantitative forecasts from estimated models. We discuss the following modelling and forecasting methodologies: simple extrapolation/trend models, exponential smoothing, decomposition, ARIMA, ARIMAX and the ARCH/GARCH family of specifications.
The module also considers associated topics such as intervention analysis, applications involving programming and multivariate ARCH/GARCH models. .
This module builds on the topics introduced in EC7021 Econometrics, which is a pre-requisite. Inter alia this course discusses techniques used in microeconometrics and provides an introduction to panel data analysis. It also considers Monte Carlo simulation methods, the autoregressive distributed lag modelling approach and allowing for asymmetries and nonlinearities in models.
In the microeconometric topic we study limited dependent variable models. We analyse discrete choice models, where the outcome takes on two (binary choice) or more outcomes (count data models). We also consider models reflecting the fact that optimizing behaviour of individuals (families or firms) may give rise to a corner solution for some nontrivial fraction of the population (censored regression model). Finally, we consider models to deal with the fact that we may observe a non-random sample from the underlying population (sample selection model).
The autoregressive distributed lag (ARDL) methodology and its relation with cointegration and error-correction modelling is considered including bounds tests for cointegration and extensions to allow for asymmetric behaviour. More general consideration of asymmetric and nonlinear modelling, as relevant to financial and macroeconomic data, is also covered.
Monte Carlo simulation methods to obtain critical values for tests with non standard distributions and evaluate size and power properties will be considered. We will look at the specification of data generation processes and the simulation process to, for example, verify the unbiasedness property of the OLS estimator, obtain critical values for the augmented Dickey-Fuller test and ascertain the size of a test. This will involve the theoretical specification of the problems and the writing of computer programs.
In panel data the same individual (family or firm) is followed across time which allows for a much richer analysis than pure cross-sectional or time series data. The econometric analysis of such data needs to recognize that observations cannot be treated as being independently distributed across time. Unobserved factors (such as ability) that affect a person’s wage in one year will affect that person’s wage in another year as well. Appropriate models that deal with this individual heterogeneity are discussed such as the fixed effects model and the random effects model.
You will have the opportunity to study a foreign language, free of charge, during your time at the University as part of the Kingston Language Scheme. Options currently include: Arabic, French, German, Italian, Japanese, Mandarin, Portuguese, Russian and Spanish.