DOCSLIB.ORG

Variance, Skewness & Kurtosis: Results from the APM Cluster

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text

Load more

a Mon Not R Astron So c Printed January MN LT X style le v E Variance Skewness Kurtosis results from the APM Cluster Redshift Survey and mo del predictions E Gazta naga RAC Croft G B Dalton Department of Physics Astrophysics University of Oxford Nuclear Astrophysics Laboratory Keble Road Oxford OX RH CSIC Centre dEstudis Avancatsde Blanes Astrophysics Blanes Girona Spain ABSTRACT We estimate the variance the skewness and the kurtosis in the distribution of 2 3 4 density uctuations in a complete sample from the APM Cluster Redshift Survey with 1 clusters and a mean depth D h Mp c We are able to measure the statistics 1 of uctuations in spheres of radius R h Mp c with reasonable errorbars The J 1 statistics in the cluster distribution follow the hierarchical pattern S with J J 2 S roughly constant S and S We analyse the distribution of clusters taken J 3 4 from Nb o dy simulations of dierent dark matter mo dels The results are compared with an alternative metho d of simulating clusters which uses the truncated Zeldovich approximation We argue that this alternative metho d is not reliable enough for making The Nb o dy simulation results follow similar hierarchical quantitative predictions of relations to the observations with S almost unaected by redshift distortions from J p eculiar motions The standard Cold Dark Matter CDM mo del is inconsistent with either the second third or fourth order statistics at all scales However b oth a hybrid Mixed Dark Matter mo del and a low density CDM variant agree with the J observations Key words Largescale structure of the universe galaxies clustering metho ds numerical statistical limited samples of Xray clusters Romer et al give INTRODUCTION mutually consistent results The p oint correlation of the Rich clusters of galaxies can b e used to trace the structure APM Automatic Plate Measuring machine Cluster Red of the Universe on very large scales Although the identi shift Survey has already b een used to constrain dark matter astro-ph/9501115 31 Jan 1995 cation of clusters is more dicult than that of galaxies mo dels Dalton et al b clusters have several imp ortant advantages as tracers b oth Previous analysis of p oint and p oint statistics in observations and in simulations Since clusters represent and of cluster samples have b een based on Ab ells cat high densities in the galaxy distributio n it is p ossible to alogue and extensions by Ab ell Corwin Olowin use them to map out the distant Universe in a systematic eg Jing Zhang TothHollosi Szalay Jing and economical way As clusters are asso ciated with high Valdarnini Plionis Valdarnini Cappi Mau p eaks in the mass distribution cluster correlations evolve rogordato These samples seem to repro duce the hi very slowly with time and so the clustering statistics deter J found in the galaxy distribu erarchical relation J mined from numerical simulations are relatively insensitive tion eg Groth Peebles Fry Peebles Sharp to the ep o ch at which the simulations are analysed Croft et al Szapudi Szalay Boschan Meiksin Sza Efstathiou a An additional advantage is that clusters pudi Szalay Bouchet et al Gazta naga are strongly correlated so that even a small sample can b e These observations could provide imp ortant insights used to estimate the correlation functions into the underlying matter distribution and the formation of The p oint correlation function for galaxy clusters was clusters However there is evidence that the clustering mea rst analysed in Ab ells catalogue eg Hauser sured from Ab ells catalogue is aected by inhomogeneities Peebles Bahcall Soneira Klypin Kopylov in the selection of clusters eg Efstathiou et al and Recent determinations based on automated cluster it is not clear how this could aect the ab ove mentioned catalogues Dalton et al Nichol et al and on ux estimates of and Here we estimate the variance the skewness and Throughout the pap er we use H h kmsMp c the kurtosis in the distribution of density uctuations 2 Gazta naga Croft Dalton sen to b e h Mp c and the mean intercluster separation in a sample from the APM Cluster Redshift Survey The h Mp c Croft Efstathiou a selection of clusters in the APM Cluster Catalogue is based on a computer algorithm applied to digitised photographic We have also tested an alternative scheme for mak ing theoretical predictions ab out cluster clustering This plates and so provides a more uniform sample of clusters than Ab ells We compare the results with previous analysis scheme advocated by Plionis et al makes use of the and with dierent cluster simulations Zeldovich approximation to dynamicall y evolve the density eld b efore clusters are selected In order to assess In next section we review the APM cluster data and the validity of results derived using this metho d we have Nb o dy mo dels We present the metho d of estimation and the results in Section Sections and are devoted to examined clustering in the Zeldovich approximation and carried out a comparison with Nb o dy clusters These tests the discussion and conclusions In an app endix we consider are detailed in App endix A the prop erties of simulations generated using the truncated Zeldovich approximation CLUSTER CORRELATIONS CLUSTER SAMPLES Metho d of estimation The APM Cluster Survey We apply a variant of the metho d used by Efstathiou et al We use sample B from Dalton et al b which contains which accounts for selection eects automatically by clusters with mean redshift and spacedensity estimating the mean density and its uctuations in shells h Mp c The clusters were selected from the of mean constant redshift We divide space in the redshift APM Galaxy Survey Maddox et al bc using an au catalogue into a series of concentric shells of width R cen tomatic selection algorithm describ ed in detail by Dalton tred on the observer which are further sub divided into M s et al in preparation We have cut the sample to redshifts spherical cells of radius R For each shell s we compute the larger than kms to avoid uncertainties in the selec mean number of particles N R in a cell and the moments s tion function at low redshift and imp osed an upp er limit of m R J kms There are only clusters outside this range so that the total number in the selected region is N As M s X J the redshifts are obtained from only a few galaxies p er clus m N N J c s ter the individu al cluster redshifts are uncertain to within c kms Dalton et al a th where N is the number of galaxies in the c cell The cor c J N corrected for resp onding connected moments k J s J Simulations Poisson shotnoise see section are given by The cluster simulations are those generated by Croft Efs k m N s tathiou ab using a particleparticl eparti cl emesh N N k m m s b o dy co de Efstathiou Eastwoo d Efstathiou et al k m m m m N s and are the same as those used by Dalton et al b We consider three dierent mo dels First the stan see eg Gazta naga Thus k N is the variance s dard CDM mo del which is scaleinvariant and has k N is the skewness and k N also called s s hereafter h CDM Second a scaleinvariant the kurtosis lowdensity CDM universe with h and a cos To estimate the mean value of h i for the whole sample mological constant such that H we introduce the standard Gaussian likeliho o d function so CDM The p ower sp ectrum for these two mo dels is taken that the value of k in each shell is weighted by the inverse from Efstathiou Bond White Third a scale of its variance Vark s invariant mo del containing b oth a massive neutrino com p onent eV neutrinos and CDM such that total k k k k with h Mixed Dark Matter Vark CDM s M s MDM from Klypin et al The mo dels have b een nor malised to b e consistent with the amplitude of uctuations where we have made the assumption that the cells are in in the microwave background Wright et al so that dep endent Thus the weighting comp ensates for the eects the value of the linear theory rms mass uctuations in h of uctuations from b oth shotnoise imp ortant for distant Mp c spheres for the two CDM mo dels and for diluted shells or from having a small number of cells imp or MDM realisations of each mo del were run using dierent tant in the nearby shells We have tried two prescriptions random phases in b oxes of side length h Mp c with to estimate this variance The rst one uses the higher order 6 particles We have also run realisations of the moments m in the shell to estimate k in equation In J J the second prescription we assume a Gaussian distribution CDM with the slightly higher normalisation Clus ters of particles are lo cated using a p ercolation algorithm to estimate k from m as in Efstathiou et al Both J the cluster mass is dened within a metric radius and the give similar results richness limit of the cluster sample is chosen to match the For higher order moments we use the same weights for observed cluster space density The metric radius is cho each shell as the ones used to estimate so that in general Variance Skewness Kurtosis from galaxy clusters 3 Figure The variance R estimated from countsincells in Figure The skewness R estimated from countsincells in a simulated mo ck catalogue of clusters symbols with errorbars a simulated mo ck catalogue of clusters compared with the full compared with the values from clusters selected from the same simulation as in Figure realization
Recommended publications
  • 1. How Different Is the T Distribution from the Normal?

    1. How Different Is the T Distribution from the Normal?

    Statistics 101–106 Lecture 7 (20 October 98) c David Pollard Page 1 Read M&M §7.1 and §7.2, ignoring starred parts. Reread M&M §3.2. The eects of estimated variances on normal approximations. t-distributions. Comparison of two means: pooling of estimates of variances, or paired observations. In Lecture 6, when discussing comparison of two Binomial proportions, I was content to estimate unknown variances when calculating statistics that were to be treated as approximately normally distributed. You might have worried about the effect of variability of the estimate. W. S. Gosset (“Student”) considered a similar problem in a very famous 1908 paper, where the role of Student’s t-distribution was first recognized. Gosset discovered that the effect of estimated variances could be described exactly in a simplified problem where n independent observations X1,...,Xn are taken from (, ) = ( + ...+ )/ a normal√ distribution, N . The sample mean, X X1 Xn n has a N(, / n) distribution. The random variable X Z = √ / n 2 2 Phas a standard normal distribution. If we estimate by the sample variance, s = ( )2/( ) i Xi X n 1 , then the resulting statistic, X T = √ s/ n no longer has a normal distribution. It has a t-distribution on n 1 degrees of freedom. Remark. I have written T , instead of the t used by M&M page 505. I find it causes confusion that t refers to both the name of the statistic and the name of its distribution. As you will soon see, the estimation of the variance has the effect of spreading out the distribution a little beyond what it would be if were used.
  • On the Meaning and Use of Kurtosis

    On the Meaning and Use of Kurtosis

    Psychological Methods Copyright 1997 by the American Psychological Association, Inc. 1997, Vol. 2, No. 3,292-307 1082-989X/97/$3.00 On the Meaning and Use of Kurtosis Lawrence T. DeCarlo Fordham University For symmetric unimodal distributions, positive kurtosis indicates heavy tails and peakedness relative to the normal distribution, whereas negative kurtosis indicates light tails and flatness. Many textbooks, however, describe or illustrate kurtosis incompletely or incorrectly. In this article, kurtosis is illustrated with well-known distributions, and aspects of its interpretation and misinterpretation are discussed. The role of kurtosis in testing univariate and multivariate normality; as a measure of departures from normality; in issues of robustness, outliers, and bimodality; in generalized tests and estimators, as well as limitations of and alternatives to the kurtosis measure [32, are discussed. It is typically noted in introductory statistics standard deviation. The normal distribution has a kur- courses that distributions can be characterized in tosis of 3, and 132 - 3 is often used so that the refer- terms of central tendency, variability, and shape. With ence normal distribution has a kurtosis of zero (132 - respect to shape, virtually every textbook defines and 3 is sometimes denoted as Y2)- A sample counterpart illustrates skewness. On the other hand, another as- to 132 can be obtained by replacing the population pect of shape, which is kurtosis, is either not discussed moments with the sample moments, which gives or, worse yet, is often described or illustrated incor- rectly. Kurtosis is also frequently not reported in re- ~(X i -- S)4/n search articles, in spite of the fact that virtually every b2 (•(X i - ~')2/n)2' statistical package provides a measure of kurtosis.
  • Volatility Modeling Using the Student's T Distribution

    Volatility Modeling Using the Student's T Distribution

    Volatility Modeling Using the Student’s t Distribution Maria S. Heracleous Dissertation submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Economics Aris Spanos, Chair Richard Ashley Raman Kumar Anya McGuirk Dennis Yang August 29, 2003 Blacksburg, Virginia Keywords: Student’s t Distribution, Multivariate GARCH, VAR, Exchange Rates Copyright 2003, Maria S. Heracleous Volatility Modeling Using the Student’s t Distribution Maria S. Heracleous (ABSTRACT) Over the last twenty years or so the Dynamic Volatility literature has produced a wealth of uni- variateandmultivariateGARCHtypemodels.Whiletheunivariatemodelshavebeenrelatively successful in empirical studies, they suffer from a number of weaknesses, such as unverifiable param- eter restrictions, existence of moment conditions and the retention of Normality. These problems are naturally more acute in the multivariate GARCH type models, which in addition have the problem of overparameterization. This dissertation uses the Student’s t distribution and follows the Probabilistic Reduction (PR) methodology to modify and extend the univariate and multivariate volatility models viewed as alternative to the GARCH models. Its most important advantage is that it gives rise to internally consistent statistical models that do not require ad hoc parameter restrictions unlike the GARCH formulations. Chapters 1 and 2 provide an overview of my dissertation and recent developments in the volatil- ity literature. In Chapter 3 we provide an empirical illustration of the PR approach for modeling univariate volatility. Estimation results suggest that the Student’s t AR model is a parsimonious and statistically adequate representation of exchange rate returns and Dow Jones returns data.
  • Calculating Variance and Standard Deviation

    Calculating Variance and Standard Deviation

    VARIANCE AND STANDARD DEVIATION Recall that the range is the difference between the upper and lower limits of the data. While this is important, it does have one major disadvantage. It does not describe the variation among the variables. For instance, both of these sets of data have the same range, yet their values are definitely different. 90, 90, 90, 98, 90 Range = 8 1, 6, 8, 1, 9, 5 Range = 8 To better describe the variation, we will introduce two other measures of variation—variance and standard deviation (the variance is the square of the standard deviation). These measures tell us how much the actual values differ from the mean. The larger the standard deviation, the more spread out the values. The smaller the standard deviation, the less spread out the values. This measure is particularly helpful to teachers as they try to find whether their students’ scores on a certain test are closely related to the class average. To find the standard deviation of a set of values: a. Find the mean of the data b. Find the difference (deviation) between each of the scores and the mean c. Square each deviation d. Sum the squares e. Dividing by one less than the number of values, find the “mean” of this sum (the variance*) f. Find the square root of the variance (the standard deviation) *Note: In some books, the variance is found by dividing by n. In statistics it is more useful to divide by n -1. EXAMPLE Find the variance and standard deviation of the following scores on an exam: 92, 95, 85, 80, 75, 50 SOLUTION First we find the mean of the data: 92+95+85+80+75+50 477 Mean = = = 79.5 6 6 Then we find the difference between each score and the mean (deviation).
  • Variance Difference Between Maximum Likelihood Estimation Method and Expected a Posteriori Estimation Method Viewed from Number of Test Items

    Variance Difference Between Maximum Likelihood Estimation Method and Expected a Posteriori Estimation Method Viewed from Number of Test Items

    Vol. 11(16), pp. 1579-1589, 23 August, 2016 DOI: 10.5897/ERR2016.2807 Article Number: B2D124860158 ISSN 1990-3839 Educational Research and Reviews Copyright © 2016 Author(s) retain the copyright of this article http://www.academicjournals.org/ERR Full Length Research Paper Variance difference between maximum likelihood estimation method and expected A posteriori estimation method viewed from number of test items Jumailiyah Mahmud*, Muzayanah Sutikno and Dali S. Naga 1Institute of Teaching and Educational Sciences of Mataram, Indonesia. 2State University of Jakarta, Indonesia. 3 University of Tarumanegara, Indonesia. Received 8 April, 2016; Accepted 12 August, 2016 The aim of this study is to determine variance difference between maximum likelihood and expected A posteriori estimation methods viewed from number of test items of aptitude test. The variance presents an accuracy generated by both maximum likelihood and Bayes estimation methods. The test consists of three subtests, each with 40 multiple-choice items of 5 alternatives. The total items are 102 and 3159 respondents which were drawn using random matrix sampling technique, thus 73 items were generated which were qualified based on classical theory and IRT. The study examines 5 hypotheses. The results are variance of the estimation method using MLE is higher than the estimation method using EAP on the test consisting of 25 items with F= 1.602, variance of the estimation method using MLE is higher than the estimation method using EAP on the test consisting of 50 items with F= 1.332, variance of estimation with the test of 50 items is higher than the test of 25 items, and variance of estimation with the test of 50 items is higher than the test of 25 items on EAP method with F=1.329.
  • Analysis of Variance with Summary Statistics in Microsoft Excel

    Analysis of Variance with Summary Statistics in Microsoft Excel

    American Journal of Business Education – April 2010 Volume 3, Number 4 Analysis Of Variance With Summary Statistics In Microsoft Excel David A. Larson, University of South Alabama, USA Ko-Cheng Hsu, University of South Alabama, USA ABSTRACT Students regularly are asked to solve Single Factor Analysis of Variance problems given only the sample summary statistics (number of observations per category, category means, and corresponding category standard deviations). Most undergraduate students today use Excel for data analysis of this type. However, Excel, like all other statistical software packages, requires an input data set in order to invoke its Anova: Single Factor procedure. The purpose of this paper is therefore to provide the student with an Excel macro that, given just the sample summary statistics as input, generates an equivalent underlying data set. This data set can then be used as the required input data set in Excel for Single Factor Analysis of Variance. Keywords: Analysis of Variance, Summary Statistics, Excel Macro INTRODUCTION ost students have migrated to Excel for data analysis because of Excel‟s pervasive accessibility. This means, given just the summary statistics for Analysis of Variance, the Excel user is either limited to solving the problem by hand or to solving the problem using an Excel Add-in. Both of Mthese options have shortcomings. Solving by hand means there is no Excel counterpart solution that puts the entire answer „right there in front of the student‟. Using an Excel Add-in is often not an option because, typically, Excel Add-ins are not available. The purpose of this paper, therefore, is to explain how to accomplish analysis of variance using an equivalent input data set and also to provide the Excel user with a straight-forward macro that accomplishes this technique.
  • The Multivariate Normal Distribution

    The Multivariate Normal Distribution

    Multivariate normal distribution Linear combinations and quadratic forms Marginal and conditional distributions The multivariate normal distribution Patrick Breheny September 2 Patrick Breheny University of Iowa Likelihood Theory (BIOS 7110) 1 / 31 Multivariate normal distribution Linear algebra background Linear combinations and quadratic forms Definition Marginal and conditional distributions Density and MGF Introduction • Today we will introduce the multivariate normal distribution and attempt to discuss its properties in a fairly thorough manner • The multivariate normal distribution is by far the most important multivariate distribution in statistics • It’s important for all the reasons that the one-dimensional Gaussian distribution is important, but even more so in higher dimensions because many distributions that are useful in one dimension do not easily extend to the multivariate case Patrick Breheny University of Iowa Likelihood Theory (BIOS 7110) 2 / 31 Multivariate normal distribution Linear algebra background Linear combinations and quadratic forms Definition Marginal and conditional distributions Density and MGF Inverse • Before we get to the multivariate normal distribution, let’s review some important results from linear algebra that we will use throughout the course, starting with inverses • Definition: The inverse of an n × n matrix A, denoted A−1, −1 −1 is the matrix satisfying AA = A A = In, where In is the n × n identity matrix. • Note: We’re sort of getting ahead of ourselves by saying that −1 −1 A is “the” matrix satisfying
  • Relationship Between Two Variables: Correlation, Covariance and R-Squared

    Relationship Between Two Variables: Correlation, Covariance and R-Squared

    FEEG6017 lecture: Relationship between two variables: correlation, covariance and r-squared Markus Brede [email protected] Relationships between variables • So far we have looked at ways of characterizing the distribution of a single variable, and testing hypotheses about the population based on a sample. • We're now moving on to the ways in which two variables can be examined together. • This comes up a lot in research! Relationships between variables • You might want to know: o To what extent the change in a patient's blood pressure is linked to the dosage level of a drug they've been given. o To what degree the number of plant species in an ecosystem is related to the number of animal species. o Whether temperature affects the rate of a chemical reaction. Relationships between variables • We assume that for each case we have at least two real-valued variables. • For example: both height (cm) and weight (kg) recorded for a group of people. • The standard way to display this is using a dot plot or scatterplot. Positive Relationship Negative Relationship No Relationship Measuring relationships? • We're going to need a way of measuring whether one variable changes when another one does. • Another way of putting it: when we know the value of variable A, how much information do we have about variable B's value? Recap of the one-variable case • Perhaps we can borrow some ideas about the way we characterized variation in the single-variable case. • With one variable, we start out by finding the mean, which is also the expectation of the distribution.
  • Volatility and Kurtosis of Daily Stock Returns at MSE

    Volatility and Kurtosis of Daily Stock Returns at MSE

    Zoran Ivanovski, Toni Stojanovski, and Zoran Narasanov. 2015. Volatility and Kurtosis of Daily Stock Returns at MSE. UTMS Journal of Economics 6 (2): 209–221. Review (accepted August 14, 2015) VOLATILITY AND KURTOSIS OF DAILY STOCK RETURNS AT MSE Zoran Ivanovski1 Toni Stojanovski Zoran Narasanov Abstract Prominent financial stock pricing models are built on assumption that asset returns follow a normal (Gaussian) distribution. However, many authors argue that in the practice stock returns are often characterized by skewness and kurtosis, so we test the existence of the Gaussian distribution of stock returns and calculate the kurtosis of several stocks at the Macedonian Stock Exchange (MSE). Obtaining information about the shape of distribution is an important step for models of pricing risky assets. The daily stock returns at Macedonian Stock Exchange (MSE) are characterized by high volatility and non-Gaussian behaviors as well as they are extremely leptokurtic. The analysis of MSE time series stock returns determine volatility clustering and high kurtosis. The fact that daily stock returns at MSE are not normally distributed put into doubt results that rely heavily on this assumption and have significant implications for portfolio management. We consider this stock market as good representatives of emerging markets. Therefore, we argue that our results are valid for other similar emerging stock markets. Keywords: models, leptokurtic, investment, stocks. Jel Classification: G1; G12 INTRODUCTION For a long time Gaussian models (Brownian motion) were applied in economics and finances and especially to model of stock prices return. However, in the practice, real data for stock prices returns are often characterized by skewness, kurtosis and have heavy tails.
  • The Multivariate Normal Distribution

    The Multivariate Normal Distribution

    The Multivariate Normal Distribution Why should we consider the multivariate normal distribution? It would seem that applied problems are so complex that it would only be interesting from a mathematical perspective. 1. It is mathematically tractable for a large number of problems, and, therefore, progress towards answers to statistical questions can be provided, even if only approximately so. 2. Because it is tractable for so many problems, it provides insight into techniques based upon other distributions or even non-parametric techniques. For this, it is often a benchmark against which other methods are judged. 3. For some problems it serves as a reasonable model of the data. In other instances, transfor- mations can be applied to the set of responses to have the set conform well to multivariate normality. 4. The sampling distribution of many (multivariate) statistics are normal, regardless of the parent distribution (Multivariate Central Limit Theorems). Thus, for large sample sizes, we may be able to make use of results from the multivariate normal distribution to answer our statistical questions, even when the parent distribution is not multivariate normal. Consider first the univariate normal distribution with parameters µ (the mean) and σ (the variance) for the random variable x, 2 1 − 1 (x−µ) f(x)=√ e 2 σ2 (1) 2πσ2 for −∞ <x<∞, −∞ <µ<∞,andσ2 > 0. Now rewrite the exponent (x − µ)2/σ2 using the linear algebra formulation of (x − µ)(σ2)−1(x − µ). This formulation matches that for the generalized or Mahalanobis squared distance (x − µ)Σ−1(x − µ), where both x and µ are vectors.
  • Variance, Covariance, Correlation, Moment-Generating Functions

    Variance, Covariance, Correlation, Moment-Generating Functions

    Math 461 Introduction to Probability A.J. Hildebrand Variance, covariance, correlation, moment-generating functions [In the Ross text, this is covered in Sections 7.4 and 7.7. See also the Chapter Summary on pp. 405–407.] • Variance: – Definition: Var(X) = E(X2) − E(X)2(= E(X − E(X))2) – Properties: Var(c) = 0, Var(cX) = c2 Var(X), Var(X + c) = Var(X) • Covariance: – Definition: Cov(X, Y ) = E(XY ) − E(X)E(Y )(= E(X − E(X))(Y − E(Y ))) – Properties: ∗ Symmetry: Cov(X, Y ) = Cov(Y, X) ∗ Relation to variance: Var(X) = Cov(X, X), Var(X +Y ) = Var(X)+Var(Y )+2 Cov(X, Y ) ∗ Bilinearity: Cov(cX, Y ) = Cov(X, cY ) = c Cov(X, Y ), Cov(X1 + X2,Y ) = Cov(X1,Y ) + Cov(X2,Y ), Cov(X, Y1 + Y2) = Cov(X, Y1) + Cov(X, Y2). Pn Pm Pn Pm ∗ Product formula: Cov( i=1 Xi, j=1 Yj) = i=1 y=1 Cov(Xi,Yj) • Correlation: – Definition: ρ(X, Y ) = √ Cov(X,Y ) Var(X) Var(Y ) – Properties: −1 ≤ ρ(X, Y ) ≤ 1 • Moment-generating function: tX – Definition: M(t) = MX (t) = E(e ) – Computing moments via mgf’s: The derivates of M(t), evaluated at t = 0, give the successive “moments” of a random variable X: M(0) = 1, M 0(0) = E(X), M 00(0) = E(X2), M 000(0) = E(X3), etc. – Special cases: (No need to memorize these formulas.) t2 x ∗ X standard normal: M(t) = exp{ 2 } (where exp(x) = e ) 2 σ2t2 ∗ X normal N(µ, σ ): M(t) = exp{µt + 2 } ∗ X Poisson with parameter λ: M(t) = exp{λ(et − 1)} λ ∗ X exponential with parameter λ: M(t) = λ−t for |t| < λ.
  • Mean-Variance-Skewness Portfolio Performance Gauging: a General Shortage Function and Dual Approach

    Mean-Variance-Skewness Portfolio Performance Gauging: a General Shortage Function and Dual Approach

    EDHEC RISK AND ASSET MANAGEMENT RESEARCH CENTRE 393-400 promenade des Anglais 06202 Nice Cedex 3 Tel.: +33 (0)4 93 18 32 53 E-mail: [email protected] Web: www.edhec-risk.com Mean-Variance-Skewness Portfolio Performance Gauging: A General Shortage Function and Dual Approach September 2005 Walter Briec Maître de conférences, University of Perpignan Kristiaan Kersten Chargé de recherche, CNRS-LABORES, IESEG Octave Jokung Professor, EDHEC Business School Abstract This paper proposes a nonparametric efficiency measurement approach for the static portfolio selection problem in mean-variance-skewness space. A shortage function is defined that looks for possible increases in return and skewness and decreases in variance. Global optimality is guaranteed for the resulting optimal portfolios. We also establish a link to a proper indirect mean-variance-skewness utility function. For computational reasons, the optimal portfolios resulting from this dual approach are only locally optimal. This framework makes it possible to differentiate between portfolio efficiency and allocative efficiency, and a convexity efficiency component related to the difference between the primal, non-convex approach and the dual, convex approach. Furthermore, in principle, information can be retrieved about the revealed risk aversion and prudence of investors. An empirical section on a small sample of assets serves as an illustration. JEL: G11 Keywords: shortage function, efficient frontier, mean-variance-skewness portfolios, risk aversion, prudence. EDHEC is one of the top five business schools in France. Its reputation is built on the high quality of its faculty (110 professors and researchers from France and abroad) and the privileged relationship with professionals that the school has cultivated since its establishment in 1906.