Skip to main content

Über dieses Buch

This book presents recent research on robustness in econometrics. Robust data processing techniques – i.e., techniques that yield results minimally affected by outliers – and their applications to real-life economic and financial situations are the main focus of this book. The book also discusses applications of more traditional statistical techniques to econometric problems.
Econometrics is a branch of economics that uses mathematical (especially statistical) methods to analyze economic systems, to forecast economic and financial dynamics, and to develop strategies for achieving desirable economic performance. In day-by-day data, we often encounter outliers that do not reflect the long-term economic trends, e.g., unexpected and abrupt fluctuations. As such, it is important to develop robust data processing techniques that can accommodate these fluctuations.



Keynote Addresses


Robust Estimation of Heckman Model

We first review the basic ideas of robust statistics and define the main tools used to formalize the problem and to construct new robust statistical procedures. In particular we focus on the influence function, the Gâteaux derivative of a functional in direction of a point mass, which can be used both to study the local stability properties of a statistical procedure and to construct new robust procedures. In the second part we show how these principles can be used to carry out a robustness analysis in [13] model and how to construct robust versions of Heckman’s two-stage estimator. These are central tools for the statistical analysis of data based on non-random samples from a population.

Elvezio Ronchetti

Fundamental Theory


Sequential Monte Carlo Sampling for State Space Models

The aim of these notes is to revisit sequential Monte Carlo (SMC) sampling. SMC sampling is a powerful simulation tool for solving non-linear and/or non-Gaussian state space models. We illustrate this with several examples.

Mario V. Wüthrich

Robustness as a Criterion for Selecting a Probability Distribution Under Uncertainty

Often, we only have partial knowledge about a probability distribution, and we would like to select a single probability distribution $$\rho (x)$$ρ(x) out of all probability distributions which are consistent with the available knowledge. One way to make this selection is to take into account that usually, the values x of the corresponding quantity are also known only with some accuracy. It is therefore desirable to select a distribution which is the most robust—in the sense the x-inaccuracy leads to the smallest possible inaccuracy in the resulting probabilities. In this paper, we describe the corresponding most robust probability distributions, and we show that the use of resulting probability distributions has an additional advantage: it makes related computations easier and faster.

Songsak Sriboonchitta, Hung T. Nguyen, Vladik Kreinovich, Olga Kosheleva

Why Cannot We Have a Strongly Consistent Family of Skew Normal (and Higher Order) Distributions

In many practical situations, the only information that we have about the probability distribution is its first few moments. Since many statistical techniques requires us to select a single distribution, it is therefore desirable to select, out of all possible distributions with these moments, a single “most representative” one. When we know the first two moments, a natural idea is to select a normal distribution. This selection is strongly consistent in the sense that if a random variable is a sum of several independent ones, then selecting normal distribution for all of the terms in the sum leads to a similar normal distribution for the sum. In situations when we know three moments, there is also a widely used selection—of the so-called skew-normal distribution. However, this selection is not strongly consistent in the above sense. In this paper, we show that this absence of strong consistency is not a fault of a specific selection but a general feature of the problem: for third and higher order moments, no strongly consistent selection is possible.

Thongchai Dumrongpokaphan, Vladik Kreinovich

Econometric Models of Probabilistic Choice: Beyond McFadden’s Formulas

Traditional decision theory assumes that for every two alternatives, people always make the same (deterministic) choice. In practice, people’s choices are often probabilistic, especially for similar alternatives: the same decision maker can sometimes select one of them and sometimes the other one. In many practical situations, an adequate description of this probabilistic choice can be provided by a logit model proposed by 2001 Nobelist D. McFadden. In this model, the probability of selecting an alternative a is proportional to $$\exp (\beta \cdot u(a))$$exp(β·u(a)), where u(a) is the alternative’s utility. Recently, however, empirical evidence appeared that shows that in some situations, we need to go beyond McFadden’s formulas. In this paper, we use natural symmetries to come up with an appropriate generalization of McFadden’s formulas.

Olga Kosheleva, Vladik Kreinovich, Songsak Sriboonchitta

How to Explain Ubiquity of Constant Elasticity of Substitution (CES) Production and Utility Functions Without Explicitly Postulating CES

In many situations, the dependence of the production or utility on the corresponding factors is described by the CES (Constant Elasticity of Substitution) functions. These functions are usually explained by postulating two requirements: an economically reasonable postulate of homogeneity (that the formulas should not change if we change a measuring unit) and a less convincing CSE requirement. In this paper, we show that the CES requirement can be replaced by a more convincing requirement—that the combined effect of all the factors should not depend on the order in which we combine these factors.

Olga Kosheleva, Vladik Kreinovich, Thongchai Dumrongpokaphan

How to Make Plausibility-Based Forecasting More Accurate

In recent papers, a new plausibility-based forecasting method was proposed. While this method has been empirically successful, one of its steps—selecting a uniform probability distribution for the plausibility level—is heuristic. It is therefore desirable to check whether this selection is optimal or whether a modified selection would like to a more accurate forecast. In this paper, we show that the uniform distribution does not always lead to (asymptotically) optimal estimates, and we show how to modify the uniform-distribution step so that the resulting estimates become asymptotically optimal.

Kongliang Zhu, Nantiworn Thianpaen, Vladik Kreinovich

Structural Breaks of CAPM-type Market Model with Heteroskedasticity and Quantile Regression

In this study we analyze the market beta coefficients of two large capitalization stocks, American International Group (AIG) and Citigroup, from 2005 to 2016 based on a capital asset pricing model (CAPM). Since the daily returns of stock prices experience structural changes in their underlying CAPM-type models, we detect the number and locations of change employing the residual-based cumulative sum (CUSUM) of squares test and then estimate the parameters for each sub-period to evaluate market risk. Moreover, using the quantile regression method, we explore the different behaviors of the market beta and lagged autoregressive effects for different sub-periods and quantile levels. Our final result pertains to the relationship between time-varying betas and structural breaks.

Cathy W. S. Chen, Khemmanant Khamthong, Sangyeol Lee

Weighted Least Squares and Adaptive Least Squares: Further Empirical Evidence

This paper compares ordinary least squares (OLS), weighted least squares (WLS), and adaptive least squares (ALS) by means of a Monte Carlo study and an application to two empirical data sets. Overall, ALS emerges as the winner: It achieves most or even all of the efficiency gains of WLS over OLS when WLS outperforms OLS, but it only has very limited downside risk compared to OLS when OLS outperforms WLS.

Martin Sterchi, Michael Wolf

Prior-Free Probabilistic Inference for Econometricians

The econometrics literature is dominated by the frequentist school which places primary emphasis on the specification of methods that have certain long-run frequency properties, mostly disavowing any notion of inference based on the given data. This preference for frequentism is at least partially based on the belief that probabilistic inference is possible only through a Bayesian approach, the success of which generally depends on the unrealistic assumption that the prior distribution is meaningful in some way. This paper is intended to inform econometricians that an alternative inferential model (IM) approach exists that can achieve probabilistic inference without a prior and while enjoying certain calibration properties essential for reproducibility, etc. Details about the IM construction and its properties are presented, along with some intuition and examples.

Ryan Martin

Robustness in Forecasting Future Liabilities in Insurance

The Gaussian distribution has been widely used in statistical modelling. Being susceptible to outliers, the distribution hampers the robustness of statistical inference. In this paper, we propose two heavy-tailed distributions in the normal location-scale family and show that they are superior to the Gaussian distribution in the modelling of claim amount data from multiple lines of insurance business. Moreover, they also enable better forecasts of future liabilities and risk assessment and management. Implications on risk management practices are also discussed.

W. Y. Jessica Leung, S. T. Boris Choy

On Conditioning in Multidimensional Probabilistic Models

Graphical Markov models, and above all Bayesian networks have become a very popular tool for multidimensional probability distribution representation and processing. The technique making computation with several hundred dimensional probability distribution possible was suggested by Lauritzen and Spiegelhalter. However, to employ it one has to transform a Bayesian network into a decomposable model. This is because decomposable models (or more precisely their building blocks, i.e., their low-dimensional marginals) can be reordered in many ways, so that each variable can be placed at the beginning of the model. It is not difficult to show that there is a much wider class of models possessing this property. In compositional models theory we call these models flexible. It is the widest class of models for which one can always restructure the model in the way that any variable can appear at the beginning of the model. But until recently it had been an open problem whether this class of models is closed under conditioning; i.e., whether a conditional of a flexible model is again flexible. In the paper we will show that this property holds true, which proves the importance of flexible models for practical applications.

Radim Jiroušek

New Estimation Method for Mixture of Normal Distributions

Normal mixture models are widely used for statistical modeling of data, including classification and cluster analysis. However the popular EM algorithms for normal mixtures may give imprecise estimates due to singularities or degeneracies. To avoid this, we propose a new two-step estimation method: first truncate the whole data set to tail data sets that contain points belonging to one component normal distribution with very high probability, and obtain initial estimates of parameters; then upgrade the estimates to better estimates recursively. The initial estimates are simply Method of Moments Estimates in this paper. Empirical results show that parameter estimates are more accurate than that with traditional EM and SEM algorithms.

Qianfang Hu, Zheng Wei, Baokun Li, Tonghui Wang

EM Estimation for Multivariate Skew Slash Distribution

In this paper, the class of multivariate skew slash distributions under different type of setting is introduced and its density function is discussed. A procedure to obtain the Maximum Likelihood estimators for this family is studied. In addition, the Maximum Likelihood estimators for the mixture model based on this family are discussed. For illustration of the main results, we use the actual data coming from the Inner Mongolia Academy of Agriculture and Animal Husbandry Research Station to show the performance of the proposed algorithm.

Weizhong Tian, Guodong Han, Tonghui Wang, Varith Pipitpojanakarn

Constructions of Multivariate Copulas

In this chapter, several general methods of constructions of multivariate copulas are presented, which are generalizations of some existing constructions in bivariate copulas. Dependence properties of new families are explored and examples are given for illustration of our results.

Xiaonan Zhu, Tonghui Wang, Varith Pipitpojanakarn

Plausibility Regions on the Skewness Parameter of Skew Normal Distributions Based on Inferential Models

Inferential models (IMs) are new methods of statistical inference. They have several advantages: (1) They are free of prior distributions; (2) They rely on data. In this paper, $$100(1-\alpha )\%$$100(1-α)% plausibility regions of the skewness parameter of skew-normal distributions are constructed by using IMs, which are the counterparts of classical confidence intervals in IMs.

Xiaonan Zhu, Ziwei Ma, Tonghui Wang, Teerawut Teetranont

International Yield Curve Prediction with Common Functional Principal Component Analysis

We propose an international yield curve predictive model, where common factors are identified using the common functional principal component (CFPC) method that enables a comparison of the variation patterns across different economies with heterogeneous covariances. The dynamics of the international yield curves are further forecasted based on the data-driven common factors in an autoregression framework. For the 1-day ahead out-of-sample forecasts of the US, Sterling, Euro and Japanese yield curve from 07 April 2014 to 06 April 2015, the CFPC factor model is compared with an alternative factor model based on the functional principal component analysis.

Jiejie Zhang, Ying Chen, Stefan Klotz, Kian Guan Lim

An Alternative to p-Values in Hypothesis Testing with Applications in Model Selection of Stock Price Data

In support of the American Statistical Association’s statement on p-value in 2016, see [8], we investigate, in this paper, a classical question in model selection, namely finding a “best-fit” probability distribution to a set of data. Throughout history, there have been a number of tests designed to determine whether a particular distribution fit a set of data, for instance, see [6]. The popular approach is to compute certain test statistics and base the decisions on the p values of these test statistics. As pointed out numerous times in the literature, see [5] for example, p values suffer serious drawbacks which make it untrustworthy in decision making. One typical situation is when the p value is larger than the significance level $$\alpha $$α which results in an inconclusive case. In many studies, a common mistake is to claim that the null hypothesis is true or most likely whereas a big p value merely implies that the null hypothesis is statistically consistent with the observed data; there is no indication that the null hypothesis is “better” than any other hypothesis in the confidence interval. We notice this situation happens a great deal in testing goodness of fit. Therefore, hereby, we propose an approach using the Akaike information criterion (AIC) or the Bayesian information criterion (BIC) to make a selection of the best fit distribution among a group of candidates. As for applications, a variety of stock price data are processed to find a fit distribution. Both the p value and the new approach are computed and compared carefully. The virtue of our approach is that there is always a justified decision made in the end; and, there will be no inconclusiveness whatsoever.

Hien D. Tran, Son P. Nguyen, Hoa T. Le, Uyen H. Pham

Confidence Intervals for the Common Mean of Several Normal Populations

This paper proposes a novel approach for confidence interval estimation for the common mean of several normal populations. This will be achieved by using the concept of an adjusted method of variance estimates recovery approach. The Monte Carlo simulation was used to evaluate the coverage probability and average length. Simulation results are presented to compare the performance from the proposed approach with that of existing approaches. The promising simulation results indicated that the proposed approach should be considered as an alternative to the interval estimation for the common mean.

Warisa Thangjai, Sa-Aat Niwitpong, Suparat Niwitpong

A Generalized Information Theoretical Approach to Non-linear Time Series Model

The limited data will bring about an underdetermined, or ill-posed problem for the observed data, or for regressions using small data set with limited data and the traditional estimation techniques are difficult to obtain the optimal solution. Thus the approach of Generalized Maximum Entropy (GME) is proposed in this study and applied it to estimate the kink regression model under the limited information situation. To the best of our knowledge, the estimation of kink regression model using GME has been not done yet. Hence, we extend the entropy linear regression to non-linear kink regression by modifying the objective and constraint functions under the context of GME. We use both Monte Carlo simulation and real data study to evaluate the performance of our estimation from Kink regression and found that GME estimator performs slightly better compared to the traditional Least squares and Maximum likelihood estimators.

Songsak Sriboochitta, Woraphon Yamaka, Paravee Maneejuk, Pathairat Pastpipatkul

Predictive Recursion Maximum Likelihood of Threshold Autoregressive Model

In the threshold model, it is often the case that an error distribution is not easy to specify, especially when the error has a mixture distribution. In such a situation, standard estimation yields biased results. Thus, this paper proposes a flexible semiparametric estimation for Threshold autoregressive model (TAR) to avoid the specification of error distribution in TAR model. We apply a predictive recursion-based marginal likelihood function in TAR model and maximize this function using hybrid PREM algorithm. We conducted a simulation data and apply the model in the real data application to evaluate the performance of the TAR model. In the simulation data, we found that hybrid PREM algorithm is not outperform Conditional Least Square (CLS) and Bayesian when the error has a normal distribution. However, when Normal-Uniform mixture error is assumed, we found that the PR-EM algorithm produce the best estimation for TAR model.

Pathairat Pastpipatkul, Woraphon Yamaka, Songsak Sriboonchitta

A Multivariate Generalized FGM Copulas and Its Application to Multiple Regression

We introduce a class of multivariate non-exchangeable copulas which generalizes many known bivariate FGM type copula families. The properties such as moments, affiliation, association, and positive lower orthant dependent of the proposed class of copula are studied. The simple-to-use multiple regression function and multiple dependence measure formula for this new class of copulas are derived. Several examples are given to illustrate the main results obtained in this paper.

Zheng Wei, Daeyoung Kim, Tonghui Wang, Teerawut Teetranont



Key Economic Sectors and Their Transitions: Analysis of World Input-Output Network

In the modern society, all major economic sectors have been connected tightly in an extremely complicated global network. In this type of network, a small shock occurred at certain point can be spread instantly through the whole network and may cause catastrophe. Production systems, traditionally analyzed as almost independent national systems, are increasingly connected on a global scale. The world input-output database, only recently becoming available, is one of the first efforts to construct the global and multi-regional input-output tables. The usual way of identifying key sectors in an economy in Input-output analysis is using Leontief inverse matrix to measure the backward linkages and the forward linkages of each sector. In other words, evaluating the role of sectors is performed by means of their centrality assessment. Network analysis of the input-output tables can give valuable insights into identifying the key industries in a world-wide economy. The world input-output tables are viewed as complex networks where the nodes are the individual industries in different economies and the edges are the monetary goods flows between industries. We characterize a certain aspect of centrality or status that is captured by the network measure. We use an $$\alpha $$α-centrality modified method to the weighted directed network. It is used to identify both how a sector could be affected by other sectors and how it could infect the others in the whole economy. The data used is the world input-output table, part of the world input-output database (WIOD) funded by European Commission from 1995 to 2011. We capture the transition of key industries over years through the network measures. We argue that the network structure captured from the input-output tables is a key in determining whether and how microeconomic expansion or shocks propagate throughout the whole economy and shape aggregate outcomes. Understanding the network structure of world input-output data can better inform on how the world economy grows as well as how to prepare for and recover from adverse shocks that disrupt the global production chains. Having analyzed these results, the trend of these sectors in that range of time will be used to reveal how the world economy changed in the last decade.

T. K. Tran, H. Sato, A. Namatame

Natural Resources, Financial Development and Sectoral Value Added in a Resource Based Economy

This chapter vestigates the effects of natural resource dependence and financial development on the sectoral value added in a resource based economy, Yemen. We allow the effect of these two factors to be different for the growth of agricultural, manufacturing and service sectors respectively. We remark on one hand that natural resource curse hypothesis is strongly supported. The agricultural and manufacturing sectors are affected by this phenomenon which implies the existence of Dutch disease symptoms in Yemen. On the other hand, financial sector development does not play an important role in fostering real sectors activities. The service sector is the only sector that benefit from the financial sector development in Yemen. This finding opens up a new insight for Yemeni economy to sustain sectoral growth by controlling the level of natural resource dependence and proactiveness sectoral strategy for financial sector development.

Ramez Abubakr Badeeb, Hooi Hooi Lean

Can Bagging Improve the Forecasting Performance of Tourism Demand Models?

This study examines the forecasting performance of the general-to-specific (GETS) models developed for Hong Kong through the bootstrap aggregating method (known as bagging). Although the literature in other research areas shows that bagging can improve the forecasting performance of GETS models, the empirical analysis in this study does not confirm this conclusion. This study is the first attempt to apply bagging to tourism forecasting, but additional effort is needed to examine the effectiveness of bagging in tourism forecasting by extending the models to cover more destination-source markets related to destinations other than Hong Kong.

Haiyan Song, Stephen F. Witt, Richard T. Qiu

The Role of Asian Credit Default Swap Index in Portfolio Risk Management

This paper aims at evaluating the performance of Asian Credit Default Swap (CDS) index in risk measurement and portfolio optimization by using several multivariate copulas-GARCH models with Expected Shortfall and Sharpe ratio. Multivariate copula-GARCH models consider the volatility and dependence structures of financial assets so that they are conductive to accurately predict risk and optimal portfolio. We find that vine copulas have better performance than other multivariate copulas in model estimation, while the multivariate T copulas have better performance than other kinds of copulas in risk measurement and portfolio optimization. Therefore, the model estimation, risk measurement, and portfolio optimization in empirical study should use different copula models. More importantly, the empirical results give evidences that Asian CDS index can reduce risk.

Jianxu Liu, Chatchai Khiewngamdee, Songsak Sriboonchitta

Chinese Outbound Tourism Demand to Singapore, Malaysia and Thailand Destinations: A Study of Political Events and Holiday Impacts

This chapter investigates the effects of Thailand’s political turmoil and the Chinese Spring Festival on the dynamic dependence between the Chinese outbound tourism demand for Singapore, Malaysia and Thailand (SMT) using the bivariate and multivariate dynamic copula-based ARMAX-APARCH model with skewed Student’s t-distribution and normal inverse Gaussian marginals. We selected political events and the Chinese Spring Festival as the forcing variables to explain the time-varying dependences, and also proposed a dynamic multivariate Gaussian copula to capture the dependence between the Chinese outbound tourism demand for Singapore, Malaysia and Thailand. The main empirical results show that Thailand’s political turmoil and the Chinese Spring Festival, respectively, have negative and positive effects on Chinese tourist arrivals to SMT. Also, there does exist a high degree of persistence pertaining to the dependence structure among SMT. In addition, both the lagged one period of Thailand’s political turmoil and the Chinese Spring Festival are found to have a positive influence on time-varying dependences. Lastly, we found that substitute effects exist between Thailand and Malaysia, while complementary effects prevail between Thailand and Singapore, and Singapore and Malaysia. The findings of this study have important implications for destination managers and travel agents as they help them to understand the impact of political events and holidays on China outbound tourism demand and provide them with a complementary academic approach on evaluating the role of dependencies in the international tourism demand model.

Jianxu Liu, Duangthip Sirikanchanarak, Jiachun Xie, Songsak Sriboonchitta

Forecasting Asian Credit Default Swap Spreads: A Comparison of Multi-regime Models

This paper aims to explore the best forecasting model for predicting the Credit Default Swap (CDS) index spreads in emerging markets Asia by comparing the forecasting performance between the multi-regime models. We apply threshold, Markov switching, Markov switching GARCH and simple least squares for structural and autoregressive modeling. Both in- and out-of-sample forecasts are conducted to compare the forecasting performance between models. The results suggest that Markov switching GARCH(1,1) structural model presents the best performance in predicting Asian Credit Default Swap (CDS) index spreads. We also check the preciseness of our selected model by employing the robustness test.

Chatchai Khiewngamdee, Woraphon Yamaka, Songsak Sriboonchitta

Effect of Helmet Use on Severity of Head Injuries Using Doubly Robust Estimators

Causal inference based on observational data can be formulated as a missing outcome imputation and an adjustment for covariate imbalance models. Doubly robust estimators–a combination of imputation-based and inverse probability weighting estimators–offer some protection against some particular misspecified assumptions. When at least one of the two models is correctly specified, doubly robust estimators are asymptotically unbiased and consistent. We reviewed and applied the doubly robust estimators for estimating causal effect of helmet use on the severity of head injury from observational data. We found that helmet usage has a small effect on the severity of head injury.

Jirakom Sirisrisakulchai, Songsak Sriboonchitta

Forecasting Cash Holding with Cash Deposit Using Time Series Approaches

The levels of cash holding and cash deposit for Thai banks have significantly increased over the past 10 years. This paper aims to forecast cash holding by using cash deposit. For banks, cash holding partially is from the cash deposited. In addition, accurate prediction on the cash holding would provide valuable information and indicators supervising bankers to control the levels of both cash holding and cash deposit effectively. In addition, the empirical relevance of cash holding and cash deposit is examined with three different models; linear model, ARIMA model and state space model. Experimental results with real data sets illustrate that state space model tends be the most accurate model compared to the other two models for prediction.

Kobpongkit Navapan, Jianxu Liu, Songsak Sriboonchitta

Forecasting GDP Growth in Thailand with Different Leading Indicators Using MIDAS Regression Models

In this study, we compare the performance between three leading indicators, namely, export, unemployment rate, and SET index in forecasting QGDP growth in Thailand using the mixed-frequency data sampling (MIDAS) approach. The MIDAS approach allows us to use monthly information of leading indicators to forecast QGDP growth without transforming them into quarterly frequency. The basic MIDAS model and the U-MIDAS model are considered. Our findings show that unemployment rate is the best leading indicator for forecasting QGDP growth for both MIDAS settings. In addition, we investigate the forecast performance between the basic MIDAS model and the U-MIDAS model. The results suggest that the U-MIDAS model can outperform the basic MIDAS model regardless of leading indicators considered in this study.

Natthaphat Kingnetr, Tanaporn Tungtrakul, Songsak Sriboonchitta

Testing the Validity of Economic Growth Theories Using Copula-Based Seemingly Unrelated Quantile Kink Regression

The distinct points of view about factors driving economic growth are introduced all the time in which some effectively useful suggestions then become the growth theories, which in turn lead to various researches on economic growth. This paper aims to examine the joint validity of the growth theories using our introduced model named copula based seemingly unrelated quantile kink regression as a key tool in this work. We concentrate exclusively on the experience of Thailand and found that the growth models can prove their validities for the Thai economy through this experiment.

Pathairat Pastpipatkul, Paravee Maneejuk, Songsak Sriboonchitta

Analysis of Global Competitiveness Using Copula-Based Stochastic Frontier Kink Model

The competitiveness is a considerable issue for nations who rely on the international trade and hence leads to the competitiveness evaluation. This paper suggests considering a country’s productive efficiency to reflect the competitive ability. We introduce the copula-based nonlinear stochastic frontier model as a contribution to the competitiveness evaluation due to a special concern about the difference among countries in terms of size and structure of the economies. As a specific capability of this proposed model, we are able to find the different impact of inputs on output from the group of small countries to the group of large countries. Finally, this paper provides the efficiency scores according to our analysis and the overall ranking of global competitiveness.

Paravee Maneejuk, Woraphon Yamaka, Songsak Sriboonchitta

Gravity Model of Trade with Linear Quantile Mixed Models Approach

The Thai economy has mostly depended on exports, which has significantly declined in recent years. Hence, this paper is to investigate the determinants affecting Thailands exports with its top ten trading partners by using gravity model approach along with panel data. In panel data, there are different characteristics between entities that account for unobserved individual effects. Previous studies have only focused on estimating mean effect. Mixed models are relatively selected as additional approach for panel data that accounts for individual heterogeneity in terms of variance components. Another advantage is that they are suitable for dependent data which are likely to be similar as collected repeatedly on the same country. We also take an interest in studying different magnitudes and directions of the effects of determinants on different parts of the distribution of export values. Meanwhile, Quantile regression (QR) allows study of different quantiles of the conditional distribution. In this study we combine the benefits of both mixed effects and quantile estimator to study Thai exports and employ linear quantile mixed models (LQMMs) with gravity model.

Pathairat Pastpipatkul, Petchaluck Boonyakunakorn, Songsak Sriboonchitta

Stochastic Frontier Model in Financial Econometrics: A Copula-Based Approach

This study applies the principle of stochastic frontier model (SFM) to calculate the optimal frontier of the stock prices in a stock market. We use copula to measure dependence between the error terms in SFM by examining several stocks in Down Jones industrial. The results show that our modified stochastic frontier model is more applicable for financial econometrics. Finally, we use AIC for model selection.

P. Tibprasorn, K. Autchariyapanitkul, S. Sriboonchitta

Quantile Forecasting of PM10 Data in Korea Based on Time Series Models

In this chapter, we analyze the particulate matter PM10 data in Korea using time series models. For this task, we use the log-transformed data of the daily averages of the PM10 values collected from Korea Meteorological Administration and obtain an optimal ARMA model. We then conduct the entropy-based goodness of fit test for the obtained residuals to check the departure from the normal and skew-t distributions. Based on the selected skew-t ARMA model, we obtain conditional quantile forecasts using the parametric and quantile regression methods. The obtained result has a potential usage as a guideline for the patients with some respiratory disease to pay more attention to health care when the conditional quantile forecast is beyond the limit values of severe health hazards.

Yingshi Xu, Sangyeol Lee

Do We Have Robust GARCH Models Under Different Mean Equations: Evidence from Exchange Rates of Thailand?

This study investigates the exchange rate volatility of Thai baht using GARCH, TGARCH, EGARCH and PGARCH models and examines the robustness of these models under different mean equation specifications. The data consisted of monthly exchange rate of Thai baht with five currencies of leading trade partners during January 2002–March 2016. The results show that the GARCH model is well-fitted for Chinese yuan and US dollar exchange rate, while TGARCH model is suitable to be selected for Japanese yen, Malaysian ringgit and Singapore dollar. For the model sensitivity, the findings indicate that the GARCH model is robust for the cases of Chinese yuan and US dollar, while TGARCH model is robust only for Malaysian ringgit. Therefore, We conclude that the selection of GARCH models is sensitive to mean equation specification. This confirms that researchers should pay attention to mean equation specifications when it comes to volatility modelling.

Tanaporn Tungtrakul, Natthaphat Kingnetr, Songsak Sriboonchitta

Joint Determinants of Foreign Direct Investment (FDI) Inflow in Cambodia: A Panel Co-integration Approach

Globalization and modernization have generated the new opportunities for Multinational Enterprises (MNEs) to invest in foreign countries. Especially, many emerging and developing countries are making efforts actively to attract foreign direct investment (FDI) inflow in the purpose of boosting economic growth and development. This paper investigates the determinants of Cambodia’s inward FDI within the time interval from 1995 to 2014. Panel co-integration approach, namely Full Modified Ordinary Least Square (FMOLS) and Dynamic Ordinary Least Square (DOLS) are proposed to estimate the long run coefficients. Our analysis shows that most of the variables are statistically significant except for population growth rate. Market size and financial development are, as expected, positively correlated whereas macroeconomic instability and cost of living are negatively associated but poor institution is, as unexpected, positively associated to inward FDI. The sign of ECT $$(\mathrm {t}-1)$$(t-1) coefficient from panel causality analysis is significantly negative for GDP to FDI equation. It is indicated that economic growth and FDI is bidirectional causal relationship in the short run and the long run. The result from measurement predictive accuracy obtained from out of sample ex-post forecasting (2013–2014) confirmed that panel DOLS has a good predictive power to apply the long run ex-ante forecasting of Cambodia’s inward FDI. Thus, our findings suggest that improving macroeconomic indicators, administrative barrier and financial instrument and development are the crucial policies to attract more inward FDI in the upcoming period.

Theara Chhorn, Jirakom Sirisrisakulchai, Chukiat Chaiboonsri, Jianxu Liu

The Visitors’ Attitudes and Perceived Value Toward Rural Regeneration Community Development of Taiwan

The purpose of the rural regeneration plan carried out for years is mainly for rural sustainable development, which makes communities change and indirectly attracts many tourists. Especially the rural experience tourism emerged recently drives the rural economy grow entirely, enriches the rural environment and style, and also increases many job opportunities and accelerates the prosperity of local communities. Although the booming tourism increases the number of travelers and facilitates the local development, it has the cognitive deficiency in the aspect of ecology maintenance. As a result, the conservation and the economic development fail to reach a balance.In this study, we will take the Wu Mi Le community of Tainan as an example to analyze the cognitive elements of the rural regeneration, and use the cluster analysis to discuss the preference of difference groups to travel experience. In addition, we will further use the contingent valuation method (CVM) to measure the willingness to pay (WTP) of tourists to the rural maintenance and the tourist activities in this study.The research results are summarized as below: 1. The environment conservations cognition is firstly considered for tourists to the rural regeneration communities; 2. The multi-existence group has a higher contribution in rural development; 3. Tourists think the maintained value is higher than the recreation value.

Wan-Tran Huang, Chung-Te Ting, Yu-Sheng Huang, Cheng-Han Chuang

Analyzing the Contribution of ASEAN Stock Markets to Systemic Risk

In this paper, seven stock markets from six countries (Thailand, Malaysia, Indonesia, Vietnam, the Philippines, and Singapore) and their risk contribution to ASEAN stock system are investigated using the Component Expected Shortfall approach. Prior to computing this systemic risk measure, we need to compute a dynamic correlation, thus the study proposes a Markov Switching copula with time varying parameter to measure the dynamic correlation between each pair of stock market index and ASEAN stock system. The empirical results show that Philippines stock index contributed the highest risk to the ASEAN stock system.

Roengchai Tansuchat, Woraphon Yamaka, Kritsana Khemawanit, Songsak Sriboonchitta

Estimating Efficiency of Stock Return with Interval Data

Existing studies on capital asset pricing model (CAPM) have basically focused on point data which may not concern about the variability and uncertainty in the data. Hence, this paper suggests the approach that gains more efficiency, that is, the interval data in CAPM analysis. The interval data is applied to the copula-based stochastic frontier model to obtain the return efficiency. This approach has proved its efficiency through application in three stock prices: Apple, Facebook and Google.

Phachongchit Tibprasorn, Chatchai Khiewngamdee, Woraphon Yamaka, Songsak Sriboonchitta

The Impact of Extreme Events on Portfolio in Financial Risk Management

We use the concept of copula and extreme value theory to evaluate the impact of extreme events such as flooding, nuclear disaster, etc. on the industry index portfolio. A t copulas based on GARCH model is applied to explain a portfolio risk management with high-dimensional asset allocation. Finally, we calculate the condition Value-at-Risk (CVaR) with the hypothesis of t joint distribution to construct the potential frontier of the portfolio during the times of crisis.

K. Chuangchid, K. Autchariyapanitkul, S. Sriboonchitta

Foreign Direct Investment, Exports and Economic Growth in ASEAN Region: Empirical Analysis from Panel Data

The major purpose of this research study is twofold. Firstly, to examine the causal relationship among foreign direct investment (FDI), exports, and economic growth of ASEAN economy comprising Cambodia, Lao PDR, Malaysia, Philippines, Singapore, Thailand, and Vietnam, by using panel VECM covering from 2000 to 2014. Secondly, to estimate the impact of FDI and exports on ASEAN economy. The dummy variable representing the financial crisis in 2008 is used to see the real effect in this study. The empirical results indicate that bidirectional causal relation between economic growth and exports is found in ASEAN association while there are two unidirectional causal linkages between FDI-economic growth and FDI-exports as the causal direction running from FDI to economic growth and running from FDI to exports in ASEAN economy. Based on the findings from panel dynamic ordinary least square (DOLS) and fully modified ordinary least square (FMOLS) methods, the elasticity of GDP with respect to FDI is 0.048 and 0.044% and respect to exports is 0.547 and 0.578%. Therefore, it can be concluded that FDI and exports are significant aspects which positively impact on ASEAN economic development.

Pheara Pheang, Jianxu Liu, Jirakom Sirisrisakulchai, Chukiat Chaiboonsri, Songsak Sriboonchitta
Weitere Informationen

Premium Partner