I completed my PhD in Economics with specialisation in Econometrics at Boston College in 2003. Prior to joining UNSW, I held several teaching and research appointments at Boston College and University of Montreal. I have also been a visiting academic at the Risk Management Institute in Singapore, Department of Economics in San Diego USA, Department of Economics at University College Dublin, Ireland, and European Centre for Applied Research in Economics and Statistics (ECARES) in Brussels.
My main research expertise is on modelling, estimating and making correct statistical analysis in big systems. Big-data presents technical challenges (hence curse of dimensionality) to the existing statistical tools that are used in economics and finance. My work falls in the new line of research that aims at turning this curse into a blessing. I have been published in top tier field journals. My contribution to the theoretical developments in econometrics has direct empirical implications. For example, a hedge fund manager wants to know what are the drivers of the market risk. Once these risk factors are identified, the hedge fund manager can build a portfolio to diversify away this risk. In big-data world, there are potentially hundreds/thousands of sources of systematic risk. The statistical tool I develop use this large volumes of information and selects the main key risk drivers without the fund manager having to make an uninformed ad-hoc guess. The technique is accurate and reliable and can be applied in many scenarios in Finance and Economics. Currently, I am working on identifying key drivers of growth in the real estate market in Australian Capital cities. For the Sydney area, the work is building a big data spatial econometric model to uncover what drives the high premium some suburbs earn their homeowners.
In another. area interest, I have published in A* journal in Finance where I establish new evidence of dependence of risk aversion on the Business cycle. Aggregate risk aversion it seems does vary with periods of economic booms and busts. Consumers perception of risk and wiliness to engage in risky ventures is conditional of the health of the economy.
Statistical inference sometimes has to be performed in small samples and asymptotic tools are no longer reliable. I have expertise in using simulation methods like the Bootstrap to study the statistical properties of key estimators like the Generalized Method of Moments estimator in models of practical importance in consumer behaviours. These models include the rational expectation model of the consumption asset pricing model.
- Econometric Theory: Statistical Inference in High-Dimensional Factor Models, Bootstrap Methods, GMM Estimation, Bayesian Econometrics
- Applied Econometrics: Forecasting with large number of predictors
- Financial Econometrics: Arbitrage pricing theory model, Consumption CAPM