Find us on Facebook Find us on LinkedIn Follow us on Twitter Subscribe to our YouTube channel

2012 Seminars

10th Feb 2012 - 11:00 am

Speaker:

Associate Professor Jean Imbs,

Affiliation:

Paris School of Economics

Venue:

Room 498, Merewether Building

Title:

Economic Integration and Structural Change

Abstract:

We show the dynamics of sectoral production - structural change - come with systematic changes in the geographic dispersion of activity. In developing countries, sectoral diversification is accompanied by geographic agglomeration, and regions become heterogeneous. In advanced economies, sectoral specialization is accompanied by geographic dispersion, and regions become homogeneous. We argue that developing countries diversify because their constituent regions integrate with each other, and can specialize regionally as a result. Advanced economies specialize because they integrate internationally and their constituent regions all produce according to the global pattern of comparative advantage. We .nd systematic support for these claims in international data on sectoral production at the regional level, including in the US, China and India, but no such evidence once the samples focus on non-traded sectors or relatively closed regions. Economic areas formed by specialized, regionally homogeneous countries tend to diversify and agglomerate, as if their constituent countries were integrating.

24th Feb 2012 - 11:00 am

Speaker:

Nikky Kortbeek,

Affiliation:

University of Twente

Venue:

Room 498, Merewether Building

Title:

Balancing appointments and walk-ins in healthcare

Abstract:

Outpatient clinics and diagnostic testing facilities traditionally provide patients with individual appointments to balance workload. Disadvantages however, include patients needing to revisit the hospital, an involved planning process and potentially long access times. In this talk I discuss a study in which we explore the viability of walk-in based policies. We introduce a stochastic method that finds the mixed strategy that balances the benefits and drawbacks of pure appointment and walk-in policies.

9th Mar 2012 - 11:00 am

Speaker:

Dr Valentyn Panchenko,

Affiliation:

School of Economics, UNSW

Venue:

Room 498, Merewether Building (H04)

Title:

Accuracy of Copula-Based Multivariate Density Forecasts in Selected Regions of Support

Abstract:

This paper develops a testing framework for comparing the predictive accuracy of copula-based multivariate density forecasts, focusing on a specific part of the joint distribution. The test is framed in the context of the Kullback-Leibler Information Criterion, and using (out-of-sample) conditional likelihood and censored likelihood in order to restrict the evaluation to the region of interest. Monte Carlo simulations show that the resulting test statistics have satisfactory size and power properties in small samples. In an empirical application to daily exchange rate returns we find evidence that the dependence structure varies with the sign and magnitude of returns, such that different parametric copula models achieve superior forecasting performance in different regions of the copula support. Our analysis highlights the importance of allowing for lower and upper tail dependence for accurate forecasting of common extreme appreciation and depreciation of different currencies.

16th Mar 2012 - 11:00 am

Speaker:

Dr Andrey Vasnev,

Affiliation:

University of Sydney Business School

Venue:

Room 498, Merewether Building (H04)

Title:

Practical Use of Sensitivity in Econometrics with an Illustration for Forecast Combinations

Abstract:

Jan R. Magnus and Andrey L. Vasnev

We consider practical use of the sensitivity measure studied by Magnus and Vasnev (2007). For this purpose we distinguish between absolute and relative sensitivity
and highlight the context dependent nature of the sensitivity analysis. Relative sensitivity is then applied in the context of forecast combination and sensitivity based
weights are introduced. All concepts are illustrated with the help of the European
yield curve example. In this context it is natural to look at sensitivity to autocorrelation
and normality assumptions. Different forecasting models are combined with equal, fit based and sensitivity based weights and compared against the multivariate and random walk benchmarks. We show that the fit based weights and the sensitivity based weights are complimentary.

30th Mar 2012 - 11:00 am

Speaker:

Dr Marco Reale,

Affiliation:

University of Canterbury

Venue:

Room 498, Merewether Building (H04)

Title:

Graphical modelling representation of multivariate time series

Abstract:

Graphical modelling is very effective for identification and estimation of sparse structural multivariate time series model. After an introduction to graphical modelling and some of its useful properties there will be a discussion of the issues pertaining the identification of sparse structural vector autoregressions and how graphical modelling can assist. Extensions will be presented to I(1) time series and structural vector autoregression moving average models.

20th Apr 2012 - 11:00 am

Speaker:

Professor Gianni Amisano,

Affiliation:

ECB Research

Venue:

Room 498, Merewether Building H04

Title:

Prediction with Macroeconomic Models

Abstract:

There are several relevant layers of uncertainty that characterise econometric models routinely used for policy making. First and foremost, there is intrinsic uncertainty about the future conditional on a model and parameters. Then there is extrinsic uncertainty about model parameters conditional on a model. Then there is uncertainty about models conditional on a set of models. In addition there is unconditional uncertainty, when all models considered are false.

In this paper we incorporate all four levels of uncertainty and we assess the improvements in the quality of prediction that we get by doing so. We provide a practical example based on the joint combination of a DSGE model, a Bayesian VAR and a dynamic factor model for a set of US macroeconomic time series.

In our paper we find that:

  1. Taking into consideration parameter uncertainty is most relevant in periods of unusual data
  2. A pool with equal weights provides predictions of superior quality with respect to prediction obtained with individual models
  3. We introduce a measure of value of each of the models being combined that can be decomposed across sub-periods and this measure provides important indication regarding the usefulness of the individual models.

27th Apr 2012 - 11:00 am

Speaker:

Professor Eddie Anderson,

Affiliation:

University of Sydney

Venue:

Room 498, Merewether Building (H04)

Title:

Ranking games and gambling: When to quit when you're ahead

Abstract:

It is common for rewards to be given on the basis of a rank ordering, so that relative performance amongst a cohort is the criterion. In this paper we formulate an equilibrium model in which an agent makes successive decisions on whether or not to gamble and is rewarded on the basis of a rank ordering of the final position amongst competing players. One application of this model is to the behavior of mutual fund managers who are paid depending on funds under management, which in turn are greatly influenced by annual or quarterly rank orderings. We model a situation in which fund managers can elect either to pick stocks or to use a market tracking strategy. In equilibrium the distribution of the final position will have a negative skew. We explore how this distribution depends on the number of players, the probability of success when gambling, the structure of the rewards, and on information regarding the performance of other players.

18th May 2012 - 11:00 am

Speaker:

Dr Jing Zhao,

Affiliation:

La Trobe University

Venue:

Room 498, Merewether Building

Title:

A Hidden Markov Process Approach to Information-Based Trading

Abstract:

This paper proposes a novel approach to information-based trading, incorporating the dynamics and serial correlation of trading activities.  Unlike the existing approaches of sequential trading modeling, it updates the prior belief of information states using newly observed order flows and identifies trading motives in a data-driven manner.  It allows the set of information states to vary across time and companies.  Extensive simulation demonstrates that the proposed approach can generate dynamic daily measures of information-based trading in high accuracy.  Based on a sample of 30 NYSE stocks, we provide evidence of the significant explanatory power of information-based trading on return volatility.

15th Jun 2012 - 11:00 am

Speaker:

Professor Daewon Sun,

Affiliation:

University of Notre Dame, Mendoza College of Business

Venue:

Room 498, Merewether Building

Title:

Competition and Coordination in Online Marketplaces

Abstract:

Online marketplaces, such as those operated by Amazon, have seen rapid growth in recent years. These marketplaces serve as an intermediary, matching buyers with sellers, while control of the good is left to the seller. In some cases, e.g., the Amazon marketplace system, the firm that owns and manages the marketplace system will also sell competing products through the marketplace system. This creates a new form of channel conflict, which is a focus of this paper. We consider a setting in which a marketplace firm operates an online marketplace through which retailers can sell their products directly to consumers. We consider a single retailer, who currently sells its product only through its own website, but who may choose to contract with Amazon to sell its product through the marketplace system. Selling the product through the marketplace expands the available market for the retailer, but comes at some expense, e.g., a fixed participation fee or a revenue sharing requirement. Thus, a key question for the retailer is whether she should choose to sell through the marketplace system, and if so, at what price. We analyze the optimal decisions for both the retailer and the marketplace firm and characterize the system equilibrium.

18th Jun 2012 - 11:00 am

Speaker:

Professor Michael H.Y. Wong,

Affiliation:

The Chinese University of Hong Kong

Venue:

Room 498, Merewether Building

Title:

Application and Implication of Cointegration in Asset Pricing

Abstract:

Cointegration is a useful econometric tool for identifying assets which share a common equilibrium. The importance of cointegration has become recognized and resulted in a Nobel Prize in Economics for Granger in 2003. In this talk, I will report several recent advances of asset pricing theories based on continuous-time cointegration dynamics. It covers cointegrated pairs-trading using classical mean-variance portfolio theory, cointegration option pricing with stochastic correlations using Fourier analysis, and (if time allows) the hedging with mortality risk in insurance products. Our theories predict that 1. if cointegrated assets are liquidly traded, then there exists a statistical arbitrage opportunity; 2. If the assets are not traded or not liquidly traded, their corresponding derivatives securities, in particular futures contracts, exhibit stochastic convenient yields which are partially driven by cointegrating factors; and 3. As human mortality is not traded by its nature and the national mortality rate is cointegrated with the mortality rate of an individual insurance company¿¿¿s client pool, cointegration techniques enhance the hedging of mortality risk with national mortality bonds. Empirical studies are performed to validate the use of the developed theories and numerical methods. (The talk is based several joint papers with M.C. Chiu, T.W. Wong and J. Zhao)

19th Jun 2012 - 11:00 am

Speaker:

Professor Simon Jackman,

Affiliation:

Stanford University

Venue:

Room 498, Merewether Building

Title:

How Does Obama Match Up?

Abstract:

Abstract: Barack Obama won a convincing victory in the 2008 U.S. presidential election.   But would the outcome of the election had been different had Obama not been the nominee?   In this paper we analyze a unique data set in which over 14,000 survey respondents were presented with numerous head-to-head match-ups between Democratic and Republican candidates (both real and hypothetical), at various time points over the 2008 election campaign.  We use these data to estimate Obama's exceptionalism as a candidate, via hierarchical logistic regression models, with each hypothetical matchup generating a distinct set of parameters for predictors of vote choice.    We find that "old-fashioned" racial stereotyping is uniquely important in shaping decisions about Obama in 2008, relative to its role in past elections or in 2008 choices substituting Hillary Clinton or John Edwards for Obama.  A similar pattern emerges for other measures of racial prejudice.   Inspection of the posterior predictive densities from the model indicate that the Democrats would have won the 2008 election regardless of who they nominated; but the average Democratic party nominee from the last 16 years, and either of Edwards or Hillary Clinton, would have done better against McCain than Obama, although in Clinton's case, not by much.   We conclude with what these results portend for the 2012 election, given that the electoral context is less favorable for Democrats -- and for Obama in particular -- than in 2008.

3rd Aug 2012 - 11:00 am

Speaker:

Nuttanan (Nate) Wichitaksorn,

Affiliation:

University of Sydney Business School

Venue:

Room 489 Merewether Building (H04)

Title:

An Alternative Class of Skew Distributions and Parametric Quantile Regression Models?

Abstract:

This paper proposes a method to construct a univariate skew distribution
through the mixture of two scaled normal distributions. As a result, we obtain
an alternative class of skew distributions where the skewness parameter value is defined in the ]0,1[ interval and this allows us to have an application on parametric quantile regression. By expressing a skew distribution as a scale mixture of normal, it can facilitate a flexible parameter estimation procedure via the Bayesian Markov Chain Monte Carlo methods. In addition, the proposed distribution has a closed-form probability density function and we can perform statistical inference via likelihood-based approaches such as maximum likelihood. The performance of the proposed distributions is demonstrated in two simulation studies on (i) regression models with skewed error distribution and (ii) parametric quantile regression models. In empirical studies, we analyse the U.S. market return data for skew error regression and the U.S. infant birthweight data for parametric quantile regression.

8th Aug 2012 - 02:00 pm

Speaker:

Associate Professor Artem Prokhorov,

Affiliation:

Concordia University, Montreal

Venue:

Room 489 Merewether Building (H04)

Title:

Efficient Estimation of Parameters in Marginals in Semiparametric Multivariate Models

Abstract:

Recent literature on semiparametric copula models focused on the situation when the marginals are speci
ed nonparametrically and the copula function is given a parametric form. For example, this setup is used in Chen, Fan and Tsyrennikov (2006) [Efficient Estimation of Semiparametric Multivariate Copula Models, JASA] who focus on
efficient estimation of copula parameters. We consider a reverse situation when the marginals are speci
ed parametrically and the copula function is modelled nonparametrically. This setting is no less relevant in applications. We use the method of sieve for efficient estimation of parameters in marginals, derive its asymptotic distribution and show that the estimator is semiparametrically
efficient. Simulations suggest that the sieve MLE can be up to 70% more efficient relative to QMLE depending on the strength of dependence between the marginals. An application using insurance company loss and expense data demonstrates empirical relevance of this setting.

9th Aug 2012 - 01:00 pm

Speaker:

Dr Xinyu Zhang,

Affiliation:

Academy of Mathematics and Systems Science, Chinese Academy of Sciences

Venue:

Room 489 Merewether Building (H04)

Title:

Adaptively Combined Forecasting for Discrete Responses

Abstract:

Adaptive combining is generally desirable for forecasts. In this paper, we propose an adaptively combined forecasting method for discrete response time series data. We demonstrate in theory that the proposed forecast is of the desired adaptation with respect to the widely used squared risk and other significant risk functions under mild conditions. Furthermore, we study the adaptation of the proposed forecasting method with a model screening step that is often useful in applications. Our simulation study evidently illustrates the superiority of the proposed approach, with two real-world data examples further demonstrated.

10th Aug 2012 - 10:00 am

Speaker:

Assistant Professor Joaquim Ramalho,

Affiliation:

Universidade de Evora, Portugal

Venue:

Room 489 Merewether Building (H04)

Title:

Hedonic functions, hedonic methods, estimation methods and Dutot and Jevons house price indexes: are there any links?

Abstract:

Hedonic methods are a prominent approach in the construction of house price indexes. This paper investigates in a comprehensive way whether or not there exists any kind of link between the type of price index to be computed (Dutot or Jevons) and the form of hedonic functions, hedonic methods and estimation methods, with a link being defined as a specific combination of price indexes, functions and methods that simplifies substantially the calculations required to compute hedonic price indexes. It is found that: (i) there is a link between Dutot indexes, exponential hedonic functions and the Poisson pseudo maximum likelihood estimator, on the one hand, and Jevons indexes, log-linear hedonic functions and ordinary least squares, on the other hand; and (ii) unlike implicitly assumed in the hedonic literature, there is no link between Jevons indexes and the time dummy variable method, since in this context quality-adjusted Dutot price indexes may also be simply computed as the exponential transformation of a time dummy variable coefficient, provided that an exponential hedonic function is used. A Monte Carlo simulation study illustrates both the convenience of the links identified and the biases that result from overlooking them or implementing bias corrections based on invalid assumptions.

7th Sep 2012 - 11:00 am

Speaker:

Dr SeoJeong (Jay) Lee,

Affiliation:

Australian School of Business, UNSW

Venue:

Room 498 Merewether Building H04

Title:

Asymptotic Refinements of a Misspecification-Robust Bootstrap for GMM Estimators

Abstract:

I propose a nonparametric iid bootstrap that achieves asymptotic refinements for t tests and confidence intervals based on the generalized method of moments (GMM) estimators even when the model is misspecified. In addition, my bootstrap does not require recentering the bootstrap moment function, which has been considered as a critical procedure for bootstrapping GMM. Regardless of whether the assumed model is correctly specified or not, the proposed bootstrap achieves the same sharp magnitude of refinements as the conventional bootstrap methods which establish asymptotic refinements by recentering in the absence of misspecification. The key procedure is to use a misspecification-robust variance estimator for GMM of Hall and Inoue (2003, Journal of Econometrics 114, 361-394) in constructing the t statistic. Examples of overidentified and possibly misspecified moment condition models with Monte Carlo simulation results are provided: (i) Combining data sets, and (ii) invalid instrumental variables.

14th Sep 2012 - 11:00 am

Speaker:

Dr David Ubilava,

Affiliation:

Department of Agricultural and Resource Economics The University of Sydney

Venue:

Room 498 Merewether Building H04

Title:

El Nino Southern Oscillation and Primary Commodity Prices: Causal Inferences from Smooth Transition Models

Abstract:

Global climate anomalies affect world economies and primary commodity prices. One of the more pronounced climate anomalies is El
Ni¿¿o Southern Oscillation (ENSO). In this study I examine the relationship between ENSO and world commodity prices using monthly
time series of the sea-surface temperature anomalies in the Nino 3.4 region, and the real prices of the twenty four food and agricultural
commodities. I apply smooth transition auoregressive (STAR) modelling techniques to assess causal inferences that could potentially be
camouflaged in the linear setting. I consider out-of-sample causality based one-step-ahead forecasts applied to expanding windows. I
illustrate dynamics of ENSO and commodity price behavior using generalized impulse-response functions.

21st Sep 2012 - 11:00 am

Speaker:

Professor Martin Savelsbergh,

Affiliation:

University of Newcastle

Venue:

Room 498 Merewether Building H04

Title:

Incremental network design problems

Abstract:

Network infrastructures are a common phenomenon. Network upgrades and expansions typically occur over time due to budget constraints.
We introduce a class of incremental network design problems that allow investigation of many of the key issues related to the choice and timing of infrastructure expansions and their impact on the costs of the activities performed on that infrastructure. We focus on two variants: incremental network design with shortest paths and incremental network design with maximum flows. We consider the complexity of the problem, we analyze the performance of natural heuristics, we derive approximation algorithms, and we study integer program formulations.

28th Sep 2012 - 11:00 am

Speaker:

Vinit Mishra,

Affiliation:

Discipline of Business Analytics, University of Sydney Business School

Venue:

Room 498 Merewether Building H04

Title:

The Cross Moment Model of Choice

Abstract:

Several generalizations of Chebyshev-type (1874) inequalities were published in 1950s and 1960s that proposed tight upper or lower bounds on the expectation of functions of random variables given moments information. When mean-covariance information is given, a tight upper bound on the expectation of highest order statistic can be found using a semidefinite program. Empowered with this result and an extreme point argument, I will propose a method of finding choice probabilities in discrete choice.  This new method, known as Cross Moment Model (CMM), generates choice probabilities using a convex semidefinite program and avoids the evaluation of multidimensional integrals as is typically done in choice models such as multinomial probit. Several simple examples will illustrate power of the proposed approach and beauty of this deterministic way of solving problems with random parameters. 

19th Oct 2012 - 11:00 am

Speaker:

Dr Erick Li,

Affiliation:

Discipline of Business Analytics The University of Sydney

Venue:

Room 498 Merewether Building H04

Title:

What Can Headquarters Do If Divisions Prefer Capital-Intensive Projects?

Abstract:

We analyse the friction between the headquarters and the divisional manager in the project selection process. The divisional manager who prefers capital-intensive projects manipulates the selection process by presenting the project that maximizes the managerial utility rather than the added value. Because the headquarters cannot observe all available projects, an ex ante standard for project approval must be specified. We derive the optimal standard for the headquarters. The optimal standard can be described by a threshold function that has a complex form. We also study several heuristics that are commonly used in practice. Our analysis indicates that the heuristic that imposes a hurdle rate on the project return rate gives the headquarters almost the same payoff as the optimal standard does. However, the headquarters¿¿¿ payoff reduces substantially if a threshold is put on the required capital or the expected return. Our findings underscore the importance of using appropriate criterion in project selection process to mitigate the organizational challenge caused by misaligned objectives and information asymmetry. Although NPV is the best statistics to measure added values, it is not the best instrument to increase the added values.

16th Nov 2012 - 11:00 am

Speaker:

Dr Boris Choy,

Affiliation:

Discipline of Business Analytics, University of Sydney Business School

Venue:

Room 498 Merewether Building H04

Title:

Some New Developments of the Student-t Distribution

Abstract:

This talk presents some well-known and new properties of the Student-t distribution. Some of the properties improve the model ability of the distribution and some make the model implementation of the Student-t distribution as easy as the normal distribution using Bayesian computational methods. Actuarial and financial applications will be given.

23rd Nov 2012 - 11:00 am

Speaker:

Associate Professor Tommaso Proietti,

Affiliation:

Discipline of Business Analytics, University of Sydney Business School

Venue:

Room 498 Merewether Building H04

Title:

The Generalised Autocovariance Function

Abstract:

The generalised autocovariance function is defined for a stationary stochastic process as the inverse Fourier transform of the power transformation of the spectral density function. Depending on the value of the transformation parameter, this function nests the inverse and the traditional autocovariance functions. A frequency domain non-parametric estimator based on the power transformation of the pooled periodogram is considered and its asymptotic distribution is derived. The results are employed to construct classes of tests of the white noise hypothesis, for clustering and discrimination of stochastic processes and to introduce a novel feature matching estimator of the spectrum.

Keywords: Stationary Gaussian processes. Non-parametric spectral estimation. White noise tests. Feature matching. Discriminant Analysis.

10th Dec 2012 - 11:00 am

Speaker:

Professor Yeh Lam,

Affiliation:

Department of Statistics and Actuarial Science, University of Hong Kong

Venue:

Room 498 Merewether Building H04

Title:

Geometric Process Maintenance Model

Abstract:

In this talk, as an important application of geometric process (GP) which was introduced by Lam (1988), we shall study the geometric process maintenance model (GPMM).  According to different values of parameters in the GP, a deteriorating GPMM and an improving GPMM are studied respectively. A policy N is applied by which the system will be replaced following the Nth failure. Then an optimal policy N* is determined analytically and the monotone properties of N* are then studied. Hence, the GPMM is a more reasonable model and a simpler model than the classical maintenance models such as the perfect repair model and the minimal repair model. Some generalization of the GPMM is also introduced.