Sains Malaysiana 53(2)(2024): 461-476

http://doi.org/10.17576/jsm-2024-5302-18

Examining Tail Index Estimators in New Pareto Distribution: Monte Carlo Simulations and Income Data Applications

(Menyemak Penganggar Indeks Ekor dalam Taburan Pareto Baharu: Simulasi Monte Carlo dan Aplikasi Data Pendapatan)

MUHAMMAD ASLAM MOHD SAFARI^1,2,*, NURULKAMAL MASSERAN³ & MOHD AZMI HARON⁴

¹Department of Mathematics and Statistics, Faculty of Science, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

²Institute for Mathematical Research, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

³Department of Mathematical Sciences, Faculty of Science and Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia

⁴Institute of Mathematical Sciences, Faculty of Science, Universiti Malaya, 50603 Kuala Lumpur, Malaysia

Diserahkan: 22 Julai2023/Diterima: 16 Januari2024

Abstract

An evolved form of Pareto distribution, the new Pareto-type distribution, offers an alternative model for data with heavy-tailed characteristics. This investigation examines and discusses fourteen diverse estimators for the tail index of the new Pareto-type, including estimators such as maximum likelihood, method of moments, maximum product of spacing, its modified version, ordinary least squares, weighted least squares, percentile, Kolmogorov-Smirnov, Anderson-Darling, its modified version, Cramér-von Mises, and Zhang's variants of the previous three. Using Monte Carlo simulations, the effectiveness of these estimators is compared both with and without the presence of outliers. The findings show that, without outliers, the maximum product of spacing, its modified version, and maximum likelihood are the most effective estimators. In contrast, with outliers present, the top performers are Cramér-von Mises, ordinary least squares, and weighted least squares. The study further introduces a graphical method called the new Pareto-type quantile plot for validating the new Pareto-type assumptions and outlines a stepwise process to identify the optimal threshold for this distribution. Concluding the study, the new Pareto-type distribution is employed to model the high-end household income data from Italy and Malaysia, leveraging all the methodologies proposed.

Keywords: Estimation techniques; heavy-tailed data; income data modelling; Monte Carlo analysis; pareto distribution; robustness

Abstrak

Satu taburan Pareto yang berkembang iaitu taburan jenis Pareto baharu, menawarkan model alternatif untuk data dengan ciri ekor berat. Kajian ini meneliti dan membincangkan empat belas penganggar yang pelbagai bagi indeks ekor jenis Pareto baharu, termasuk penganggar seperti kebolehjadian maksimum, kaedah momen, produk jarak maksimum bersama versi yang diubah suai, kuasa dua terkecil biasa, kuasa dua terkecil berwajaran, persentil, Kolmogorov-Smirnov, Anderson-Darling Bersama versi yang diubah suai, Cramér-von Mises, dan varian Zhang bagi Kolmogorov-Smirnov, Anderson-Darling serta Cramér-von Mises. Dengan menggunakan simulasi Monte Carlo, keberkesanan penganggar ini dibandingkan dengan kehadiran dan tanpa kehadiran titik terpencil. Hasil kajian menunjukkan bahawa, tanpa titik terpencil, produk jarak maksimum bersama versi yang diubah suai dan kebolehjadian maksimum adalah penganggar yang paling berkesan. Sebaliknya, dengan kehadiran titik terpencil, penganggar terbaik adalah Cramér-von Mises, kuasa dua terkecil biasa dan kuasa dua terkecil berwajaran. Kajian ini seterusnya memperkenalkan kaedah grafik yang disebut sebagai plot kuantil jenis Pareto baharu untuk mengesahkan andaian jenis Pareto baharu dan menggariskan proses bertahap untuk mengenal pasti ambang optimum untuk taburan ini. Mengakhiri kajian, taburan jenis Pareto baharu digunakan untuk memodelkan data pendapatan isi rumah kelas atas dari Itali dan Malaysia, memanfaatkan semua kaedah yang dicadangkan.

Kata kunci: Analisis Monte Carlo; data ekor berat; kaedah penganggaran; keteguhan; pemodelan data pendapatan; taburan Pareto

RUJUKAN

Abd Raof, A.S., Haron, M.A., Safari, M.A.M. & Siri, Z. 2022. Modeling the incomes of the upper-class group in Malaysia using new Pareto-type distribution. Sains Malaysiana 51(10): 3437-3448.

Alfons, A., Templ, M. & Filzmoser, P. 2013. Robust estimation of economic indicators from survey samples based on Pareto tail modelling. Journal of the Royal Statistical Society. Series C: Applied Statistics 62(2): 271-286.

Amoroso, L. 1938. Vilfredo Pareto. Econometrica: Journal of the Econometric Society 6(1): 1-21.

Banca d’Italia. 2008. Survey of Household Income and Wealth (SHIW) of the Bank of Italy. https://www.bancaditalia.it/pubblicazioni/indagine-famiglie/index.html

Bee, M., Riccaboni, M. & Schiavo, S. 2019. Distribution of city size: Gibrat, Pareto, Zipf. In The Mathematics of Urban Morphology, edited by D’Acci L. Modeling and Simulation in Science, Engineering and Technology. Birkhäuser, Cham. pp. 77-91. https://doi.org/10.1007/978-3-030-12381-9_4

Beirlant, J., Vynckier, P. & Teugels, J.L. 1996. Tail index estimation, pareto quantile plots regression diagnostics. Journal of the American Statistical Association 91(436): 1659-1667.

Bourguignon, M., Saulo, H. & Fernandez, R.N. 2016. A new Pareto-type distribution with applications in reliability and income data. Physica A: Statistical Mechanics and Its Applications 457: 166-175.

Cheng, R.C.H. & Amin, N.A.K. 1983. Estimating parameters in continuous univariate distributions with a shifted origin. Journal of the Royal Statistical Society: Series B (Methodological) 45(3): 394-403.

Cheng, R.C.H. & Stephens, M.A. 1989. A goodness-of-fit test using Moran’s statistic with estimated parameters. Biometrika 76(2): 385-392.

Cirillo, P. 2013. Are your data really Pareto distributed? Physica A: Statistical Mechanics and Its Applications 392(23): 5947-5962.

Cirillo, P. & Hüsler, J. 2009. On the upper tail of Italian firms’ size distribution. Physica A: Statistical Mechanics and Its Applications 388(8): 1546-1554.

Coronel-Brizio, H.F. & Hernandez-Montoya, A.R. 2005. On fitting the Pareto–Levy distribution to stock market index data: Selecting a suitable cutoff value. Physica A: Statistical Mechanics and Its Applications 354: 437-449.

Department of Statistics Malaysia. 2017. Household Income and Basic Amenities Survey Report 2016.

Díaz, J.D., Cubillos, P.G. & Griñen, P.T. 2021. The exponential Pareto model with hidden income processes: Evidence from Chile. Physica A: Statistical Mechanics and Its Applications 561: 125196.

Dunford, R., Su, Q. & Tamang, E. 2014. The pareto principle. The Plymouth Student Scientist 7(1): 140-148.

Filimonov, V. & Sornette, D. 2015. Power law scaling and 'Dragon-Kings' in distributions of intraday financial drawdowns. Chaos, Solitons & Fractals 74: 27-45.

Gabaix, X. 2009. Power laws in economics and finance. Annu. Rev. Econ. 1(1): 255-294.

García, I.G. & Caballero, A.M. 2021. Models of wealth and inequality using fiscal microdata: Distribution in Spain from 2015 to 2020. Mathematics 9(4): 377.

Giesen, K., Zimmermann, A. & Suedekum, J. 2010. The size distribution across all cities - Double Pareto lognormal strikes. Journal of Urban Economics 68(2): 129-137.

Giorgi, G.M. & Gigliarano, C. 2017. The Gini concentration index: A review of the inference literature. Journal of Economic Surveys 31(4): 1130-1148.

Hlasny, V. & Verme, P. 2018. Top incomes and the measurement of inequality in Egypt. The World Bank Economic Review 32(2): 428-455.

Jiang, R. 2013. A modified MPS method for fitting the 3-parameter Weibull distribution. 2013 International Conference on Quality, Reliability, Risk, Maintenance, and Safety Engineering (QR2MSE). pp. 983-985.

Kao, J.H.K. 1958. Computer methods for estimating Weibull parameters in reliability studies. IRE Transactions on Reliability and Quality Control. pp. 15-22.

Luceño, A. 2008. Maximum likelihood vs. maximum goodness of fit estimation of the three-parameter Weibull distribution. Journal of Statistical Computation and Simulation 78(10): 941-949.

Luceño, A. 2006. Fitting the generalized Pareto distribution to data using maximum goodness-of-fit estimators. Computational Statistics & Data Analysis 51(2): 904-917.

Lux, T. & Alfarano, S. 2016. Financial power laws: Empirical evidence, models, and mechanisms. Chaos, Solitons & Fractals 88: 3-18. https://doi.org/https://doi.org/10.1016/j.chaos.2016.01.020

Majid, M.H.A. & Ibrahim, K. 2021. Composite Pareto distributions for modelling household income distribution in Malaysia. Sains Malaysiana 50(7): 2047-2058.

Majid, M.H.A., Ibrahim, K. & Masseran, N. 2023. Three-part composite Pareto modelling for income distribution in Malaysia. Mathematics 11(13): 2899.

Meyer, S. & Held, L. 2014. Power-law models for infectious disease spread. Annals of Applied Statistics 8(3): 1612-1639.

Newman, M.E.J. 2005. Power laws, Pareto distributions and Zipf’s law. Contemporary Physics 46(5): 323-351.

Oancea, B., Andrei, T. & Pirjol, D. 2017. Income inequality in Romania: The exponential-Pareto distribution. Physica A: Statistical Mechanics and Its Applications 469: 486-498.

Pinto, C.M.A., Lopes, A.M. & Machado, J.A.T. 2012. A review of power laws in real life phenomena. Communications in Nonlinear Science and Numerical Simulation 17(9): 3558-3578.

Safari, M.A.M., Masseran, N. & Ibrahim, K. 2018a. A robust semi-parametric approach for measuring income inequality in Malaysia. Physica A: Statistical Mechanics and Its Applications 512: 1-13.

Safari, M.A.M., Masseran, N. & Ibrahim, K. 2018b. Optimal threshold for Pareto tail modelling in the presence of outliers. Physica A: Statistical Mechanics and Its Applications 509: 169-180.

Safari, M.A.M., Masseran, N., Ibrahim, K. & Hussain, S.I. 2021. Measuring income inequality: A robust semi-parametric approach. Physica A: Statistical Mechanics and Its Applications 562: 125359.

Safari, M.A.M., Masseran, N., Ibrahim, K. & AL-Dhurafi, N.A. 2020. The power-law distribution for the income of poor households. Physica A: Statistical Mechanics and Its Applications 557: 124893.

Safari, M.A.M., Masseran, N., Ibrahim, K. & Hussain, S.I. 2019. A robust and efficient estimator for the tail index of inverse Pareto distribution. Physica A: Statistical Mechanics and Its Applications 517: 431-439.

Sarabia, J.M., Jordá, V. & Prieto, F. 2019. On a new Pareto-type distribution with applications in the study of income inequality and risk analysis. Physica A: Statistical Mechanics and Its Applications 527: 121277.

Soriano-Hernández, P., del Castillo-Mussot, M., Córdoba-Rodríguez, O. & Mansilla-Corona, R. 2017. Non-stationary individual and household income of poor, rich and middle classes in Mexico. Physica A: Statistical Mechanics and Its Applications 465: 403-413.

Xu, Y., Wang, Y., Tao, X. & Ližbetinová, L. 2017. Evidence of Chinese income dynamics and its effects on income scaling law. Physica A: Statistical Mechanics and Its Applications 487: 143-152.

*Pengarang untuk surat-menyurat; email: aslam.safari@upm.edu.my