DOCSLIB.ORG

Arxiv:2009.14354V2 [Cond-Mat.Stat-Mech] 2 Jan 2021

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Arxiv:2009.14354V2 [Cond-Mat.Stat-Mech] 2 Jan 2021 Skewed Thermodynamic Geometry and Optimal Free Energy Estimation Steven Blaber∗ and David A. Sivaky Dept. of Physics, Simon Fraser University, Burnaby, British Columbia V5A 1S6, Canada (Dated: January 5, 2021) Free energy differences are a central quantity of interest in physics, chemistry, and biology. We develop design principles that improve the precision and accuracy of free energy estimators, which has potential applications to screening for targeted drug discovery. Specifically, by exploiting the connection between the work statistics of time-reversed protocol pairs, we develop near-equilibrium approximations for moments of the excess work and analyze the dominant contributions to the precision and accuracy of standard nonequilibrium free-energy estimators. Within linear response, minimum-dissipation protocols follow geodesics of the Riemannian metric induced by the Stokes' friction tensor. We find the next-order contribution arises from the rank-3 supra-Stokes' tensor that skews the geometric structure such that minimum-dissipation protocols follow geodesics of a generalized cubic Finsler metric. Thus, near equilibrium the supra-Stokes' tensor determines the leading-order contribution to the bias of bidirectional free-energy estimators. I. INTRODUCTION (equaling the dissipation), and hence the error, can be reduced by following a different path through control- parameter space and varying the velocity along the path Free energy differences determine the equilibrium 14{16 phases of thermodynamic systems as well as the relative while keeping the protocol duration fixed. Near equi- reaction rates and binding affinities of chemical species.1 librium, this can be mapped onto the thermodynamic- Computational and experimental techniques which accu- geometry framework: for the mean-work estimator, the rately and precisely predict free energy differences are bias and variance from finite-time protocols is improved therefore highly desirable. One important application is by following geodesics of a thermodynamic metric, the in pharmaceutical drug discovery, where computation of force-variance (FV) metric, a Riemannian metric de- fined by the covariance matrix of conjugate-force fluctu- free energy differences can aid in the identification and 17{21 design of ligands for targeted protein binding.2{5 Current ations; similarly, for BAR, the variance (but not the methods rely primarily on costly and time-consuming ex- bias) is minimized if the protocol follows geodesics of the FV metric.22 This has been used to improve the precision perimentation, which can be reduced through screening 23{26 with efficient computational techniques.1{8 of calculated binding potentials of mean force. Nonequilbrium measurements of free energy differences The FV metric only minimizes the variance of the are often estimated by measuring the work incurred from mean-work estimator and BAR if the relaxation time a protocol (a dynamic variation in control parameters) is independent of the control parameters. The friction- 27,28 that drives the system between control-parameter end- tensor metric, the product of the covariance and inte- points corresponding to each state. A unidirectional es- gral relaxation time of conjugate-force fluctuations (19), timator calculates the free energy difference using only is more general than the FV and provides a relatively sim- the work done by a protocol driving from initial to final ple prescription for reducing excess work in more general states (forward protocol), while a bidirectional estimator near-equilibrium processes, where minimum-dissipation also uses a protocol that drives from final to initial states protocols follow geodesics of the Riemannian metric in- (reverse protocol). The mean-work estimator9 equates duced by the friction tensor. For common unidirectional the mean work with the free energy difference and yields estimators near equilibrium, the bias and variance are a biased estimate for any non-quasistatic (finite-speed) proportional to the first moment (mean) of the excess protocol. The Jarzynski estimator for free energy differ- work, and hence minimum-dissipation protocols defined ences (derived from the Jarzynski equality relating the by the friction tensor minimize both the bias and vari- exponentially averaged work to the free energy change10) ance. is unbiased for a large number of samples. The mean- We show that for bidirectional estimators, the variance work and Jarzynski estimators can be used as either uni- is proportional to the sum of the second moments of the arXiv:2009.14354v2 [cond-mat.stat-mech] 2 Jan 2021 or bidirectional estimators; however, if bidirectional data excess work from forward and reverse protocols (26a), is available the maximum log-likelihood estimate is Ben- while the bias is proportional to the difference of the nett's acceptance ratio (BAR)11 which yields (for a large first moments (27a).29 With this in mind, we extend the number of samples) the minimum variance of any unbi- thermodynamic-geometry framework to higher-order mo- ased estimator.12,13 ments of the excess work and beyond the leading-order Regardless of the estimator, non-quasistatic control- friction-tensor approximation. We find that for bidi- parameter protocols result in excess work (work in ex- rectional estimators near equilibrium, minimum-variance cess of the free energy difference), which increases the protocols follow geodesics of the Riemannian metric in- bias and variance (decreases the accuracy and precision, duced by the friction tensor, while minimum-bias pro- respectively) of free energy estimates. This excess work tocols follow geodesics of a cubic Finsler metric. For 2 the simple model system of a Brownian particle in a In analogy with fluid dynamics, this rank-two tensor is quadratic trap with time-dependent stiffness (a breathing the Stokes' friction, since it produces a drag force that de- harmonic trap), the minimum-variance and minimum- pends linearly on velocity. We denote the Stokes' friction bias protocols can improve variance by a factor of 3 4 with superscript (1) since it is the leading-order contri- and bias by over a factor of 10 (Fig. 2). − bution to dissipation.27 Following parallel arguments, we approximate the third moment of the excess-work distribution as II. DERIVATION 3 d W Λ 3 (2) ex _ _ _ h i 2 ζijk[λ(t)]λi(t)λj(t)λk(t) ; (6) Consider a system in the canonical ensemble, in dt ≈ β thermal equilibrium with a heat bath at temper- for the rank-three tensor ature T . The probability distribution over mi- crostates x at control parameters λ is π(x λ) = ζ(2)[λ(t)] (7) β F λ E x; λ E x; λ j ijk ≡ exp [ ( ) ( )], with energy ( ) and free en- Z 1 Z 1 − P 2 00 000 00 000 ergy F (λ) kBT ln x exp [ βE(x; λ)]. Here β β dt dt δf (0)δf (t )δf (t ) : −1 ≡ − − ≡ i j k λ(t) (kBT ) for Boltzmann's constant kB. We define the − 0 0 h i work done in a single realization of an external agent changing the control parameters λ according to a proto- (The factor of three in (6) results from grouping the in- R t 0 _ 0 dex permutations ijk; jik; kij into one term, using the col Λ as W dt fiλi(t ), which implies the average f g ≡ − 0 invariance of the sum under exchange of indices, e.g., excess work (Wex W ∆F ) is (2) (2) _ _ _ _ _ _ ≡ − ζijkλiλjλk = ζjikλiλjλk.) We call the rank-three tensor Z t [Eq. (7)] the supra-Stokes' tensor and index it by super- 0 _ 0 Wex Λ = dt δfi Λλi(t ) ; (1) script (2), as it corresponds to the leading-order correc- h i − 0 h i tion to dissipation beyond the Stokes' friction (15). where we adopt the Einstein summation convention of For fourth and higher moments, Appendix B shows implied summation over repeated indices. A dot denotes that the integration bounds must remain finite since the the time derivative λ_ i dλi=dt, fi @λ U is the gen- n-time covariance functions do not decay to zero, so there ≡ ≡ − i eralized force conjugate to λi, and δfi fi fi eq is the is no clear analogy to frictional drag forces. Nevertheless, difference from the equilibrium average.≡ Angle−h i brackets we can still approximate them by Λ denote an average over the nonequilibrium ensem- h· · · i n n ble of system responses to control-parameter protocol Λ. d Wex Λ (n−1) Y h i n ν ···ν [λ(t); t] λ_ ν (t) ; (8) The time derivative of the second moment of the dt ≈ C 1 n i excess-work distribution is i=1 where we use index notation ν ; ν ; ν instead of i; j; k, d W 2 Z t 1 2 3 ex Λ _ 0 0 _ 0 and define integral n-time covariance··· functions h i = 2λi(t) dt δfi(t)δfj(t ) Λ λj(t ) : (2) dt 0 h i (n−1) ν ···ν [λ(t); t] (9) For a sufficiently slow protocol, we replace the nonequi- C 1 n ≡ n t * n + librium average Λ with the equilibrium average Y Z Y h· · · i ( β)n dt δf (0)δf (t ) : λ(t) at fixed control parameters λ(t): i ν1 νj j − 0 h· · · i i=2 j=2 λ(t) 2 Z t d Wex Λ 00 00 h i 2λ_ i(t)λ_ j(t) dt δfi(0)δfj(t ) ; (3) This implies that higher-order moments of the excess dt ≈ h iλ(t) 0 work are higher order in control-parameter velocity and where we used the stationarity of the equilibrium aver- are therefore smaller for slow protocols. The approxima- age, defined t00 t0 t, and assumed smooth protocols tions of (4), (6), and (8) are the leading-order contribu- to expand the control-parameter≡ − velocity to zeroth or- tions to each moment of the excess work.
Load more

Recommended publications
  • Basic Statistics Objective
    Basic Statistics Objective Interpret and apply the mean, standard deviation, and variance of a random variable. Calculate the mean, standard deviation, and variance of a discrete random variable. Interpret and calculate the expected value of a discrete random variable. Calculate and interpret the covariance and correlation between two random variables. Calculate the mean and variance of sums of variables. Describe the four central moments of a statistical variable or distribution: mean, variance, skewness, and kurtosis. Interpret the skewness and kurtosis of a statistical distribution, and interpret the concepts of coskewness and cokurtosis. Describe and interpret the best linear unbiased estimator. riskmacro.com 3 Basic Statistics ➢ Introduction: ➢The word statistic is used to refers to data and the methods we use to analyze data. ➢Descriptive statistics are used to summarize the important characteristics of large data sets. ➢Inferential statistic, pertain to the procedures used to make forecasts, estimates, or judgments about a large set of data on the basis of the statistical characteristics of a smaller set ( a sample). ➢A population is defined as the set of all possible members of a stated group. ➢Measures of central tendency identify the center, or average, of a data set. Median Mode Population Mean Sample Mean Example: Calculate Mean, Mode & Median from the following data set: 12%, 25%, 34%, 15%, 19%, 44%, 54%, 34%, 22%, 28%, 17%, 24%. riskmacro.com 4 Basic Statistics ➢ Introduction: Geometric Mean of Returns Example: Geometric mean return For the last three years, the returns for Acme Corporation common stock have been -9.34%, 23.45%, and 8.92%. compute the compound rate of return over the 3-year period.
    [Show full text]
  • Idiosyncratic Coskewness and Equity Return Anomalies by Fousseni Chabi-Yo and Jun Yang
    Working Paper/Document de travail 2010-11 Idiosyncratic Coskewness and Equity Return Anomalies by Fousseni Chabi-Yo and Jun Yang Bank of Canada Working Paper 2010-11 May 2010 Idiosyncratic Coskewness and Equity Return Anomalies by Fousseni Chabi-Yo1 and Jun Yang2 1Fisher College of Business Ohio State University Columbus, Ohio, U.S.A. 43210-1144 [email protected] 2Financial Markets Department Bank of Canada Ottawa, Ontario, Canada K1A 0G9 [email protected] Bank of Canada working papers are theoretical or empirical works-in-progress on subjects in economics and finance. The views expressed in this paper are those of the authors. No responsibility for them should be attributed to the Bank of Canada. 2 ISSN 1701-9397 © 2010 Bank of Canada Acknowledgements We are grateful to Bill Bobey, Oliver Boguth, Jean-Sébastien Fontaine, Scott Hendry, Kewei Hou, Michael Lemmon, Jesus Sierra-Jiménez, René Stulz, Ingrid Werner, and seminar participants at the Bank of Canada, the Ohio State University, and the Northern Finance Association 2009 conference. We thank Kenneth French for making a large amount of historical data publicly available in his online data library. We welcome comments, including references to related papers we have inadvertently overlooked. Fousseni Chabi-Yo would like to thank the Dice Center for Financial Economics for financial support. ii Abstract In this paper, we show that in a model where investors have heterogeneous preferences, the expected return of risky assets depends on the idiosyncratic coskewness beta, which measures the co-movement of the individual stock variance and the market return. We find that there is a negative (positive) relation between idiosyncratic coskewness and equity returns when idiosyncratic coskewness betas are positive (negative).
    [Show full text]
  • The Matrix Cookbook
    The Matrix Cookbook Kaare Brandt Petersen Michael Syskind Pedersen Version: February 16, 2006 What is this? These pages are a collection of facts (identities, approxima- tions, inequalities, relations, ...) about matrices and matters relating to them. It is collected in this form for the convenience of anyone who wants a quick desktop reference . Disclaimer: The identities, approximations and relations presented here were obviously not invented but collected, borrowed and copied from a large amount of sources. These sources include similar but shorter notes found on the internet and appendices in books - see the references for a full list. Errors: Very likely there are errors, typos, and mistakes for which we apolo- gize and would be grateful to receive corrections at [email protected]. Its ongoing: The project of keeping a large repository of relations involving matrices is naturally ongoing and the version will be apparent from the date in the header. Suggestions: Your suggestion for additional content or elaboration of some topics is most welcome at [email protected]. Keywords: Matrix algebra, matrix relations, matrix identities, derivative of determinant, derivative of inverse matrix, differentiate a matrix. Acknowledgements: We would like to thank the following for contribu- tions and suggestions: Christian Rishøj, Douglas L. Theobald, Esben Hoegh- Rasmussen, Lars Christiansen, and Vasile Sima. We would also like thank The Oticon Foundation for funding our PhD studies. 1 CONTENTS CONTENTS Contents 1 Basics 5 1.1 Trace and Determinants . 5 1.2 The Special Case 2x2 . 5 2 Derivatives 7 2.1 Derivatives of a Determinant . 7 2.2 Derivatives of an Inverse .
    [Show full text]
  • The Matrix Cookbook [ ]
    The Matrix Cookbook [ http://matrixcookbook.com ] Kaare Brandt Petersen Michael Syskind Pedersen Version: November 14, 2008 What is this? These pages are a collection of facts (identities, approxima- tions, inequalities, relations, ...) about matrices and matters relating to them. It is collected in this form for the convenience of anyone who wants a quick desktop reference . Disclaimer: The identities, approximations and relations presented here were obviously not invented but collected, borrowed and copied from a large amount of sources. These sources include similar but shorter notes found on the internet and appendices in books - see the references for a full list. Errors: Very likely there are errors, typos, and mistakes for which we apolo- gize and would be grateful to receive corrections at [email protected]. Its ongoing: The project of keeping a large repository of relations involving matrices is naturally ongoing and the version will be apparent from the date in the header. Suggestions: Your suggestion for additional content or elaboration of some topics is most welcome at [email protected]. Keywords: Matrix algebra, matrix relations, matrix identities, derivative of determinant, derivative of inverse matrix, differentiate a matrix. Acknowledgements: We would like to thank the following for contributions and suggestions: Bill Baxter, Brian Templeton, Christian Rishøj, Christian Schr¨oppel Douglas L. Theobald, Esben Hoegh-Rasmussen, Glynne Casteel, Jan Larsen, Jun Bin Gao, J¨urgen Struckmeier, Kamil Dedecius, Korbinian Strim- mer, Lars Christiansen, Lars Kai Hansen, Leland Wilkinson, Liguo He, Loic Thibaut, Miguel Bar˜ao,Ole Winther, Pavel Sakov, Stephan Hattinger, Vasile Sima, Vincent Rabaud, Zhaoshui He. We would also like thank The Oticon Foundation for funding our PhD studies.
    [Show full text]
  • Conditional Skewness in Asset Pricing Tests
    THE JOURNAL OF FINANCE • VOL. LV, NO. 3 • JUNE 2000 Conditional Skewness in Asset Pricing Tests CAMPBELL R. HARVEY and AKHTAR SIDDIQUE* ABSTRACT If asset returns have systematic skewness, expected returns should include re- wards for accepting this risk. We formalize this intuition with an asset pricing model that incorporates conditional skewness. Our results show that conditional skewness helps explain the cross-sectional variation of expected returns across assets and is significant even when factors based on size and book-to-market are included. Systematic skewness is economically important and commands a risk premium, on average, of 3.60 percent per year. Our results suggest that the mo- mentum effect is related to systematic skewness. The low expected return momen- tum portfolios have higher skewness than high expected return portfolios. THE SINGLE FACTOR CAPITAL ASSET PRICING MODEL ~CAPM! of Sharpe ~1964! and Lintner ~1965! has come under recent scrutiny. Tests indicate that the cross- asset variation in expected returns cannot be explained by the market beta alone. For example, a growing number of studies show that “fundamental” variables such as size, book-to-market value, and price to earnings ratios account for a sizeable portion of the cross-sectional variation in expected returns ~see, e.g., Chan, Hamao, and Lakonishok ~1991! and Fama and French ~1992!!. Fama and French ~1995! document the importance of SMB ~the dif- ference between the return on a portfolio of small size stocks and the return on a portfolio of large size stocks! and HML ~the difference between the return on a portfolio of high book-to-market value stocks and the return on a portfolio of low book-to-market value stocks!.
    [Show full text]
  • Downside Risk and the Momentum Effect
    NBER WORKING PAPER SERIES DOWNSIDE RISK AND THE MOMENTUM EFFECT Andrew Ang Joseph Chen Yuhang Xing Working Paper 8643 http://www.nber.org/papers/w8643 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge, MA 02138 December 2001 The authors would like to thank Brad Barber, Alon Brav, Geert Bekaert, John Cochrane, Randy Cohen, Kent Daniel, Bob Dittmar, Cam Harvey, David Hirschleiffer, Qing Li, Terence Lim, Bob Stambaugh, Akhtar Siddique and Zhenyu Wang for insightful discussions. We especially thank Bob Hodrick for detailed comments. We thank seminar participants at Columbia University and USC for helpful comments. This paper is funded by a Q-Group research grant. The views expressed herein are those of the authors and not necessarily those of the National Bureau of Economic Research. © 2001 by Andrew Ang, Joseph Chen and Yuhang Xing. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit, including © notice, is given to the source. Downside Risk and the Momentum Effect Andrew Ang, Joseph Chen and Yuhang Xing NBER Working Paper No. 8643 December 2001 JEL No. C12, C15, C32, G12 ABSTRACT Stocks with greater downside risk, which is measured by higher correlations conditional on downside moves of the market, have higher returns. After controlling for the market beta, the size effect and the book-to-market effect, the average rate of return on stocks with the greatest downside risk exceeds the average rate of return on stocks with the least downside risk by 6.55% per annum. Downside risk is important for explaining the cross-section of expected returns.
    [Show full text]
  • Skewness and Cokurtosis
    Is RMB a Safe Haven Currency? Evidence from Conditional Coskewness and Cokurtosis • Xin Cheng1 , Hongyi Chen2, Yinggang Zhou1* • Xiamen University, SOE & WISE • Hong Kong Monetary Authority Safe Haven Currencies Habib and Stracca (2012) •Safe-haven currencies are good hedge against financial stress •Traditional safe haven currencies: US dollar, Japanese Yen. Motivation US dollar’s performance in the event of the crisis on Sep. 2008 Motivation Japanese Yen’s performance in the event of the crisis on Sep. 2008 Motivation RMB’s performance in the event of the crisis on Sep. 2008 Motivation Campbell, Medeiros and Viceira (2010) •US dollar, Euro and Swiss franc moved against world equity markets •Their findings are based on correlation only Co-movement is not just correlation •Correlation is linear co-movement Asset returns have long (usually left tail) and fat tails, especially in the case of extreme events •Coskewness and Cokurtosis is nonlinear co- movement Motivation Investors have preferences beyond mean & variance •Skewness preference is about “prudence” (Kimball, 1990) •A prudent investor will seek higher (positive) skewness (Rubinstein, 1973). •Kurtosis preference is about “Temperance” (Denuit and Eeckhoudt, 2010) •A temperate investor will seek lower (negative) kurtosis (Kraus and Litzenberger, 1976) Motivation Higher-order-moment CAPM for stock market •Rubinstein (1973) , Kraus and Litzenberger (1976) •Harvey and Siddque (2000), Dittmar (2002) Recent renewed interest and extension to (international) stock, bond and/or option markets
    [Show full text]
  • Coskewness Risk Decomposition
    Coskewness Risk Decomposition Leon Zolotoy Melbourne Business School, University of Melbourne Petko S. Kalev Centre for Applied Financial Studies, University of South Australia Business School This draft 10 September 2013 Abstract This paper investigates the relative importance of aggregate cash–flow and discount–rate news components in the context of higher–moment asset pricing model of Kraus and Litzenberger (1976) which includes coskewness risk. We break down coskewness of the stock with the market into the coskewness with the cash–flow news, coskewness with the discount–rate news and the covariance with the product of the two — the covariation risk. Using US data from 1963 to 2010, we observe that stocks with high past returns have higher exposure to the covariation risk in comparison to the stocks with low past returns. We also document positive and economically significant market price of the covariation risk, while no such evidence is found for the cash–flow and discount–rate components of the coskewness risk. When augmented with the covariation risk component, the two–beta model of Campbell and Vuolteenaho (2004a) captures about 73% of the return variation across size/book–to– market, size/momentum and industry–sorted portfolios. In addition, we complement our analysis by examining the effects of the coskewness component on the expected rate of return of individual stocks using the implied cost of equity capital approach. Once again, we find the covariation risk to be the dominant factor that drives the positive relation between the expected returns and coskewness risk. JEL classification: G12, G14 Keywords: Asset pricing; Coskewness risk; Cash–flow news; Discount–rate news; Momentum.
    [Show full text]
  • Portfolio Selection with Higher Moments
    Portfolio Selection With Higher Moments Campbell R. Harvey∗ Duke University, Durham, NC 27708 USA John C. Liechty Pennsylvania State University, University Park, PA 16803 USA Merrill W. Liechty Drexel University, Philadelphia, PA 19104 USA Peter M¨uller University of Texas M.D. Anderson Cancer Center, Houston, TX 77030 USA Abstract We propose a method for optimal portfolio selection using a Bayesian decision the- oretic framework that addresses two major shortcomings of the Markowitz approach: the ability to handle higher moments and estimation error. We employ the skew normal distribution which has many attractive features for modeling multivariate returns. Our results suggest that it is important to incorporate higher order moments in portfolio selection. Further, our comparison to other methods where parameter uncertainty is either ignored or accommodated in an ad hoc way, shows that our approach leads to higher expected utility than the resampling methods that are common in the practice of finance. Keywords: Bayesian decision problem, multivariate skewness, parameter uncer- tainty, optimal portfolios, utility function maximization. ∗Campbell R. Harvery is J. Paul Sticht Professor of International Business, Fuqua School of Business, Duke University, Durham NC 27708 and Research Associate, National Bureau of Economic Research, Cam- bridge, MA 02138 (E-mail: [email protected]). John C. Liechty is Assistant Professor, Department of Marketing, Smeal College of Business, and Department of Statistics, Pennsylvania State University, Uni- versity Park, PA 16803 (E-mail: [email protected]). Merrill W. Liechty is Assistant Professor, Department of Decision Sciences, LeBow College of Business, Drexel University, Philadelphia, PA 19104 (E-mail: mer- [email protected]).
    [Show full text]
  • Magisterarbeit
    Magisterarbeit Titel der Magisterarbeit “A Discussion of Data Enhancement and Optimization Techniques for a Fund of Hedge Funds Portfolio” Verfasser Wolfgang Kappel, bakk.rer.soc.oec. angestrebter akademischer Grad Magister (Mag.rer.soc.oec.) Wien, im August 2008 Studienkennzahl lt. Studienblatt: A 066 915 Studienrichtung lt. Studienblatt: Magisterstudium Betriebswirtschaft Betreuer: Univ.-Prof. Dr. Dr.h.c. Josef Zechner 1 2 Acknowledgements The topic of this thesis was motivated by my colleagues at Benchmark Capital Management. I want thank them for the helpful suggestions, discussions and checks for plausibility, especially Karin Kisling, Florian Haberfelner and Olivier Chevaux. Important advice came from Prof. Josef Zechner, Prof. Thomas Dangl, Michael Halling and Arne Westerkamp. With Ronald Hochreiter I had a lot of very interesting and creative discussions. Thanks a lot! Many thanks go to my family and friends for support, especially to my mother, Simona, Sandra and Gerhard. 3 4 Inhaltsverzeichnis Acknowledgements ............................................................................................................................. 2 1 Introduction ................................................................................................................................... 10 1.1 German Abstract ................................................................................................................... 10 1.2 English Abstract ....................................................................................................................
    [Show full text]
  • Performance Evaluation Considring the Coskewness. a Stochastic
    s ai Performancewww emeraldinsight com/ 07-4358.htm evaluation considering the coskewness A stochastic discount factor framework David Moreno and Rosa Rodrı´guez Dpto. Economı´a de la Empresa, Universidad Carlos III de Madrid, Madrid, Spain Abstract Purpose The paper aims to examine the performance of Spanish mutual funds between 1999 and 2003. Design/methodology/approach The methodolgy uses the stochastic discount factor (SDF) framework across a variety of models developed in the recent asset pricing literature. This approach is a fairly recent innovation in the evaluation of investment performance. Findings The present work complements the research of Farnworth et al. and Fletcher and Forbes, adding a new issue to the SDF, the third co moment of asset returns. Recent asset pricing studies show the relevance of the component of an asset’s skewness related to the market portfolio’s skewness, the coskewness, and how it helps to explain the time variation of ex ante market risk premiums. It is found that the effects of adding coskewness to evaluate the performance is significant even when factors based on size, book to market and momentum are included. Practical implications The omission of a coskewness factor may lead to erroneous evaluations of a fund’s performance, and therefore, issues such as the persistence of performance should be revised. Originality/value This paper explores, for the first time, the effects of incorporating a coskewness factor in the analysis of investment performance, both in an unconditional and a conditional framework using SDF models. Keywords Spain, Fund management, Capital asset pricing model Paper type Research paper 1.
    [Show full text]
  • Default Risk, Idiosyncratic Coskewness and Equity Returns1
    Default Risk, Idiosyncratic Coskewness and Equity Returns1 Fousseni Chabi-Yo2 Ohio State University Jun Yang3 Bank of Canada October 13, 2009 Abstract In this paper, we intend to explain an empirical ¯nding that distressed stocks delivered anomalously low returns. We show that in a model with heterogeneous investors where idiosyncratic skewness is priced, the expected return of risky assets depends on idiosyn- cratic coskewness betas, which measure the covariance between idiosyncratic variance and the market return. We ¯nd that there is a negative (positive) relation between idiosyncratic coskewness and equity returns when idiosyncratic coskewness betas are positive (negative). We construct two idiosyncratic coskewness factors to capture market-wide e®ect of idiosyn- cratic coskewness betas. When we control for these two idiosyncratic coskewness factors, the return di®erence for distress-sorted portfolios becomes insigni¯cant. High stressed ¯rms earn low returns because high stressed ¯rms have high (low) idiosyncratic coskewness betas when idiosyncratic coskewness betas are positive (negative). JEL Classi¯cation Codes: G11, G12, G14, G33 Keywords: Financial distress, Higher moment returns, Idiosyncratic coskewness 1We are grateful to Bill Bobey, Oliver Boguth, Kewei Hou, Michael Lemmon, Rene Stulz, Ingrid Werner, and seminar participants at the Bank of Canada, The Ohio State University, and the Northern Finance Association 2009 conference. We thank Kenneth French for making a large amount of historical data publicly available in his online data library. We welcome comments, including references to related papers we have inadvertently overlooked. Fousseni Chabi-Yo would like to thank the Dice Center for Financial Economics for ¯nancial support. 2Fisher College of Business, Ohio State University, 840 Fisher Hall, 2100 Neil Avenue, Columbus, OH 43210{1144.
    [Show full text]