Introduction to modeling and generating probabilistic input processes for simulation

Kuhl,; Lada,; Wagner, Verena; Ivy,; Steiger,; Wilson, F. F.

doi:10.1109/wsc.2007.4419589

Cited by 21 publications

(10 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…My purpose in this tutorial is to discuss some issues and approaches for effective representation and generation of uncertain inputs when building and exercising computer-simulation models. I won't be going into a lot of technical or mathematical detail, as such issues are very well covered elsewhere (see, for example, Kuhl et al 2008, and elsewhere in these Proceedings as well).…”

Section: Introductionmentioning

confidence: 99%

Representing and generating uncertainty effectively

Kelton

2009

Proceedings of the 2009 Winter Simulation Conference (WSC)

View full text Add to dashboard Cite

Stochastic simulations involve random inputs, so produce random outputs too. This introductory tutorial is meant to call attention to the need to model and generate such inputs in ways that may not be the standard or defaults in simulation-modeling software, yet can be critical to model validity (a.k.a. getting right rather than wrong answers). There are both dangers involved with doing this inappropriately, as well as opportunities to do things better, making for more accurate and more precise results from simulations. Specific issues include possible dependence across and within random inputs, use of empirical distributions even if a "standard" fits the data, and non-default use of the underlying random-number generator. Suggestions for novel ways of implementing some of these ideas in simulation-modeling software are offered.

show abstract

Section: Introductionmentioning

confidence: 99%

Representing and generating uncertainty effectively

Kelton

2009

Proceedings of the 2009 Winter Simulation Conference (WSC)

View full text Add to dashboard Cite

show abstract

“…Thus, the flaws and inadequacies of the normal distribution in describing returns are generic to financial assets, REITs and similar trusts. Some recent applications to capture non-normality in returns have based on generalised Pareto distribution (GPD), generalised extreme value distribution (GEVD), Student t-distribution, unbounded Johnson family (Johnson, 1949), normal inverse Gaussian distribution, generalised lambda distribution (GLD) (Freimer et al, 1988;Ramberg and Schmeiser, 1974), generalised beta family of distributions (GBD) (Kuhl et al, 2006) and others.…”

Section: Brief Literature Review and Related Gapsmentioning

confidence: 99%

Modelling return behaviour of global real estate investment trusts equities

Owusu

Tweneboah

Ijasan

et al. 2019

JERER

View full text Add to dashboard Cite

Purpose This paper aims to contribute to knowledge by investigating the return behaviour of seven global real estate investment trusts (REITs) with respect to the appropriate distributional fit that captures tail and shape characteristics. The study adds to the knowledge of distributional properties of seven global REITs by using the generalised lambda distribution (GLD), which captures fairly well the higher moments of the returns. Design/methodology/approach This is an empirical study with GLD through three rival methods of fitting tail and shape properties of seven REIT return data from January 2008 to November 2017. A post-Global Financial Crisis (GFC) (from July 2009) period fits from the same methods are juxtaposed for comparison. Findings The maximum likelihood estimates outperform the methods of moment matching and quantile matching in terms of goodness-of-fit in line with extant literature; for the post-GFC period as against the full-sample period. All three methods fit better in full-sample period than post-GFC period for all seven countries for the Region 4 support dynamics. Further, USA and Singapore possess the strongest and stronger infinite supports for both time regimes. Research limitations/implications The REITs markets, however, developed, are of wide varied sizes. This makes comparison less than ideal. This is mitigated by a univariate analysis rather than multivariate one. Practical implications This paper is a reminder of the inadequacy of the normal distribution, as well as the mean, variance, skewness and kurtosis measures, in describing distributions of asset returns. Investors and policymakers may look at the location and scale of GLD for decision-making about REITs. Originality/value The novelty of this work lies with the data used and the detailed analysis and for the post-GFC sample.

show abstract

“…For the purpose of this paper, it suffices to note that for both of these aspects, a separate probability density function (pdf) should be chosen. The generalised beta distribution [38] is an appropriate choice for both [15]. We chose to let a parameter vector ω = (a, b, m, μ) characterise the pdf for a random variable that is distributed according to the generalised beta distribution (β(ω)).…”

Section: Activity Duration Distributions Modellingmentioning

confidence: 99%