% COPYRIGHT NOTICE:
% 
% CAMPBELL, et al.: The Economics of Financial Markets
% is published by Princeton University Press and copyrighted, (c) 1996, by 
% Princeton University Press. All rights reserved. This text may be used 
% and shared in accordance with the fair-use provisions of US copyright law, 
% and it may be archived and redistributed in electronic form, provided that 
% this entire notice is carried and provided that Princeton University Press 
% is notified and no fee is charged for access. Archiving, redistribution, 
% or republication of this text on other terms, in any medium, requires the 
% consent of Princeton University Press.
% 
% For COURSE PACK PERMISSIONS, refer to entry on previous menu. For 
% more information, send e-mail to permissions@pupress.princeton.edu
%
\documentstyle[chapbib,clm,newbas,campbell,rotate,psfig]{book}

\makeatother
\catcode`@=11
\def\@pnumwidth{1.55em}
\def\@tocrmarg {3.55em}
\def\@dotsep{4.5}
\def\@dottedtocline#1#2#3#4{
  \vskip \z@ plus .2pt
  {\hangindent #1\relax \rightskip \@tocrmarg \parfillskip -\rightskip
    \parindent #1\relax
   \interlinepenalty\@M
   \leavevmode 
    #3\nobreak\leaders\hbox{$\m@th \mkern \@dotsep mu.\mkern \@dotsep 
       mu$}\hfill \nobreak \hbox to\@pnumwidth{\hfil\rm #4}\par}}
\def\l@section{\@dottedtocline{1.5em}{2.3em}}
\def\l@figure{\@dottedtocline{0em}{2.3em}}
\def\l@table{\@dottedtocline{0em}{2.3em}}
\def\l@subsection{\@dottedtocline{4.5em}{2.3em}}
\def\l@subsubsection{\@dottedtocline{1.5em}{2.3em}}
%\def\l@subsection#1#2{}
\def\l@subsubsection#1#2{}
\def\contentsline#1{\csname l@#1\endcsname}
\def\l@chapter#1#2{
 \vskip 1.0em plus 1pt \begingroup\let\numberline\snumberline
 \parindent \z@ \rightskip \@pnumwidth 
 \parfillskip -\@pnumwidth 
 \bf \leavevmode #1\hfil \hbox to\@pnumwidth{\hss #2}\par
 \endgroup}
\def\l@part#1#2{
 \vskip 1.0em plus 1pt \begingroup
 \parindent \z@ \rightskip \@pnumwidth 
 \parfillskip -\@pnumwidth 
 \bf \leavevmode \hfil PART \uppercase{#1}\hfil \hbox to\@pnumwidth{\hss #2}\par
 \endgroup}
\def\l@app#1#2{
 \vskip 1.0em plus 1pt \begingroup
 \parindent \z@ \rightskip \@pnumwidth 
 \parfillskip -\@pnumwidth 
 \bf \leavevmode \hfil\uppercase{#1}\hfil \hbox to\@pnumwidth{\hss #2}\par
 \endgroup}
\def\snumberline#1{\advance\hangindent 1.5em \hbox to 1.5em{#1\hfil}}
\def\numberline#1{\advance\hangindent 3em \hbox to 3em{#1\hfil}}
\catcode`@=12 
\makeatletter
\parindent=0pt
\font\elevenbf=NewBasBol at 11pt

\begin{document}

\chapter*{\normalsize\rm The Econometrics of Financial Markets}\thispagestyle{empty}

\newpage
\setcounter{page}{5}
\chapter*{The Econometrics of Financial Markets}\thispagestyle{empty}
\medskip
$\mbox{}$\hfill {\elevenbf John Y.~Campbell}\\[8pt]
%$\mbox{}$\hfill {\normalsize\it Department of Economics}\\
%$\mbox{}$\hfill {\normalsize\it Harvard University}\\[10pt]
$\mbox{}$\hfill {\elevenbf Andrew W.~Lo}\\[8pt]
%$\mbox{}$\hfill {\normalsize\it Sloan School of Management}\\
%$\mbox{}$\hfill {\normalsize\it Massachusetts Institute of Technology}\\[10pt]
$\mbox{}$\hfill {\elevenbf A.~Craig MacKinlay}\\[8pt]
%$\mbox{}$\hfill {\normalsize\it Department of Finance}\\
%$\mbox{}$\hfill {\normalsize\it Wharton School}\\
%$\mbox{}$\hfill {\normalsize\it University of Pennsylvania}\\
% added by lori in revisions

\vfill

{\elevenbf Princeton University Press}\hfill $\hbox{}$

\vskip 2pt

{\elevenbf Princeton, New Jersey}\hfill $\hbox{}$


\newpage

{\thispagestyle{empty}
\footnotesize 

$\mbox{$\,$}$

\vspace*{7.25pc}

\vbox{
\noindent Copyright \copyright\ 1997 by Princeton University Press 

\noindent Published by Princeton University Press, 41 William Street,

\noindent Princeton, New Jersey 08540

\noindent In the United Kingdom: Princeton University Press, Chichester,\\
West Sussex

\vskip 12pt

\noindent All Rights Reserved

\vskip 12pt

\noindent {\bf Library of Congress Cataloging-in-Publication Data}

\vskip 12pt

\ifnum 1=0
\vbox to 74pt{
\vfil
\centerline{\bf [CIP TO COME]}
\vfil
}
\else
\noindent Campbell, John Y.

\noindent \hspace*{2pc} The econometrics of financial markets / John Y.\ Campbell, Andrew\par
 W.\ Lo, A. Craig MacKinlay.

\noindent \hspace*{3pc} p.~~~~cm.

\noindent \hspace*{2pc} Includes bibliographical references and index.

\noindent \hspace*{2pc} ISBN 0-691-04301-9 (cloth~~:~~alk.\ paper)

\noindent \hspace*{2pc} 1.~~Capital market---Econometric models. ~~I.~~Lo, Andrew W. (Andrew\par
Wen-Chuan).~~II.~~MacKinlay, Archie Craig, 1955--~~~~~.~~III.~~Title.\par

\noindent  HG4523.C27~~~1997

\noindent  332$'$.09414---dc20\hfill 96--27868

% $\mbox{ }$\hfill CIP~~
\fi

\vskip 30pt

\noindent This book was composed in ITC New Baskerville with \LaTeX\ by Archetype Publishing
Inc.,\\ 15 Turtle Pointe Road, Monticello, IL 61856.
\vskip 12pt

\noindent Princeton University Press books are printed on acid-free paper and
meet the guidelines for permanence and durability of the Committee on Production 
Guidelines for Book Longevity of the Council on Library Resources.

\vskip 12pt

\noindent Printed in the United States of America

\vskip 12pt

\noindent 10\hskip 1em9\hskip 1em8\hskip 1em7\hskip 1em6\hskip 1em5\hskip 1em4\hskip 1em3\hskip 1em2\hskip 1em1
}
}

\chapter*{{\normalsize\it To Susanna, Nancy, and Tina}}
\thispagestyle{empty}
\newpage
$\mbox{}$
\thispagestyle{empty}
\chapter*{Contents}\markboth{Contents}{Contents}
\pagenumbering{roman}
\setcounter{page}{7}
\setcounter{tocdepth}{3}

\contentsline{chapter}{\numberline{}List of Figures}{xiii	}
\contentsline{chapter}{\numberline{}List of Tables}{xv}
\contentsline{chapter}{\numberline{}Preface}{xix}
\contentsline{chapter}{\numberline{1}Introduction}{3}
\contentsline{section}{\numberline{1.1}Organization of the Book}{4}
\contentsline{section}{\numberline{1.2}Useful Background}{6}
\contentsline{subsection}{\numberline{1.2.1}Mathematics Background}{6}
\contentsline{subsection}{\numberline{1.2.2}Probability and Statistics Background}{6}
\contentsline{subsection}{\numberline{1.2.3}Finance Theory Background}{7}
\contentsline{section}{\numberline{1.3}Notation}{8}
\contentsline{section}{\numberline{1.4}Prices, Returns, and Compounding}{9}
\contentsline{subsection}{\numberline{1.4.1}Definitions and Conventions}{9}
\contentsline{subsection}{\numberline{1.4.2}The Marginal, Conditional, and Joint Distribution of Returns}{13}
\contentsline{section}{\numberline{1.5}Market Efficiency}{20}
\contentsline{subsection}{\numberline{1.5.1}Efficient Markets and the Law of Iterated\protect\\ Expectations}{22}
\contentsline{subsection}{\numberline{1.5.2}Is Market Efficiency Testable?}{24}
\contentsline{chapter}{\numberline{2}The Predictability of Asset Returns}{27}
\contentsline{section}{\numberline{2.1}The Random Walk Hypotheses}{28}
\contentsline{subsection}{\numberline{2.1.1}The Random Walk 1: IID Increments}{31}
\contentsline{subsection}{\numberline{2.1.2}The Random Walk 2: Independent Increments}{32}
\contentsline{subsection}{\numberline{2.1.3}The Random Walk 3: Uncorrelated Increments}{33}
\contentsline{section}{\numberline{2.2}Tests of Random Walk 1: IID Increments}{33}
\contentsline{subsection}{\numberline{2.2.1}Traditional Statistical Tests}{33}
\contentsline{subsection}{\numberline{2.2.2}Sequences and Reversals, and Runs}{34}
\contentsline{section}{\numberline{2.3}Tests of Random Walk 2: Independent Increments}{41}
\contentsline{subsection}{\numberline{2.3.1}Filter Rules}{42}
\contentsline{subsection}{\numberline{2.3.2}Technical Analysis}{43}
\contentsline{section}{\numberline{2.4}Tests of Random Walk 3: Uncorrelated Increments}{44}
\contentsline{subsection}{\numberline{2.4.1}Autocorrelation Coefficients}{44}
\contentsline{subsection}{\numberline{2.4.2}Portmanteau Statistics}{47}
\contentsline{subsection}{\numberline{2.4.3}Variance Ratios}{48}
\contentsline{section}{\numberline{2.5}Long-Horizon Returns}{55}
\contentsline{subsection}{\numberline{2.5.1}Problems with Long-Horizon Inferences}{57}
\contentsline{section}{\numberline{2.6}Tests For Long-Range Dependence}{59}
\contentsline{subsection}{\numberline{2.6.1}Examples of Long-Range Dependence}{59}
\contentsline{subsection}{\numberline{2.6.2}The Hurst-Mandelbrot Rescaled Range Statistic}{62}
\contentsline{section}{\numberline{2.7}Unit Root Tests}{64}
\contentsline{section}{\numberline{2.8}Recent Empirical Evidence}{65}
\contentsline{subsection}{\numberline{2.8.1}Autocorrelations}{66}
\contentsline{subsection}{\numberline{2.8.2}Variance Ratios}{68}
\contentsline{subsection}{\numberline{2.8.3}Cross-Autocorrelations and Lead-Lag Relations}{74}
\contentsline{subsection}{\numberline{2.8.4}Tests Using Long-Horizon Returns}{78}
\contentsline{section}{\numberline{2.9}Conclusion}{80}
\contentsline{chapter}{\numberline{3}Market Microstructure}{83}
\contentsline{section}{\numberline{3.1}Nonsynchronous Trading}{84}
\contentsline{subsection}{\numberline{3.1.1}A Model of Nonsynchronous Trading}{85}
\contentsline{subsection}{\numberline{3.1.2}Extensions and Generalizations}{98}
\contentsline{section}{\numberline{3.2}The Bid-Ask Spread}{99}
\contentsline{subsection}{\numberline{3.2.1}Bid-Ask Bounce}{101}
\contentsline{subsection}{\numberline{3.2.2}Components of the Bid-Ask Spread}{103}
\contentsline{section}{\numberline{3.3}Modeling Transactions Data}{107}
\contentsline{subsection}{\numberline{3.3.1}Motivation}{108}
\contentsline{subsection}{\numberline{3.3.2}Rounding and Barrier Models}{114}
\contentsline{subsection}{\numberline{3.3.3}The Ordered Probit Model}{122}
\contentsline{section}{\numberline{3.4}Recent Empirical Findings}{128}
\contentsline{subsection}{\numberline{3.4.1}Nonsynchronous Trading}{128}
\contentsline{subsection}{\numberline{3.4.2}Estimating the Effective Bid-Ask Spread}{134}
\contentsline{subsection}{\numberline{3.4.3}Transactions Data}{136}
\contentsline{section}{\numberline{3.5}Conclusion}{144}
\contentsline{chapter}{\numberline{4}Event-Study Analysis}{149}
\contentsline{section}{\numberline{4.1}Outline of an Event Study}{150}
\contentsline{section}{\numberline{4.2}An Example of an Event Study}{152}
\contentsline{section}{\numberline{4.3}Models for Measuring Normal Performance}{153}
\contentsline{subsection}{\numberline{4.3.1}Constant-Mean-Return Model}{154}
\contentsline{subsection}{\numberline{4.3.2}Market Model}{155}
\contentsline{subsection}{\numberline{4.3.3}Other Statistical Models}{155}
\contentsline{subsection}{\numberline{4.3.4}Economic Models}{156}
\contentsline{section}{\numberline{4.4}Measuring and Analyzing Abnormal Returns}{157}
\contentsline{subsection}{\numberline{4.4.1}Estimation of the Market Model}{158}
\contentsline{subsection}{\numberline{4.4.2}Statistical Properties of Abnormal Returns}{159}
\contentsline{subsection}{\numberline{4.4.3}Aggregation of Abnormal Returns}{160}
\contentsline{subsection}{\numberline{4.4.4}Sensitivity to Normal Return Model}{162}
\contentsline{subsection}{\numberline{4.4.5}CARs for the Earnings-Announcement Example}{163}
\contentsline{subsection}{\numberline{4.4.6}Inferences with Clustering}{166}
\contentsline{section}{\numberline{4.5}Modifying the Null Hypothesis}{167}
\contentsline{section}{\numberline{4.6}Analysis of Power}{168}
\contentsline{section}{\numberline{4.7}Nonparametric Tests}{172}
\contentsline{section}{\numberline{4.8}Cross-Sectional Models}{173}
\contentsline{section}{\numberline{4.9}Further Issues}{175}
\contentsline{subsection}{\numberline{4.9.1}Role of the Sampling Interval}{175}
\contentsline{subsection}{\numberline{4.9.2}Inferences with Event-Date Uncertainty}{176}
\contentsline{subsection}{\numberline{4.9.3}Possible Biases}{177}
\contentsline{section}{\numberline{4.10}Conclusion}{178}
\contentsline{chapter}{\numberline{5}The Capital Asset Pricing Model}{181}
\contentsline{section}{\numberline{5.1}Review of the CAPM}{181}
\contentsline{section}{\numberline{5.2}Results from Efficient-Set Mathematics}{184}
\contentsline{section}{\numberline{5.3}Statistical Framework for Estimation and Testing}{188}
\contentsline{subsection}{\numberline{5.3.1}Sharpe-Lintner Version}{189}
\contentsline{subsection}{\numberline{5.3.2}Black Version}{196}
\contentsline{section}{\numberline{5.4}Size of Tests}{203}
\contentsline{section}{\numberline{5.5}Power of Tests}{204}
\contentsline{section}{\numberline{5.6}Nonnormal and Non-IID Returns}{208}
\contentsline{section}{\numberline{5.7}Implementation of Tests}{211}
\contentsline{subsection}{\numberline{5.7.1}Summary of Empirical Evidence}{211}
\contentsline{subsection}{\numberline{5.7.2}Illustrative Implementation}{212}
\contentsline{subsection}{\numberline{5.7.3}Unobservability of the Market Portfolio}{213}
\contentsline{section}{\numberline{5.8}Cross-Sectional Regressions}{215}
\contentsline{section}{\numberline{5.9}Conclusion}{217}
\contentsline{chapter}{\numberline{6}Multifactor Pricing Models}{219}
\contentsline{section}{\numberline{6.1}Theoretical Background}{219}
\contentsline{section}{\numberline{6.2}Estimation and Testing}{222}
\contentsline{subsection}{\numberline{6.2.1}Portfolios as Factors with a Riskfree Asset}{223}
\contentsline{subsection}{\numberline{6.2.2}Portfolios as Factors without a Riskfree Asset}{224}
\contentsline{subsection}{\numberline{6.2.3}Macroeconomic Variables as Factors}{226}
\contentsline{subsection}{\numberline{6.2.4}Factor Portfolios Spanning the Mean-Variance\protect\\ Frontier}{228}
\contentsline{section}{\numberline{6.3}Estimation of Risk Premia and Expected Returns}{231}
\contentsline{section}{\numberline{6.4}Selection of Factors}{233}
\contentsline{subsection}{\numberline{6.4.1}Statistical Approaches}{233}
\contentsline{subsection}{\numberline{6.4.2}Number of Factors}{238}
\contentsline{subsection}{\numberline{6.4.3}Theoretical Approaches}{239}
\contentsline{section}{\numberline{6.5}Empirical Results}{240}
\contentsline{section}{\numberline{6.6}Interpreting Deviations from Exact Factor Pricing}{242}
\contentsline{subsection}{\numberline{6.6.1}Exact Factor Pricing Models, Mean-Variance Analysis, and the Optimal Orthogonal Portfolio}{243}
\contentsline{subsection}{\numberline{6.6.2}Squared Sharpe Ratios}{245}
\contentsline{subsection}{\numberline{6.6.3}Implications for Separating Alternative Theories}{246}
\contentsline{section}{\numberline{6.7}Conclusion}{251}
\contentsline{chapter}{\numberline{7}Present-Value Relations}{253}
\contentsline{section}{\numberline{7.1}The Relation between Prices, Dividends, and Returns}{254}
\contentsline{subsection}{\numberline{7.1.1}The Linear Present-Value Relation with Constant Expected Returns}{255}
\contentsline{subsection}{\numberline{7.1.2}Rational Bubbles}{258}
\contentsline{subsection}{\numberline{7.1.3}An Approximate Present-Value Relation with Time-Varying Expected Returns}{260}
\contentsline{subsection}{\numberline{7.1.4}Prices and Returns in a Simple Example}{264}
\contentsline{section}{\numberline{7.2}Present-Value Relations and US Stock Price Behavior}{267}
\contentsline{subsection}{\numberline{7.2.1}Long-Horizon Regressions}{267}
\contentsline{subsection}{\numberline{7.2.2}Volatility Tests}{275}
\contentsline{subsection}{\numberline{7.2.3}Vector Autoregressive Methods}{279}
\contentsline{section}{\numberline{7.3}Conclusion}{286}
\contentsline{chapter}{\numberline{8}Intertemporal Equilibrium Models}{291}
\contentsline{section}{\numberline{8.1}The Stochastic Discount Factor}{293}
\contentsline{subsection}{\numberline{8.1.1}Volatility Bounds}{296}
\contentsline{section}{\numberline{8.2}Consumption-Based Asset Pricing with Power Utility}{304}
\contentsline{subsection}{\numberline{8.2.1}Power Utility in a Lognormal Model}{306}
\contentsline{subsection}{\numberline{8.2.2}Power Utility and Generalized Method of\protect\\ Moments}{314}
\contentsline{section}{\numberline{8.3}Market Frictions}{314}
\contentsline{subsection}{\numberline{8.3.1}Market Frictions and Hansen-Jagannathan\protect\\ Bounds}{315}
\contentsline{subsection}{\numberline{8.3.2}Market Frictions and Aggregate Consumption\protect\\ Data}{316}
\contentsline{section}{\numberline{8.4}More General Utility Functions}{326}
\contentsline{subsection}{\numberline{8.4.1}Habit Formation}{326}
\contentsline{subsection}{\numberline{8.4.2}Psychological Models of Preferences}{332}
\contentsline{section}{\numberline{8.5}Conclusion}{334}
\contentsline{chapter}{\numberline{9}Derivative Pricing Models}{339}
\contentsline{section}{\numberline{9.1}Brownian Motion}{341}
\contentsline{subsection}{\numberline{9.1.1}Constructing Brownian Motion}{341}
\contentsline{subsection}{\numberline{9.1.2}Stochastic Differential Equations}{346}
\contentsline{section}{\numberline{9.2}A Brief Review of Derivative Pricing Methods}{349}
\contentsline{subsection}{\numberline{9.2.1}The Black-Scholes and Merton Approach}{350}
\contentsline{subsection}{\numberline{9.2.2}The Martingale Approach}{354}
\contentsline{section}{\numberline{9.3}Implementing Parametric Option Pricing Models}{355}
\contentsline{subsection}{\numberline{9.3.1}Parameter Estimation of Asset Price Dynamics}{356}
\contentsline{subsection}{\numberline{9.3.2}Estimating $\sigma $ in the Black-Scholes Model}{361}
\contentsline{subsection}{\numberline{9.3.3}Quantifying the Precision of Option Price\protect\\ Estimators}{367}
\contentsline{subsection}{\numberline{9.3.4}The Effects of Asset Return Predictability}{369}
\contentsline{subsection}{\numberline{9.3.5}Implied Volatility Estimators}{377}
\contentsline{subsection}{\numberline{9.3.6}Stochastic Volatility Models}{379}
\contentsline{section}{\numberline{9.4}Pricing Path-Dependent Derivatives Via Monte Carlo Simulation}{382}
\contentsline{subsection}{\numberline{9.4.1}Discrete Versus Continuous Time}{383}
\contentsline{subsection}{\numberline{9.4.2}How Many Simulations to Perform}{384}
\contentsline{subsection}{\numberline{9.4.3}Comparisons with a Closed-Form Solution}{384}
\contentsline{subsection}{\numberline{9.4.4}Computational Efficiency}{386}
\contentsline{subsection}{\numberline{9.4.5}Extensions and Limitations}{390}
\contentsline{section}{\numberline{9.5}Conclusion}{391}
\contentsline{chapter}{\numberline{10}Fixed-Income Securities}{395}
\contentsline{section}{\numberline{10.1}Basic Concepts}{396}
\contentsline{subsection}{\numberline{10.1.1}Discount Bonds}{397}
\contentsline{subsection}{\numberline{10.1.2}Coupon Bonds}{401}
\contentsline{subsection}{\numberline{10.1.3}Estimating the Zero-Coupon Term Structure}{409}
\contentsline{section}{\numberline{10.2}Interpreting the Term Structure of Interest Rates}{413}
\contentsline{subsection}{\numberline{10.2.1}The Expectations Hypothesis}{413}
\contentsline{subsection}{\numberline{10.2.2}Yield Spreads and Interest Rate Forecasts}{418}
\contentsline{section}{\numberline{10.3}Conclusion}{423}
\contentsline{chapter}{\numberline{11}Term-Structure Models}{427}
\contentsline{section}{\numberline{11.1}Affine-Yield Models}{428}
\contentsline{subsection}{\numberline{11.1.1}A Homoskedastic Single-Factor Model}{429}
\contentsline{subsection}{\numberline{11.1.2}A Square-Root Single-Factor Model}{435}
\contentsline{subsection}{\numberline{11.1.3}A Two-Factor Model}{438}
\contentsline{subsection}{\numberline{11.1.4}Beyond Affine-Yield Models}{441}
\contentsline{section}{\numberline{11.2}Fitting Term-Structure Models to the Data}{442}
\contentsline{subsection}{\numberline{11.2.1}Real Bonds, Nominal Bonds, and Inflation}{442}
\contentsline{subsection}{\numberline{11.2.2}Empirical Evidence on Affine-Yield Models }{445}
\contentsline{section}{\numberline{11.3}Pricing Fixed-Income Derivative Securities}{455}
\contentsline{subsection}{\numberline{11.3.1}Fitting the Current Term Structure Exactly}{456}
\contentsline{subsection}{\numberline{11.3.2}Forwards and Futures}{458}
\contentsline{subsection}{\numberline{11.3.3}Option Pricing in a Term-Structure Model}{461}
\contentsline{section}{\numberline{11.4}Conclusion}{464}
\contentsline{chapter}{\numberline{12}Nonlinearities in Financial Data}{467}
\contentsline{section}{\numberline{12.1}Nonlinear Structure in Univariate Time Series}{468}
\contentsline{subsection}{\numberline{12.1.1}Some Parametric Models}{470}
\contentsline{subsection}{\numberline{12.1.2}Univariate Tests for Nonlinear Structure}{475}
\contentsline{section}{\numberline{12.2}Models of Changing Volatility}{479}
\contentsline{subsection}{\numberline{12.2.1}Univariate Models}{481}
\contentsline{subsection}{\numberline{12.2.2}Multivariate Models}{490}
\contentsline{subsection}{\numberline{12.2.3}Links between First and Second Moments}{494}
\contentsline{section}{\numberline{12.3}Nonparametric Estimation}{498}
\contentsline{subsection}{\numberline{12.3.1}Kernel Regression}{500}
\contentsline{subsection}{\numberline{12.3.2}Optimal Bandwidth Selection}{502}
\contentsline{subsection}{\numberline{12.3.3}Average Derivative Estimators}{504}
\contentsline{subsection}{\numberline{12.3.4}Application: Estimating State-Price Densities}{507}
\contentsline{section}{\numberline{12.4}Artificial Neural Networks}{512}
\contentsline{subsection}{\numberline{12.4.1}Multilayer Perceptrons}{512}
\contentsline{subsection}{\numberline{12.4.2}Radial Basis Functions}{516}
\contentsline{subsection}{\numberline{12.4.3}Projection Pursuit Regression}{518}
\contentsline{subsection}{\numberline{12.4.4}Limitations of Learning Networks}{518}
\contentsline{subsection}{\numberline{12.4.5}Application: Learning the Black-Scholes Formula}{519}
\contentsline{section}{\numberline{12.5}Overfitting and Data-Snooping}{523}
\contentsline{section}{\numberline{12.6}Conclusion}{524}
\contentsline{chapter}{\numberline{}Appendix}{527}
\contentsline{section}{\numberline{A.1}Linear Instrumental Variables}{527}
\contentsline{section}{\numberline{A.2}Generalized Method of Moments}{532}
\contentsline{section}{\numberline{A.3}Serially Correlated and Heteroskedastic Errors}{534}
\contentsline{section}{\numberline{A.4}GMM and Maximum Likelihood}{536}
\contentsline{chapter}{\numberline{}References}{541}
\contentsline{chapter}{\numberline{}Author Index}{587}
\contentsline{chapter}{\numberline{}Subject Index}{597}

\chapter*{List of Figures}\markboth{List of Figures}{List of Figures}
\contentsline{figure}{\numberline{1.1}Dividend Payment Timing Convention}{12}
\contentsline{figure}{\numberline{1.2}Comparison of Stable and Normal Density Functions}{18}
\addvspace{8pt}
\contentsline{figure}{\numberline{3.1}Nontrading-Induced Autocorrelations}{96}
\contentsline{figure}{\numberline{3.2}Histogram of Daily Price Fractions and Price Changes for Five NYSE Stocks from January 2, 1990 to December 31,\protect\break 1992}{111}
\contentsline{figure}{\numberline{3.3}2-Histories of Daily Stock Returns for Five NYSE Stocks from January 2, 1990 to December 31, 1992}{113}
\contentsline{figure}{\numberline{3.4}The Ordered Probit Model}{125}
\addvspace{8pt}
\contentsline{figure}{\numberline{4.1}Time Line for an Event Study}{157}
\contentsline{figure}{\numberline{4.2}(a) Plot of Cumulative Market-Model Abnormal Return for Earning Announcements;
(b) Plot of Cumulative Con-\protect\break stant-Mean-Return-Model Abnormal Return for Earning\protect\break Announcements}{165}
\contentsline{figure}{\numberline{4.3}Power of Event-Study Test Statistic $J_1$ to Reject the Null Hypothesis that the Abnormal Return Is Zero, When the Square Root of the Average Variance of the Abnormal Return Across Firms is (a)\penalty \@M \ 2\% and (b)\penalty \@M \ 4\%}{171}
\contentsline{figure}{\numberline{4.4}Power of Event-Study Test Statistic $J_1$ to Reject the Null Hypothesis that the Abnormal Return is Zero, for Different Sampling Intervals, When the Square Root of the Average Variance of the Abnormal Return Across Firms Is 4\% for the Daily Interval}{176}
\addvspace{8pt}
\contentsline{figure}{\numberline{5.1}Minimum-Variance Portfolios Without Riskfree Asset}{187}
\contentsline{figure}{\numberline{5.2}Minimum-Variance Portfolios With Riskfree Asset}{189}
\addvspace{8pt}
\contentsline{figure}{\numberline{6.1}Distributions for the CAPM Zero-Intercept Test Statistic for Four Hypotheses}{250}
\addvspace{8pt}
\contentsline{figure}{\numberline{7.1}Log Real Stock Price and Dividend Series, Annual US Data, 1872 to 1994}{282}
\contentsline{figure}{\numberline{7.2}Log Real Stock Price and Estimated Dividend Component, Annual US Data, 1876 to 1994}{283}
\contentsline{figure}{\numberline{7.3}Log Dividend-Price Ratio and Estimated Dividend Component, Annual US data, 1876 to 1994}{284}
\addvspace{8pt}
\contentsline{figure}{\numberline{8.1}(a) Mean-Standard Deviation Diagram for Asset Returns; (b)\penalty \@M \ Implied Standard Deviation-Mean Diagram for Stochastic Discount Factors}{299}
\contentsline{figure}{\numberline{8.2}(a) Mean-Standard Deviation Diagram for a Single Excess Asset Return; (b)\penalty \@M \ Implied Standard Deviation-Mean Diagram for Stochastic Discount Factors}{302}
\contentsline{figure}{\numberline{8.3}Feasible Region for Stochastic Discount Factors Implied by Annual US Data, 1891 to 1994}{303}
\addvspace{8pt}
\contentsline{figure}{\numberline{9.1}Sample Path of a Discrete-Time Random Walk}{342}
\contentsline{figure}{\numberline{9.2}Sample Path and Conditional Expectation of a Brownian Motion with Drift}{345}
\addvspace{8pt}
\contentsline{figure}{\numberline{10.1}Zero-Coupon Yield and Forward-Rate Curves in January\protect\break 1987}{398}
\contentsline{figure}{\numberline{10.2}Cash Flows in a Forward Transaction}{400}
\contentsline{figure}{\numberline{10.3}Calculation of Duration for a Coupon Bond}{402}
\contentsline{figure}{\numberline{10.4}The Price-Yield Relationship}{407}
\contentsline{figure}{\numberline{10.5}Short- and Long-Term Interest Rates 1952 to 1991}{416}
\addvspace{8pt}
\contentsline{figure}{\numberline{11.1}Change in Short Rate Divided by Short Rate to the Power $\gamma$}{450}
\contentsline{figure}{\numberline{11.2}Sample and Theoretical Average Forward-Rate Curves}{454}
\addvspace{8pt}
\contentsline{figure}{\numberline{12.1}The Tent Map}{474}
\contentsline{figure}{\numberline{12.2}Monthly Excess Log US Stock Returns, 1926 to 1994}{482}
\contentsline{figure}{\numberline{12.3}Shifted and Tilted Absolute-Value Function}{486}
\contentsline{figure}{\numberline{12.4}Simulation of $Y_t=\mathop {\rm Sin}(X_t)+0.5\epsilon _t$}{501}
\contentsline{figure}{\numberline{12.5}Kernel Estimator}{503}
\contentsline{figure}{\numberline{12.6}Bullish Vertical Spread Payoff Function and Smoothed\protect\\ Version}{510}
\contentsline{figure}{\numberline{12.7}Binary Threshold Model}{513}
\contentsline{figure}{\numberline{12.8}Comparison of Heaviside and Logistic Activation Functions}{513}
\contentsline{figure}{\numberline{12.9}Multilayer Perceptron with a Single Hidden Layer}{514}
\contentsline{figure}{\numberline{12.10}MLP(1,5) Model of $Y_t=\mathop {\rm Sin}(X_t)+0.5\epsilon _t$}{516}
\contentsline{figure}{\numberline{12.11}Typical Simulated Training Path}{520}
\contentsline{figure}{\numberline{12.12}Typical Behavior of Four-Nonlinear-Term RBF Model}{521}

\chapter*{List of Tables}\markboth{List of Tables}{List of Tables}
\addvspace{8pt}
\contentsline{table}{\numberline{1.1}Stock market returns, 1962 to 1994.}{21}
\addvspace{8pt}
\contentsline{table}{\numberline{2.1}Classification of random walk and martingale hypotheses.}{29}
\contentsline{table}{\numberline{2.2}Expected runs for a random walk with drift $\mu $.}{40}
\contentsline{table}{\numberline{2.3}Autocorrelation function for fractionally differenced process.}{61}
\contentsline{table}{\numberline{2.4}Autocorrelation in daily, weekly, and monthly stock index returns.}{67}
\contentsline{table}{\numberline{2.5}Variance ratios for weekly stock index returns.}{69}
\contentsline{table}{\numberline{2.6}Variance ratios for weekly size-sorted portfolio returns.}{71}
\contentsline{table}{\numberline{2.7}Variance ratios for weekly individual security returns.}{73}
\contentsline{table}{\numberline{2.8}Cross-autocorrelation matrices for size-sorted portfolio returns.}{75}
\contentsline{table}{\numberline{2.9}Asymmetry of cross-autocorrelation matrices.}{77}
\addvspace{8pt}
\contentsline{table}{\numberline{3.1}Summary statistics for daily returns of five NYSE stocks.}{109}
\contentsline{table}{\numberline{3.2}Relative frequencies of price changes for tick data of five\protect\\ stocks.}{112}
\contentsline{table}{\numberline{3.3a}Expected upper bounds for discreteness bias: daily returns.}{118}
\contentsline{table}{\numberline{3.3b}Expected upper bounds for discreteness bias: monthly returns.}{119}
\contentsline{table}{\numberline{3.3c}Expected upper bounds for discreteness bias: annual returns.}{120}
\contentsline{table}{\numberline{3.4}Autocorrelation matrices for size-sorted portfolio returns.}{131}
\contentsline{table}{\numberline{3.5}Estimates of daily nontrading probabilities.}{132}
\contentsline{table}{\numberline{3.6}Nontrading-implied weekly index autocorrelations.}{133}
\contentsline{table}{\numberline{3.7}Summary statistics for transactions data of six stocks.}{138}
\contentsline{table}{\numberline{3.8a}Estimates of ordered probit partition boundaries.}{141}
\contentsline{table}{\numberline{3.8b}Estimates of ordered probit ``slope'' coefficients.}{142}
\addvspace{8pt}
\contentsline{table}{\numberline{4.1}Abnormal returns for an event study of the information content of earnings announcements.}{164}
\contentsline{table}{\numberline{4.2}Power of event-study test statistic $J_1$ to reject the null hypothesis that the abnormal return is zero.}{170}
\addvspace{8pt}
\contentsline{table}{\numberline{5.1}Finite-sample size of tests of the Sharpe-Lintner CAPM using large-sample test statistics.}{205}
\contentsline{table}{\numberline{5.2}Power of $F$-test of Sharpe-Lintner CAPM using statistic $J_1$}{207}
\contentsline{table}{\numberline{5.3}Empirical results for tests of the Sharpe-Lintner version of the CAPM.}{214}
\addvspace{8pt}
\contentsline{table}{\numberline{6.1}Summary of results for tests of exact factor pricing using zero-intercept $F$-test.}{241}
\addvspace{8pt}
\contentsline{table}{\numberline{7.1}Long-horizon regressions of log stock returns on the log dividend-price ratio.}{269}
\contentsline{table}{\numberline{7.2}Long-horizon regressions of log stock returns on the stochastically detrended short-term interest rate.}{270}
\addvspace{8pt}
\contentsline{table}{\numberline{8.1}Moments of consumption growth and asset returns.}{308}
\contentsline{table}{\numberline{8.2}Instrumental variables regressions for returns and consumption growth.}{312}
\addvspace{8pt}
\contentsline{table}{\numberline{9.1}Multiplication rules for stochastic differentials.}{347}
\contentsline{table}{\numberline{9.2a}Asymptotic standard errors for $\mathaccent "705E\relax {\alpha }$.}{365}
\contentsline{table}{\numberline{9.2b}Asymptotic standard errors for $\mathaccent "705E\relax {\sigma }^2$.}{366}
\contentsline{table}{\numberline{9.3}Cutoff values for comparative statics of $V_f$.}{369}
\contentsline{table}{\numberline{9.4}Asymptotic variances of Black-Scholes call price sensitivity estimators.}{370}
\contentsline{table}{\numberline{9.5}Option prices on assets with negatively autocorrelated\protect\\ returns.}{376}
\contentsline{table}{\numberline{9.6}Monte Carlo estimation of lookback option price.}{386}
\addvspace{8pt}
\contentsline{table}{\numberline{10.1}Macaulay's and modified duration for selected bonds.}{404}
\contentsline{table}{\numberline{10.2}Means and standard deviations of term-structure variables.}{415}
\contentsline{table}{\numberline{10.3}Regression coefficients $\mathaccent "705E\relax {\beta }_n$ and $\hat{\gamma}_n$.}{420}

\chapter*{\normalsize\rm The Econometrics of Financial Markets}\thispagestyle{empty}

\newpage
$\mbox{}$
\thispagestyle{empty}

\chapter*{Preface}\markboth{Preface}{Preface}
\parindent=18pt

The seeds of this book were planted over fifteen years ago, at the very start
of our professional careers.  While studying financial economics, and as we
began to teach it, we discovered several excellent textbooks for financial
theory---Duffie (1992), Huang and Litzenberger (1988), and Ingersoll (1987),
for example---but no equivalent textbook for empirical methods.

During the same period, we participated in research conferences on Financial
Markets and Monetary Economics, held under the auspices of the National
Bureau of Economic Research in Cambridge, Massachusetts.  Many of the papers
that captured our attention at these meetings involved new econometric
methods or new empirical findings in financial economics.  We felt that this
was some of the most exciting research being done in finance, and that
students should be exposed to this material at an early stage.

In 1989 we began to discuss the idea of writing a book that would cover
econometric methods as applied to finance, along with some of the more
prominent empirical results in this area.  We began writing in earnest in
1991, completing this arduous project five years and almost six hundred pages
later.  This book is considerably longer than we had originally planned, but
we have finally overcome the temptation to include just one more new topic,
and have put our pens to rest.  Of course, the academic literature has
evolved rapidly while we have been writing, and it continues to do so
even as this book goes to press.  We have attempted to provide broad
coverage, but even so, there are many subjects that we do not touch upon, and
many others that we can only mention in passing.

We owe many more debts---personal and intellectual---than we can possibly
acknowledge.  Throughout our professional careers our colleagues and mentors
have offered us advice, debate, inspiration, and friendship; we wish to thank
in particular Andy Abel, Ben Bernanke, Steve Cecchetti, John Cox, Angus
Deaton, Gene Fama, Bruce Grundy, Jerry Hausman, Chi-fu Huang, Mervyn King,
Nobu Kiyotaki, Pete Kyle, Greg Mankiw, Bob Merton, Whitney Newey, Bob
Shiller, Jim Tobin, and Arnold Zellner.

Many individuals have also provided us with invaluable comments and
discussions regarding the contents and exposition of this book.  We thank
David Backus, Nick Barberis, David Barr, Dimitris Bertsimas, Tim Bollerslev,
George Constantinides, John Cox, Xavier Gabaix, Lars Hansen, Campbell Harvey,
John Heaton, Ludger Hentschel, Roger Huang, Ravi Jagannathan, Gautam Kaul,
Dan Nelson, Amlan Roy, Bob Shiller, Tom Stoker, Jean-Luc Vila, Jiang Wang,
and the Ph.D.\ students at Harvard, MIT, Princeton, and Wharton on whom this
material was ``test-marketed'' and refined.

We have relied heavily on the able research assistance of Petr Adamek,
Sangjoon Kim, Martin Lettau, Terence Lim, Constantin Petrov, Chunsheng Zhou,
and particularly Matt Van Vlack and Luis Viceira, who undertook the difficult
tasks of proofreading the manuscript and preparing the index.

We are grateful to Stephanie Hogue for her great skill and care in preparing
the electronic version of this manuscript, and the typesetters at Archetype
for producing the final version of the book.

We thank Peter Dougherty, our editor at Princeton University Press, for his
patience, encouragement, and support throughout this project.

Several organizations provided us with generous support during various stages
of this book's gestation; in particular, we thank Batterymarch Financial
Management, the National Bureau of Economic Research, the National Science
Foundation, the John M.\ Olin Foundation, the Alfred P.\ Sloan Foundation,
and research centers at Harvard, MIT, Princeton, and Wharton.

And finally, we owe more than we can say to the support and love of our
families.

\vskip 18pt\noindent
\begin{tabular}{@{}l}
JYC\\
AWL\\
ACM
\end{tabular}
\newpage
\end{document}