Forecasting when model stability is uncertain. Forecasts of macroeconomic and financial variables play a crucial role in forward planning undertaken by government and financial institutions, but the predictability of these series is often context and time specific, making standard forecasting techniques unreliable. This project aims to develop new modelling and forecasting techniques that can adapt to structural changes in the model soon after they occur. It aims to derive relevant econometric t ....Forecasting when model stability is uncertain. Forecasts of macroeconomic and financial variables play a crucial role in forward planning undertaken by government and financial institutions, but the predictability of these series is often context and time specific, making standard forecasting techniques unreliable. This project aims to develop new modelling and forecasting techniques that can adapt to structural changes in the model soon after they occur. It aims to derive relevant econometric theory, use simulations to study the properties of the proposed techniques, as well as apply these new techniques to observed data.Read moreRead less
Inference in partially non-stationary time series models. Economic theories typically specify the long-run relationship between economic variables. However, researchers usually examine the long-run features of the data by fitting a restrictive class of models using criteria that have only proven useful for short-term forecasting. In this project we consider alternative models and modelling strategies that are appropriate for the study of the long-run. We also develop computer intensive (bootstra ....Inference in partially non-stationary time series models. Economic theories typically specify the long-run relationship between economic variables. However, researchers usually examine the long-run features of the data by fitting a restrictive class of models using criteria that have only proven useful for short-term forecasting. In this project we consider alternative models and modelling strategies that are appropriate for the study of the long-run. We also develop computer intensive (bootstrap) methods, which will provide a much-needed improvement over the existing (asymptotic) methods for making inference about the long-run. Our research will lead to more reliable models for long-term planning in business, industry and government.Read moreRead less
Vector ARMA Models and Macroeconomic Modelling: Some New Methodology and Algorithms. Economic variables are strongly related to each other, as well as being strongly related to their recent history. As a result, good dynamic multivariate models are crucial for effective policy making and forecasting in areas of vital national importance such as monetary and fiscal policy, environmental policy and tourism. Our project advances the frontiers of knowledge in multivariate time series modelling. The ....Vector ARMA Models and Macroeconomic Modelling: Some New Methodology and Algorithms. Economic variables are strongly related to each other, as well as being strongly related to their recent history. As a result, good dynamic multivariate models are crucial for effective policy making and forecasting in areas of vital national importance such as monetary and fiscal policy, environmental policy and tourism. Our project advances the frontiers of knowledge in multivariate time series modelling. The outcome of this project will be immediately useful for macroeconomic policy makers such as the Reserve Bank of Australia and the Treasury, and for industry bodies such as Tourism Australia. Read moreRead less
Macroeconomic forecasting in a 'Big Data' world. This project will develop methods for forecasting important macroeconomic variables where a large set of predictors is available. As well as raw variables and composite indices such as principal components. This project will also include various lags and nonlinear functions of potential predictors. The project will adapt Bayesian statistical methods for selecting these predictors so that they can be applied to time series data, thus developing inn ....Macroeconomic forecasting in a 'Big Data' world. This project will develop methods for forecasting important macroeconomic variables where a large set of predictors is available. As well as raw variables and composite indices such as principal components. This project will also include various lags and nonlinear functions of potential predictors. The project will adapt Bayesian statistical methods for selecting these predictors so that they can be applied to time series data, thus developing innovative forecasting methods that can be used on a range of important problems involving 'Big Data'. The project will compare forecasts from different methods using simulated and empirical data from the US and Australia. For the latter an outcome will be an online handbook of available Australian economic data for public use.Read moreRead less
Non-parametric estimation of forecast distributions in non-Gaussian state space models. The production of accurate forecasts is arguably one of the most challenging tasks in economics, business and finance, where data often assume strictly positive, integer or binary values, or are characterized by many extreme values far from the average. This project will produce new, state-of-the-art statistical methods for generating accurate estimates of the probabilities attached to different possible futu ....Non-parametric estimation of forecast distributions in non-Gaussian state space models. The production of accurate forecasts is arguably one of the most challenging tasks in economics, business and finance, where data often assume strictly positive, integer or binary values, or are characterized by many extreme values far from the average. This project will produce new, state-of-the-art statistical methods for generating accurate estimates of the probabilities attached to different possible future values of such variables. Although far-ranging in scope, the techniques advocated will have particular impact in the financial sphere, where the concept of future risk is inextricably linked to the probability of occurrence of extreme values and, hence, to the future probability distribution of the financial variable. Read moreRead less
New Approaches to the Analysis of Count Time Series. The focus of this proposal is on the analysis of data that enumerate events over time. Occurrences of such count data abound in economics and business, examples being observations on insurance claims, loan defaults and individual product demand. This project develops a suite of innovative methods for modelling and predicting event counts. The methods explicitly accommodate both the discreteness of the data and possible complexities in its evo ....New Approaches to the Analysis of Count Time Series. The focus of this proposal is on the analysis of data that enumerate events over time. Occurrences of such count data abound in economics and business, examples being observations on insurance claims, loan defaults and individual product demand. This project develops a suite of innovative methods for modelling and predicting event counts. The methods explicitly accommodate both the discreteness of the data and possible complexities in its evolution over time. In so doing, they enable both accurate inferences regarding the dynamic structure of the data to be drawn and accurate forecasts of future event counts to be produced.Read moreRead less
Fractional Integration, Power Laws and Econometric Models: Some Methodological and Theoretical Developments. The fundamental objectives of this project are to: (i) Extend
current econometric practice and consider the use of power laws as
a basis for the construction of a more flexible and realistic
class of models for the analysis of economic and financial time
series. (ii) To develop inferential techniques appropriate for the
modelling of dynamic econometric systems that incorporate
struc ....Fractional Integration, Power Laws and Econometric Models: Some Methodological and Theoretical Developments. The fundamental objectives of this project are to: (i) Extend
current econometric practice and consider the use of power laws as
a basis for the construction of a more flexible and realistic
class of models for the analysis of economic and financial time
series. (ii) To develop inferential techniques appropriate for the
modelling of dynamic econometric systems that incorporate
structure characterized by power laws. This will be achieved by
building upon the class of fractionally integrated processes. New
econometric models and methodologies for the analysis of
non-stationarity series will be developed, along with the
associated theoretical results.Read moreRead less
Semi-parametric bootstrap-based inference in long-memory models. Given the long lead times involved in implementing economic decisions, a clear understanding of the long-term dynamics driving key variables is crucial. This project will produce significant advances in the analysis of long-range dependence, with decisions underpinned by more accurate and robust statistical information as a consequence.
Approximate Bayesian computation in state space models. Economic and financial data frequently exhibit dynamic patterns, driven by unobserved processes that relate to the behaviour of economic agents, or to institutional and technological change. To gain insight into such 'latent' processes is of paramount importance in terms of both understanding the economy and producing accurate, readily up-dated, forecasts of its future performance. Using a Bayesian approach, new simulation-based statistical ....Approximate Bayesian computation in state space models. Economic and financial data frequently exhibit dynamic patterns, driven by unobserved processes that relate to the behaviour of economic agents, or to institutional and technological change. To gain insight into such 'latent' processes is of paramount importance in terms of both understanding the economy and producing accurate, readily up-dated, forecasts of its future performance. Using a Bayesian approach, new simulation-based statistical methods for analysing latent variable models are proposed. Emphasis is given to the development of relatively simple techniques that are applicable to a wide range of empirically relevant models, with a view to improving the access of non-specialists to this powerful form of statistical analysis.Read moreRead less
Advanced Computational and Analytic Studies in Lattice Statistical Mechanics and Applications. Lattice Statistical Mechanics is one of the current success stories of Australian Science with a significant international presence. The applicants represent a centre of excellence, particularly in the area of combining computational and analytic studies for maximum scientific benefit. The programme of research maximises Australia's investment in this human resource by focussing on an integrated set of ....Advanced Computational and Analytic Studies in Lattice Statistical Mechanics and Applications. Lattice Statistical Mechanics is one of the current success stories of Australian Science with a significant international presence. The applicants represent a centre of excellence, particularly in the area of combining computational and analytic studies for maximum scientific benefit. The programme of research maximises Australia's investment in this human resource by focussing on an integrated set of projects comprising a diverse and innovative group of applications in areas such as polymer science, DNA denaturation, combinatorics and the study of traffic flows. The underlying theme is always the utility of lattice statistical mechanics in 21st century science.Read moreRead less