Citation
Bayesian prediction in mixed linear models with applications in small area estimation

Material Information

Title:
Bayesian prediction in mixed linear models with applications in small area estimation
Creator:
Datta, Gauri Sankar, 1962-
Publication Date:
Language:
English
Physical Description:
ix, 214 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Linear models ( jstor )
Logical givens ( jstor )
Matrices ( jstor )
Modeling ( jstor )
Population estimates ( jstor )
Population mean ( jstor )
Statism ( jstor )
Statistical discrepancies ( jstor )
Statistical estimation ( jstor )
Statistics ( jstor )
Bayesian statistical decision theory ( lcsh )
Dissertations, Academic -- Statistics -- UF
Estimation theory ( lcsh )
Linear models (Statistics) ( lcsh )
Population forecasting ( lcsh )
Statistics thesis Ph. D

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 207-212)
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Gauri Sankar Datta.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
001687933 ( ALEPH )
25116672 ( OCLC )
AHZ9963 ( NOTIS )

Downloads

This item has the following downloads:


Full Text













BAYESIAN PREDICTION
WITH APPLICATIONS IN


IN MIXED LINEAR MODELS
SMALL AREA ESTIMATION


GAURI


SANKAR


DATTA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY


































to my parents and


teachers,


with


regards




















ACKNOWLEDGEMENTS


would


like


to express


sincere


gratitude


Professor Malay Ghosh


being my advisor,


for originally


proposing the


problem and


for


the attention


received


from


him for the


past


five


years.


Without


enormous


patience,


encouragement


guidance,


it would


not


have


been


possible


to complete the


work.


Throughout my years


graduate


program,


he has


been


my friend


, philosopher


guide ,


consider myself


extremely


lucky to get


as my dissertation


advisor.


would


like


thank


Professors


Michael


DeLorenzo


Ronald


Randles


for


serving on my


committee


Also,


am grateful


to Professors


Ramon


Littell,


Kenneth M.


Portier


P.V.


for


being


on my Part C and


oral


defense


committees.


special


Richard


thanks


to Professor Ghosh


Scheaffer for their genuine


interest,


Professor


incessant


efforts


unlimited


energy which


made


it possible


remove


stumbling stone out


of my way to


join


University


of Florida.


would


also


like


to express my


irrat itude


to mv


respected


teachers.


especially to











were


crucial


to my


coming to


United


States.


feel


very fortunate


being able


turn


them whenever


necessary.


would


also


like


acknowledge


the


help and


support


received


from Krishnandu


Ghosh


preparing me


come


United States.


am al


so grateful


highly


indebted


to my


Alma


Mater RamaKrishna Mission


Residential


College,


Narendrapur,


West


Bengal,


India,


for


support


received.


not


been


admitted


to Narendrapur,


would


have


never


pursued my


studies


in statistics.


this


respect,


will


always


remember our


Principal,


Respected Swami


Suparnananda


Maharaj,


our Head


Department,


P.K.


Giri,


their


care


concern


about me.


would


like


to offer my


humble


regards


them.


will


take


this opportunity to


express my appreciation


to Professor


Ashok Kumar Hazra


for


initiative


is a great


took


pleasure


introducing me


to acknowledge


to Narendrapur.


Professor Uttam


Bandyopadhyay to whom


definitely


owe


a lot


for my basic


understanding of


statistics as an


undergraduate.


interest


was further


stimulated


by the


insightful


teaching


of Professor S.K.


Chatterjee of


Calcutta University when


was a master's


student.


would


nona for me.


like


thank my parents for


indebted


they


to my numerous


have


well-


stei n dee I


*


R











Bibekananda Nandi,


Professor


K.M.


Senapati,


Mrs.


Durga


Senapati,


K.C.


Ghosh


and Mrs.


Bimala Ghosh,


who


have


always


considered me


consider myself


as a part


extremely


of their family.


lucky to


get the


also


affection ate


concern


of Mrs.


Senapati,


who


has always


treated


like


her own


son.


My heartfelt


thanks are


for my


"unofficial"


host


family


in Gainesville,


Malay Ghosh


his wife,


our


beloved Dolad i.


would


like


thank


A.P


. Reznek of the


United


States Census


Bureau


for providing me


with


computing


facilities during my


stay


in the


Bureau


as an


ASA/NSF


Research


Associate.


Last


least


would


like to


thank Ms.


Cindy Zimmerman


for


her skillful


typing


putting a scribbled manuscript


into


final


form.




















TABLE OF CONTENTS


Page


ACKNOWLEDGEMENTS


ABSTRACT


a.. I .a .a... .a .a a .. ... a


v111


CHAPTERS


ONE


INTRODUCTION


Literature
The Subject


Review .. ... ...
of This Dissertation


. 10


TWO


BAYESIAN PREDICTION
GENERAL CASE ..


OF MEANS


IN LINEAR MODELS
. .a .a ..a ..a. .a. S


..15


Introduction


Desc
Mode
Hier
Appl
Anal
Hier
Popu
Leve


ript
1 wi
arch
icat
ysis
arch
lati
1 Ob


Cr --.I


It! --


ion Or me nierarcnic
th Examples ..
ical Bayes Analysis .
ions of Hierarchical

ical Bayes Prediction
on Mean Vector in Abs
servations .


Bayes


Baye

of
ence


* .. 18
.. ... 31


.*


. .. ... 37
Finite
of Unit
S ......57


THREE


OPTIMALITY OF BAYES PREDICTORS
A SPECIAL CASE .... ... .


FOR MEANS


IN
S. ... .. 65


Introduction
The Hierarchi


. .. 65


Bayes


Predictor


3.3

3.4

3.5


Spec
Best
Domi
Best
Domi
Best


ial
Unb
nati
Unb
nati
Equ


ase
ase
n i
ase
n i
var


Predict
Small A
Predict
Infinit
ant Pred


ion and
rea Est
ion and
e Popul
iction


Stochast
imation
Stochast
action ...
in Small


... 67
ic


C


. .88


*


--










FOUR


ASYMPTOTIC
PREDICTORS


OPTIMALITY
FOR MEANS


OF HIERARCHICAL


BAYES


Introduct i
Model. Los


4.2.2


>n ....... ....... 114
i, Prior and Predictors ....... 117
ral Expressions for Bayes
s Difference. .. ... 119
om Regression Coefficients


Error


Regr


Model


S. .120
S. 129


4.3


4.4


Optimal ity


with


Stage Variance Component:
Model ....................
Asymptotic Optimality with
Variance Components ......


Known
Fay-H


First
erriot


Unknown


. 137


. 153


FIVE


SIMULTANEOUS BAYESIAN
AREA VARIANCES ......


ESTIMATION


OF SMALL


Introduction


5.2


5.3


5.4


Baye
when
Know
Asym
Regr
Vari
Asym
Regr


Estimation of a
Ratios of Varian


Quad rati
ce Compon


in N
Known


sted
Ratio


in Ne
Unkno


Error
of


S. 173
Error


Variance


Components


SIX


SUMMARY


FUTURE


RESEARCH


Summary
Future


Research


... 189
... 190


APPENDICES


PROOF


OF THEOREM


2.3.1


. 194


AN INDEPENDENCE RESULT
ELLIPTICALLY SYMMETRIC


IN A FAMILY OF
DISTRIBUTIONS


S.. 203


BIBLIOGRAPHY


. .. .. .. .. .. .. .. 207


BIOGRAPHICAL


SKETCH


. ...... .. 213

















Abstract of Dissertation Presented


the Graduate School


the University


Requirements


of Florida


for the Degree


in Partial


of Doctor of


Fulfillment of


Philosophy


BAYESIAN PREDICTION


IN MIXED


LINEAR MODELS


WITH APPLICATIONS


GAURI


IN SMALL AREA ESTIMATION


SANKAR DATTA


August,


Malay Ghosh


Major Department


Statistics


Small


area estimation


is gaining


increasing popularity


recent


times.


Government


agencies


United


States


Canada have


been


involved


estimating unemployment


per


for many


capital


state


income,

local


crop yield,

government r


etc.

-egions


simultaneously

Typically,


only


a few samples are


available


from an


individual


area.


Consequently


reliable


estimators of


"parameters,


" such


the mean


or the


variance


for the


area,


need


"borrow


strength"


from similar


neighboring areas


implicitly


explicitly through


a model.


Such


estimators


usually have


smaller mean


squared


error of


prediction


than


survey


estimators.


this dissertation


a general


hierarchical


Bayes


Chai rman:


1990


rates,


Sa a











regression


coefficients model ,


authors are seen


to be special


etc.

cases


considered


earl ier


proposed


general


model.


predictive distribution


characteristic


inte rest


for the


unsampled


population


units


is found


given


the observations on


sampled


units


is used


estimators of


to draw


several


inference


small 1


particular,


area means and


simultaneous


variances are


developed.


A mixed


linear model


with


noninformative


prior


regression


coefficients


(or fixed


effects)


independent gamma priors


(possibly noninformative)


for the


inverse


variance


components


is used.


a special


case


this


HB analysis,


when


vector


ratios of


variance


components


is known,


predictor of


sampling


is shown


vector of


means


possess some


finite


frequentist


population


optimal


properties


(such


as best


unbiased


predictor,


best


equivariant


predictor,


etc .)


basically


under the


elliptical


symmetry


assumptions.


Performance of


this


HB predictor


is evaluated


comparing


Bayes


risk with


that


of subjective


Bayes


predictor with

superpopulation


"true"


or


"elicited"


parameters.


prior for the


is shown


that,


unknown


under a


balanced


one-way


random effects model 1


with


covariates and


ax'sP maeI~


niua Ired


errnr


1 nAs -


the difference


I-E t-,ICc


Raves


.



















CHAPTER


ONE


INTRODUCTION


Literature


Review


linear models


astronomers


for predicting the


positions of


Starting from


celestial


these


bodies


days,


goes


use


back several


of model-based


centuries.


inference


prediction


received


considerable


attention.


part icular,


animal


plant


breeders


have


used


such models


predicting


some


characteristics of


future


progeny.


Starting with


pioneering work of Henderson


(1953),


considerable


attention


been


devoted


this


problem.


refer


to Gianola and


Fernando


(1986)


Harville


press)


where


other


references are


cited.


other


hand,


survey


analysts


have


used


the model-based


approach


finite


population


sampling with


goal


predicting


certain


characteristics of


population


basis


unsampled


units


observed sample.


Early work


this


topic may


found


in Cochran


(1939,


1946)


where


finite


population


is viewed


as a realization


from


hypothetical


superpopulation.














small 1


area statistics


was


in existence


as early


as the


llth


century


England


17th


century


in Canada


(see


Brackstone,


1987).


However,


these


early


small


area


statistics


were based


on data obtained


complete


enumeration.


availability


limited


resources


advent


of sophisticated


statistical


methodologies,


past


few decades,


sample


surveys


for most


purposes,


have


been


widely


used


as the


means of


data collection


contrast


to complete enumeration.


data collected


from


these


surveys


have


been


very


effectively


used


to provide


suitable


statistics


national


state


levels on


regular

sublevel


basis.


However,


below the


state


use of


level


(for


survey


data


example,


county


other subdi


vision)


was


limited


because


the estimates


these


small 1


areas


usually were


based


on small


samples


produced


unacceptably


large


standard


errors


coefficients


of variation.


improve


reliability


small


area statistics,


is necessary


to have


a much


larger


sample


size


an


individual


area


than


can


afforded


with


limited


resources


available.


Consequently,


use


of survey data


(possibly


or














During the


last


United States and


few years,

Canada, h


many


iave


countries,


recognized


including


the


importance of


small


area estimation.


Recently


there


growing concern among several


governments


with


the


issues


of distribution,


equity


and disparity.


There may


exist


subgroups


within


a given


population


which


are


far


below the


average


certain


respects,


thereby necessitating


remedial


action


on the


part


government.


Before


taking such


an action, there

accordingly, the


a need


statistical


identify


data at


such


subgroups,


relevant


subgroup


levels must


available.


different


government


agencies


like


the Census


Bureau,


Bureau


of Labor


Statistics,


Stat ist i cs Canada and


Central


Bureau


Statistics


of Norway


have


been


involved


obtaining


estimates of


population


counts,


adjustment


factors


census


counts


, unemployment


rates,


per


capital


income,


etc.


state


local


government


areas.


techniques


face

have


this


emerged


problem,


that


small 1


area estimation


"borrow strength"


from similar


neighboring areas


for estimation


prediction


purposes.


Through


use


some


appropriate model


auxiliary


information


(possibly


obtained


through


complete















over the


survey


est imators


a good


review of


small


area estimation


literature one may


refer to Ghosh


(1990).


necessity


"borrowing strength"


been


real ized


by many


stat i st i c i ans.


Ericksen


(1974)


advocated


use


regression


method


for estimating population


changes


local


areas.


Herriot


(1979)


proposed


adaptation


James-Stein


estimator to


survey


estimates


income


small


areas.


Survey


estimates


being


based


a small


sample


size


(which


is usually 20 percent


population


of size


less


than


1000)


usually


have


large


standard


this,


errors


these


coefficients of


authors f i rst


variation.


a regression


To rectify


equation


census sample


estimates,


using


as


independent


variables


county values,


tax


return


data for


year


1969


data


housing from the


1970


census.


The


estimate


they


provided


each


place


was


a weighted


ave rage


sample

Harter

under


estimate


and

corn


Fuller


(1988)


soybeans


regression

considered


estimate.

prediction


12 counties


Battese,


areas


north-central


Iowa based


satellite data


1978 June E

Battese,


:numerat ive


Harter and


survey

Fuller


nd LANDSAT

(BHF) used














sampled


counties.


Fuller


Harter


(1987)


also considered


a multivariate


extension


this model.


There


is a similar problem of


prediction


faced


by the


animal


breeders-.


purpose


of selecting the


best


animals


for


future


breeding,


they need


come


up with


index for


each


animal


under


consideration.


Henderson


(1953,


1975)


predictor


advocated


(BLUP)


certain


use


best


linear


linear unbiased


combinations


of fixed


random


effects


using a mixed


linear model.


Harville


press)


used


a mixed


linear model


for predicting the


ave rage


sires


weight


of single-birth male


belonging


to different


lambs which


population


are


lines


progeny


dams


belonging to different


categories.


Harvilie


Fenech


(1985)


considered


this


example


for estimating the


heritabilit


other


problems


based


on a


linear model


come


varietal


trials


comparative


experiments


comparative experiments,


several


treatments


have


to be


compared


their effects


or some


suitable


contrasts


have


to be estimated.


Multicentered


clinical


trials


are


good


examples of


comparative experiments


(see


Fleiss,


1986).


Problems


this


type


ones mentioned


in the














The methods


that


have


usually been


proposed


in model-


based


inference


use


either


a variance


components approach


or an


empirical


Bayes


(EB)


approach,


although


as pointed


out


by Harville (1988,


press),


the distinction


between


the


two


is often


superfluous.


Both


these


procedures


use


certain

First,


mixed


linear models for


assuming the


variance


prediction


components to


purposes.

be known,


certain


BLUPs or


EB predictors


are


obtained


for the


unknown


parameters of


interest.


Then


unknown


variance


components are

fitting of con


estimated


stands or the


typically


by Henderson's method


restricted max i mum


likelihood


(REML)


method.


resulting estimators,


which


can


called


estimated


BLUPs or


EBLUPs


(see


Harville,


1977) ,


are


used


for final


prediction


purposes.


Empirical


Bayes


approach


in small


area estimation


was


first given


Herriot


(1979)


later


also


used


by Ghosh


Meeden


(1986),


Ghosh


Lahiri


(1987a,


1988)


among others.


According to


this


procedure,


first


a Bayes


estimate of


unknown


parameter of


interest


is obtained


by using a normal


prior or using a


linear Bayes


argument


(Hartigan,


1969).


unknown


parameters of


prior


are


then


estimated


some


classical


methods


like


method














Although


above


approach


of EBLUP


or EB


is usually


quite


satisfactory for


point


prediction,


1s very


difficult


to estimate


standard


errors associated


with


these


predictors.


This


is primarily


due


the


lack of


closed


form expressions


for the


mean


squared


errors


(MSEs)


(1984)


EBLUPs or the

suggested an


EB predictors.


approximation


Kackar


the


MSEs


Harvilie


(also


Harville,


1985,


1988


, in


press;


Harvilie


Jeske,


1989).


Prasad


and Rao


approximate


MSEs


(1990)


proposed


in three


estimates of


specific


mixed


these


linear models.


these


approximations


rest


heavily on


normal ity


assumption.


Recently,


Lah iri


Rao


(1990)


considered


this


problem,


relaxing the


normality


assumption,


assuming


some moment


conditions


without


presence of


auxiliary


information.


work of Prasad


(1990)


suggests


that


their


approximations


work well


when


number of


small


areas


is sufficiently


large.


not


clear


though


how these approximations


fare


for a small


or even


moderately


large


number


of small


areas.


Ghosh


as an


Lahiri


alternative


press)


EBLUP


proposed


or the


an HB procedure


EB procedure.


HB procedure,


one


uses


posterior mean


for


estimating














often


complicated,


can


found


exactly via numerical


integration


without


approximation.


The model


considered


by Ghosh


Lahiri


press)


was,


however,


only


a special


case


so-called nested


error


regression model,


also


used


by BHF


A similar model


was


considered


by Stroud


(1987),


general


analysis


was


performed


only for the


balanced


case


that


is when


number of


samples


was


same


for


each


stratum.


Other models


have


also


been


proposed.


a recent


article,


Choudhry


(1988)


considered


five


specific


models


small


area estimation


not


included


earlier work of


(1979)


Prasad


Cumberland


(1990).


(1989)


Recently,


considered


Royal 1


certain


cross-c


lassificatory models for


small 1


area estimation.


latter


carried


out


a Bayesian


analysis assuming the


degeneracy


certain


terms


an usual


two-way


linear


model.


For a Bayesian


analysis


context


animal


breed ing,


one may


refer to Gianola and


Fernando


(1986).


However,


they


consider the


HB analysis.


They


used


subjective


informative


priors


which


are


constructed


from


previous data and


experiments.


Also


, they showed














important


special


case


which


arises


in the


above


approaches which


is also


important


the theory


least


squares.


When


ratios of


variance


components


are


known,


predictors


(least


squares,


empirical


Bayes or


hierarchical


Bayes)


are


BLUPs


(Henderson,


1963).


related


BLUP


results


predicting scalars


in finite


population


Cumberland


sampling one may


(1989),


refer to


Prasad


Royall 1


(1990)


(1979),


several


others.


Harvilie


(1985,


1988,


in press)


pointed


out


BLUP properties of


Bayesian


scalars


in general


mixed


linear models


(see


also


Harville,


1976).


Ghosh


Lahiri


press)


have


extended


Henderson


others scalar


BLUP


notion


to show the


Bayesian


predictor of


vector of


finite


population


mean


is BLUP


To conclude


this discussion,


we will


briefly mention


another problem.


far,


we have


considered


problem of


estimating the


mean


in finite


population


sampling.


Another


important


problem


finite


population


sampling


is estimating the


finite


population


variance.


Ericson


(1969)


found


Bayes estimator of


finite


population


Empirical


variance


Bayes


under


estimation


a normal


of finite


theory


set


population


variance














Subject


of This


D i ssertat i on


this dissertation,


we present a unified


Bayesian


prediction


theory for


linear models


small


area


estimation


context


finite


population


sampling.


general


Bayesian model


as an extension


is presented


ideas of


which

Lindley


can


regarded


and Smith


(1972)


to prediction.


This


general


model


can


also


be applied


infinite


population


situations,


for


example,


animal


breeding


other


applications


where


a mixed


linear


model


used.


In Chapte


r Two,


introduce


a general


HB model


use


this model


for simultaneous estimation


of several


small


area means


in finite


population


sampling.


Some of


widely used models


small


area estimation


including the


nested


error


regression model


(Battese


al.,


1988;


Prasad


Rao,


1990;


Stroud,


1987;


Ghosh


Lah i ri,


press),


random


regression


coefficients model


(Dempster et


al .,


1981


Prasad


Rao,


1990),


cross-classificat o ry


models


stage


(Royall,


1979;


sampling models


Cumberland,


(Ghosh


Lahiri,


1989)


1988;


multi-


Malec


Sedransk,


1985;


Scott


Smith,


1969)


can


regarded


special


cases


of our model.


posterior distribution














population,


given


the data,


conditional


distribution


conditional


mean


variance


vector of


effects are


provided.


These


two


analyses are


applied


two


real


data sets.


is worthwhile


to mention


that


Bayesian


analysis


linear models was


initiated


by Hill


(1965).


also


Hill


(1977,


1980).


For a good


exposition


HB analysis


see


Berge r


(1985).


In Chapter


discussed


this model,


which


Three,


previous

assumes


a special


chapter

known


case


of HB models


is considered.


rat ios


Based


of variance


components,


certain


optimal


properties


of the


HB predictors


proposed


within

appeal


this


a Bayesian


also


chapter


are


framework,


frequentists.


proved.


these

The


Although,


results


should


BLUP notion


developed


be of

real


for


valued


parameters


is extended


vector valued


parameters,


is shown


that


Bayesian


predictors derived


this


chapter


are


indeed


BLUPs.


From


this,


as a special


case,


follows


that


Bayesian


predictors of


finite


population mean


vector and


other


linear


parameters are


BLUPs


as well.


BLUP


result


for the


finite


population


mean


vector unifies


a number


similar


results derived


under specific models


(e.g. ,


Royall,


1979;


Ghosh


on














suitable


subclass


elliptically


symmetric


distributions,


including but


not


limited


normal,


the


HB predictors


are shown


best


unbiased;


that


they


have


smallest


variance-covariance matrix within


class of


unbiased

been able


predictors.

to show that


Also,

the


following Hwang


BLUPs also


(1985),


"universally"


we have

(or


stochasticallyy")


for elliptically


dominate


linear unbiased


symmetric distributions.


The


predictors


notion


"universal" and "s

precise in Chapter


tochastic"

Three. A


dominant ion


lso,


will


be made


is established


that


under


a suitable


group of


transformations,


pred ictors


are


best


within


class


of all


equivariant


predictors for


elliptically


symmetric


distributions.


Jeske


Harvilie


(1987)


have


shown


that


scalar BLUPs are


best


equivariant


within


class of


linear equivariant


predictors


to our


without


knowledge,


any

the


distributional

equivariance re


assumption.


:suits


However,


for vector


valued


predictors


have


not


been addressed


before


this


context


their


full


generality.


In Chapter


results


Four,


regarding the


we have

Bayes r


established


isk


some


performance


asymptotic

certain H


predictors of


finite


population


mean


vector.


We have














that


under


average


squared


error


loss


Bayes


risk


difference


between


HB predictors and


subjective


Bayes


predictors for


"true"


prior


goes


to zero as


number of


small


areas


goes


infinity.


This


shows our HB


predictors


are


asymptoticallyy


optimal"


(A.O.)


sense


of Robbins


(1955).


The


A.O.


property


certain


predictors


arising naturally


context


finite


population


sampling


was


proved


in Ghosh


Meeden


(1986),


Ghosh


Lahiri


(1987a)


Ghosh,


Lahiri


and Tiwari


(1989).


Chapter

of several s


special


Five


trata


cases


is devoted

variances.


nested


the simultaneous

have considered


error


estimation

the


regression model


considered


in detail.


property


by Ghosh


in Chapter


these


Lah i r i


Four,


predictors.


press)


we have


Ghosh


and Stroud


proved


Lahiri


(1987)


A.O.


(1987b)


Lahiri


Tiwari


press)


have


proved


A.O.


property


certain


EB predictors


finite


population


variances.


reemphasize


that


present dissertation


provides


a unified


Bayesian


analysis


both


finite


infinite


population


framework.


For finite


population,


we


unify


number of models


considered


earl ier


by different authors.














to our


knowledge,


estimates


of MSEs or good


approximations


thereof


are


not


available


except


for a few


specific models.


Bayesian


procedures


this dissertation,


the other


hand,


can


serve


as a general


recipe


to handle


a greater


variety


of problems.


Also,


inferential


methods of


following


chapters are


implementable


for data analysis,


especially


in these


days of


sophisticated


computing


facilities.


















CHAPTER TWO


BAYESIAN PREDICTION


OF MEANS


IN LINEAR MODELS:


GENERAL CASE


Introduction


this


chapter we


will


consider two similar


different


small


population


prediction


problems


simultaneously.


area estimation


sampling and


problem


problem


other


comparative


problem


context


problem deals


experiments


refers to


finite


with


the


context


of ANOVA,


ANOCOVA


or


linear


regression


infinite


population


situation.


In both


these


cases,


a mixed


linear


model


used.


first


case,


we are


interested


predicting some


finite


population


characteristic


(e.g.,


finite


population


totals or means)


whereas


second


case


we


are


interested


in predicting


linear functions of


fixed


random effects.


finite


population


sampling set


we


assume


that


there


are m strata,


stratum Ui


containing a


finite


number of


units


with


units


labe lled


TUT


UiN.


Let


Y.j denote


with


some


characteristic


unit


stratum


interest associated


1,...,














some


finite


cost.


are


interested


predicting some


linear


combinations of


these observables


(like the


finite


population


total


or mean


for each


small


area or domain)


using a quadratic


loss.


notational


convenience,


we will


denote


a sample of


size


from the


stratum


the other


' Yi2

hand


,9 .


in.


infinite


population


set


are


interested


particular


predicting


contrasts)


fixed


near

and


combinations (

random effects.


Note


that


this set


up these quantities are


observables.


For this


problem


too,


use


a quadratic


loss


function.


We will


quantities


use


we want


word


predictands to


to predict


both


refer to


problems.


analysis


assume


will


be done


ratios of


in two


stages;


variance


components


first

are


stage we

known


whereas


situation


the

where


second


stage


we consider the


variance


more


components are


general

unknown.


In Section


a general


HB model


will


be described


a number


interesting


examples arising


finite


population


will


sampling or


considered.


in the


Some


of the


infinite


population


existing models


set

used


context


of finite


population


sampling are shown


to be














Realizing the


importance of


problem,


most


general


unknown


situation


will


where


be considered


the variance


this chapter,


components are


whereas


known


ratios of -variance


components


will


be considered


next


Section


chapter.


2.3.


This


A general


general


mixed


situation


linear model


is considered


is considered


some


prior distribution


parameters which


the variance


consist of


components.


is assigned


vector of


first


to all


fixed


unknown


effects


part of Section


2.3,


for the model


introduced


in Section


2.2,


have


found


posterior


(predictive)


distribution


of the


characteristic


interest of


nonsampled


population


units


given


values of


that


characteristic for the


sample


units


finite


population


sampling.


Also


posterior mean


vector


posterior variance-covari ance matrix


corresponding to


characteristic vector of


nonsampled


units are


obtained


from this


predictive distribution.


In particular,


posterior means


variances of


finite


population


means


small


areas are obtained.


In the


second


half


of Section


2.3,


have


obtained


posterior


distribution


the


vector of


fixed


random effects


for


the model


introduced


in Section


2.2.


particular,














In Section


2.4,


we have


applied


results


of Section


2.3 to


some


actual


data sets.


First,


we shall


consider the


corn and


soybeans data


which


appeared


in Battese,


Harter


Fuller


(1988).


Using the


HB analysis developed


Section


2.3


we have derived


the


posterior means


posterior standard deviations for the


12 small


area


(county)


means.


second


data set containing the weights


of 62 single-birth


lambs appeared


Harvilie


press).


This


set


is analyzed


HB methods


up developed


second


for


half


infinite


of Section


population


2.3.


Finall


in Section


an HB analysis of


the model


considered


by Carter


Rolph


(1974)


subsequently


Herriot


(1979)


to estimate


per


capital


income


small


places


is considered.


this


situation


unit


level


observations are


not


available


we are


interested


predicting the


finite


population mean


for each


small


area.


Here


sampling variances are different and


are


assumed


to be known;


also,


a uniform


prior


on the


regression


coefficients and


a gamma prior


(proper or


improper)


on the


inverse


prior variance.


Description


Hierarchical


Bayes


Model


with


Examples













r- )


(B) conditional


(C)B,


R and


N(O

have


r- D(A));

a certain


joint


prior distribution


proper


improper.


Stages


of the model


can


identified


as a


general


mixed


linear model.


see


this,


write


(2.2.1)


where


is the


vector


fixed


effects,


are


mutually


independent


with


N(O ,


r-19) and


N(O,


r-1D(A))


are


known


design matrices,


is a


known


positive definite


(p.d.)


matrix,


while


is a p.d.


matrix which


is structurally


known


except


possibly for some


unknown


examples


follow,


involves


ratios


variance


components.


context


of small


area estimation,


part it ion


Y(NTXl),


X (NTxp),


Z(NTxq)


e(NTXl)


with


conformity


rewrite


model


given


2.2.


= IL)


(2)
+ e(2).


2.22)


as


N(Xb +


Xb +


12 (x)


(1)

(2)


2(1)

(2)














S(2)((NT-nT) xl) corresponds


vector of


unsampled


units.

(.(1)T


We will


y(1)T )


, .


further partition


where


into


y(1)T


ni-component


vector


(1)
Yi


Yil


,. .


small 1

((2)T


Yin. )T


area.


is the


Similarly


(2)T)
Ym ,


S...


vector of

, y(2)T


where


sampled


can


(Ni-ni)-component


units


from the


partitioned


into


vector


(2)
xi


(Yi +l' "
the ith sma


Yi ,N
1,N)


Lii


area.


vector


of unsampled


our primary


units


objectives


for

in small


area estimation


vector


(71 ..


to estimate


ym)T


where


finite
N.


population mean


>2YiJ/ Ni


1 .


More


generally,


we may


interested


predicting the


vector


(say)


this


linear


combinations


known matrices


purpose


distribution


it suff

of Y(2)


ices


given


, y2))


+ c (2)
+ CY


A(uxnT)


find


C(ux(NT-nT)).

predictive


y(1)


next


section


this


will


accomplished


by using model-based


approach


survey sampling.


Before we


consider the


other problem


infinite


population


introduced


set


identify


small


some


area estimation


existing models


several


authors


special


cases


(2.2.2).


In what


follows,


we shall


use


ident-.^ t.' mn+-r iv


notat. nfl


mnr


(1)T


(1)
AY


E(Y(1)


I1Y1I


Il,


I II


ff


I -


1 11| -














Also,


col
1

(B.)


denote


the matrix B ,...,


Z)T


k
SA.
i= 1


denote


the matrix


First,


consider the


nested


error


regression model


= jb
- v
- ^ii-


+ eij


(j = 1,..., N.; i = 1,..., m).


(2.2.3)


The model


was


considered


by Battese,


Harter


Fuller


(1988).


They


assumed


to be mutually


independent


N(O,


with


this


case


N(0,

X(1)


(ArY,'),


col
1

col
1

col col
1

T
:ij


DA)


= A-1


m
i=l 1i

Im In


z(2)


the


m
=i1
i=l-Ni-ni

further


special


case


of Ghosh


Lahiri


press)


, xi
"1


for


every


1,...,


1,...,


Note


that


a ratio of


V(eij)/V(vi),


variance


components.


The


random


regression


coefficients model


of Dempster,


Rubin


Tsutakawa


(1981)


(also


Prasad


Rao,


1990)


also a special


case


ours.


this set


up,


are


same


in the


nested


error


regression


... A


T
~ij


D(A)


,-1>


,(2)


(1)


XI
"1


(1)


Xt














Some


the models of


Choudhry and


(1988)


can


also


treated


as special


cases


ours.


For example,


one of


their models


given


bxi1


1/2
+ eijij


(j = 1,..., Ni


(2.2.4)


with


N(O,


(rA)


-1) and


N(0,


r-1).


Here


= 1,


are


-V)


same


in the


nested


error


regression


model


with


vector xi
"1]


replaced


scalar


5(2)


are


-F'.
ii


same


in the


nested


error


regression


model,


= Diag(x1l,...,


X1N1


, .


xml,'..


XmNm)


Another model


considered


by these authors


similar to


one


given


(2.2.4)


with


replacing x1/2
replacing x _1


as multipliers


eij.


another model


considered


Choudhry


(1988)


bxij


1/2
V1 ij


1/2
+ e .jx.
1JJ 1J


1,...,


.2.5)


with


having the


same distribution


previous model.


Here


= Diag(x11,...,


X1N1


xlii, ,.. .


, t


XmN),'


z(1)


m (1)
i= u
i- ~1


S(2)


m (2)
G u


with


"(1)
u.i


( 1/2
I 1 ,...


(2)
, u.


1x/2 T
x, )


1/2
-- IX i ,...,


xY/ 2


are


. 9


b,


i 1,, m)


.(2)


2(1)


o(x>


(1)


(2)
X


,


---














linear model.


For example,


suppose


there


are m small


areas


labelled


Within


1 ,...,


each


small


area,


units are


further


classified


into


c subgroups


(socioeconomic


class,


age,


etc )


labelled


1 ,..


The


cell


sizes


1 ,...,

1 ,...,


- 1,...,


Nij) denote
121


assumed


to be known.


the measurement on


Yijk


individual


(i,j)th


cell.


Conditional


r and


suppose


Yijk
ijk


= xT .b +
-1J-


+ eijk


(2.2.6)


= 1,...,


iid N(O,


N(O,


Nij;


mutually

(A3r)-1),


(A1r)-1).


1,...,


independent


N(0,


this


1 ,...,


with


with


N(O,


(A2r)-) and


case


col col col
1 *J


Y.,
ilk


col
1

( col
c \n. +l

Y.,
13k


col
1


col
1i

ol ij Tj)x,
j~c ~ j-1


c -Ni.-ni j)}


r-1),


(l)


(1)


.(2)

















col
1

ln.i}'
1~v i


-'3


m
i=lj


c
e In.
=1-


is a matrix similar to


with


(N~~-n~~)


replacing n
lJ


in defining the


dimensions of


vectors.


Also,


(r ,...


, rm


r~-' .


Y 1 *''*


7mc)


= 3


, A


(A1,


pm~


A21 Ic


= Diag(AllIm ,


A31 Imc).


Special


cases


of this model


have


been


considered


several


others.


Cumberland


7. -are degenerate


(1989)


considered


zeroes.


Also,


a model


they


where


assumed


variance


rat io


to be known


in deriving their


estimators,


and did


not


address


issue


unknown


appropriately.


Next


we show that


two stage


sampling model


with


covariates and


m strata


is a special


case


our


general


linear model.


Suppose


that


stratum


contains


primary


units.


Suppose


also


that


primary


unit


within


denote


stratum


value of


contains


subunits.


characteristic


interest


Yijk


for the


subunit


within


primary


unit


from


stratum


1,...,


1 ,...,


1,...,


From the


stratum,


a sample


primary units


taken.


selected


orimarv unit


within


stratum.


z
~2


'A'


,(2)


2(1)


u ir~


uiir. I













Assume


conditional


r and


Yijk


= xijb +


(k= 1,..., Nij; j = 1,..., Li
1 1


i = 1,..., m),


(2.2.7)


where


', ij and


are mutually


independent


with (i


N(O,


(A1r)


"ili


N(O,


(Ar) -1),


N(O,


col
1

col { col
1_ lle A Jn


col
1

col
1

col
u .. 1J 1J


+ nij [j

1,...,


= s(


2)T
W2)


col
1

(Q


ol col
i

('i~))


col col
1 LI 1i


Al sn.


let+


he defined~


i1 m 1 lnrl


v(i)


+ eijk


r-1)


('i~~)


eijk


(Yij k)


.(1)


(Yij


(2)


ilY
















col col (1 x. ,
1

where


- uij


C'')),


m
1- ini
i=l1


S ni.


E nij
j=- J


W(1)


= w(1


V where


(1)
W1
-1


m i
e *e in.
i 3 =n.1.
i- j-i "ij


(1)
V2


m


= "TOL.


m
Li ,'


z(2)


= (S 2)


m
e 1
* i ~=.


Li
l Ni.j-ni.


L.
m i
i= 1r. .
-~~~ 1'


Here

with


= 2,


= (A,
L.
INi,
JE Nj.


2)T ,

The


, D(A)


ideas


can


= Diag(AllIm,


be extended


A21IL.)


directly


to multistage


sampling with


more


complicated


notations.


may mention


here


that


Bayesian


analysis


for two stage


sampling


was


introduced


first


by Scott


Smith


(1969)


a much


simpler framework.


A multistage


analog of


their


work was


provided


by Malec


Sedransk


(1985).


Ghosh


= N.3


= (1)


m
E.i'
i=l


(2)


s


2(1)


"T
1


W(2)),


(2)
S


(2)














Now we


will


consider the


infinite


population


set


this


context,


we will


use


the model


given


(A),


with


mixed


linear model


representation


given


(2.2.1)


Here


we will


assume


the data


vector


nTxl


associated design matrices X


are


nTxp and


nT X


respectively 1


Without


loss of


generality,


also


assume


that


rank(X


- p.


Our objective


is to predict


(say)


on the


basis of


Y where S(uxp)


T(uxq)


are


known matrices.


Following the model-based


inference, it suffi

distribution of ()
V~


conditional


ces

= W


find


(say)


distribution


posterior


given


= y.


is provided


(conditional)


This


next


section.


We will


conclude


this


section


discussing a few


spec if ic


models


context


of comparative


trials


an imal


breeding which


are


special


cases


general


model


proposed


(2.2.1).


First


consider multicentered


clinical


trial


which


conducted


participating


clinics


to compare


two


treatments,


one


already


existing


the market


other


newly


developed.


Suppose


there


are


subjects


receiving the


treatment


jth


clini


Some of


nij. could
IJ


zero.


are


interested


in estimating the


SL? +














treatment


participating


clinic.


consider the


model:


+ Tij + eijk


Yijk


(2.2.8)


2 ,...


n.i ,
13


1,...,


where


mutually


independent


with


subject effects eijk
SJ k


are


N(O,

N(O,


treatment-clinic


(A2r)-1)


clinic


interaction


effects


, (A1r)-1)


the effect due


treatment.


Now we


will


write


down


for


(2.2.8).


ease


presentation,


assume


> 0


1 ,...,


Then


writing


S(Y1-11 '"


Y1in11


S. .


Y1cl'"'


Y1n1


~2c1'--


Y2cn2c)T


(l'I,


S711' **


2
s In.
i=l1


71c -* *


C
j=l n


C
) In .


el S..


elcl,...,


etnl


e2cn2c)T


A21 I2zC)~


In ,
- 7


(A1,


is clear that


A2)T


and D(A)


(2.2.8)


is a special


case


(2.2.1)


2
E "ij,'
* = _-


where


observations.


above


c and


want


C
="ij '


c
E nij
j=1


to estimate


= total


which


number


is a


* "


Diag(Alic,


, **


x,


Z
1


I-- 1>


pa~T


(C1,...,


TZc)T


(ellll,


"2 )


e2cl'''''


CI1-Cc2














used.


Once


again,


one


can


use


linear model


given


(2.2.8)


inferential


purpose.


case one may


interested


predicting some


suitable


linear functions of


random quantities,


c and
&J


7ij,


known


as the


breeding


values.


These


predicted


breeding values can


used


as a


selection


index for


selecting the


most


suitable


breeds for


future


breeding purpose.


a concrete


example


in animal


breeding we


will


discuss


example


considered


by Harville


press)


which


involves


prediction


the average


birth


weights of


infinite


number of


offspring of


single-birth male


different


sires


lambs


in different


that


are


population


lines.


The data consist of


weights


birth)


of 62 single-


birth


male


lambs,


caine


from five distinct


population


lines


Eac h


1amb


was


progeny of


one of


rams,


each


lamb


a different dam.


Age of


dam was


recorded


as belonging to one


three


categories,


numbered


(1-2


years),


(2-3


years


(over 3


years)


Yijkd


represent


weight


birth)


the dth


those


lambs


that


are


offspring of


sire


population


line


a dam


belonging to


age


category.


Following Harville


press),


we will


use














where d


1 ,...,


nijk


I. ,.,


= 1,


3 and


1 ,...,


5 where


nijk


is the


number of


lambs whose dams


belong


age


category when


population


line


and


si re


is k and m.
j


is the


total


number of


lambs


whose


sires are


from


population


line


Here


effects


line


effects


are


, .


considered


fixed


effects


sire


(within


line)


effects


Sjk


are


iid N(O,


(rA)-1) and


independent


error variables eijkd


which


are


N(O,


r-1)


To make


the design


matrix


associated


with


fixed


effects


full


rank we


can


take 63


= 0


= r5


which


usual


formulation


needed


for GLM


Procedures


in SAS.


= E(Yij kd)


=14+


+ and
kJ


there


n"i


observations


will


corresponding to


interested


category.


predicting


Wjk


pijni.


rj +


(2.2.10)


where


The


value


Wjk


can


interpreted


3
i n i .


average


birth


weight


infinite


number of


male


lambs


that


are


offspring of


sire


line.


as


(sl'














2.3


Hierarchical


Bayes Analysis


this


section,


for the


finite


population


sampling we


provide


predictive distribution


of Y


given


y(1)
y


and for the


infinite


population


set


up we


provide


posterior distribution


vector of


effects


, vT)T


given


= y.


We will


use


following notations


label


certain


distributions


used


this


section.


A random


variable


is said


have


a gamma(a,


distribution


f(z)


= [exp(-oz


random vector


a/(0) I[z>OJ]


, ,


Tp)T


is said


(2.3.1)


to have


mu It ivar i ate


t-distribution


with


location


parameter


scale


parameter c,


a p.d.


pxp matrix and


degrees


freedom


(d.f


v if


g(t)


- II)l~


cx ~ [


) J4(V+p)


(2.3.2)


(see


Zelliner,


1971,


383,


or Press,


1972,


136) .


Here


E denotes


the determinant of


a square matrix E.


Assume


v > 2.


Then


E(T)


=/"


V(T)


= (/(v-2)).


a nr-,e- I f


onnditi onsI


driven


at the


(1)


)~P-1


Sll...~


lr n.~nl


. .1A .i .i













(Cl)


AL ~


AtR are


independently


distributed


with


uniform(RP) ,


~ gamma 4a0,


gamma( ai,


2gi) with


> 0,


19...,


Allowing a&


some


zero


some


improper gamma


distributions are


included


as a possibility


our prior.


Before


stating the


predictive


distribution


given


notations.


we


write


need


E(A)


introduce


a few matrix


= + D(A)zT


partition


into


Also,


22.


-22


21 1112;
^l2C1112


-11


.3.3)


.3.4)


-~11-i -11~ )


--2


(X(1)T


V1 x(1))
11-


(1)Ti -1


(2.3.5)


- 22


-2


- 2
-2


S(-1 ))1)
1-11-


(2.3.6)


Now the


predictive


distribution


of Y(2)


given


given


in the


following theorem


in two


steps


250 '


of Y(2)


Y


+(x(


g0


(1)


x (1)TE
-Ili


-1
(l))(,(l)TC"`11


(1)














Theorem


2.3.1.


Consider


model


given


.2.2)


(cl).


Assume


t
Egi
i=O


that


Then


conditional


on A


multivariate


distribution


with


d.f


t
Z gi
i=0


- pt


location


parameter


scale


t
Ea1 1


parameter


+ (1)T (1)
J


ti
i=O


Also,


-1
-P


conditional


distribution


given


y (1))


u Cx


-A, (1)T


-11--


t
i ai.A +
i1-1


(1)TKy (1)


moments


of a multivariate


t-distribution


iterated


formulas


conditional


expectations


variances


follows


from


above


theorem


that


y (1))


-E(M


y(1))(1)


.3.8)


(1))


(1)


t
i=l


1
1sr
A!


t
i 0 i
i-0


Y


(2>


"T


My (1)


(1)


1
-i (nT+


ag+


,(y(2)


,(y(2)












= V My l Y + nT +


t
Egi
i=0


-2


t + (1)T (1)
i=l 1 1
1 .^^-


(1)


(2.3.9)


Using (2.3.8) and (2.3.9),


it is possible to find the


posterior mean and variance of


, Y(2))


4Y(1)


+ CY 2)


where A and C are known matrices.


The Bayes


estimate of


(1)


, (2)) under any quadratic


loss


posterior mean, and


is given by


BFsing (2.3.8)


using (2.3.8).


= A + MIy(1)(1)


Similarly,


(2.3.10)


using (2.3.9), one may obtain


= CV(Y y (2) ))CT


. (2.3.11)


Note that when A


m T
i=l 1


and C


,(2))


im T
i= lNi-ni


reduces to the vector of finite population totals for the m


smal1 areas,


il1i /N, ( Y(1)
1. N.-n. 1! -


areas


for the choice A


.ilTin /Ni) and C
J ^nn 1


S(2)) reduces to the vector of finite


population means for the m small areas.

Now we will get back to the infinite population set up


to provide the posterior distribution of W


,T)T
v )


t S S I S ** U I *


V ( 1)


(


I(u(')


~(y (1)

















theorem.


proof


similarity to


this

the


theorem will


proof


be omitted


of Theorem 2.3.1.


because

We will


consider the model


given


of Section


(cl).


Recall


from the


middle


of Section


that


have


redefined


the dimensions of Y


e appearing there


Y(n xl),


X(nTxp) ,


Z(nTxq)


e(nTxl)


Also


we have


assumed


rank(X)


= p.


Now we will


state


theorem.


Theorem


2.3.2.


Consider the model


stated


above


assume


that


t
E gi
i=O


- p


Then,


conditional


has multivariate


t-distribution


with


d.f.


nT +


nT +


t
E gi
i=O
t
i=gi
i=1


location


- p,


parameter


ai.Ai


scale


parameter


where


x(xTx-1X -1)xTr-1;
XX X X *


(2.3.12)


_= [-1x(xTy-1X)


QZD


2.3.


(xTE-1X)

-DZT -1X(x T


-(XTE-X)


-1)


xTE-1ZD


- DZTQZD


S C


2.3.


on


= -1


Y,


c-


x,


+ yT9y















f( iy)


I! xTslx


i=l
2'-.-
[{ A -


a.A.
1 1


yTQy


- C


(2.3.15)


Again


using the moments


iterated


formulas


of multivariate


for expectation


t-distribution,


variance,


above


theorem can


used


find


the


computational


formulas for


E(W|y)


V(Wly)


as


(2.3


(2.3.9).


Similarly,


one


can


find


eBI(


(say)


(2.3.10)


V( (b

Sb +


known matrices S(uxp)


.3.11)


where


and T(uxq).


Applications of


these


two


theorems will


considered


in Section


some


actual


data sets.


There


we will


carry


out


an


HB analysis of


data sets


which


appeared


Battese et


(1988)


Harvilie


press)


Before


we conclude


this


section ,


we will


make


a final


observation.


comparison


(2.3.4)


.3.7)


with


(2.3.12)


(2.3.15)


reveals


that


replace


by y(1)


X by


c by


f( ly(1))


as given


f(Aly)


(2.3.7).


(2.3.15)


we obtain


This observation


will


referred


in Section


2.4.


vi)


C(b,


x (1)














Applications of Hierarchical


Bayes


Analysis


This


section


concerns


analysis of


two


real


data


sets


using the


HB procedures


suggested


in Section


2.3.


first data set


soybeans for


related


12 counties


prediction


north-central


corn and


Iowa based


1978 June


Enumerative


Survey


as we 11


as LANDSAT


satellite


data .


It appeared


in Battese,


Harter


Fuller


(BHF)


who


conducted


a variance


components


analysis for thi


problem.


second


data


set


original 1


appeared


Harville


Fenech


(1985)


reappeared


in Harville


press)


where


he conducted


a variance


components


as well


as an HB


analysis


predict


Wjk


, given


(2.2.10)


ave rage


weight of


an


infinite


number


of single-birth male


lambs


that are offspring of


sire


population


line.


We will


first


consider


data set.


start


with


briefly


give


a background


this


problem.


USDA Statistical


Reporting Service


field


staff


determined


area of


corn


soybeans


in 37


sample


segments


(each


segment


about


hectares)


12 counties


in north-central


Iowa by


interviewing farm operators.


Based


on LANDSAT


readings obtained during August


September


1978,


USDA














number


June


of hectares of


Enumerative


corn


Survey),


soybeans


number of


reported


pixels


classified


as corn


soybeans for each sample


segment,


county mean


number of


pixels classified as


corn


soybeans


(the


total


number of


pixels


classified


as that


crop divided


number of


segments


that


county)


are


reported


in Table


of BHF


ready


reference,


reproduced


comparable


Table


that


2.1.


orde r to make


of BHF


second


our


segment


results


Hard in


county


was


ignored


model


considered


by BHF


bl 1 ij


+ b x2ij


+ e,.,
1J


(2.4.1)


where


a subscript


for the


county,


a subscript


a segment


within


given


county


number


segments


county,


...


12) .


Here


Xl ii


number of


pixels


corn


the


number


of pixels


of soybeans


for the


segment


in the


county.


They


assumed


our


notations)


E(vi)


= E(eij)


= 0,


Cov(vi 1


V(vi)


(Ar)


v.,)


, Cov(vi,


Cov(e1J,


i .1')


eij)


= 0,


= 0


, j').


are


interested


oo. ,


bg


Vt













Table


2.1


Survey
in 12


and Satellite
Iowa Counties


Data for Corn


Soybeans


Segments


Reported
hectares


No.
in


of pixels
sample


segments


Mean no. of
pixels per
segments


County


Sample


County


Corn


Soybean Corn Soybean Corn


Soybean


Cerro Gordo
Hamilton
Worth
Humboldt


Franklin


Pocahontas


Winnebago


Wright


Webster


374
209
253
432
367
361
288
369
206
316
145
355
295
223
459
290
307
252
293
206
302
313
246
353
271
237
221
369
343
342
294
220
340
355


Hancock


Kossuth


Hard in


55
218
250
96
178
137
206
165
218
221
338
128
147
204
77
217
258
303
221
222
274
190
270
172
228
297
167
191
249
182
179
262
87
160


295.29
300.40
289.60
290.74

318.21


257.17


291.77


301.26


262.17



314.28




298.65




325.99


189
196
205
220


188.06


247.13


185.37


221.36


247.09



198.66




204.61




177.05
















can


be written


as


where


N.
-1 1
= Ni E eij
*j=l ~


+ bRip)
1 li(p)


+ b2R2i(p)


Xli(p)


N N.
-1 1
= N. xlij and
j=-1


N.
= Nix2ij.
J=l


Under the


assumptions of


model


(2.4


.1),


can


interpreted


as the


conditional


mean


hectares of


corn


(or soybeans)


per segment,


given


realized


county


effect


values of


satellite


data.

because


Clearly,


average


segments


county


mean


not


equivalent


over the


not


finite


identically 0.


population


However,


either


if N


i are


m) are


appropriate


conditions


large


small,


predictor


appears


or


then


of VY.

to be


sampling


rates


predictor of


this example


true .


n i/N


i is an

either


predicting


first


assuming


A and


r known,


obtained


BLUPs


12) .


Then,


using


Henderson 's Method


, they


obtained


estimates of


variance


components,


final


predictors


involved


estimated


variance


components.


Henderson 's method


being an


ANOVA method


could


lead


to negative


estimates


this were


case,


set


it equal


to zero.


This


phenomenon


likely to


I,. n C


nfl r+- 4 i-nt 1r ttn r .,tI


~~~~11


- ~ -


r~~~~~~ ~~ n1 a np it 'i f-'T U u r..4I


. .


+ 73i


nllm kar nC


h Yna 1













special


case


of nested


error


regression


model,


we will


now develop


expressions


for the


posterior distribution


for the


posterior means


given


(2.3


expressions


variances of Y


Here,


we have


DO)


'Lu,


Then


m In
e(Ini


A'1J~.)


so that 111

n.


+ ni)/A}


Also,


writing


= nil


1,...,


where xij
-ii


one


gets


x(1)T -1 (1)
x r ,x


nm 2n
- i= I 1 1


_
Si


-T
i (s)- 1 (s)


= H(A)


say


(2.4.2)


Next


writing


i(s)


-1
1


one


gets


(1)T


n
=m .
= i= Ij


- n
i=1 j


AS'!' ("


=1-X ij ij


x H-1m
1


=1 j=1-


- "i(n


-1
A) y
i(s)


-T

Yi(s)


=1 x; jX


n
illl((x


Xlij Y2ij~T


(1)
Ky


- "i("i












f(1))


1 1 1
-2(m+ )-1 m "2 2
Aml)- n 0(A + ni) H(A)
I 1||H|--


- (nT+g+g -p))


a + alA + QO(A))


(2.4.4)


Next writing fi


- (N.
1


- ni)/Ni


/ \ 1 N.
= tN.-n. E. x..
1 j=ni+1-ij


posterior means, variances and covariances of the finite

population means are given by



1


- fi(s) + fiE


ni(ni + A)


+ f.E
1


{-i


- n.(ni + A)
1 11


T
-i (s)} ()


Xn.
x Em Tijpx x j(Yij


- ni(n. + A)
~1


- eHB


(say


(2.4.5)


Ni Ni (1)


= fV


ni(n.+A)


fi(s) + {T-ni(ni+A)
1(s) 1 1 1


-i(s)}


i(s>















-n.(n.+A)


iT


say


=2
SHB


.4.6)


it


Cov


N1
1


Ej:*'Yii


-1k .
k j=l kjy-


f.f iCov
1 k


ni(n.+A)
I~1


+ {8T-ni(ni+A)


x H-(A)



nk(nk+A)


-n.(n +A)
1 ~11


*~


-nk(nk+A)


H-1()
H1 (A)


x k=l


"kj(


-nk (nk+A)


(nT+go+g1


-p-2)


-1
fifkE


(ao+alA+Q o(A))


-ni(ni+A)


H-(A:)(


-nk (nk+A)


x (Ni+A)Ni(n,+A)


x H-1)(


2


I I


~T-n i(ni+~>


"ij(yij


r













covariances which may


be necessary for providing


simultaneous


confidence


set


finite


population mean


vector,


we will


not


use


it here.


Before


we find


posterior means and


variances of


in the


infinite


population


set


we give


a general


discussion


comparing HB


predictors with


EB predictors.


Writing


N.

j=l 1J


- f.)
1


i(s)


f. ^N
i ~it


fini(nl
+ -1

+ A)-1


A)-1
Yi(s)


)H-1 )


i(s


x m

i1l


- in


= gl(A)


(say),


(2.4.8)


1 j= 1J
N~Yi.


(1)
-y


(nT + go


- p


- 2)-lf{aO


+ A +(


f


- "i("i


j(Y;j














= g2()


(say),


(2.4.9)


we have


from


(2.4.5),


(2.4.6),


(2.4.8)


(2.4.9)


that


SEgi(A)yl ,]


(2.4.10)


= v[gl()y(1),


(2.4.11)


(2.4.12)


In EB analysis,


to obtain


EB predictor,


usually


replace


Egl (A)I) (1)


by gl(A)


= eEB


(say)


where


is some


estimate


of A,


which


can


be ML,


REML or


ANOVA


estimate


report


a naive measure


of posterior variance


g2(A)


(say).


Usually,


point


estimates


eHB


are


not


too


apart.


to measure


posterior variance


2
5EB'


we may


underestimate


actual


measure


because


failure


to account


for the estimation


We may


grossly underestimate


actual


measure


if gl(A)


varies


too much


within


body


posterior distribution


this


case


will


be significantly


large.


We will


see


this example


that


for some


of the


counties


= Eg (A)y(1) .













2B
SHB


usually


increases with


relative difference


between


eHB


eEB.


Now to develop


expressions


for the


posterior means


and variances for


. we use
1


Theorem 2.3.2.


Note


that


observation made


posterior distribution


of Section


given


2.3,


is given


f(Ajy),


(2.4.4).


After


considerable


simplifications


this


particular


case,


we obtain


ECUi lIJ


= E ni(ni


Af-


S)Y


{ i(p)


- "i(n1


A)-x. }
/ ~i(s)


H-1()
-H (A)


- ni~e1


- eHB


(say)


(2.4.13)


VP ilR


ni(n.


-1 (s)
A1 sf


+ X. (
Si(p)


- ni(ni


Af 1-1


i=1


(nT +


- Ti.n


-p -2)-1E


A) -I


a{o


+ alA


m
i=


+ O


v


+ E


j(yij


+ gl
















_ H*2
- SHB


(say)


(2.4.14)


where


2i(p))T


, ()


as given


(2.4.2)


%Q(A)


xli(p)'

as given


(2.4.3).


Note


that


since


>K
-x .


- 0


as N.


can


seen


informal 1


that


rhs of


(2.4.5)


(2.4.6)


approach


limit


(2.4


(2.4.14),


respectively.


We will


now get


back to


actual


data analysis of


data set


given


in Table


2.1.


use


formulas


(2.4.5)


, (2.4.6)


, (2.4.8),


(2.4.9),


(2.4.13)


(2.4.14)


to obtain


HB and


posterior means


and variances


population means


for the


12 counties.


HB approach


eliminates the


possibility


obtaining zero estimates of


variance


components.


A number of


different


priors for


R and


were


tried;


both


informative


noninformative


results


for the


posterior means


were


quite


similar


whereas


posterior variances


varied


approximately


much


as


10%.


i 1 1 lustration


purpose,


have decided


report


our analysis


for the


prior with a0


= 0.005,


= 0,


= 0.005 and


= 0.


since


choice


= 0


gives


improper posterior distribution


we took a1












Table


Predicted


Hectares of


Corn


and Associated


Standard


Errors


.005


.005


County eHB eEB eBHF sHB SEB SBHF


Cerro Gordo 122.1 122.2 122.2 9.3 9.4 10.3
Franklin 143.6 144.2 145.3 6.9 6.4 6.7
Hamilton 126.2 126.2 126.5 9.2 9.3 10.1
Hancock 124.6 124.4 124.2 5.3 5.3 5.5
Hardin 142.6 143.0 143.5 5.8 5.6 5.8
Humboldt 108.9 108.5 107.7 8.2 7.9 8.4
Kossuth 107.7 106.9 106.1 5.8 5.2 5.4
Pocahontas 111.8 112.1 112.9 6.6 6.4 6.8
Webster 114.9 115.3 116.0 5.9 5.7 6.0
Winnebago 113.3 112.8 112.1 6.6 6.4 6.8
Worth 107.1 106.8 105.6 9.9 9.1 10.0
Wright 122.0 122.0 122.1 6.4 6.5 6.9















eBHF and


respective associated


standard


errors


sHB'


SBHF


for the


corn


data.


Table


2.3


provides


the values


eHB,


eEB,


eBHF and


eHB


for the


soybeans data


for the


same


choice of


prior


hyperparameters,


whereas


Table


2.4


provides


their


respective


standard


errors along with


components


of SHB.


Values of


eBHF and


SBHF


presented


in Tables


-2.4


are


computed


using FORTRAN


from the


formulas given


paper


are


slightly


different


from


values


reported


Battese


(1988).


From


Tables


2.3,


for


predicting corn


soybeans,


one


can


see


that


eHB


, eHB,


eBHF are quite


close


to each


other.


From


Tables


2.2


2.4,


SEB and


appear to


be smaller


than


5BEIF


But since


sEB


is naive


posterior


s.d. ,


is probably


underestimate


true


measure.

difference


From


Tables


either


2.3 and


between


eHB


we find


eB
eHB


hardly


or between


any

their


standard


errors


SHB.


This


is what


we anticipated


for this data.


To draw a clear


comparison


between


HB and


EB procedures,


we


added


one


extra column


the end


Tables


2.3


2.4.


last


column


of Table


2.3 measures


percent


relative


difference


x IeHB


- eEBI/


between


EB and


HB predicted


values


whereas


last column












Table


2.3 The


Predicted


Hectares of


Soybeans Obtained


Using Different


.005


Procedures


.005


County eHB eHB eEB eBHF l eEB/eHB xl00%


Cerro Gordo 78.8 78.8 78.2 77.5 0.78
Franklin 67.1 67.1 65.9 64.8 1.80
Hamilton 94.4 94.4 94.6 96.0 0.21
Hancock 100.4 100.4 100.8 101.1 0.40
Hardin 75. 4 75.4 75 .1 74.9 0.39
Humboldt 81.9 82.0 80.6 79.2 1.71
Kossuth 118.2 118.2 119.2 120.2 0.84
Pocahontas 113.9 113.9 113.7 113.8 0.18
Webster 110.0 110.0 109.7 109.6 0.37
Winnebago 97.3 97.3 98.0 98.7 0.72
Worth 87.8 87.8 87.2 86.6 0.68
Wright 111.9 111.9 112.4 112.9 0.45













Table


2.4


Standard


Errors Associated


with


Different


Predictors


of Hectares


of Soybeans


.005


.005


County SHB SHB SEB SBHF V1 V2 V1/(V1+V2)xl00%


Cerro Gordo 11.7 11.7 11.6 12.7 7.67 128.59 5.1
Franklin 8.2 8.2 7.5 7.8 11.94 54.92 18.0
Hamilton 11.2 11.2 11.4 12.4 1.97 123.61 1.6
Hancock 6.2 6.3 6.1 6.3 1.35 37.59 3.4
Hardin 6.5 6.5 6.5 6.6 0.37 41.84 0.9
Humboldt 10.4 10.4 9.9 10.0 22.62 85.40 20.9
Kossuth 6.6 6.7 6.0 6.2 7.99 36.23 18.1
Pocahontas 7.5 7.5 7.5 7.9 0.06 55.98 0.1
Webster 6.6 6.7 6.6 6.8 0.64 43.51 1.5
Winnebago 7.7 7.8 7.5 7.9 4.11 55.70 6.9
Worth 11.1 11.1 11.1 12.1 4.06 118.17 3.3
Wright 7.7 7.7 7.6 8.0 1.62 57.48 2.7














contribution


of V1


usually


increases with


the


relative


difference


In particular,


for


counties


Franklin,


Humboldt


and Kossuth


1.80%,


these


1.71% and


relative differences are


.84% and


as high


corresponding contributions


of V1

made


are


as nonnegligible


SEB much


smaller than


18.0%,


SHB


20.9% and


for these


18.1%.


This


counties.


one


uses


a naive


EB or


estimated


BLUP


approach,


he will


tend


to underestimate


mean


squared


error


(MSE)


prediction.


should


note


that


though


BHF


used


estimated


BLUP


, they tried


to account


for the


uncertainty


involved


estimation


their


approximations


of MSE.


Similar approximations


of MSE of


prediction


have


been


suggested


Kackar


Harville


(1984),


Prasad


(1990)


Lahiri


(1990).


Now we


will


consider the


lamb-weight data set


Harville


given


press)


example


The


background


presented


the data set


of Section


2.2.


We will


use


a model 1


similar to


the one


given


(2.2.9)


analyze


the data set.


There


assumed,


following Harville


press)


population


line


effects


as fixed.


For the


purpose


will


illustration


assume


with


population


three


effects


variance


random.


components,


we


This would














(age


of dam)


full


column


rank,


we will


write


Now we


have


following mixed


linear model


Yijkd


+ sjk


=pi


+ eijkd


(2.4.15)


1 ,...,


r1ijk'


3 and


1 ,...,


1,...,


where


"ijk


are


are


and mj

, (rA1)

, r-1)


are


-1),


same


are


Moreover


as


(2.2.9)


N(O,


assume


(rA2-1) and


eijkd


are


eijkd


assumed


to be mutually


independent


We want


to predict


given


2.2.


10).


Using


(2.4.15) ,


we will


rewrite


~~1


ni..
1 .*


+ rj + sjk


(2.4.16)


where


nT are


given


(2.2.10).


We will


carry


out


a noninformative


Bayesian


analysis


using


a uniform(R3)


prior


independent


gamma( ao,


gamma( a,


~51) and


gamma( La2,


1~2


priors


RA2


respectively.


Using Theorem


fixed


E(wjk


2.3.2,


random effect


= eHB


(say)


being a


we can


linear


find


combination


posterior mean


posterior variance


V(wjk


- --


-~~ -- -


r I T a 1 14 1. .


p + 6i


as


- S


C -


Wi k


Wi k


pg~T


(Ccl,


40)


C 1














example


choice


0 of


hype rparamete rs


the variance


components .


gives a noninformative


choice


prior for


= 0 or


= 0 will


give


improper posterior distribution of


(A1,


we tried


several


combinations of


these


hyperparameters


which


are small


positive


numbers.


Our


findings


for this data set,


provided


Table


2.5,


are


not


different


from


data set.


report our


analysis


- 0.0005,


- 0.05


- 0.01


in Table


2.6.


The estimated


BLUPs


for w13


w56


reported


in Harville


press)


are


10.98 and


10.29


respectively,


whereas


corresponding values we


obtained


using a


noninformative


HB analysis are


11.0


10.4


respectively.


agreement


between


two


sets


of estimates


remarkably


close


considering the


fact


that


underlying


models


2.2.


(2.4.15)


are


not


identical.


Harvilie


press)


also


estimated


the difference


- w56


associated


MSE of


prediction


using


both


variance


components approach


HB approach.


The


estimated


given


MSE of


(0.955)2


- w56


whereas


naive


for


EBLUP


Kackar


approach


Harvilie


was


(1984)


approximation it


was


(1.053)


for Prasad


Rao


(1990)


"1.


&2


g0


gl


82


gl


g2













Table


Birth Weights (in pounds) of Lambs


Sire Dam Age Weight Sire Dam Age Weight


Line 1


Line 4


6.2
13.0
9.5
10.1
11.4
11.8
12.9
13.1
10.4


9.2
10.6
10.6
7.7
10.0
11.2
10.2
10.9
11.7
9.9


8.5


Line
1
2


Line


13.5
10.1
11.0
14.0
15.5
12.0


Line 3


11.7
12.6
9.0
11.0
9.0
12.0
9.9
13.5
10.9
5.9
10.0
12.7
13.2
13.3
10.7
11.0
12.5
9.0
10.2


10.0













Table


Predicted
Standard


Birth
Errors


Weights


of Lambs and


Associated


aO = .0005


g0=O


a = .05


gl=0


a2 = .01
21


g2=0


Line Sire e* s*
eHB 9HB


10.1
10.9
11.0
10.4


11.9
11.7


10.8
10.8
11.3
11.1

10.2
10.5
10.5


0.90
0.86
0.62
0.80


0.88
0.71
0.87
0.80


0.70
0.53
0.75
0.82

0.62
0.80
0.79


11.2
10.7
10.8
10.8
11.3
10.4
11.4
10.8














error.


HB estimate of


Wi3


in this


case


reported


1.042.


as 0.69


posterior s.d.


corresponding values


obtained


reported


using our


approach


are


0.60


0.99


respectively.


To conclude


this


section,


we can


recommend


from


whatever


sets


that


have


learned


from


noninformative


analysis of


HB method


these data


is clearly


a viable


alternative


usual


EB or variance


components


approach,


should


given


every


serious


consideration


prediction


both


finite


population


sampling and


infinite


population


situation.


Hierarchical


Bayes


Prediction


Finite


Population


Mean


Vector


in Absence of


Unit


Level


Observations


Sometimes


is either difficult


impossible


obtain


information


unit


level


for


small


areas.


this


section


we will


derive


predictor


finite


population


mean


vector when


we do


not


have


observations


unit


level.


1 ,...,


small


area with


units,


assume


that


based


on a sample of


size


know


only


sample mean


characteristic


inte rest,


sample mean


auxiliary variables.


Also


vYi(s)
vector


we


have


i(s)


(pxl)


information


on the


we


- "56














population mean
N.


Nl1 Y
1 j=(s)"'


based


vector


( 1 ...,


on (s)'""


Vm)T


Vm(s))


= 7 where


(say),


m(s)T
-rm(s))


5 ,)


-m(p))


YiN.)
1


1,...,


Consider the


following model.


Conditional


on ??


(N.xl),


(pxl)


RilIN.i)


, 9


independently,


where


(ii)


are


Conditional


known


sampling variances.


b and


lIiNi)


independently


(iii)B


A are


independent


uniform(RP) and


a prior

gamma( a,


with


Z-g


Combining


(ii)


we


have


conditional


b and


~ N(X ib,
~1-


(R +6-1)IN
iI


independently.


Carter


Rolph


(1974)


introduced


this


type


of model


Herriot


(1979)


considered


EB approach


this


problem


a special


case


- e1^


assumed


place


(ii)


that


conditional


S1,...,


independently.


Subsequently,


place


(i')


they


assumed


that


conditional


, Y.
-1


NT/ /..TL \


fllT -I


- 1 -


, .


1;' _


9


(Yilr.


N(?i.


N(Xib ,


("iTb


n-lr


ffi














variance,


they


estimated


iteratively


applying


generalized


least


squares


procedure


to Y. (s)
i(s)


~ N(xTb,


-1


+ 6-1


1,...,


independently where


Yi(s)


is the


sample mean


based


on ni


units.


They


estimated


= 1i'


- 1,...,


based


on their superpopulation model


whereas


are


interested


predicting the


finite


population mean


1,...,


based


on


(i')


(iii).


Now we


will


back to our


problem.


For the


sake


notational


simplicity,


we will


assume without


any


loss


generality that


units


is given


sample mean


by Pi(s)


Yi(s)
n.


=- ni1.


is based


Now


first


define


S


-"~i.-


N.
.E
j=ni.+1
+1


N1

j=ni+1


T
"i1)


Y1j'


have


-T
-i(s)


"n
-1= X
n1 1.2i,'
1 ---4 1


also defined


-T(
-.,u)


in Section


that


1 ,...,


Since


vector 7,


- ni)/Ni

f .(u)

is enough


known,


find


to predict


predictor of


Y(u) "


Ym(u))
enough


(say).


find


to predict


the


vector


predictive distribution


given


sample mean


vector


(say).


any


quadrat i


OSS,


predictor c


is given


we


E(Pi)


fi)pi (












and its posterior variance is given by


= f f(a IY)( .


(2.5.2)


Now from (i), given b and 6, (s)'""


m(u))


m(s)'


is multivariate normal (MVN) with mean 4


and variance Diag(a-,..., 2rm) where i


= X-(s)b,' i+m


-T b 2
"i(u)~' i


= (R1 + 6-1n. and a2
0i+m


(R1 + (N


i = 1,..., m. From this, it is easy to derive that given

b and 6, a is MVN with


E("il (s) b 6)


=x (u)b
-i u)


(2.5.3)


Cov(ai, aki(s), b, 6)


= +mik
i +m i k


(2.5.4)


where 6;i


the Kronecker delta which is 1 if i


= k and


zero otherwise.

Using the iterative formulas for expectation and


variance we have from (2.5.1)


E(ri z(s))


- (2.5.4)


- fi)Y(s) + fEs)' b (s)


- fi)i() + fE i(u by ))


- fi)i(s) + fiT(u)Ebl_(s))


(2.5.5)


Y1 Y


i>













v ri (s))


=1 1 (ils))


=f2
-


E V(a i s)


b 6)s)+
- -(s)-


VE a. V ,b,
~ -(s)- (s


i t E+m i(s) (u) V"( I(s))Si (u)


(2.5.6)


Note


that


from


(ii),


have


(iv)


given


b and


' (s)


where


(2.5.7)


Xrn(


2 m)


2.5.8)


from


(iii)


we can


write


given


uniform(RP).


From


(iv)


we have


joint


of Y( )
~(s)


B given


(s)'


b16)


i (1/-)exp 1y( -Ab)V-(
=1


(2.5.9)


assume


rank(A)


p and


define


= (ATv-1A)-1ATv-1


-T/ 2
Xi
(s)1-x1


-1
m
E-1


(2.5.10)


m
\-1


N(Ab,


(Xl(s>'''''


Diag(~ 2~,,,,


(,)-Ab~.


'i(s)"i(9)/D?


(2.5.10)












(2.5.9) can be written as


fy -AbV- 1 -Ab
S(s) -- -(s) ~~


(-bT(AV-1A )(b-b) + sQy(
-b~T(4 ()-s


(2.5.12)


From (2.5.9) and


(2.5.12)


, it follows that given


Y(s) and


(ATV-1A)-1)
(\ ^ _)- i)


Note that in (2.5.10) that b


depends on ? ( and 6 since a2 depend on 6.

Again using the iterative formulas for expectation and

variance we have


E[ Iv(


-1
*ER. ()&(s)1 /) 10i/<1 1
1 1 / \ -X /


Y(s)


(2.5.13)


= EEVTy l)-1 + VE(By ()
-- ( s ) (+


T 2
(s) i(s)/ 8i


mi


N(1?,


E[E(BIP~,). 6)e(


E(B IP~s))


E


\r(B IP(9))


E













to evaluate


E(il Ys)) and


it follows


from


(2.5.5),


(2.5.6)


, (2.5.13)


(2.5.14)


that


enough


to evaluate


E(1+ml(s)) and


the quantities that


appear

evaluate


on the

e them,


(2.5.13)


we need


find


and

the


(2.5.14).


conditional


order


distribution


of A given


From


(iii),


(iv)


(v),


joint


of (s)


is given


f (s)'


o c (1/.i)exp -l(
[?ii "o]^


( -AbV)Ty(-1 ( Ab) 62
(s) ~~ ~ -(s) -~~


exp(- ab .


.15)


Using


~ N(b,


2.5.


(ATVl1A )31


fact


that


given


integrating out


f(s)
from


(2.5.15)


that


joint


of Y(


is given


m 1{
i=l A ~


S-1 1
62 A I V-12
s) ~


2.5.


Since


f(s))


cx (P9. )


the


conditional


IO~s)) ~


f(y(s) .


9s
s>


f(y(g) ,
f(v















.5.14)


are


accomplished


now


using


.16)


typically


some


numerical


integration


techniques.


















CHAPTER THREE
OPTIMALITY OF BAYES PREDICTORS


FOR MEANS


3.1


IN A SPECIAL CASE


Introduction


In Chapter


Two


a hierarchical


Bayes


procedure was


introduced


prediction


in mixed


linear models,


Section


results


were


utilized


for


prediction


purpose


both


population


set


finite


in th


population

e presence


sampling and


auxiliary


infinite

information.


There


considered


general


case


unknown


variance


components and


derived


posterior distributions of


interest

fixed ef


assigning


fects and


independent


gamma priors


uniform


prior


inverse


variance


components.


this


chapter,


we will


consider


a special


case.


assume


that


ratios


variance


components


are


known.


We derive


HB predictors


for the mean


vector


prove


some


optimal


properties of


this


predictor.


In Section


(2.2.2)


3.2,


of Section


we consider the


with


normal


vector of


linear model


ratios


variance


components,


known.


assign


a uniform


prior to













posterior distribution


nonsampled


units given


sampled units

we derive the


finite


population


HB predictor of


sampling and

e finite pop


from this


ulation mean


vector.


Later


this


section,


infinite


population


situation,


posterior distribution


vector of


fixed


random effects and


HB predictors


for


linear


combinations


fixed


random effects are determined.


Our


approach


these


problems


can


regarded


extensions of


ideas


Lindley


Smith


(1972)


prediction.


Although


developed


within


a Bayesian


framework,


our


results


should


be of


appeal


also


frequentists.


both


problems,


BLUP notion


for


real


valued


parameters


(see,


for example,


Henderson


, 1963;


Royal 1,


1976)


extended


in Sections 3.3 and


3.4


to vector valued


parameters,


is shown


that


Bayesian


predictors


Section


are


indeed


BLUP


Like


other


related


papers,


our


BLUP


results do


not


require


any


normality


assumption.


With


added


assumption


of normality,


BLUPs


indeed


turn


out


best


unbiased


predictors


(BUPs)


within


the


class of a

that these


unbiased


Bayes


predictors.


predictors are


addition,


BUPs even


is shown


for some


nonnormal


distributions.


In these


sections.


we have


also













distributions.


In Sections 3.5 and 3.6 we


have


shown


that


these


Bayes


predictors


are


best


equivariant


predictors


for


both


the matrix


loss


(or standardized matrix


loss)


quadratic


loss


(or standardized


quadratic


loss)


under


suitable


groups of


elliptically


transformations for


symmetric


distributions,


a broad


class of


including but


not


limited


normal


distribution.


We conclude


this


section


introducing a few


notations.


a square matrix T


(txt),


tr(T)


denotes


trace.


a symmetric


nonnegative


definite


(n.n.d.)


matrix


is a symmetric


n .n.d.


matrix such


1 1
that T2T2


for


a symmetric


p.d.


matrix T


,1
, T 2


is a symmetric


p.d.


matrix such


that


= (T1


The


Hierarchical


Bayes


Predictor


a Special


Case


We will


assume


normal


linear model


(2.2.2)


Section


when A

known,


2.2.


the

while


We consider


vector of


B and


R are


this section


ratios of va

independently


special


r ance


case


components,


distributed


with


uniform(R )


- gamma( aO,


Here we


will


cons


ider the


case


of finite


population


sampling


in details


1I U Cl 'A I... t... 1


i, \
~s0j


I I


.... 11


F


1


I












still


interested


finding the


predictor of


(recall 1

suffices


that (y(1)


SAy(1)


find


+ C (2)),


predictive distribution


id for this ii

of Y given


Recall


notations K,


G given


(2.3.4)


- (2.3.6).


following


Theorem


Since


3.2.


is known


instead


this case,


we have


of Theorem 2.3.1.


The


proof


of Theorem 3.2.1


is similar to


that of Theorem 2.3.1


is omitted.


Theorem 3.2.1.


Assume


that


Then


under


the model


given


independent


prior for R,


uniform(R )


in Section


prior for B and


predictive distribution


with


known,


gamma( ao,


of Y(2)


given


8g0)

y(1)


is multivariate


t-distribution


with


d.f.


nT +go -P'


location


paramet


er My(1)


scale


parameter (n + go


-K


x (ao


y(1)TKy(1))G


Using the


properties of


the multivariate


distribution,


possible


now to


obtain


closed


form


expressions


for


= y1) and


= y(1).


particular,


Bayes


estimate


of


, (2))


(1)
( y


under


any


quadratic


loss


is now


w1Ven


= (1)


S(1)(


, y(2))


_ y (1)


y(2))


(1>


+ g0


~(y(2)


U











We may note


that


predictor


eBF(()) given


(3.2.1)


is the


outcome


the model


given


and


with


known,


and the


not depend


use of


on the


uniform(R )


choice


the


prior on


prior


B and


(proper)


it does


distribution


of R.


This


can


formally seen,


assuming all


the


expectations


appearing


below


exist,


as follows:


Y R


+(x(2)


+(x


- 52


I. (1)


'ic;1X')


1 (1))
E X^
-11~


(3.2.2)


qualities,


second


equality


follows


from


fact


that


conditional


R=r


(1)b), r-1


-I-


S22
-2^


the fourth


foll


ows


from the


fact


that


conditional


on R


E Y(2)Y())


E{E(y2) B,


E X (2)B +


E 11,


-2


x ((1)T


= My( )


where


in the


above


string of


y(1)


(1)


.1 ,


Cz


E


''')


Y


X


~1(y(l)


(1)


''')


L~Z1C;:11


(1)


-1 (1)
1C11Y


-1 (1>
1C11Y


(1>T _1 (1.)
x C,,y


(2>
b+ C21E;ll\y-


(2)














definition

e* (Y(1)
~BFl ~


of M,


robust


given


(2.3.5).


against


Thus,


choice


the


of prior


predictor

hs for R.


There


are


alternate ways


to generate


same


predictor eBF(Y (1) of


^ w(1)


Suppose,


for


example,


one


assumes onl


be known).


Then


and

best


with


predi


ctor


known

(best


(r may

linear


or may not

predictor


without


normality


assumption)


,(2))


of ((1)


sense


of having the


smallest mean


squared


error matrix


given


SAy(1)
-AY


+ (x(2)


-1 1
~ 21-11-() 3


a.e.,


(3.2.3)


where


(b).


say that


for two


symmetric


matrices


if F


is n.n.d


If b


unknown,


then


one


replaces


UMVUE


(BLUE without

x(1)T _y1 (1)
X E ,Y


normality


assumption)


resulting


predictor of


1.II -


, (2)) turns out


1 a1 an anf


to be e F((1))


PR nredic tnr of


y(2))~


1(Y(1)


i c[EzlE -1.i:Y (1)


(''')


(X(1)TC,1 X''')


f(Y(1)


"Pnnp p













Similarly,


in this special


case,


one


can


derive


predictor


(for quadratic


loss)


of C(b,


context


infinite


population


set


Denoting this


HB predictor


y)IY)


e*I(Y),
*B


one


can


see


that


the


arguments


leading to


empirical


Baye s


interpretation


of eBF(y())


work equally well


to show that


ei~1(Y)


also


possesses the


empirical


Bayes


interpretation.


Harvilie


(1985,


1988,


press)


recognized


this


for


predicting


scalars.


next


four


sections,


we will


discuss


a few


frequentist


3.3 we


show


properties

e* () i
~BFI" ) -


of egF(y(1)) and


best


unbiased


eBI(Y).


Section


predictor


consider


its


stochastic domination


whereas


in Section 3.4


we consider these


properties


for egI(Y).


In Section 3.5 we


show that e YF(1)y )


is best equivariant


predictor of


, y(2))


under suitable


groups of


transformations,


whereas


property


in Section


of eoI(Y)


3.6 we


under


consider the


same


best


groups


equ variance


transforma-


tions.

scalar


Jeske

BLUPs


and

are


Harville


best


(1987)


equivariant


have

with


shown

n the


that

class


the

of all


linear


equ variant


predictors


without any


distributional


assumption.


However,


to our


knowledge,


the equivariance


results


for vector valued


predictors


have


not


been


E((b,


i(y(f )











Best


Unbiased


Prediction


and Stochastic


Domination


in Small


Area Estimation


this


section,


we


assume


normal


linear model


(2.2.2)


with


known.


No prior distribution


for


B and


assumed,


, r)T


is treated


as an


unknown


parameter.


within


First,


class of


prove


unbiased


optimality


of e y(1)


predictors of


, (2))


Next,


we dispense


with


the


normality


assumption


eF(1y()) within


and

the


e, and prove

class of all


optimality


linear


unbiased


predictors


(LUPs).


We start


with


following definition


of a best


unbiased


predictor


(BUP).


Definition


3.3.1.


A predictor T(Y(1))


said


to be a BUP


, (2))


if E [T(Y ')


for every predictor


b(Y()) of


.(1)


= O for


, Y (2)) satisfying


E[6(Y<(1))


0 ''


'p


2))


Y (2))]
, Yovided

provided


n. n.d.


- V T(Y )


(1)


the quantities are


(2))]

finite.


following general


lemma plays a key


proving the


best


unbiasedness of


predictor


eBF(. ) of


,., I 1 I.1 -


of ((Y)


E(yf i)


-~(Y(1)


Y(


(,(1)

_ f(u(')


(2)\


.1! .. L


Ir I


vii











assume


that


each


component


of g(Y)


has a finite


second


moment


Denote


by Ug ,


class


of all


unbiased


predictors


6Yl) of


g(Y)


with


each


component of


6(Y(1)) having a


finite


real


second moment.


valued


Also,


statistics (i


*e,


denote


class of


functions of Y (1)) with


finite


second


moments


having


zero


expectations


identically


in 9.


Lemma 3.3


A predictor T(Y (1))


E Ug


is BUP for g(Y)


only


Cov[T(Y(1))


=9


.3.1)


for


every


Proof


Lemma


.3.1


TT
Tu(Y(1~)


et T(Y)

6u(Y1))


(_)_


If Y(1))


is another pre


I-dictor


then


(Y(1))


a)


+ Cov T(Y(1))


+ Cov ((Y(1)


-g(Y)


(3.3.2)


*..., .


~


~(''',3


(y(l)~


VB~(Y(1))


VB~(y(l))


6(Y (1))


T(Y(1))


-T(r(l))~












Cov T(Y


-6(Y(1))


= 0.


(3.3.3)


From


(3.3.2)


(3.3.3)


follows


that


V0[ (Y(1)


Sv0[r(x<1))


- g(Y) +


- TY( ) ,


(3.3.4)


for all


Hence


T(Y(1))


is BUP for


Only


Given


that


is BUP


we will


show that


condition


3.3.


true.


First


we will


show that

>t UiY (1)


is BUP for gi(Y)


for


every


1,...,


any


unbiased


predictor for


Then 6y Y(1)),


a u-component


column


vector with


component


equal


to U (1)),


belongs


to Ug.


Then


is n.n.d.


So we


have


V Ui((Y'1))


- i(Y)]


- i(Y)


> o,


S ,


consequently


is BUP for


Now following


usual


Lehmann-Scheffe


(1950)


technique


(also


Rao,


Ti(Y(I))


- g(Y)


g(y;]


Vl~(y(l))


g(y>


T(u(l))


Ti(U(1))


gi(y>


g(Y ~


V


~(1''',


g(Y ~


Ve[Ti(Y('))


Ti(Y(1))


gi(y>











Hence,


(3.3.1)


holds,


the


proof


lemma


is complete.


Remark 3.3.1.


follows from the


above


lemma


(see


(3.3.4))


that


if-T) and T2(Y(1)) are


both


BUPs of


then


r6'''IC


= CovO Tl(Y('))


C- ov T2(Y(1))


= 0


(Y),


- g(Y)


(3.3.1)


- 2(Y('))


T2(Y (1))


Ilu~)


for all


- 2(Y(1)


Remark


3.3.2.


is also


clear that


technique of


above


lemma can


applied


more


general


contexts.


We will


use


above


lemma


prove


BUP


property


of e*F(y)


following theorem.


Recal 1


from


(3.2.1)


that


eBF(yl))


Theorem 3.3.


Under


normal


linear model


(2.2.2),


*- a -


i.e.,


e; (Y)


(Tl(y(l))


Tl(u(l))


Pb~l(Y(1))


+ CM)Y


,, (2)\


f LI.











Proof


of Theorem


3.3.


view


of Lemma 3.3.1,


suffices to


show that


for every m(Y(1)


Cov[ e(Y(1))


that


is EG C(MY
O ~- --


-y (2))(Y (1))]


= f'
-


= O for


or all


Since,


under the model


2.2.2


- E4Y~2~


(- -l


, using


E,.(Y (1))


it suff


ices


to show that


(3.3.6)


Since


E,(Y(1))


= 0


-x())


s-i


- 1)2


- 0,


differentiating


both


sides


of this


equation


w.r.t


one


gets


(see


318 of Rao,


1973)


/X (1)T-1 (1)


-1 (1)
-2r (Jl


- X b)


11 (1)
E11 y


d(1)
dy


= 0.


(3.3.7)


2


E (X (1)Ty (1) m(y (1)


/o(


x exp


- f(r(l)


t u (2))


MY(1)


x(l>TC-1. (1)
Y11Y


,e11 X(1))


(X(2)


(1)
dy


_L2 r(,(l)


(l))exp/


b,


X (I)b)m(y(l))


X (l)b)













Remark 3.3.3.


Equation


(3.3.6)


can


be alternatively proved


r-1E
--11)


following way.

(x1)T (1)


Note


that


, Y(1)TKy (1))


since


is complete


suffic


ient


for 0


Hence (1)T1 (1)
Hence X' E Y


must


have Q covariance


vector


with


every


zero estimator m(Y ()),


i.e.,


E[(X (1)T(1))m()]


=9.


Next


we show that


conclusion


of Theorem 3.3.1


continues


to hold


even


certain


nonnormal


distributions.


Suppose


, T)T


that


= Diag(D,


Assume


that


given


N(O,


r-1A),


while


the df


of R


is an


arbitrary member of


family


is absolutely


continuous with


f(r)


= 0 for


r < O}.


denote


subfamily


of 1 such


that each


component


of egF((1)) and


(.71)


Sy(2))


finite


second


moment


under the model


(2.2.2)


joint distribution


of e*


now


prove


following theorem.


Theorem 3.3.2.


model


(2.2.2),


eBF(Y())


is BUP


~ N(O,


of t

r-1a)


, (2)) under the


R has


a df


from


Proof


of Theorem 3.3.2.


Using


Lemma 3.3.1,


following


proof


of Theorem 3.3.1,


it suffices


to show that


r,


N(X(l)t2


y (1)


(y (1)











(3.3.9)


Eb ,F[(1))


Eb, m2( (1)


00 for all


b and


Consider


subfamily


Sgamma( c


2} of


, d)


Since


(3.3.9)


holds


for this


subfamily fl,


EkF^m(1)


36 gives


exp(


1 (n +d)-1
-2c r


x exp


- Xb)T


-1 (1)
E ny


- (~b


LIr(
-2^


xrnY~))v'~ d


= 0


(3.3.10)


c >


0 and


> 2.


using the


uniqueness


property


of Laplace


transforms,


follows


from


(3.3.10)


that


5(nT+d)-1


exp


- X()b)T


1-
--r
2


- X b)


x m(


a.e.


Lebesgue


Jexp


1,
-2


r > O and


- X b)T
- -


- (~b


x m(y


(1) dy=


(3.3.11)


*e,


(


c > 0,


ci:(r


(I)),y


C 1 l\y_












simplifications


using


(3.3.11)


lead


x exp


SX(1)b)T


_1-ry


-111,1


- X ( b)


(3.3.12)


a.e.


Lebesgue


0 and


Multiplying both


sides of


(3.3.12)


1


integrating with


respect


dF(r)


where


one


gets


(3.3.8).


Remark 3.3.4.


Since


does


not


contain


the degenerate


distributions of


R on


(0,oo) ,


Theorem 3.3.1


does


not


follow


from


Theorem 3.3.2.


Remark 3.3.5.


In Theorem 3.3.2


we take


for F*


,we


see


that


marginal


distribution


of Y


given


by the


family


of distributions


(c/d)


> 0,


this


> 2}


family where


distribution


with


eBF (Y)

N*I|NT, Xl

location


BUP for


(c/d)E,


parameter


, (2)) for


is NT-variate


scale


parameter


(c/d)


and d.f.


Next


will

Y(1)


show that


predictor e~F(Y() (which


linear


a best


linear


unbiased


predictor


/(X(1)TFi:r fl))m(y(l))


x dy


Xb,


~(U(')


t -











, we say


that (Y(1))


is a LUP


,(2))


of (Y1)


need the


following definition.


Definition 3


.3.2.


A LUP Py(


, (2))


of (( )


is said


be a BLUF

V(HY1)
VHY
- --


n.n.d.


for


every


LUP HY(1)


, y (2))


, Y<2-


of ((1)


S- -(


Sv(py1
a ~~


for all


now


prove


the BLUP property


of eBF(y1) for


predicting


(1)


, (2))


this


end,


we will


state


lemma whose


proof


is similar to


proof


of Lemma 3


.3.1


hence


proof


will


be omitted.


Lemma


3.3


A LUP


, y(2))


of ((1)


- V(1


a BLUP


if and


only


, (2))


Cov0PY
O -


, Ty())
mY}


=9


.3.13)


for all


every


known


nTxl


vector m


satisfying


E(mTy (1))


= 0 for


The


following theorem


provides


BLUP


property


-BF(- )) for


predicting f(Y


, y(2)).


proving this


BLUP


property


we do


not


need


any


distributional


assumption


on e


*. We only


assume


=9


ae~


) = r-14


_~(u(l)


_ E(Y(I)


Eg(e













Proof


of Theorem 3.3.3.


If E0,(mTy())


= mTx (l)b


= 0 for


mTX<(1)


= OT


Hence,


, (2))


Cov Y ))


= CovC( MY (1)


-y (2)),


Ty (1)]


for all


last


two qualities


follow from


definition


of M and


from


fact mnTx(1)


Applying


Lemma 3.3.2,


Remark 3.3.6.


result


follows.


already mentioned,


normal ity


assumption

* Y(1).


is not


needed


Theorem 3.3.3


proving the


unifies


BLUP property


extends


available


BLUP


results


related


estimation


finite


population mean


vector


under different models


(cf.


Ghosh


Lahiri,


in press;


Royall


, 1976;


others)


Remark 3.3.1


one


can


prove


that


BLUP


unique


with


probability


one.


, Ty(1)


S


C(X


~(y(l)


C21)m


C (MFI1


-1 x (1))(X (1>T -1 X''')
- C21C11 C11


,(1)T


T
O












- a)(


(3.3.14)


the model


(2.2.2)


without any


distributional


assumption


on e


The


optimality


of eF(y (1) within


LUPs


holds


a fortiori


under


quadrati


loss


2
-in


- C


tr[gLo (,


(3.3.15)


where


is a n.n. d.


matr ix.


Such


loss


will,


henceforth,


be referred


to as generalized


Euclidean


error w.r.t.


optimality


Theorem 3.3.2


results


under the


carry


added


over via


Theorem 3.3.1


distributional


assumption


(which


not


necessarily normality


assumption)


on e


natural


question


to ask


now


is whether the


risk optimality


of eF(Y (1))

predictors,


holds


or at


with in


least


class of


within


class


unbiased


LUPs


under


certain


other


criterion


a broader family


distributions of


notions of


"un iversal"


investigate


this question,


"stochast ic"


we need


domination,


their


interrelationship


as given


Hwang


(1985).


It. (R F


= p a I I-ti''--


risk


16


Lo


Li(rl


-~)Tn(h


u(2)\12\


- FIV(1)


IF; L~IIC


I


.


I












w.r.t.


O for


some


function


The


following definition


adapted


from Hwang


(1985).


Definition 3.3.3.

dominates 2(Y (1))


w.r.t.


An estimator


(under the


for every ,


general ized


6l(Y () universally


Euclidean error


every nondecreasing


loss


function


holds and


for


particular


RLo(,

loss,


risk functions are


not


identical.


Hwang


(1985)


shown


that


(see


Theorem


2.3)


universally


dominates


under the


generalized


Euclidean


error w.r.t


2
(2))I
' -Y


if 6(Y(1))


stochastically


smaller than (62(Y )


2
Y(2)
, Y-


say that


a random variable


is stochastically


smaller


than


if Po(Z1


> x)


PO (Z2


> x)


for


for some


have distinct distributions.


next


theorem shows


that


for


a general


ass


elliptically symmetric distributions of


, eBF(1)


universal ly


dominates


every


, (2)) under


every


generalized Euclidean


error w.r. t.


a n .n.d.


Assume


that


has an


elliptically symmetric


given


Ir-1 -Afre*T A-le*


(3.3.16)


h (e* IA,


RL(8


- f(Y(1)


(U(')


HY (1)


o,














q
|vi. +
i=1


NT
i=le i +


f(re*TA-le
f( -


*)de*


(3.3.17)


where


(vl,...,


vq)T


(el,...,


eNT)T


We will


denote


this distribution


by 8f(O,


r-1A)


where


n*<2)


denotes


the distribution


whose


is given


k(tl|,


Qa *2 f((t


- ) Tg*-l(t


- )/o2)


(3.3.18)


where


are


, g*(pxp)


is p.d.


Note


that


normality


with


mean


variance-covariance matrix


r-1A


sufficient


not


necessary for


(3.3.16)


(3.3.17)


to hold.


follows


from


(3.3.17)


Note


that


that


**


exists


1)

with


distribution


.1
2*+


from


(3.3.16)


has a spherically


characteristic


function


symmetric


(c.f.)


E exp(iuT e


= c (uu)


some


function


(see


Kelker,


1970)


where


=47j,


(Ul,...,


= NT


qv)


+ q.


Hence


c.f.


given


E[exp(iuT e


= c(r-1uTAu).


(3.3.19)


(4' *t I


B)


"f( II ,


gf(0


,,~














where


+ ZDZT


Comparing


(3.3.16)


(3.3.19)


one


can


see


from


(3.3.20)


that


has also an


elliptically


symmetric distribution


with


given


h(w*I ,


Ir-1 2 (rw


*TE-lw*).


(3.3.21)


Theorem 3.3.4.


Under the model


(2.2.2),


(3.3.16)


(3.3.17),


eBF (1))


universally


dominates every


LUP


Y, 2)) for


of (y1)


(y(l))


every p.d.


Remark


3.3.7.


Theorem 3.3.4 does


not


contain


Theorem


3.3.3


since


Theorem 3.3.4


distribution


of W*


requires


, while


the elliptical


other does


symmetry


not.


of the


It should


noted


though


that


model


assumption made


(3.3.16)


not


necessarily


stronger than


usual


assumption


finiteness of


certain moments.


assumptions of Theorem 3.3.4


This


hold


even


is because


if a distribution


infinite


second


moment


(e.g.,


for


certain multivariate


BLUP property


is meaningless


in such


instance.


Now we will


Theorem 3.3.4


state


rests


prove


crucially


lemma.


on this


proof


lemma.


Lemma 3.3.3.


If W(NTXl)


then


e


r-1C)


gf(0,


HY(I)


)













Proof


of Lemma


3.3


proof


follows


arguments


Hwang


(1985)


From


(3.3


.19)


follows


that


E[exp( itTW)


= c(r


-it T


Hence


E[exp( itTLW)]


= c(r


-1 TLL
~1--


Ttl)


(3.3


.22)


where


is a uxl


vector


Next


using


(3.3


.22),


exp(


T
it (LL
~1(- -


^T) u1
T)2w


1
T)' (Iu


expl itT(LL
-~1 (


= c(r


-1iT -(L


1(
T- (Iu


O)(Iu


O)T(LL


-tT (LLT)tl)
~1-Wi-


(3.3.23)


so that


lemma


follows


from


(3.3


.22)


.3.23).


follows


as a consequence


of Lemma


3.3.3


that


WTLTLW


(LW) T(LW)


1
T) Wu


(,


T) w


= WTLL W .


(3.3


.24)


We shall


use


.3.24)


repeatedly


for


proving


Theorem


3.3.4.


'-i'


= c(r


) tl1


E


5


o)w












- A)


- CX


2) (3


.3.25)


Writing


_1
= E 2W*


using


.3.24)


.3.25)


one


HY(i)


2T
2))]


(y(l)


Y


_HY_(1)
SHY


(y(1)


- cT [H


-C *


= T
a.


1


c]Tn{H


- C]


1T
WOH CIL


- C]


.3.26)


Similarly,


*
i3F(Y('))


(y(1)


2))]


eBF(Y())


(y(1)


= W*T CM


= wT
a.


E- CM


- CT nCM


- C-T -CM


W1
WTQ2[ucM..


1
CT 2V .


- C]


.3.27)


Write


- CM


Then,


.3.26)


.3.27)


T+ W
wuo FEr


rT2W- u


.3.28)


a.cC~


- C *


(2
Y


Ut


Y(2











T
- C)Ef )
O )


-l 11 T
s21


Ss21) T


- C(MEi


-1 1T-1
- a x Yx 0^ x
~21-11- A- ~11~


x(1)TET


(3.3.29)


using


(3.3.25),


-CM)(1)x
- -ll~


-MX(1))


= c(x(2)


= 0.


(3.3.30)


Theorem 3.3.4


follows


now from


(3.3.26)


- (3.3.28).


Also,


since


1 is positive


definite,


follows


from


(3.3.28)


that


rhs of


(3.3.26)


rhs of


(3.3.27)


only


=9,


that


is H


(cf.


Hwang,


1985).


3.4


Best


Unbiased Prediction


and Stochastic


Domination


Infinite


Population


In this section


will


briefly


consider


a few


optimal


properties


which


of ei (Y)


are


similar to


those of


eBF (Y) following closely Section 3.3.


First,


we note


that


-eBI(Y)


is optimal


within


class


of all


unbiased


predictors of


under


normal


linear model


with


known.


finite


population


case,


no prior


distribution


B and


is assigned,


, r)T


treated


as an unknown


parameter.


Next,


dispensing with


= c(x 2)


if r


c(M


rX(1)


A + CM














Definition 3.4.1.


be an


unbiased


A predictor 6(Y)


predictor


if E [6(Y)


of C(b,

- C(b ,


is said


= 0 for all


An unbiased


predictor U(Y)


of C(b


, v)


is said


to be


for


v0e (Y)


-V


every unbiased


V- -U(Y)
O-V-


predictor


- c(b,


6(Y)


of C(b,


is n.n.d.


for


provided


the quantiti


exist


finitely.


Recall


that


-0b


following


lemma


is analogous


to Lemma 3.3.1,


concerns


characterization


of a BUP


based


on


for some


known


function


g where each


component


has a


finite


second moment.


Lemma 3.4.1


An


Es[T (Y)U (Y)
0- -


unbiased


predictor U(Y)


BUP for


g(W)


of g(W)


with


only


Cov{U(Y)

statist i


m(Y)]


m(Y)


such


that


every


Eo(m(Y))


0 and


E0[m2(Y)


oo00 for


Lemma 3.4.1


can


proved


similarly


Lemma 3.3.1


proof


is omitted.


We will


use


this


lemma


to sketch


a proof


of the


following theorem which


concerns


best


unbiased


prediction


of C(b,


Theorem 3.4.1.


Under


normal


linear model


(2.2.1),


SBI(Y)


is the


BUP


of C(b,


of g(W)


- g(W)












v)}m(Y)~


=
J-


for


(3.4.1)


Note,


however that with


PO-probabi 1 ity


E[SBI(Y)


-eiI(Y)


TDZT-1 (y
- -


- _Xb)


S(XT


- TDZT


1X


(3.4.2)


From


(3.4.1)


(3.4


it suffices to show that


E [(XT


= 0 for


This


is proved


similar to


(3.3.6).


Remark 3.4.1.


conclusion


of the


above


theorem holds


even


certain


nonnormal


distributions.


Theorem


3.3.2,


one


can


show that


e;I(Y)


of C(b


v, )


under the


model


(2.2.1)


where e*


~N(O,


r-1A)


R has a df


from r*


where e*


are


same


in Theorem


3.3


Next,


note


that


predictor eiI(Y)


linear


can


proved


as


in Theorem 3.3.3 that


is BLUP


under


linear model


(2.2


without


any


E {eBI (Y)


, v)


sb


xb)


ly),(y~












Now we


will


show that


131(Y)


dominates


universally


of ((b,


an elliptically symmetric distribution


Consider the


generalized


Euclidean


error


loss


w.r.t.


a uxu


p.d.


matrix 0


2
- cIn


- ).


- C(b,


v


(3.4.3)


risk


function


predictor 6


for


predicting (


under a


loss


function


which


is a


function


of generalized


Euclidean


error


w.r.t.


O for


some


function


following definition


similar


to Definition


3 3.3.


Definition


3.4.2.


An estimator


61(Y)


universally


dominates


another


estimator


62(Y)


(under the


generalized


Euclidean


error w.r.t.


for every 0


every nondecreasing


function


, C;


holds and


for a


particular

Now we


RL( ,

loss,

will


the r

state


isk functions are


not


identical


following theorem on


stochastic


domination


of e*I(Y)


proof


will


be omitted


because


its similarity to


Theorem 3.3.4.


of e*


16


L1


C>To (


(y)


RL(B


Ee


RL (B




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E37PE6NDK_36QGW0 INGEST_TIME 2017-07-11T21:20:23Z PACKAGE AA00002109_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

%$<(6,$1 35(',&7,21 ,1 0,;(' /,1($5 02'(/6 :,7+ $33/,&$7,216 ,1 60$// $5($ (67,0$7,21 %< *$85, 6$1.$5 '$77$ $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$ /LFLWR 81,9(56,7< 2) )/25,'$

PAGE 2

WR P\ SDUHQWV DQG WHDFKHUV ZLWK UHJDUGV

PAGE 3

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

PAGE 4

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fV VWXGHQW ZRXOG OLNH WR WKDQN P\ SDUHQWV IRU DOO WKH\ KDYH GRQH IRU PH DP GHHSO\ LQGHEWHG WR P\ QXPHURXV ZHOO ZLVKHUV DQG IULHQGV ZLWKRXW ZKRVH JHQHURXV VXSSRUW LW ZRXOG KDYH EHHQ LPSRVVLEOH IRU PH WR SXUVXH P\ KLJKHU VWXGLHV ZRXOG OLNH WR H[SUHVV P\ VLQFHUH JUDWLWXGH HVSHFLDOO\ WR ,9

PAGE 5

0U %LEHNDQDQGD 1DQGL 3URIHVVRU .0 6HQDSDWL 0UV 'XUJD 6HQDSDWL 0U .& *KRVK DQG 0UV %LPDOD *KRVK ZKR KDYH DOZD\V FRQVLGHUHG PH DV D SDUW RI fWKHLU IDPLO\ DOVR FRQVLGHU P\VHOI H[WUHPHO\ OXFN\ WR JHW WKH DIIHFWLRQDWH FRQFHUQ RI 0UV 6HQDSDWL ZKR KDV DOZD\V WUHDWHG PH OLNH KHU RZQ VRQ 0\ KHDUWIHOW WKDQNV DUH IRU P\ fXQRIILFLDOf KRVW IDPLO\ LQ *DLQHVYLOOH 'U 0DOD\ *KRVK DQG KLV ZLIH RXU EHORYHG 'RODGL ZRXOG OLNH WR WKDQN 'U $3 5H]QHN RI WKH 8QLWHG 6WDWHV &HQVXV %XUHDX IRU SURYLGLQJ PH ZLWK WKH FRPSXWLQJ IDFLOLWLHV GXULQJ P\ VWD\ LQ WKH %XUHDX DV DQ $6$16) 5HVHDUFK $VVRFLDWH /DVW EXW QRW OHDVW ZRXOG OLNH WR WKDQN 0V &LQG\ =LPPHUPDQ IRU KHU VNLOOIXO W\SLQJ LQ SXWWLQJ D VFULEEOHG PDQXVFULSW LQWR ILQDO IRUP Y

PAGE 6

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

PAGE 7

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f f 9 %,2*5$3+,&$/ 6.(7&+

PAGE 8

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fSDUDPHWHUVf VXFK DV WKH PHDQ RU WKH YDULDQFH IRU WKH DUHD QHHG WR fERUURZ VWUHQJWKf IURP VLPLODU QHLJKERULQJ DUHDV LPSOLFLWO\ RU H[SOLFLWO\ WKURXJK D PRGHO 6XFK HVWLPDWRUV XVXDOO\ KDYH D VPDOOHU PHDQ VTXDUHG HUURU RI SUHGLFWLRQ WKDQ WKH VXUYH\ HVWLPDWRUV ,Q WKLV GLVVHUWDWLRQ D JHQHUDO KLHUDUFKLFDO %D\HV +%f PRGHO LV FRQVLGHUHG IRU VPDOO DUHD HVWLPDWLRQ 6RPH RI WKH ZLGHO\ XVHG PRGHOV LQ VPDOO DUHD HVWLPDWLRQ LQFOXGLQJ WKH QHVWHG HUURU UHJUHVVLRQ PRGHO UDQGRP Y L L

PAGE 9

UHJUHVVLRQ FRHIILFLHQWV PRGHO HWF FRQVLGHUHG E\ HDUOLHU DXWKRUV DUH VHHQ WR EH VSHFLDO FDVHV RI WKH SURSRVHG JHQHUDO PRGHO 7KH SUHGLFWLYH GLVWULEXWLRQ RI D FKDUDFWHULVWLF RI LQWHUHVW IRU WKH XQVDPSOHG SRSXODWLRQ XQLWV LV IRXQG JLYHQ WKH REVHUYDWLRQV RQ WKH VDPSOHG XQLWV DQG LV XVHG WR GUDZ LQIHUHQFH ,Q SDUWLFXODU VLPXOWDQHRXV HVWLPDWRUV RI VHYHUDO VPDOO DUHD PHDQV DQG YDULDQFHV DUH GHYHORSHG $ PL[HG OLQHDU PRGHO ZLWK QRQ LQIRUPDWLYH SULRU IRU UHJUHVVLRQ FRHIILFLHQWV RU IL[HG HIIHFWVf DQG LQGHSHQGHQW JDPPD SULRUV SRVVLEO\ QRQLQIRUPDWLYHf IRU WKH LQYHUVH RI WKH YDULDQFH FRPSRQHQWV LV XVHG ,Q D VSHFLDO FDVH RI WKLV +% DQDO\VLV ZKHQ WKH YHFWRU RI WKH UDWLRV RI WKH YDULDQFH FRPSRQHQWV LV NQRZQ WKH +% SUHGLFWRU RI WKH YHFWRU RI PHDQV LQ ILQLWH SRSXODWLRQ VDPSOLQJ LV VKRZQ WR SRVVHVV VRPH IUHTXHQWLVW RSWLPDO SURSHUWLHV VXFK DV EHVW XQELDVHG SUHGLFWRU EHVW HTXLYDULDQW SUHGLFWRU HWFf EDVLFDOO\ XQGHU WKH HOOLSWLFDO V\PPHWU\ DVVXPSWLRQV 3HUIRUPDQFH RI WKLV +% SUHGLFWRU LV HYDOXDWHG E\ FRPSDULQJ LWV %D\HV ULVN ZLWK WKDW RI VXEMHFWLYH %D\HV SUHGLFWRU ZLWK fWUXHf RU fHOLFLWHGf SULRU IRU WKH XQNQRZQ VXSHUSRSXODWLRQ SDUDPHWHUV ,W LV VKRZQ WKDW XQGHU D EDODQFHG RQHZD\ UDQGRP HIIHFWV PRGHO ZLWK FRYDULDWHV DQG DYHUDJH VTXDUHG HUURU ORVV WKH GLIIHUHQFH LQ WKH %D\HV ULVNV RI WKH +% SUHGLFWRU DQG WKH fWUXHf %D\HV SUHGLFWRU RI WKH ILQLWH SRSXODWLRQ PHDQ RU YDULDQFH YHFWRU DSSURDFKHV ]HUR DV WKH QXPEHU RI VPDOO DUHDV EHFRPHV LQFUHDVLQJO\ DUJH ,;

PAGE 10

&+$37(5 21( ,1752'8&7,21 /LWHUDWXUH 5HYLHZ 8VH RI OLQHDU PRGHOV E\ DVWURQRPHUV IRU SUHGLFWLQJ WKH SRVLWLRQV RI FHOHVWLDO ERGLHV JRHV EDFN VHYHUDO FHQWXULHV 6WDUWLQJ IURP WKHVH GD\V WKH XVH RI PRGHOEDVHG LQIHUHQFH IRU SUHGLFWLRQ KDV UHFHLYHG FRQVLGHUDEOH DWWHQWLRQ ,Q SDUWLFXODU DQLPDO DQG SODQW EUHHGHUV KDYH XVHG VXFK PRGHOV IRU SUHGLFWLQJ VRPH FKDUDFWHULVWLFV RI WKH IXWXUH SURJHQ\ 6WDUWLQJ ZLWK WKH SLRQHHULQJ ZRUN RI +HQGHUVRQ f FRQVLGHUDEOH DWWHQWLRQ KDV EHHQ GHYRWHG WR WKLV SUREOHP :H UHIHU WR *LDQROD DQG )HUQDQGR f DQG +DUYLOOH LQ SUHVVf ZKHUH RWKHU UHIHUHQFHV DUH FLWHG 2Q WKH RWKHU KDQG VXUYH\ DQDO\VWV KDYH XVHG WKH PRGHOEDVHG DSSURDFK LQ ILQLWH SRSXODWLRQ VDPSOLQJ ZLWK WKH JRDO RI SUHGLFWLQJ FHUWDLQ FKDUDFWHULVWLFV RI WKH XQVDPSOHG XQLWV LQ WKH SRSXODWLRQ RQ WKH EDVLV RI WKH REVHUYHG VDPSOH (DUO\ ZRUN RQ WKLV WRSLF PD\ EH IRXQG LQ &RFKUDQ f ZKHUH WKH ILQLWH SRSXODWLRQ LV YLHZHG DV D UHDOL]DWLRQ IURP D K\SRWKHWLFDO VXSHUSRSXODWLRQ ,Q UHFHQW \HDUV VPDOO DUHD GRPDLQf HVWLPDWLRQ KDV JURZQ LQWR DQ LPSRUWDQW WRSLF LQ VXUYH\ VDPSOLQJ 8VH RI

PAGE 11

VPDOO DUHD VWDWLVWLFV ZDV LQ H[LVWHQFH DV HDUO\ DV WKH WK FHQWXU\ LQ (QJODQG DQG WKH WK FHQWXU\ LQ &DQDGD VHH %UDFNVWRQH f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f ZDV OLPLWHG EHFDXVH WKH HVWLPDWHV IRU WKHVH VPDOO DUHDV XVXDOO\ ZHUH EDVHG RQ VPDOO VDPSOHV DQG SURGXFHG XQDFFHSWDEO\ ODUJH VWDQGDUG HUURUV DQG FRHIILFLHQWV RI YDULDWLRQ 7R LPSURYH WKH UHOLDELOLW\ RI WKH VPDOO DUHD VWDWLVWLFV LW LV QHFHVVDU\ WR KDYH D PXFK ODUJHU VDPSOH VL]H IRU DQ LQGLYLGXDO DUHD WKDQ FDQ EH DIIRUGHG ZLWK WKH OLPLWHG UHVRXUFHV DYDLODEOH &RQVHTXHQWO\ WKH XVH RI VXUYH\ GDWD SRVVLEO\ LQ DVVRFLDWLRQ ZLWK WKH FHQVXV GDWDf LQ SURGXFLQJ UHOLDEOH VPDOO DUHD VWDWLVWLFV GLG QRW UHFHLYH PXFK DWWHQWLRQ

PAGE 12

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fERUURZ VWUHQJWKf IURP VLPLODU QHLJKERULQJ DUHDV IRU HVWLPDWLRQ DQG SUHGLFWLRQ SXUSRVHV 7KURXJK XVH RI VRPH DSSURSULDWH PRGHO DQG DX[LOLDU\ LQIRUPDWLRQ SRVVLEO\ REWDLQHG WKURXJK FRPSOHWH HQXPHUDWLRQ IRU H[DPSOH FHQVXV RU VDWHOOLWHf VPDOO DUHD HVWLPDWRUV RI WKH SDUDPHWHUV RI LQWHUHVW VXFK DV WKH ILQLWH SRSXODWLRQ PHDQ YDULDQFH HWFf XVXDOO\ LPSURYH

PAGE 13

RYHU WKH VXUYH\ HVWLPDWRUV )RU D JRRG UHYLHZ RI WKH VPDOO DUHD HVWLPDWLRQ OLWHUDWXUH RQH PD\ UHIHU WR *KRVK DQG 5DR f 7KH QHFHVVLW\ RI fERUURZLQJ VWUHQJWKf KDV EHHQ UHDOL]HG E\ PDQ\ VWDWLVWLFLDQV (ULFNVHQ f DGYRFDWHG WKH XVH RI UHJUHVVLRQ PHWKRG IRU HVWLPDWLQJ SRSXODWLRQ FKDQJHV RI ORFDO DUHDV )D\ DQG +HUULRW f SURSRVHG DQ DGDSWDWLRQ RI WKH -DPHV6WHLQ HVWLPDWRU WR VXUYH\ HVWLPDWHV RI LQFRPH IRU VPDOO DUHDV 6XUYH\ HVWLPDWHV EHLQJ EDVHG RQ D VPDOO VDPSOH VL]H ZKLFK LV XVXDOO\ SHUFHQW RI SRSXODWLRQ RI VL]H OHVV WKDQ f XVXDOO\ KDYH ODUJH VWDQGDUG HUURUV DQG FRHIILFLHQWV RI YDULDWLRQ 7R UHFWLI\ WKLV WKHVH DXWKRUV ILUVW ILW D UHJUHVVLRQ HTXDWLRQ WR WKH FHQVXV VDPSOH HVWLPDWHV XVLQJ DV LQGHSHQGHQW YDULDEOHV WKH FRXQW\ YDOXHV WD[ UHWXUQ GDWD IRU WKH \HDU DQG GDWD IRU KRXVLQJ IURP WKH FHQVXV 7KH HVWLPDWH WKH\ SURYLGHG IRU HDFK SODFH ZDV D ZHLJKWHG DYHUDJH RI WKH VDPSOH HVWLPDWH DQG WKH UHJUHVVLRQ HVWLPDWH %DWWHVH +DUWHU DQG )XOOHU f FRQVLGHUHG SUHGLFWLRQ RI DUHDV XQGHU FRUQ DQG VR\EHDQV IRU FRXQWLHV LQ QRUWKFHQWUDO ,RZD EDVHG RQ -XQH (QXPHUDWLYH 6XUYH\ DQG /$1'6$7 VDWHOOLWH GDWD %DWWHVH +DUWHU DQG )XOOHU %+)f XVHG D OLQHDU UHJUHVVLRQ PRGHO GHILQLQJ D UHODWLRQVKLS EHWZHHQ WKH VXUYH\ DQG VDWHOOLWH GDWD DQG XVHG WKLV UHODWLRQVKLS WR REWDLQ SUHGLFWRUV RI PHDQ FURS DUHDV SHU VHJPHQW LQ WKH

PAGE 14

VDPSOHG FRXQWLHV )XOOHU DQG +DUWHU f DOVR FRQVLGHUHG D PXOWLYDULDWH H[WHQVLRQ R WKLV PRGHO 7KHUH LV D VLPLODU SUREOHP RI SUHGLFWLRQ IDFHG E\ WKH DQLPDO EUHHGHUV )RU WKH SXUSRVH RI VHOHFWLQJ WKH EHVW DQLPDOV IRU IXWXUH EUHHGLQJ WKH\ QHHG WR FRPH XS ZLWK DQ LQGH[ IRU HDFK DQLPDO XQGHU FRQVLGHUDWLRQ +HQGHUVRQ f DGYRFDWHG WKH XVH RI EHVW OLQHDU XQELDVHG SUHGLFWRU %/83f RI FHUWDLQ OLQHDU FRPELQDWLRQV RI IL[HG DQG UDQGRP HIIHFWV XVLQJ D PL[HG OLQHDU PRGHO +DUYLOOH LQ SUHVVf XVHG D PL[HG OLQHDU PRGHO IRU SUHGLFWLQJ WKH DYHUDJH ZHLJKW RI VLQJOHELUWK PDOH ODPEV ZKLFK DUH SURJHQ\ RI VLUHV EHORQJLQJ WR GLIIHUHQW SRSXODWLRQ OLQHV DQG GDPV EHORQJLQJ WR GLIIHUHQW DJH FDWHJRULHV +DUYLOOH DQG )HQHFK f FRQVLGHUHG WKLV H[DPSOH IRU HVWLPDWLQJ WKH KHULWDE L LW\
PAGE 15

7KH PHWKRGV WKDW KDYH XVXDOO\ EHHQ SURSRVHG LQ PRGHO EDVHG LQIHUHQFH XVH HLWKHU D YDULDQFH FRPSRQHQWV DSSURDFK RU DQ HPSLULFDO %D\HV (%f DSSURDFK DOWKRXJK DV SRLQWHG RXW E\ +DUYL Hn LQ SUHVVf WKH GLVWLQFWLRQ EHWZHHQ WKH WZR LV RIWHQ VXSHUIOXRXV %RWK WKHVH SURFHGXUHV XVH FHUWDLQ PL[HG OLQHDU PRGHOV IRU SUHGLFWLRQ SXUSRVHV )LUVW DVVXPLQJ WKH YDULDQFH FRPSRQHQWV WR EH NQRZQ FHUWDLQ %/83V RU (% SUHGLFWRUV DUH REWDLQHG IRU WKH XQNQRZQ SDUDPHWHUV RI LQWHUHVW 7KHQ WKH XQNQRZQ YDULDQFH FRPSRQHQWV DUH HVWLPDWHG W\SLFDOO\ E\ +HQGHUVRQfV PHWKRG RI ILWWLQJ RI FRQVWDQWV RU WKH UHVWULFWHG PD[LPXP OLNHOLKRRG 5(0/f PHWKRG 7KH UHVXOWLQJ HVWLPDWRUV ZKLFK FDQ EH FDOOHG HVWLPDWHG %/83V RU (%/83V VHH +DUYLOOH f DUH XVHG IRU ILQDO SUHGLFWLRQ SXUSRVHV (PSLULFDO %D\HV DSSURDFK LQ VPDOO DUHD HVWLPDWLRQ ZDV ILUVW JLYHQ LQ )D\ DQG +HUULRW f DQG ODWHU DOVR XVHG E\ *KRVK DQG 0HHGHQ f *KRVK DQG /DKLUL D f DPRQJ RWKHUV $FFRUGLQJ WR WKLV SURFHGXUH ILUVW D %D\HV HVWLPDWH RI WKH XQNQRZQ SDUDPHWHU RI LQWHUHVW LV REWDLQHG E\ XVLQJ D QRUPDO SULRU RU XVLQJ D OLQHDU %D\HV DUJXPHQW +DUWLJDQ f 7KH XQNQRZQ SDUDPHWHUV RI WKH SULRU DUH WKHQ HVWLPDWHG E\ VRPH FODVVLFDO PHWKRGV OLNH WKH PHWKRG RI PRPHQWV PHWKRG RI PD[LPXP OLNHOLKRRG RU VRPH FRPELQDWLRQ WKHUHRI 7KH UHVXOWLQJ HVWLPDWRU RI WKH SDUDPHWHU RI LQWHUHVW LV WKH VRFDOOHG (% HVWLPDWRU

PAGE 16

$OWKRXJK WKH DERYH DSSURDFK RI (%/83 RU (% LV XVXDOO\ TXLWH VDWLVIDFWRU\ IRU SRLQW SUHGLFWLRQ LW LV YHU\ GLIILFXOW WR HVWLPDWH WKH VWDQGDUG HUURUV DVVRFLDWHG ZLWK WKHVH SUHGLFWRUV 7KLV LV SULPDULO\ GXH WR WKH ODFN RI FORVHG IRUP H[SUHVVLRQV IRU WKH PHDQ VTXDUHG HUURUV 06(Vf RI WKH (%/83V RU WKH (% SUHGLFWRUV .DFNDU DQG +DUYLOOH f VXJJHVWHG DQ DSSUR[LPDWLRQ WR WKH 06(V DOVR +DUYLOOH LQ SUHVV +DUYLOOH DQG -HVNH f 3UDVDG DQG 5DR f SURSRVHG HVWLPDWHV RI WKHVH DSSUR[LPDWH 06(V LQ WKUHH VSHFLILF PL[HG OLQHDU PRGHOV $OO WKHVH DSSUR[LPDWLRQV UHVW KHDYLO\ RQ WKH QRUPDOLW\ DVVXPSWLRQ 5HFHQWO\ /DKLUL DQG 5DR f FRQVLGHUHG WKLV SUREOHP UHOD[LQJ WKH QRUPDOLW\ DVVXPSWLRQ DVVXPLQJ VRPH PRPHQW FRQGLWLRQV ZLWKRXW WKH SUHVHQFH RI DX[LOLDU\ LQIRUPDWLRQ 7KH ZRUN RI 3UDVDG DQG 5DR f VXJJHVWV WKDW WKHLU DSSUR[LPDWLRQV ZRUN ZHOO ZKHQ WKH QXPEHU RI VPDOO DUHDV LV VXIILFLHQWO\ ODUJH ,W LV QRW FOHDU WKRXJK KRZ WKHVH DSSUR[LPDWLRQV IDUH IRU D VPDOO RU HYHQ PRGHUDWHO\ ODUJH QXPEHU RI VPDOO DUHDV *KRVK DQG /DKLUL LQ SUHVVf SURSRVHG DQ +% SURFHGXUH DV DQ DOWHUQDWLYH WR WKH (%/83 RU WKH (% SURFHGXUH Q DQ +% SURFHGXUH LI RQH XVHV WKH SRVWHULRU PHDQ IRU HVWLPDWLQJ WKH SDUDPHWHU RI LQWHUHVW WKHQ D QDWXUDO HVWLPDWH RI WKH VWDQGDUG HUURU DVVRFLDWHG ZLWK WKLV HVWLPDWRU LV LWV SRVWHULRU VWDQGDUG GHYLDWLRQ VGf 7KH HVWLPDWH WKRXJK

PAGE 17

RIWHQ FRPSOLFDWHG FDQ EH IRXQG H[DFWO\ YLD QXPHULFDO LQWHJUDWLRQ ZLWKRXW DSSUR[LPDWLRQ 7KH PRGHO FRQVLGHUHG E\ *KRVK DQG /DKLUL LQ SUHVVf ZDV KRZHYHU RQO\ D VSHFLDO FDVH RI WKH VRFDOOHGQHVWHG HUURU UHJUHVVLRQ PRGHO DOVR XVHG E\ %+) $ VLPLODU PRGHO ZDV FRQVLGHUHG E\ 6WURXG f EXW KLV JHQHUDO DQDO\VLV ZDV SHUIRUPHG RQO\ IRU WKH EDODQFHG FDVH WKDW LV ZKHQ WKH QXPEHU RI VDPSOHV ZDV WKH VDPH IRU HDFK VWUDWXP 2WKHU PRGHOV KDYH DOVR EHHQ SURSRVHG ,Q D UHFHQW DUWLFOH &KRXGKU\ DQG 5DR f FRQVLGHUHG ILYH VSHFLILF PRGHOV IRU VPDOO DUHD HVWLPDWLRQ QRW LQFOXGHG LQ WKH HDUOLHU ZRUN RI 3UDVDG DQG 5DR f 5HFHQWO\ 5R\DO f DQG /XL DQG &XPEHUODQG f FRQVLGHUHG FHUWDLQ FURVVFDVV LILFDWRU\ PRGHOV IRU VPDOO DUHD HVWLPDWLRQ 7KH ODWWHU FDUULHG RXW D %D\HVLDQ DQDO\VLV DVVXPLQJ WKH GHJHQHUDF\ RI FHUWDLQ WHUPV LQ DQ XVXDO WZRZD\ OLQHDU PRGH )RU D %D\HVLDQ DQDO\VLV LQ WKH FRQWH[W RI DQLPDO EUHHGLQJ RQH PD\ UHIHU WR *LDQROD DQG )HUQDQGR f +RZHYHU WKH\ GLG QRW FRQVLGHU WKH +% DQDO\VLV 7KH\ XVHG VXEMHFWLYH LQIRUPDWLYH SULRUV ZKLFK DUH FRQVWUXFWHG IURP WKH SUHYLRXV GDWD DQG H[SHULPHQWV $OVR WKH\ VKRZHG KRZ VRPH RI WKH FODVVLFDO PHDVXUHV LQ DQLPDO EUHHGLQJ FDQ KDYH %D\HVLDQ MXVWLILFDWLRQ 7KH\ KDYH DOVR FRQVLGHUHG QRQ LQIRUPDWLYH LPSURSHU SULRU

PAGE 18

2QH LPSRUWDQW VSHFLDO FDVH ZKLFK DULVHV LQ fWKH DERYH DSSURDFKHV ZKLFK LV DOVR LPSRUWDQW LQ WKH WKHRU\ R OHDVW VTXDUHV :KHQ WKH UDWLRV RI YDULDQFH FRPSRQHQWV DUH NQRZQ WKH SUHGLFWRUV OHDVW VTXDUHV HPSLULFDO %D\HV RU KLHUDUFKLFDO %D\HVf DUH %/83V +HQGHUVRQ f )RU UHODWHG %/83 UHVXOWV IRU SUHGLFWLQJ VFDODUV LQ ILQLWH SRSXODWLRQ VDPSOLQJ RQH PD\ UHIHU WR 5R\DO f /XL DQG &XPEHUODQG f 3UDVDG DQG 5DR f DQG VHYHUDO RWKHUV +DUYLOLH LQ SUHVVf KDV SRLQWHG RXW WKH %/83 SURSHUWLHV RI %D\HVLDQ VFDODUV LQ JHQHUDO PL[HG OLQHDU PRGHOV VHH DOVR +DUYLOH f *KRVK DQG /DKLUL LQ SUHVVf KDYH H[WHQGHG +HQGHUVRQ DQG RWKHUV VFDODU %/83 QRWLRQ WR VKRZ WKH %D\HVLDQ SUHGLFWRU RI WKH YHFWRU RI ILQLWH SRSXODWLRQ PHDQ LV %/83 7R FRQFOXGH WKLV GLVFXVVLRQ ZH ZLOO EULHIO\ PHQWLRQ DQRWKHU SUREOHP 6R IDU ZH KDYH FRQVLGHUHG RQO\ WKH SUREOHP RI HVWLPDWLQJ WKH PHDQ LQ ILQLWH SRSXODWLRQ VDPSOLQJ $QRWKHU LPSRUWDQW SUREOHP LQ ILQLWH SRSXODWLRQ VDPSOLQJ LV HVWLPDWLQJ WKH ILQLWH SRSXODWLRQ YDULDQFH (ULFVRQ f IRXQG WKH %D\HV HVWLPDWRU RI ILQLWH SRSXODWLRQ YDULDQFH XQGHU D QRUPDO WKHRU\ VHW XS (PSLULFDO %D\HV HVWLPDWLRQ RI ILQLWH SRSXODWLRQ YDULDQFH LQ VPDOO DUHD HVWLPDWLRQ ZDV FRQVLGHUHG E\ *KRVK DQG /DKLUL Ef DQG /DKLUL DQG 7LZDUL LQ SUHVVf ZLWKRXW WKH SUHVHQFH RI DX[LOLDU\ LQIRUPDWLRQ

PAGE 19

7KH 6XEMHFW RI 7KLV 'LVVHUWDWLRQ ,Q fWKLV GLVVHUWDWLRQ ZH SUHVHQW D XQLILHG %D\HVLDQ SUHGLFWLRQ WKHRU\ IRU OLQHDU PRGHOV LQ VPDOO DUHD HVWLPDWLRQ LQ WKH FRQWH[W RI ILQLWH SRSXODWLRQ VDPSOLQJ $ JHQHUDO %D\HVLDQ PRGHO LV SUHVHQWHG ZKLFK FDQ EH UHJDUGHG DV DQ H[WHQVLRQ RI WKH +% LGHDV RI /LQGOH\ DQG 6PLWK f WR SUHGLFWLRQ 7KLV JHQHUDO PRGHO FDQ DOVR EH DSSOLHG LQ LQILQLWH SRSXODWLRQ VLWXDWLRQV IRU H[DPSOH LQ DQLPDO EUHHGLQJ DQG LQ RWKHU DSSOLFDWLRQV ZKHUH D PL[HG OLQHDU PRGH LV XVHG ,Q &KDSWHU 7ZR ZH LQWURGXFH D JHQHUDO +% PRGHO DQG X VH WKLV PRGHO IRU VLPXOWDQHRXV HVWLPDWLRQ RI VHYHUDO VPDOO DUHD PHDQV LQ ILQLWH SRSXODWLRQ VDPSOLQJ 6RPH RI WKH ZLGHO\ XVHG PRGHOV LQ VPDOO DUHD HVWLPDWLRQ LQFOXGLQJ WKH QHVWHG HUURU UHJUHVVLRQ PRGHO %DWWHVH HW DO DQG 5DR 6WURXG *KRVK DQG /DKLUL LQ SUHVVf WKH UDQGRP UHJUHVVLRQ FRHIILFLHQWV PRGHO 'HPSVWHU HW DO 3UDVDG DQG 5DR f WKH FURVVFODVVLILFDWRU\ PRGHOV 5R\DO /XL DQG &XPEHUODQG f DQG PXOWLn VWDJH VDPSOLQJ PRGHOV *KRVK DQG /DKLUL 0DOHH DQG 6HGUDQVN 6FRWW DQG 6PLWK f FDQ EH UHJDUGHG DV VSHFLDO FDVHV RI RXU PRGHO 7KH SRVWHULRU GLVWULEXWLRQ DV ZHOO DV WKH UHVXOWLQJ SRVWHULRU PHDQV DQG YDULDQFHV RI WKH XQREVHUYHG XQLWV LQ WKH SRSXODWLRQ JLYHQ WKH VDPSOH XQLWV DUH SURYLGHG LQ WKLV FKDSWHU $OVR IRU DQ LQILQLWH

PAGE 20

SRSXODWLRQ JLYHQ WKH GDWD WKH FRQGLWLRQDO GLVWULEXWLRQ DQG WKH FRQGLWLRQDO PHDQ DQG YDULDQFH R WKH YHFWRU R HHFWV DUH SURYLGHG 7KHVH WZR DQDO\VHV DUH DSSOLHG WR WZR UHDO GDWD VHWV ,W LV ZRUWKZKLOH WR PHQWLRQ WKDW %D\HVLDQ DQDO\VLV LQ OLQHDU PRGHOV ZDV LQLWLDWHG E\ +LOO f 6HH DOVR +LOO f )RU D JRRG H[SRVLWLRQ RQ +% DQDO\VLV VHH %HUJHU f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f /LNH RWKHU UHODWHG DUWLFOHV RXU %/83 UHVXOWV GR QRW UHTXLUH DQ\ QRUPDOLW\ DVVXPSWLRQ RI WKH PRGHO )RU D

PAGE 21

VXLWDEOH VXEFODVV RI HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV LQFOXGLQJ EXW QRW OLPLWHG WR WKH QRUPDO WKH +% SUHGLFWRUV DUH VKRZQ WR EH EHVW XQELDVHG WKDW LV WKH\ KDYH WKH VPDOOHVW YDUL DQHHFRYDULDQHH PDWUL[ ZLWKLQ WKH FODVV RI DOO XQELDVHG SUHGLFWRUV $OVR IROORZLQJ +ZDQJ f ZH KDYH EHHQ DEOH WR VKRZ WKDW WKH %/83V DOVR fXQLYHUVDOO\f RU fVWRFKDVWLFDOO\ff GRPLQDWH WKH OLQHDU XQELDVHG SUHGLFWRUV IRU HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV 7KH QRWLRQ RI fXQLYHUVDOf DQG fVWRFKDVWLFf GRPLQDWLRQ ZLOO EH PDGH SUHFLVH LQ &KDSWHU 7KUHH $OVR LW LV HVWDEOLVKHG WKDW XQGHU D VXLWDEOH JURXS RI WUDQVIRUPDWLRQV WKH +% SUHGLFWRUV DUH EHVW ZLWKLQ WKH FODVV RI DOO HTXLYDULDQW SUHGLFWRUV IRU HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV -HVNH DQG +DUYLOOH f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

PAGE 22

WKDW XQGHU DYHUDJH VTXDUHG HUURU ORVV WKH %D\HV ULVN GLIIHUHQFH EHWZHHQ WKH +% SUHGLFWRUV DQG WKH VXEMHFWLYH %D\HV SUHGLFWRUV IRU D fWUXHf SULRU JRHV WR ]HUR DV WKH QXPEHU RI VPDOO DUHDV JRHV WR LQILQLW\ 7KLV VKRZV RXU +% SUHGLFWRUV DUH fDV\PSWRWLFDOO\ RSWLPDOf $f LQ WKH VHQVH RI 5REELQV f 7KH $2 SURSHUW\ RI FHUWDLQ (% SUHGLFWRUV DULVLQJ QDWXUDOO\ LQ WKH FRQWH[W RI ILQLWH SRSXODWLRQ VDPSOLQJ ZDV SURYHG LQ *KRVK DQG 0HHGHQ f *KRVK DQG /DKLUL Df DQG *KRVK /DKLUL DQG 7LZDUL f &KDSWHU )LYH LV GHYRWHG WR WKH VLPXOWDQHRXV HVWLPDWLRQ RI VHYHUDO VWUDWD YDULDQFHV :H KDYH FRQVLGHUHG WKH VSHFLDO FDVHV RI WKH QHVWHG HUURU UHJUHVVLRQ PRGHO DV FRQVLGHUHG E\ *KRVK DQG /DKLUL LQ SUHVVf DQG 6WURXG f LQ GHWDLO $V LQ &KDSWHU )RXU ZH KDYH SURYHG WKH $2 SURSHUW\ RI WKHVH SUHGLFWRUV *KRVK DQG /DKLUL Ef DQG /DKLUL DQG 7LZDUL LQ SUHVVf KDYH SURYHG WKH $2 SURSHUW\ RI FHUWDLQ (% SUHGLFWRUV RI ILQLWH SRSXODWLRQ YDULDQFHV :H UHHPSKDVL]H WKDW WKH SUHVHQW GLVVHUWDWLRQ SURYLGHV D XQLILHG %D\HVLDQ DQDO\VLV ERWK LQ WKH ILQLWH DQG LQILQLWH SRSXODWLRQ IUDPHZRUN )RU ILQLWH SRSXODWLRQ ZH XQLI\ D QXPEHU RI PRGHOV FRQVLGHUHG HDUOLHU E\ GLIIHUHQW DXWKRUV )URP WKH DQDO\VLV RI WZR GDWD VHWV ZH XQGHUWRRN LQ &KDSWHU 7ZR LW LV FOHDU WKDW WKH SURSRVHG SURFHGXUH LV D YLDEOH DOWHUQDWLYH WR WKH DYDLODEOH (%/83 RU (% SURFHGXUHV $OVR

PAGE 23

fWR RXU NQRZOHGJH HVWLPDWHV RI 06(V RU JRRG DSSUR[LPDWLRQV WKHUHRI DUH QRW DYDLODEOH H[FHSW IRU D IHZ VSHFLILF PRGHOV 7KH %D\HVLDQ SURFHGXUHV RI WKLV GLVVHUWDWLRQ RQ WKH RWKHU KDQG FDQ VHUYH DV D JHQHUDO UHFLSH WR KDQGOH D JUHDWHU YDULHW\ RI SUREOHPV $OVR WKH LQIHUHQWLDO PHWKRGV RI WKH IROORZLQJ FKDSWHUV DUH LPSHPHQWDEH IRU GDWD DQDO\VLV HVSHFLDOO\ LQ WKHVH GD\V RI VRSKLVWLFDWHG FRPSXWLQJ IDFL LWLHV

PAGE 24

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f ZKHUHDV LQ WKH VHFRQG FDVH ZH DUH LQWHUHVWHG LQ SUHGLFWLQJ OLQHDU IXQFWLRQV RI IL[HG DQG UDQGRP HIIHFWV ,Q WKH ILQLWH SRSXODWLRQ VDPSOLQJ VHW XS ZH DVVXPH WKDW WKHUH DUH P VWUDWD WKH LAA VWUDWXP 8M FRQWDLQLQJ D ILQLWH QXPEHU RI XQLWV 1 ZLWK XQLWV ODEHOOHG 8MA /HW
PAGE 25

R WKH
PAGE 26

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f GLVWULEXWLRQ RI WKH FKDUDFWHULVWLF RI LQWHUHVW RI WKH QRQVDPSOHG SRSXODWLRQ XQLWV JLYHQ WKH YDOXHV RI WKDW FKDUDFWHULVWLF IRU WKH VDPSOH XQLWV LQ ILQLWH SRSXODWLRQ VDPSOLQJ $OVR WKH SRVWHULRU PHDQ YHFWRU DQG SRVWHULRU YDULDQFHFRYDULDQFH PDWUL[ FRUUHVSRQGLQJ WR WKH FKDUDFWHULVWLF YHFWRU RI QRQVDPSOHG XQLWV DUH REWDLQHG IURP WKLV SUHGLFWLYH GLVWULEXWLRQ ,Q SDUWLFXODU WKH SRVWHULRU PHDQV DQG WKH YDULDQFHV RI WKH ILQLWH SRSXODWLRQ PHDQV IRU DOO WKH VPDOO DUHDV DUH REWDLQHG ,Q WKH VHFRQG KDOI RI 6HFWLRQ ZH KDYH REWDLQHG WKH SRVWHULRU GLVWULEXWLRQ RI WKH YHFWRU RI IL[HG DQG UDQGRP HIIHFWV IRU WKH PRGHO LQWURGXFHG LQ 6HFWLRQ ,Q SDUWLFXODU H[SUHVVLRQV IRU WKH SRVWHULRU PHDQV DQG YDULDQFHV RI FHUWDLQ SUHGLFWDQGV DUH GHYHORSHG

PAGE 27

,Q 6HFWLRQ ZH KDYH DSSOLHG WKH UHVXOWV RI 6HFWLRQ WR VRPH DFWXDO GDWD VHWV )LUVW ZH VKDOO FRQVLGHU WKH FRUQ DQG VR\EHDQV GDWD ZKLFK DSSHDUHG LQ %DWWHVH +DUWHU DQG )XOOHU f 8VLQJ WKH +% DQDO\VLV GHYHORSHG LQ 6HFWLRQ ZH KDYH GHULYHG WKH SRVWHULRU PHDQV DQG SRVWHULRU VWDQGDUG GHYLDWLRQV IRU WKH VPDOO DUHD FRXQW\f PHDQV 7KH VHFRQG GDWD VHW FRQWDLQLQJ WKH ZHLJKWV RI VLQJOHELUWK ODPEV DSSHDUHG LQ +DUYLOOH LQ SUHVVf 7KLV LV DQDO\]HG E\ +% PHWKRGV IRU DQ LQILQLWH SRSXODWLRQ VHW XS GHYHORSHG LQ WKH VHFRQG KDOI RI 6HFWLRQ )LQDOO\ LQ 6HFWLRQ DQ +% DQDO\VLV RI WKH PRGHO FRQVLGHUHG E\ &DUWHU DQG 5ROSK f DQG VXEVHTXHQWO\ E\ )D\ DQG +HUULRW f WR HVWLPDWH WKH SHU FDSLWD LQFRPH RI VPDOO SODFHV LV FRQVLGHUHG ,Q WKLV VLWXDWLRQ XQLW OHYHO REVHUYDWLRQV DUH QRW DYDLODEOH DQG ZH DUH LQWHUHVWHG LQ SUHGLFWLQJ WKH ILQLWH SRSXODWLRQ PHDQ IRU HDFK VPDOO DUHD +HUH WKH VDPSOLQJ YDULDQFHV DUH GLIIHUHQW DQG DUH DVVXPHG WR EH NQRZQ DOVR ZH SXW D XQLIRUP SULRU RQ WKH UHJUHVVLRQ FRHIILFLHQWV DQG D JDPPD SULRU SURSHU RU LPSURSHUf RQ WKH LQYHUVH RI WKH SULRU YDULDQFH 'HVFULSWLRQ RI WKH +LHUDUFKLFDO %D\HV 0RGHO ZLWK ([DPSOHV &RQVLGHU WKH IROORZLQJ %D\HVLDQ PRGHO $f FRQGLWLRQDO RQ E EA ESf \ DQG U Y

PAGE 28

< a 1;E =Y Urf %f FRQGLWLRQDO RQ E $ DQG U OHW \ a QR U' $f A &f % 5 DQG $ KDYH D FHUWDLQ MRLQW SULRU GLVWULEXWLRQ SURSHU RU LPSURSHU 6WDJHV $f DQG %f RI WKH PRGHO FDQ EH LGHQWLILHG DV D JHQHUDO PL[HG OLQHDU PRGHO 7R VHH WKLV ZULWH < ;E I =Y H f ZKHUH E LV WKH YHFWRU RI IL[HG HIIHFWV H DQG \ DUH PXWXDOO\ LQGHSHQGHQW ZLWK H a 1A2 UAA DQG \ a 1A2 UaA'$fA ; DQG = DUH NQRZQ GHVLJQ PDWULFHV A LV D NQRZQ SRVLWLYH GHILQLWH SGf PDWUL[ ZKLOH '$f LV D SG PDWUL[ ZKLFK LV VWUXFWXUDOO\ NQRZQ H[FHSW SRVVLEO\ IRU VRPH XQNQRZQ $ ,Q WKH H[DPSOHV WR IROORZ $ LQYROYHV WKH UDWLRV RI YDULDQFH FRPSRQHQWV ,Q WKH FRQWH[W RI VPDOO DUHD HVWLPDWLRQ SDUWLWLRQ <1US;Of ;1US;Sf =1M;Tf DQG H1UA[Of ZLWK FRQIRUPLW\ DQG UHZULWH WKH PRGHO JLYHQ LQ f DV < < f f ; f ] f f ; 9a ,] Af< 9Hf f ,Q WKH DERYH SDUDJUDSK < Lf QUS[Of FRUUHVSRQGV WR WKH YHFWRU RI VDPSOHG XQLWV IURP P VPDOO DUHDV RU VWUDWD ZKLOH

PAGE 29

< f QW f§UL\!f [ M FRUUHVSRQGV WR WKH YHFWRU RI XQVDPSOHG F L P7 XQLWV :H ZLOO IXUWKHU SDUWLWLRQ
PAGE 30

$OVR OHW FRO %_f GHQRWH WKH PDWUL[ L N 7?7 %MO DQG N $ GHQRWH WKH PDWUL[ r f f R $A )LUVW FRQVLGHU WKH QHVWHG HUURU UHJUHVVLRQ PRGHO
PAGE 31

6RPH RI WKH PRGHOV RI &KRXGKU\ DQG 5DR f FDQ DOVR EH WUHDWHG DV VSHFLDO FDVHV RI RXUV )RU H[DPSOH RQH RI WKHLU PRGHOV LV JLYHQ E\
PAGE 32

OLQHDU PRGHO )RU H[DPSOH VXSSRVH fWKHUH DUH P VPDO DEH HG P :LWKLQ HDFK VPDOO DUHD XQLWV DUH IXUWKHU FODVVLILHG LQWR F VXEJURXSV VRFLRHFRQRPLF FODVV HWFf ODEHOOHG F 7KH FHOO VL]HV 1LM L ffr P M Ff DVVXPHG WR EH NQRZQ /HW
PAGE 33

]f FR ^ k ,Qff LPM Oa ]f a P L O F A ,Q M L DQG =f LV D PDWUL[ VLPLODU WR = A A ZLWK 1 Q f UHSODFLQJ Q f LQ GHILQLQJ WKH GLPHQVLRQV RI WKH YHFWRUV $OVR Y 7 U f ff P r rF AArrr 7PFf A $ 7 A MA f $f DQG $ f L DJA $ ,P A r nn $ A,PFf 6SHFLDO FDVHV RI WKLV PRGHO KDYH EHHQ FRQVLGHUHG E\ VHYHUDO RWKHUV /XL DQG &XPEHUODQG f FRQVLGHUHG D PRGHO ZKHUH DQG DUH GHJHQHUDWH DW ]HURHV $OVR WKH\ DVVXPHG WKH YDULDQFH UDWLR $ WR EH NQRZQ LQ GHULYLQJ WKHLU HVWLPDWRUV DQG GLG QRW DGGUHVV WKH LVVXH RI XQNQRZQ $ DSSURSULDWHO\ 1H[W ZH VKRZ WKDW WKH WZR VWDJH VDPSOLQJ PRGHO ZLWK FRYDULDWHV DQG P VWUDWD LV D VSHFLDO FDVH RI RXU JHQHUDO OLQHDU PRGHO 6XSSRVH WKDW WKH L f WK VWUDWXP FRQWDLQV /r SULPDU\ XQLWV 6XSSRVH DOVR WKDW WKH M f WK SULPDU\ XQLW ZLWKLQ WKH LA XUQ FRQWDLQV 1 f VXEXQLWV /HW < L GHQRWH WKH YDOXH RI WKH FKDUDFWHULVWLF RI LQWHUHVW IRU WKH NA VXEXQLW ZLWKLQ WKH SULPDU\ XQLW IURP WKH LAA VWUDWXP N 1 M M / L \ L Pf )URP WKH f WK VWUDWXP D VDPSOH RI M SULPDU\ XQLWV LV WDNHQ )RU WKH VHOHFWHG SULPDU\ XQLW ZLWKLQ WKH LAr VWUDWXP D VDPSOH RI QM f VXEXQLWV DUH VHOHFWHG )RU QRWDWLRQDO FRQYHQLHQFH GHQRWH ZLWKRXW ORVV RI JHQHUDOLW\ WKH VDPSOH YDOXHV E\
PAGE 34

$VVXPH FRQGLWLRQDO RQ E U DQG $ < f L L MN [MME r M M HLMN N f§ 1 f M r fffff / L f§ f ZKHUH UM DQG H c DUH PXWXDOO\ LQGHSHQGHQW ZLWK e LLG -. L 1 $[Uf f M M LLG 1 $Uf ff HLMN LLG 1f U f f /HW FR L P FRO FRO M A>ONQ f f &2 N1Lf A L ccL &2 L P DQG FRO L P?e $OVR OHW H EH GHILQHG VLPLODUO\ DV < L 7KHQ f FDQ EH ZULWWHQ DV f ZLWK

PAGE 35

ZKHUH Z : 1LM XL L O = f Vf Zff Vf P f Q eL + Q L M M L Z Lf Lf Lf? A nf ZAHUH 9L Lf P L O ,Q M O Zf Q7a/ f Q7 P P ( QL / L O n ( /! Af L L P ( mL L O ] f Vf Zff 6 f ,S \ 5L / L 5L 7 1 Q f L L M O DQG : f P L O /L U f f +HUH W $ f§ $ A A 7 ,W [aO 1UQ f ("af f L DJA$A[ ,P ZLWK 1 $f f r r / r f7 P L US Y < 1 f f 7KH LGHDV FDQ EH H[WHQGHG GLUHFWO\ L f§ M O WR PXOWLVWDJH VDPSOLQJ ZLWK PRUH FRPSOLFDWHG QRWDWLRQV :H PD\ PHQWLRQ KHUH WKDW %D\HVLDQ DQDO\VLV 7RU WZR VWDJH VDPSOLQJ ZDV LQWURGXFHG ILUVW E\ 6FRWW DQG 6PLWK f LQ D PXFK VLPSOHU IUDPHZRUN $ PXOWLVWDJH DQDORJ RI WKHLU ZRUN ZDV SURYLGHG E\ 0DOHH DQG 6HGUDQVN f *KRVK DQG /DKLUL f FRQVLGHUHG HPSLULFDO %D\HV HVWLPDWLRQ LQ PXOWLVWDJH VDPSOLQJ

PAGE 36

1RZ ZH ZLOO FRQVLGHU WKH LQILQLWH SRSXODWLRQ VHW XS ,Q WKLV FRQWH[W ZH ZLOO XVH WKH PRGHO JLYHQ E\ $f %f DQG &f ZLWK WKH PL[HG OLQHDU PRGHO UHSUHVHQWDWLRQ JLYHQ E\ f +HUH ZH ZLOO DVVXPH WKH GDWD YHFWRU < LV QA[O DQG WKH DVVRFLDWHG GHVLJQ PDWULFHV ; DQG = DUH QA[S DQG QUS;T UHVSHFWLYHO\ :LWKRXW ORVV RI JHQHUDOLW\ ZH DOVR DVVXPH WKDW UDQN;f 3 2XU REMHFWLYH LV WR SUHGLFW 6E 7Y &E \f VD\f RQ WKH EDVLV RI < ZKHUH 6X[Sf DQG 7X[Tf DUH NQRZQ PDWULFHV )ROORZLQJ WKH PRGHOEDVHG LQIHUHQFH LW VXIILFHV WR ILQG WKH SRVWHULRU FRQGLWLRQDOf GLVWULEXWLRQ RI M : VD\f JLYHQ < \ 7KLV FRQGLWLRQDO GLVWULEXWLRQ LV SURYLGHG LQ WKH QH[W VHFWLRQ :H ZLOO FRQFOXGH WKLV VHFWLRQ GLVFXVVLQJ D IHZ VSHFLILF PRGHOV LQ WKH FRQWH[W RI FRPSDUDWLYH WULDOV DQG DQLPDO EUHHGLQJ ZKLFK DUH VSHFLDO FDVHV RI WKH JHQHUDO PRGHO SURSRVHG LQ f )LUVW FRQVLGHU PX LFHQWHUHG FOLQLFDO WULDO ZKLFK LV FRQGXFWHG LQ F SDUWLFLSDWLQJ FOLQLFV WR FRPSDUH WZR WUHDWPHQWV RQH DOUHDG\ H[LVWLQJ LQ WKH PDUNHW DQG WKH RWKHU QHZO\ GHYHORSHG 6XSSRVH WKHUH DUH Q VXEMHFWV A UHFHLYLQJ WKH LAA WUHDWPHQW LQ WKH MAA FOLQLF 6RPH RI WKH Q M f FRXOG EH ]HUR :H DUH LQWHUHVWHG LQ HVWLPDWLQJ WKH WUHDWPHQWV GLIIHUHQFH )RU WKLV H[DPSOH ZH ZLOO XVH D PL[HG OLQHDU PRGHO IRU
PAGE 37

fWUHDWPHQW LQ WKH M f WK SDUWLFLSDWLQJ FOLQLF :H FRQVLGHU WKH PRGH < L -N X f F f f f H f L n L LMN f N } A r & A f ZKHUH F f DQG A LMN PXWXDOO\ LQGHSHQGHQW ZLWK VXEMHFW HIIHFWV H LMN DUH L LG 1 U $f WUHDWPHQWFOLQLF LQWHUDFWLRQ L LG 1 $Uf DQG FOLQLF HIIHFWV &M L LG QR $Uf A WKH HIIHFW GXH WR WKH f WK W UHDWPHQW 1RZ ZH ZLOO ZULWH GRZQ WKH < ; = E \ DQG H IRU f )RU HDVH RI SUHVHQWDWLRQ DVVXPH Qf IRU DO L M 7KHQ ZULWLQJ <


PAGE 38

XVHG 2QFH DJDLQ RQH FDQ XVH WKH OLQHDU PRGHO JLYHQ E\ f IRU LQIHUHQWLDO SXUSRVH ,Q WKLV FDVH RQH PD\ EH LQWHUHVWHG LQ SUHGLFWLQJ VRPH VXLWDEOH OLQHDU IXQFWLRQV RI UDQGRP TXDQWLWLHV F DQG f NQRZQ DV WKH EUHHGLQJ YDOXHV 7KHVH SUHGLFWHG EUHHGLQJ YDOXHV FDQ EH XVHG DV D VHOHFWLRQ LQGH[ IRU VHOHFWLQJ WKH PRVW VXLWDEOH EUHHGV IRU IXWXUH EUHHGLQJ SXUSRVH $V D FRQFUHWH H[DPSOH LQ DQLPDO EUHHGLQJ ZH ZLOO GLVFXVV WKH H[DPSOH FRQVLGHUHG E\ +DUYLOOH LQ SUHVVf ZKLFK LQYROYHV SUHGLFWLRQ RI WKH DYHUDJH ELUWK ZHLJKWV RI DQ LQILQLWH QXPEHU RI VLQJOHELUWK PDOH ODPEV WKDW DUH RIIVSULQJ RI GLIIHUHQW VLUHV LQ GLIIHUHQW SRSXODWLRQ OLQHV 7KH GDWD FRQVLVW RI WKH ZHLJKWV DW ELUWKf RI VLQJOHn ELUWK PDOH ODPEV DQG FDPH IURP ILYH GLVWLQFW SRSXODWLRQ OLQHV (DFK ODPE ZDV WKH SURJHQ\ RI RQH RI UDPV DQG HDFK ODPE KDG D GLIIHUHQW GDP $JH RI WKH GDP ZDV UHFRUGHG DV EHORQJLQJ WR RQH RI WKUHH FDWHJRULHV QXPEHUHG \HDUVf \HDUVf DQG RYHU \HDUVf /HW
PAGE 39

ZKHUH G f f f Q A M N N f§ r f f f UQ M r f§ A DQG M ZKHUH QLMN LV WKH QXPEHU R ODPEV ZKRVH GDPV EHORQJ WR WKH LWK DJH FDWHJRU\ ZKHQ WKH SRSXODWLRQ OLQH LV M DQG WKH VLUH LV N DQG P LV WKH WRWDO QXPEHU R ODPEV ZKRVH VLUHV DUH LURP SRSXODWLRQ OLQH M +HUH WKH DJH HLLHFWV A  A f DQG OLQH HLLHFWV UAf DUH FRQVLGHUHG DV LL[HG HLLHFWV DQG WKH VLUH ZLWKLQ OLQHf HLLHFWV V MN LLG 1 U$f A DQG LQGHSHQGHQW R HUURU YDULDEOHV H L MNG ZKLFK DUH LLG 1 U f 7R PDNH WKH GHVLJQ PDWUL[ DVVRFLDWHG ZLWK LL[HG HLLHFWV LXOO UDQN ZH FDQ WDNH ZKLFK LV WKH XVXDO LRUPXODWLRQ QHHGHG LRU */0 3URFHGXUHV LQ 6$6 /HW ALM (
PAGE 40

+LHUDUFKLFDO %D\HV $QDO\VLV ,Q fWKLV VHFWLRQ IRU WKH ILQLWH SRSXODWLRQ VDPSOLQJ ZH SURYLGH WKH SUHGLFWLYH GLVWULEXWLRQ RI ]!@n f $ UDQGRP YHFWRU 7 7A 7Sf LV VDLG WR KDYH D PXOWLYDULDWH WGLVWULEXWLRQ ZLWK ORFDWLRQ SDUDPHWHU L VFDOH SDUDPHWHU D SG S[S PDWUL[ DQG GHJUHHV RI IUHHGRP GIf Y LI LW KDV SGI W AWL f VHH =HOOQHU S RU 3UHVV S f +HUH _(_ GHQRWHV WKH GHWHUPLQDQW RI D VTXDUH PDWUL[ ( $VVXPH 9 7KHQ (7f L 97f >Y Lf :H DVVXPH FRQGLWLRQV $f DQG %f JLYHQ DW WKH EHJLQQLQJ RI 6HFWLRQ ,Q VWDJH &f RI WKH PRGHO LW LV DVVXPHG WKDW

PAGE 41

&Of % 5 $W5 DUH LQGHSHQGHQWO\ GLVWULEXWHG ZLWK % XQL IRUP 5Af 5 a JDPPDAD4 JJf D r J L $_5 a JDPPDADM ,JA ZLWK JM f f f W A f $OORZLQJ D4 DQG VRPH RI J fWR EH ]HUR VRPH LPSURSHU JDPPD GLVWULEXWLRQV DUH LQFOXGHG DV D SRVVLELOLW\ LQ RXU SULRU %HIRUH VWDWLQJ WKH SUHGLFWLYH GLVWULEXWLRQ RI < JLYHQ ZH QHHG WR LQWURGXFH D IHZ PDWUL[ f QRWDWLRQV ( LQWR ( Y :H ZULWH A 9 9 Y B Y "$f A ='$f DQG SDUWLWLRQ $ VR HW YB e f V V[ DYf'7VL f r r f 0 (L. ; fYOf7ABLYOf[YOf7B 9 ; < a ; 9 f a r f 9 9 [ < A LYLf9YLf9LYLK L "LLr f f 7 f 1RZ WKH SUHGLFWLYH Yf YOf GLVWULEXWLRQ RI < JLYHQ < \AA LV JLYHQ LQ WKH IROORZLQJ WKHRUHP LQ WZR $ SURRI RI WKLV WKHRUHP LV GHIHUUHG WR $SSHQGL[ $

PAGE 42

7KHRUHP &RQVLGHU WKH PRGHO JLYHQ LQ f DQG &Of $VVXPH WKDW QUS W e V L L S 7KHQ FRQGLWLRQDO RQ $ $ DQG < Lf \f
PAGE 43

9 0\ \ r Q7 I W 7 aU e L S [ ( D + DLAL L O \f7.\f!* \ Lf f 8VLQJ f DQG f LW LV SRVVLEOH WR ILQG WKH SRVWHULRU PHDQ DQG YDULDQFH RI e f ff? < $< A &< A ZKHUH $ DQG & DUH NQRZQ PDWULFHV 7KH %D\HV HVWLPDWH RI fB \f A XQGHU DQ\ TXDGUDWLF ORVV LV LWV SRVWHULRU PHDQ DQG LV JLYHQ E\ eEI\ f $ &HP \ f \ Lf f XVLQJ f 6LPLODUO\ XVLQJ f RQH PD\ REWDLQ 9 f ff < \ Lf FY\ f f \ f&7 f 1RWH WKDW ZKHQ $ P A W OQ DQG L Oa L & f < f? UHGXFHV WR WKH YHFWRU RI ILQLWH SRSXODWLRQ WRWDOV IRU WKH P VPDOO DUHDV ZKHUHDV IRU WKH FKRLFH $ L1UQ1Lf! m< ? \! 0: DQG & UHGXFHV WR WKH YHFWRU RI ILQLWH SRSXODWLRQ PHDQV IRU WKH P VPDOO DUHDV 1RZ ZH ZLOO JHW EDFN WR WKH LQILQLWH SRSXODWLRQ VHW XS WR SURYLGH WKH SRVWHULRU GLVWULEXWLRQ RI : E7 \7f7 JLYHQ < \ ZKLFK ZLOO EH XVHG WR ILQG WKH SRVWHULRU PHDQ DQG YDULDQFH RI &E \f D YHFWRU RI WKH OLQHDU FRPELQDWLRQ RI : 7KH SRVWHULRU GLVWULEXWLRQ LV JLYHQ LQ WKH IROORZLQJ

PAGE 44

WKHRUHP $ SURRI RI WKLV WKHRUHP ZLOO EH RPLWWHG EHFDXVH RI LWV VLPLODULW\ WR WKH SURRI RI 7KHRUHP :H ZLOO FRQVLGHU WKH PRGHO JLYHQ E\ $f DQG %f RI 6HFWLRQ DQG &Of 5HFDOO IURP WKH PLGGOH RI 6HFWLRQ WKDW ZH KDYH UHGHILQHG WKH GLPHQVLRQV RI < ; = DQG H DSSHDULQJ WKHUH E\
PAGE 45

f $JDLQ XVLQJ WKH PRPHQWV RI PXOWLYDULDWH WGLVWULEXWLRQ DQG WKH LWHUDWHG IRUPXODV IRU H[SHFWDWLRQ DQG YDULDQFH WKH DERYH WKHRUHP FDQ EH XVHG WR ILQG WKH FRPSXWDWLRQDO IRUPXODV IRU (:_\f DQG 9:_\f DV LQ f DQG f V%,\f VD\f DQG 6LPLODUO\ RQH FDQ ILQG (A&E \f 9A&E Yf \A DV LQ f DQG f ZKHUH eE
PAGE 46

$SSOLFDWLRQV RI +LHUDUFKLFDO %D\HV $QDO\VLV 7KLV VHFWLRQ FRQFHUQV WKH DQDO\VLV RI WZR UHDO GDWD VHWV XVLQJ WKH +% SURFHGXUHV VXJJHVWHG LQ 6HFWLRQ ILUVW GDWD VHW LV UHODWHG WR WKH SUHGLFWLRQ RI FRUQ DQG 7KH VR\EHDQV IRU FRXQWLHV LQ QRUWKFHQWUDO ,RZD EDVHG RQ -XQH (QXPHUDWLYH 6XUYH\ DV ZHOO DV /$1'6$7 VDWHOOLWH GDWD ,W DSSHDUHG LQ %DWWHVH +DUWHU DQG )XOOHU %+)f ZKR FRQGXFWHG D YDULDQFH FRPSRQHQWV DQDO\VLV IRU WKLV SUREOHP 7KH VHFRQG GDWD VHW RULJLQDOO\ DSSHDUHG LQ +DUYLOOH DQG )HQHFK f DQG UHDSSHDUHG LQ +DUYLOOH LQ SUHVVf ZKHUH KH FRQGXFWHG D YDULDQFH FRPSRQHQWV DV ZHOO DV DQ +% DQDO\VLV WR SUHGLFW Z JLYHQ LQ f WKH DYHUDJH ZHLJKW RI DQ LQILQLWH QXPEHU RI VLQJOHELUWK PDOH ODPEV WKDW DUH RIIVSULQJ RI WKH VLUH LQ WKH SRSXODWLRQ L QH :H ZLOO ILUVW FRQVLGHU WKH %+) GDWD VHW 7R VWDUW ZLWK ZH EULHIO\ JLYH D EDFNJURXQG RI WKLV SUREOHP 7KH 86'$ 6WDWLVWLFDO 5HSRUWLQJ 6HUYLFH ILHOG VWDII GHWHUPLQHG WKH DUHD RI FRUQ DQG VR\EHDQV LQ VDPSOH VHJPHQWV HDFK VHJPHQW DERXW KHFWDUHVf RI FRXQWLHV LQ QRUWKFHQWUDO ,RZD E\ LQWHUYLHZLQJ IDUP RSHUDWRUV %DVHG RQ /$1'6$7 UHDGLQJV REWDLQHG GXULQJ $XJXVW DQG 6HSWHPEHU 86'$ SURFHGXUHV ZHUH XVHG WR FODVVLI\ WKH FURS FRYHU IRU DOO SL[HOV D WHUP IRU fSLFWXUH HOHPHQWf DERXW KHFWDUHVf LQ WKH FRXQWLHV 7KH QXPEHU RI VHJPHQWV LQ HDFK FRXQW\

PAGE 47

WKH QXPEHU RI KHFWDUHV RI FRUQ DQG VR\EHDQV DV UHSRUWHG LQ WKH -XQH (QXPHUDWLYH 6XUYH\f WKH QXPEHU RI SL[HOV FODVVLILHG DV FRUQ DQG VR\EHDQV IRU HDFK VDPSOH VHJPHQW DQG WKH FRXQW\ PHDQ QXPEHU RI SL[HOV FODVVLILHG DV FRUQ DQG VR\EHDQV WKH WRWDO QXPEHU RI SL[HOV FODVVLILHG DV WKDW FURS GLYLGHG E\ WKH QXPEHU RI VHJPHQWV LQ WKDW FRXQW\f DUH UHSRUWHG LQ 7DEOH RI %+) )RU UHDG\ UHIHUHQFH LW LV UHSURGXFHG LQ 7DEOH ,Q RUGHU WR PDNH RXU UHVXOWV FRPSDUDEOH WR WKDW RI %+) WKH VHFRQG VHJPHQW LQ +DUGLQ FRXQW\ ZDV LJQRUHG 7KH PRGHO FRQVLGHUHG E\ %+) LV
PAGE 48

7DEOH 6XUYH\ DQG 6DWHOOLWH 'DWD IRU &RUQ DQG 6R\EHDQV LQ ,RZD &RXQWLHV 1R RI SL[HO V 0HDQ QR RI 1R RI 5HS RUWHG L Q VDPSOH S L[H V SH U 6HJPHQWV KHFWDUHV VHJPHQWV VHJPHQWr &RXQW\ 6DPSOH &RXQW\ &RUQ 6R\EHDQ &RUQ L6R\EHDQ &RUQ 6R\EHDQ &HUUR *RUGR L +DPLWRQ L :RUWK L +XPERGW ) UDQNLQ 3RFDKRQWDV :LQQHEDJR :ULJKW :HEVWHU +DQFRFN .RVVXWK +DUGLQ 7KH PHDQ QXPEHU RI SL[HOV RI D JLYHQ FURS SHU VHJPHQW LQ D FRXQW\ LV WKH WRWDO QXPEHU RI SL[HOV FODVVLILHG DV WKDW FURS GLYLGHG E\ WKH QXPEHU RI VHJPHQWV LQ WKDW FRXQW\

PAGE 49

< FDQ EH ZULWWHQ E EO[OLSf [LSf 1L
PAGE 50

DQG ,Q WKH VSHFLDO FDVH RI QHVWHG HUURU UHJUHVVLRQ PRGHO ZH ZLOO QRZ GHYHORS WKH H[SUHVVLRQV IRU WKH SRVWHULRU GLVWULEXWLRQ JLYHQ E\ f DQG WKH H[SUHVVLRQV IRU WKH SRVWHULRU PHDQV DQG ZH KDYH W $ $ '$f YDULDQFHV RI
PAGE 51

DA $ Q7JJLSf f 1H[W ZULWLQJ IA WKH LRU PHDQV YDULDQFHV DQG FRYDULDQFHV RI WKH ILQLWH SRSXODWLRQ PHDQV DUH JLYHQ E\ I f\ f I Q I f ( Y L\ALVf L QLQ $f \ IM( f§ H 7 Q L Q L $f ALVf` + _VL LVM LALM\LM cQM $f \ +% VD\f f I"9 Q&Q$f \ 1 [rQQ$f [ ? LY L n ffLVf L LY L aLVf [ + $f_(7 L(M LALM\LMBQLQL$f \L f \ Q7JR6OBSBf IL( _D4D$44$fM 7 f

PAGE 52

VD\f f DQG IRU L A N &RY 1L fM O
PAGE 53

FRYDULDQFHV ZKLFK PD\ EH QHFHVVDU\ IRU SURYLGLQJ VLPXOWDQHRXV FRQILGHQFH VHW IRU WKH ILQLWH SRSXODWLRQ PHDQ YHFWRU ZH ZLOO QRW XVH LW KHUH %HIRUH ZH ILQG WKH SRVWHULRU PHDQV DQG YDULDQFHV RI >L A LQ WKH LQILQLWH SRSXODWLRQ VHW XS ZH JLYH D JHQHUDO GLVFXVVLRQ FRPSDULQJ +% SUHGLFWRUV ZLWK (% SUHGLFWRUV :ULWLQJ & ILf\LVf ILQLQL $! ALUVf IL^VW Q L Q L $f 7 rL Vf` +nAf Q I P L [ L ( ; 6LM\LM X L M L QLQL $f \ L Vf 6L2f VD\f f DQG Q7 J4 J D A $ ‘‘

PAGE 54

i $f VD\f f ZH KDYH IURP f f f DQG f WKDW +% ( JL$f_\ f f 9L Y VL$f_\ Lf f DQG 9 ( J$f_\ Lf f ,Q (% DQDO\VLV fWR REWDLQ WKH (% SUHGLFWRU ZH XVXDOO\ UHSODFH ( VL$f_\ f E\ JMA$f (% VD\f ZKHUH $ LV VRPH HVWLPDWH RI $ ZKLFK FDQ EH 0/ 5(0/ RU $129$ HVWLPDWH DQG ZH UHSRUW D QDLYH PHDVXUH RI SRVWHULRU YDULDQFH E\ J$f (% VD\f 8VXDOO\ WKH SRLQW HVWLPDWHV HAJ DQG HAJ QRW WRR IDU DSDUW %XW WR PHDVXUH WKH SRVWHULRU YDULDQFH +% E\ V (% ZH PD\ XQGHUHVWLPDWH WKH DFWXDO PHDVXUH EHFDXVH RI WKH IDLOXUH WR DFFRXQW IRU WKH HVWLPDWLRQ RI $ :H PD\ JURVVO\ XQGHUHVWLPDWH WKH DFWXDO PHDVXUH LI JA$f YDULHV WRR PXFK ZLWKLQ WKH ERG\ RI WKH SRVWHULRU GLVWULEXWLRQ RI $ B LQ WKLV FDVH ZLOO EH VLJQLILFDQWO\ ODUJH :H ZLOO VHH LQ WKLV H[DPSOH WKDW IRU VRPH RI WKH FRXQWLHV FRQWULEXWHV D VLJQLILFDQW SRUWLRQ RI WKH WRWDO YDULDQFH :H ZLOO DOVR ILQG WKDW WKH SHUFHQW FRQWULEXWLRQ RI WR

PAGE 55

+% XVXDOO\ LQFUHDVHV ZLWK WKH UHODWLYH GLIIHUHQFH EHWZHHQ H77I! DQG H +% (% 1RZ WR GHYHORS WKH H[SUHVVLRQV IRU WKH SRVWHULRU PHDQV DQG YDULDQFHV IRU L ZH XVH 7KHRUHP 1RWH WKDW E\ DQ REVHUYDWLRQ PDGH DW WKH HQG RI 6HFWLRQ I$_\f WKH SRVWHULRU GLVWULEXWLRQ RI $ JLYHQ \ LV JLYHQ E\ f $IWHU FRQVLGHUDEOH VLPSOLILFDWLRQV LQ WKLV SDUWLFXODU FDVH ZH R EW DLQ P Q ( ( r c Lf§ M O -9 +% VD\f f Q7 6T 6L 3 DA $ f

PAGE 56

VI VD\f +% f ZKHUH [ 1 f 3f !‘ 7 [ f DQG 4T$f OLSff [L Sff f JLYHQ LQ f DV JLYHQ LQ 1RWH WKDW VLQFH I L DQG [ L 3f L 1 RR LW FDQ EH Q LQIRUPDOO\ WKDW WKH UKV RI f DQG f DSSURDFK LQ OLPLW f DQG f UHVSHFWLYHO\ :H ZLOO QRZ JHW EDFN WR WKH DFWXDO GDWD DQDO\VLV RI WKH %+) GDWD VHW JLYHQ LQ 7DEOH :H XVH IRUPXODV f f f f f DQG f WR REWDLQ +% DQG (% SRVWHULRU PHDQV DQG YDULDQFHV RI WKH SRSXODWLRQ PHDQV IRU WKH FRXQWLHV 2XU +% DSSURDFK HOLPLQDWHV WKH SRVVLELOLW\ RI REWDLQLQJ ]HUR HVWLPDWHV RI WKH YDULDQFH FRPSRQHQWV $ QXPEHU RI GLIIHUHQW SULRUV IRU 5 DQG 5$ ZHUH WULHG ERWK LQIRUPDWLYH DQG QRQ LQIRUPDWLYH 7KH UHVXOWV IRU WKH SRVWHULRU PHDQV ZHUH TXLWH VLPLODU ZKHUHDV WKH SRVWHULRU YDULDQFHV YDULHG DSSUR[LPDWHO\ E\ DV PXFK DV b )RU LOOXVWUDWLRQ SXUSRVH ZH KDYH GHFLGHG WR UHSRUW RXU DQDO\VLV IRU WKH SULRU ZLWK D J4 D DQG 6L %XW VLQFH WKH FKRLFH RI D JLYHV LPSURSHU SRVWHULRU GLVWULEXWLRQ RI $ ZH WRRN DA D VPDOO SRVLWLYH QXPEHU 7DEOH SURYLGHV WKH +% SUHGLFWRUV HAJ WKH (% SUHGLFWRUV HAJ WKH %+) SUHGLFWRUV

PAGE 57

7DEOH 7KH 3UHGL FWHG +HFWDUHV RI &RUQ DQG $VVRFL DWHG 6WDQGDUG (UURUV D f R ,, R E2 DO mL &RXQW\ H+% H(% H%+) 6+% 6(% 6%+) &HUUR *RUGR ) UDQNLQ +DPL WRQ +DQFRFN +DUG L Q +XPERGW .RVVXWK 3RFDKRQWDV :HEVWHU :L QQHEDJR :RUWK :ULJKW

PAGE 58

H%+) DQFA WKH UHVSHFWLYH DVVRFLDWHG VWDQGDUG HUURUV V +% } V (% DQG 6JJS IRU WKH FRUQ GDWD 7DEOH SURYLGHV WKH YDOXHV RI H +% (% %+) DQG +% IRU WKH VR\EHDQV GDWD IRU WKH VDPH FKRLFH RI SULRU K\SHUSDUDPHWHUV ZKHUHDV 7DEOH SURYLGHV WKHLU UHVSHFWLYH VWDQGDUG HUURUV DORQJ ZLWK WKH FRPSRQHQWV DQG 9A V +%r 9DOXHV RI HJJS DQG V %+) SUHVHQWHG LQ 7DEOHV DUH FRPSXWHG XVLQJ )2575$1 IURP WKH IRUPXODV JLYHQ LQ WKH %+) SDSHU DQG DUH VOLJKWO\ GLIIHUHQW IURP WKH YDOXHV UHSRUWHG LQ %DWWHVH HW DO f )URP 7DEOHV DQG IRU SUHGLFWLQJ FRUQ DQG VR\EHDQV RQH FDQ VHH WKDW HJJ H++ +% HSJ DQG H %+) DUH TXLWH FORVH WR HDFK RWKHU )URP 7DEOHV DQG 6SJ DQG 6JJ DSSHDU WR EH VPDOOHU WKDQ f %XW VLQFH 6SJ LV QDLYH (% SRVWHULRU VG LW LV SUREDEO\ DQ XQGHUHVWLPDWH RI WKH WUXH PHDVXUH )URP 7DEOHV DQG ZH ILQG KDUGO\ DQ\ GLIIHUHQFH HLWKHU EHWZHHQ H +% DQG RU EHWZHHQ WKHLU VWDQGDUG HUURUV V +% DQG +%r 7KLV LV ZKDW ZH DQWLFLSDWHG IRU WKLV GDWD 7R GUDZ D FOHDU FRPSDULVRQ EHWZHHQ +% DQG (% SURFHGXUHV ZH DGGHG RQH H[WUD FROXPQ DW WKH HQG RI 7DEOHV DQG 7KH ODVW FROXPQ RI 7DEOH PHDVXUHV WKH SHUFHQW UHODWLYH GLIIHUHQFH [ H+% f H(%_H+% R EHWZHHQ (% DQG +% SUHGLFWHG YDOXHV ZKHUHDV WKH ODVW FROXPQ RI 7DEOH PHDVXUHV WKH SHUFHQW FRQWULEXWLRQ [ 9A9A 9fbf RI 9M WRZDUGV WKH WRWDO SRVWH URU YDULDQFH V 6+% f &RPSDULVRQ RI WKHVH WZR FROXPQV LQGLFDWHV WKDW WKH

PAGE 59

7DEOH 7KH 3UHGLFWHG +HFWDUHV RI 6R\EHDQV 2EWDLQHG E\ 8VLQJ 'LIIHUHQW 3URFHGXUHV D4 J4 DA JA &RXQW\ H%+) [Ob &HUUR *RUGR ) UDQNOLQ +DPL WRQ +DQFRFN +DUGLQ +XPERGW .RVVXWK 3RFDKRQWDV :HEVWHU : L QQHEDJR :RUWK :ULJKW

PAGE 60

7DEOH 7KH 6WDQGDUG (UURUV $VVRFLDWHG ZLWK 'LIIHUHQW 3UHGLFWRUV RI +HFWDUHV RI 6R\EHDQV D4 VR DO f m &RXQW\ 6+% 6+% 6(% 6%+) 9 9 999f[F &HUUR *RUGR ) UDQNOLQ + DP L R Q +DQFRFN +DUGLQ +XPERGW .RVVXWK 3RFDKRQWDV :HEVWHU :L QQHEDJR :RUWK :ULJKW

PAGE 61

FRQWULEXWLRQ RI XVXDOO\ LQFUHDVHV ZLWK WKH UHODWLYH GLIIHUHQFH ,Q SDUWLFXODU IRU FRXQWLHV )UDQNOLQ +XPEROGW DQG .RVVXWK WKHVH UHODWLYH GLIIHUHQFHV DUH DV KLJK DV b b DQG b DQG WKH FRUUHVSRQGLQJ FRQWULEXWLRQV RI 9 DUH DV QRQQHJOLJLEOH DV b b DQG b PDGH 6JJ PXFK VPDOOHU WKDQ VAJ IRU WKHVH 7KL FRXQW LHV 6R LI RQH XVHV D QDLYH (% RU HVWLPDWHG %/83 DSSURDFK KH ZLOO WHQG WR XQGHUHVWLPDWH WKH PHDQ VTXDUHG HUURU 06(f RI SUHGLFWLRQ 2QH VKRXOG QRWH WKDW WKRXJK %+) XVHG HVWLPDWHG %/83 WKH\ WULHG WR DFFRXQW IRU WKH XQFHUWDLQW\ LQYROYHG LQ WKH HVWLPDWLRQ RI $ LQ WKHLU DSSUR[LPDWLRQV RI 06( 6LPLODU DSSUR[LPDWLRQV RI 06( RI SUHGLFWLRQ KDYH EHHQ VXJJHVWHG E\ .DFNDU DQG +DUYLOOH f 3UDVDG DQG 5DR f DQG /DKLUL DQG 5DR f 1RZ ZH ZLOO FRQVLGHU WKH ODPEZHLJKW GDWD VHW RI +DUYLOOH LQ SUHVVf 7KH EDFNJURXQG RI WKH GDWD VHW LV JLYHQ LQ WKH H[DPSOH SUHVHQWHG DW WKH HQG RI 6HFWLRQ :H ZLOO XVH D PRGHO VLPLODU WR WKH RQH JLYHQ LQ f WR DQDO\]H WKH GDWD VHW 7KHUH ZH DVVXPHG IROORZLQJ +DUYLOOH LQ SUHVVf WKH SRSXODWLRQ OLQH HIIHFWV DV IL[HG )RU WKH SXUSRVH RI LOOXVWUDWLRQ ZLWK WKUHH YDULDQFH FRPSRQHQWV ZH ZLOO DVVXPH WKH SRSXODWLRQ HIIHFWV DV UDQGRP 7KLV ZRXOG KDYH EHHQ DSSURSULDWH LI WKHVH ILYH OLQHV ZHUH UDQGRPO\ VHOHFWHG IURP D ODUJH QXPEHU RI SRSXODWLRQV OLQHV $OVR WR PDNH WKH GHVLJQ PDWUL[ DVVRFLDWHG ZLWK WKH IL[HG HIIHFWV

PAGE 62

DJH RI GDPf RI IXOO FROXPQ UDQN ZH LM 1RZ ZH KDYH WKH IROORZLQJ PL[HG ZLOO ZULWH A DV LQHDU PRGH
PAGE 63

WKH %+) H[DPSOH 7KH FKRLFH D D D 6R 6L 6 RI WKH K\SHUSDUDPHWHUV JLYHV D QRQ LQIRUPDWLYH SULRU IRU WKH YDULDQFH FRPSRQHQWV %XW WKH FKRLFH RI D RU D ZLOO JLYH DQ LPSURSHU SRVWHULRU GLVWULEXWLRQ RI $ $7 f Dn r 6R ZH WULHG VHYHUDO FRPELQDWLRQV RI WKHVH K\SHUSDUDPHWHUV ZKLFK DUH VPDOO SRVLWLYH QXPEHUV 2XU ILQGLQJV IRU WKLV GDWD VHW SURYLGHG LQ 7DEOH DUH QRW GLIIHUHQW IURP WKH %+) GDWD VHW :H UHSRUW RXU DQDO\VLV IRU D DA D DQG 6R 6L LQ 7DEOH 7KH HVWLPDWHG %/83V IRU DQG ZAJ UHSRUWHG LQ +DUYLOOH LQ SUHVVf DUH DQG UHVSHFWLYHO\ ZKHUHDV WKH FRUUHVSRQGLQJ YDOXHV ZH REWDLQHG XVLQJ D QRQLQIRUPDWLYH +% DQDO\VLV DUH DQG UHVSHFWLYHO\ 7KH DJUHHPHQW EHWZHHQ WKH WZR VHWV RI HVWLPDWHV LV UHPDUNDEO\ FORVH FRQVLGHULQJ WKH IDFW WKDW WKH XQGHUO\LQJ PRGHOV f DQG f DUH QRW LGHQWLFDO +DUYLOOH LQ SUHVVf DOVR HVWLPDWHG WKH GLIIHUHQFH Z Z DQFr 7KH DVVRFLDWHG 06( RI SUHGLFWLRQ E\ XVLQJ ERWK YDULDQFH FRPSRQHQWV DSSURDFK DQG +% DSSURDFK 7KH HVWLPDWHG 06( RI A a Z rQ QDAYH (%/83 DSSURDFK ZDV JLYHQ E\ f ZKHUHDV IRU .DFNDU DQG +DUYLOOH f R DSSUR[LPDWLRQ LW ZDV f DQG IRU 3UDVDG DQG 5DR f f f DSSUR[LPDWLRQ LW ZDV f )RU +% DSSURDFK +DUYLOOH LQ SUHVVf XVHG D XQLIRUP SULRU ERWK IRU WKH IL[HG HIIHFWV DQG WKH YDULDQFH FRPSRQHQWV DVVRFLDWHG ZLWK VLUH HIIHFW

PAGE 64

7DEOH %LUWK :HLJKWV LQ SRXQGVf RI /DPEV 6LUH 'DP $JH :HLJKW 6LUH 'DP $JH :HLJKW /LQH / L QH L / L QH / L QH /LQH L

PAGE 65

7DEOH 3UHGLFWHG %LUWK :HLJKWV R7 /DPEV DQG $VVRFLDWHG 6WDQGDUG (UURUV J4 / L QH 6LUH H+% r 6+% L L

PAGE 66

DQG HUURU 7KH +% HVWLPDWH R7 ZA f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f W K VPDOO DUHD ZLWK XQLWV ZH DVVXPH WKDW RQ D VDPSOH RI VL]H Q ZH NQRZ RQO\ WKH VDPSOH PHDQ RI WKH FKDUDFWHULVWLF RI LQWHUHVW WKH VDPSOH PHDQ YHFWRU S[Of RI WKH DX[LOLDU\ YDULDEOHV $OVR ZH KDYH LQIRUPDWLRQ RQ WKH SRSXODWLRQ PHDQ YHFWR U 3[Of RI WKH DX[LOLDU\ YDULDEOHV RI WKH XQLWV LQ WKH L f WK VPDOO DUHD RI WKH SRSXODWLRQ :H DUH LQWHUHVWHG LQ SUHGLFWLQJ WKH ILQLWH

PAGE 67

SRSXODWLRQ PHDQ YHFWRU
PAGE 68

YDULDQFH fWKH\ HVWLPDWHG LW LWHUDWLYHO\ E\ DSSO\LQJ JHQHUDOL]HG OHDVW VTXDUHV SURFHGXUH WR
PAGE 69

DQG LWV SRVWHULRU YDULDQFH LV JLYHQ E\ 9 QL" Vf I "YmLO\ Vf f 1RZ IURP Lf JLYHQ E DQG AOVfff A P Vf f
PAGE 70

f 1RWH WKDW IURP Lf DQG LLf ZH KDYH LYf JLYHQ E DQG < a 1$E 9f ZKHUH OVff f§ f aP Vff7 f f 9 'LDJA P f DQG IURP LLLf ZH FDQ ZULWH Yf JLYHQ % 3 XQLIRUP5 f )URP LYf DQG Yf ZH KDYH WKH MRLQW SGI RI DQFr JLYHQ \ 2f E_Lf D P Q ORf a8[ 2f $Ef79 Vf $Ef f 1RZ DVVXPH UDQN$f S DQG GHILQH E DWY!DfDWY"^Vf LLVfp76f$L P B f VfAL VfL ff f 4
PAGE 71

RI f FDQ EH ZULWWHQ DV $Ef7< A\A$Ef EEf7$7<$fEEf "f.Vf f )URP f DQG f LW IROORZV WKDW JLYHQ DQG 6 % a 1AE $A9A$f 1RWH WKDW LQ f WKDW E GHSHQGV RQ \ DQG VLQFH FU7 GHSHQG RQ $JDLQ XVLQJ WKH LWHUDWLYH IRUPXODV IRU H[SHFWDWLRQ DQG YDULDQFH ZH KDYH DQG f 9 ( 9 e V ?L O f

PAGE 72

6R WR HYDOXDWH (LO\6M DQG 9LO\Vfff LW` IROORZV IURP f f f DQG f WKDW LW LV HQRXJK WR HYDOXDWH A f (DLPO\Vff DQG WKH TXDQWLWLHV WKDW DSSHDU RQ WKH UKV RI f DQG f ,Q RUGHU WR HYDOXDWH WKHP ZH QHHG WR ILQG WKH FRQGLWLRQDO GLVWULEXWLRQ RI $ JLYHQ
PAGE 73

f DUH DFFRPSOLVKHG QRZ E\ XVLQJ f DQG W\SLFDOO\ VRPH QXPHULFDO LQWHJUDWLRQ WHFKQLTXHV

PAGE 74

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f RI 6HFWLRQ ZLWK $ WKH YHFWRU RI UDWLRV RI YDULDQFH FRPSRQHQWV NQRZQ :H DVVLJQ D XQLIRUP SULRU WR WKH YHFWRU RI IL[HG HIIHFWV E DQG DQ LQGHSHQGHQW JDPPD SULRU WR WKH LQYHUVH RI HUURU YDULDQFH :H ILUVW ILQG WKH

PAGE 75

SRVWHULRU GLVWULEXWLRQ RI QRQVDPSOHG XQLWV JLYHQ WKH VDPSOHG XQLWV LQ ILQLWH SRSXODWLRQ VDPSOLQJ DQG IURP WKLV ZH GHULYH WKH +% SUHGLFWRU RI WKH ILQLWH SRSXODWLRQ PHDQ YHFWRU /DWHU LQ WKLV VHFWLRQ LQ LQILQLWH SRSXODWLRQ VLWXDWLRQ WKH SRVWHULRU GLVWULEXWLRQ RI WKH YHFWRU RI IL[HG DQG UDQGRP HIIHFWV DQG WKH +% SUHGLFWRUV IRU OLQHDU FRPELQDWLRQV RI IL[HG DQG UDQGRP HIIHFWV DUH GHWHUPLQHG 2XU DSSURDFK WR WKHVH SUREOHPV FDQ EH UHJDUGHG DV H[WHQVLRQV RI WKH +% LGHDV RI /LQGOH\ DQG 6PLWK f WR SUHGLFWLRQ $OWKRXJK GHYHORSHG ZLWKLQ D %D\HVLDQ IUDPHZRUN RXU UHVXOWV VKRXOG EH RI DSSHDO DOVR WR IUHTXHQWLVWV )RU ERWK WKH SUREOHPV WKH %/83 QRWLRQ IRU UHDO YDOXHG SDUDPHWHUV VHH IRU H[DPSOH +HQGHUVRQ 5R\DO f LV H[WHQGHG LQ 6HFWLRQV DQG WR YHFWRU YDOXHG SDUDPHWHUV DQG LW LV VKRZQ WKDW WKH %D\HVLDQ SUHGLFWRUV RI 6HFWLRQ DUH LQGHHG %/83 /LNH RWKHU UHODWHG SDSHUV RXU %/83 UHVXOWV GR QRW UHTXLUH DQ\ QRUPDOLW\ DVVXPSWLRQ :LWK WKH DGGHG DVVXPSWLRQ RI QRUPDOLW\ WKH %/83V LQGHHG WXUQ RXW WR EH EHVW XQELDVHG SUHGLFWRUV %83Vf ZLWKLQ WKH FODVV RI DOO XQELDVHG SUHGLFWRUV ,Q DGGLWLRQ LW LV VKRZQ WKDW WKHVH %D\HV SUHGLFWRUV DUH %83V HYHQ IRU VRPH QRQQRUPDO GLVWULEXWLRQV ,Q WKHVH VHFWLRQV ZH KDYH DOVR VKRZQ WKDW WKH %/83V DOVR fXQLYHUVDOO\f RU fVWRFKDVWLFDOO\ff GRPLQDWH FI +ZDQJ f WKH OLQHDU XQELDVHG SUHGLFWRUV IRU HOOLSWLFDOO\ V\PPHWULF

PAGE 76

G L VW U L EXW LRQV ,Q 6HFWLRQV DQG ZH KDYH VKRZQ WKDW WKHVH %D\HV SUHGLFWRUV DUH EHVW HTXLYDULDQW SUHGLFWRUV IRU ERWK WKH PDWUL[ ORVV RU VWDQGDUGL]HG PDWUL[ ORVVf DQG TXDGUDWLF ORVV RU VWDQGDUGL]HG TXDGUDWLF ORVVf XQGHU VXLWDEOH JURXSV RI WUDQVIRUPDWLRQV IRU D EURDG FODVV RI HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV LQFOXGLQJ EXW QRW OLPLWHG WR WKH QRUPDO GLVWULEXWLRQ :H FRQFOXGH WKLV VHFWLRQ E\ LQWURGXFLQJ D IHZ QRWDWLRQV )RU D VTXDUH PDWUL[ 7 W[Wf WU7f GHQRWHV LWV WUDFH )RU D V\PPHWULF QRQQHJDWLYH GHILQLWH QQGf ,, PDWUL[ 7 7 LV D V\PPHWULF QQG PDWUL[ VXFK WKDW 77 7 DQG IRU D V\PPHWULF SG PDWUL[ 7 7 LV D V\PPHWULF L SG PDWUL[ VXFK WKDW 7 Wf N 7KH +LHUDUFKLFDO %D\HV 3UHGLFWRU LQ D 6SHFLDO &DVH :H ZLOO DVVXPH WKH QRUPDO OLQHDU PRGHO f RI 6HFWLRQ :H FRQVLGHU LQ WKLV VHFWLRQ WKH VSHFLDO FDVH ZKHQ $ WKH YHFWRU RI WKH UDWLRV RI YDULDQFH FRPSRQHQWV LV NQRZQ ZKLOH % DQG 5 DUH LQGHSHQGHQWO\ GLVWULEXWHG ZLWK % XQLIRUP5Af DQG 5 a JDPPDAD4 AJ4A +HUH ZH ZLOO FRQVLGHU WKH FDVH RI ILQLWH SRSXODWLRQ VDPSOLQJ LQ GHWDLOV DQG EULHIO\ PHQWLRQ WKH FRUUHVSRQGLQJ UHVXOWV IRU WKH LQILQLWH SRSXODWLRQ VHW XS ,Q ILQLWH SRSXODWLRQ VDPSOLQJ VLQFH $
PAGE 77

VWLOO LQWHUHVWHG LQ ILQGLQJ WKH SUHGLFWRU RI f
PAGE 78

:H PD\ QRWH fWKDW WKH SUHGLFWRU HMAS \ff JL JLYHQ LQ f LV WKH RXWFRPH RI WKH PRGHO JLYHQ E\ $f DQG %f ZLWK $ NQRZQ DQG WKH XVH RI XQLIRUP53f SULRU RQ % DQG LW GRHV QRW GHSHQG RQ WKH FKRLFH RI WKH SULRU SURSHUf GLVWULEXWLRQ RI 5 7KLV FDQ EH IRUPDOO\ VHHQ DVVXPLQJ DOO WKH H[SHFWDWLRQV DSSHDULQJ EHORZ H[LVW DV IROORZV f f ( ( ( < f 5
PAGE 79

GHILQLWLRQ RI 0 JLYHQ LQ f 7KXV WKH SUHGLFWRU %IrV UrNXVW DJDLQVW WKH FKRLFH RI SULRUV IRU 5 7KHUH DUH DOWHUQDWH ZD\V WR JHQHUDWH WKH VDPH SUHGLFWRU A A RI f ff[ < 6XSSRVH IRU H[DPSOH RQH DVVXPHV RQO\ $f DQG %f ZLWK E NQRZQ U PD\ RU PD\ QRW EH NQRZQf 7KHQ WKH EHVW SUHGLFWRU EHVW OLQHDU SUHGLFWRU ZLWKRXW WKH QRUPDOLW\ DVVXPSWLRQf RI eA
PAGE 80

6LPLODUO\ LQ WKLV VSHFLDO FDVH RQH FDQ GHULYH WKH +% SUHGLFWRU IRU TXDGUDWLF ORVVf RI &
PAGE 81

%HVW 8QELDVHG 3UHGLFWLRQ DQG 6WRFKDVWLF 'RPLQDWLRQ LQ 6PDOO $UHD (VWLPDWLRQ ,Q WKLV VHFWLRQ ZH DVVXPH WKH QRUPDO OLQHDU PRGHO LQ f ZLWK $ NQRZQ 1R SULRU GLVWULEXWLRQ IRU % DQG 5 LV DVVXPHG DQG 7 7 E Uf LV WUHDWHG DV DQ XQNQRZQ SDUDPHWHU )LUVW r f? ZH SURYH WKH RSWLPDOLW\ RI eJS< ZLWKLQ WKH FODVV RI DOO XQELDVHG SUHGLFWRUV RI Yf \ff 1H[W ZH GLVSHQVH ZLWK WKH QRUPDOLW\ DVVXPSWLRQ RI \ DQG H DQG SURYH WKH RSWLPDOLW\ RI HQA
PAGE 82

DVVXPH WKDW HDFK FRPSRQHQW R J
PAGE 83

&RY ff J
PAGE 84

IRU DOO L X +HQFH f KROGV DQG WKH SURRI RI WKH OHPPD LV FRPSOHWH 5HPDUN ff WKDW ,W IROORZV IURP WKH DERYH OHPPD VHH LI 7\ DQG 7\nnf DUH ERWK %83V RI J7L\ff JW\ff VRQ 7\ff L\ff E\ f IRU DOO LH 3MIO
PAGE 85

3URR R 7KHRUHP ,Q YLHZ R /HPPD LW VXLFHV WR VKRZ WKDW IRU HYHU\ PA< A A e 8T &RY EI\ f f  < ff .rffO IRU DOO WKDW LV (-40< f Zf:ZOf < :\ IRU DOO 6LQHH XQGHU WKH PRGHO f 0< R Z fL\! ; f ( (aA; A A [f9LAf f 9`9nnf E XV L QJ ( f LW VXLFHV WR VKRZ WKDW f LRU DOO 6 L QHH GLHUHQWLDWLQJ ERWK VLGHV R WKLV HTXDWLRQ ZUW E RQH JHWV VHH S R 5DR f Of7fL A f 8VLQJ (J f DJDLQ 7KH SURR R 7KHRUHP f IROORZV IURP f LV FRPSOHWH

PAGE 86

5HPDUN (TXDWLRQ f FDQ EH DOWHUQDWLYHO\ SURYHG LQ WKH IROORZLQJ ZD\ 1RWH WKDW VLQFH < Lf Lf E UBa IRU f7 f < fWN\f V FRPSOHWH VXIILFLHQW +HQFH PXVW KDYH FRYDULDQFH YHFWRU ZLWK HYHU\ ]HUR HVWLPDWRU P < f f r L H ( H "LL< f9 1H[W ZH VKRZ WKDW WKH FRQFOXVLRQ RI 7KHRUHP FRQWLQXHV WR KROG HYHQ IRU FHUWDLQ QRQQRUPDO GLVWULEXWLRQV 6XSSRVH WKDW H \A f DQG $ 'LDJ' Af $VVXPH WKDW JLYHQ 5 U 1 U $f ZKLOH WKH GI RI 5 LV DQ DUELWUDU\ PHPEHU RI WKH IDPLO\ ^) ) LV DEVROXWHO\ FRQWLQXRXV ZLWK SGI IUf IRU U ` /HW r GHQRWH D r f VXEIDPLO\ RI VXFK WKDW HDFK FRPSRQHQW RI f DQG A< A ? < KDV ILQLWH VHFRQG PRPHQW XQGHU WKH PRGHO f DQG WKH MRLQW GLVWULEXWLRQ RI Hr DQG 5 :H QRZ SURYH WKH IROORZLQJ WKHRUHP 7KHRUHP Hi)rff LV %83 RI f f < QGHU WKH PRGHO f Ha_5 U a 1 U $f DQG 5 KDV D GI IURP 9r 3URRI RI 7KHRUHP 8VLQJ /HPPD DQG IROORZLQJ WKH SURRI RI 7KHRUHP LW VXIILFHV WR VKRZ WKDW ( E) f7 f QLL !< f R f IRU DOO E f 5A DQG IRU DOO ) r ZKHUH

PAGE 87

R ( E) f f DQG ( E) P Y f RR IRU DOO E DQG DOO ) e r &RQVLGHU WKH VXEIDPLO\ 5 JDPPDAF Gf F G ! RI 6L QFH f KROGV IRU WKLV VXEIDPLO\ (A S f IRU DOO E DQG DO ) +! JLYHV LQ7GfO U f IRU DOO E F DQG G 1RZ XVLQJ WKH XQLTXHQHVV SURSHUW\ RI /DSODFH WUDQVIRUPV LW IROORZV IURP f WKDW DH /HEHVJXH IRU DOO U DQG DOO E LH f DH /HEHVJXH IRU DOO U DQG DOO E 'LIIHUHQWLDWLRQ RI ERWK VLGHV RI f ZLWK UHVSHFW WR E DQG VRPH

PAGE 88

VLPSOLILFDWLRQV XVLQJ f OHDG WR f DH /HEHVJXH IRU DOO U DQG DOO E 0XOWLSO\LQJ ERWK KQ7 VLGHV RI f E\ U DQG LQWHJUDWLQJ ZLWK UHVSHFW WR G)Uf ZKHUH ) e RQH JHWV f 5HPDUN 6LQFH GRHV QRW FRQWDLQ WKH GHJHQHUDWH GLVWULEXWLRQV RI 5 RQ RRf 7KHRUHP GRHV QRW IROORZ IURP 7KHRUHP 5HPDUN ,Q 7KHRUHP LI ZH WDNH b IRU ZH VHH WKDW WKH PDUJLQDO GLVWULEXWLRQ RI < LV JLYHQ E\ WKH IDPLO\ RI GLVWULEXWLRQV _7_1UM ;E FGf( Gf E e 5A F G !` DQG HJ)
PAGE 89

DOO ZH VD\ WKDW A< f? LV D /83 RI f \f :H QHHG WKH IROORZLQJ GHILQLWLRQ 'HILQLWLRQ Lf Lf $ /83 3
PAGE 90

3URRI RI 7KHRUHP ,I ( P7< f P7;AE IRU DO &RY 7;f R7 +HQFH >VEI Y2!f f < f? B7<' f p7< &RY H FP\ f B \f f P7< f J0eO[ aOfa F[ f (LLnff& f96<[W!7 P IRU DOO 7KH ODVW fWZR HTXDOLWLHV IROORZ IURP WKH GHILQLWLRQ RI 0 DQG IURP WKH IDFW 7\ f P D 7 $SSO\LQJ /HPPD WKH UHVXOW IROORZV 5HPDUN $V DOUHDG\ PHQWLRQHG WKH QRUPDOLW\ DVVXPSWLRQ LV QRW QHHGHG IRU SURYLQJ WKH %/83 SURSHUW\ RI a%IAAff 7KHRUHP XQLILHV DQG H[WHQGV WKH DYDLODEOH %/83 UHVXOWV UHODWHG WR WKH HVWLPDWLRQ RI WKH ILQLWH SRSXODWLRQ PHDQ YHFWRU XQGHU GLIIHUHQW PRGHOV FI *KRVK DQG /DKLUL LQ SUHVV 5R\DO DQG RWKHUVf $V L Q 5HPDUN RQH FDQ SURYH WKDW WKH %/83 LV XQLTXH ZLWK SUREDELOLW\ RQH ,W IROORZV DV D FRQVHTXHQFH RI 7KHRUHP WKDW a%I

PAGE 91

/e f D R D R7 f DQG WKH PRGHO f ZLWKRXW DQ\ GLVWULEXWLRQDO DVVXPSWLRQ RQ Hr 7KH RSWLPDOLW\ RI H ZLWKLQ /83V KROGV D IRUWLRUL XQGHU WKH TXDGUDWLF ORVV / W^ f ?a L? Q D RWT R WU>M4/A f ZKHUH 4 LV D QQG PDWUL[ 6XFK D ORVV ZLOO KHQFHIRUWK EH UHIHUUHG WR DV JHQHUDOL]HG (XFOLGHDQ HUURU ZUW 4 7KH RSWLPDOLW\ UHVXOWV FDUU\ RYHU YLD 7KHRUHP DQG 7KHRUHP XQGHU WKH DGGHG GLVWULEXWLRQDO DVVXPSWLRQ ZKLFK LV QRW QHFHVVDULO\ QRUPDOLW\ DVVXPSWLRQf RQ H $ QDWXUDO TXHVWLRQ WR DVN QRZ LV ZKHWKHU WKH ULVN RSWLPDOLW\ WKH FODVV RI DOO XQELDVHG SUHGLFWRUV RU DW OHDVW ZLWKLQ WKH FODVV RI /83V XQGHU FHUWDLQ RWKHU FULWHULRQ IRU D EURDGHU IDPLO\ RI GLVWULEXWLRQV RI Hr 7R LQYHVWLJDWH WKLV TXHVWLRQ ZH QHHG WKH QRWLRQV RI fXQLYHUVDOf DQG fVWRFKDVWLFf GRPLQDWLRQ DQG WKHLU LQWHUUHODWLRQVKLS DV JLYHQ LQ +ZDQJ f /HW f ( H / f f f < 4 EH WKH ULVN IXQFWLRQ RI WKH SUHGLFWRU IRU SUHGLFWLQJ e XQGHU D ORVV IXQFWLRQ ZKLFK LV D IXQFWLRQ RI JHQHUDOL]HG (XFOLGHDQ HUURU

PAGE 92

Z U W 4 IRU VRPH IXQFWLRQ / 7KH IROORZLQJ GHILQLWLRQ LV DGDSWHG IURP +ZDQJ f $Q HVWLPDWRU O< f A XQLYHUVDOO\ HUURU 'HILQLWLRQ GRPLQDWHV XQGHU WKH JHQHUDOL]HG (XFOLGHDQ ZUW ILf LI IRU HYHU\ DQG HYHU\ QRQGHFUHDVLQJ ORVV IXQFWLRQ / 5MA e Af 5AA e f KROGV DQG IRU D SDUWLFXODU ORVV WKH ULVN IXQFWLRQV DUH QRW LGHQWLFDO +ZDQJ f KDV VKRZQ WKDW VHH KLV 7KHRUHP f A XQLYHUVDOO\ GRPLQDWHV 6 XQGHU WKH JHQHUDOL]HG (XFOLGHDQ HUURU ZUW 4 LI Lff 6< Lf < f V VWRFKDVWLFDOO\ VPDOOHU WKDQ f? Z f ; < :H VD\ WKDW D UDQGRP YDULDEOH =A LV VWRFKDVWLFDOO\ VPDOOHU WKDQ MUV LI 3"= [f 3Q=f [f IRU DOO [ DQG DQG IRU VRPH =K DQG =A KDYH GLVWLQFW GLVWULEXWLRQV 7KH QH[W WKHRUHP VKRZV WKDW IRU D JHQHUDO FODVV RI HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV RI Hr VI\ff XQLYHUVDOO\ GRPLQDWHV HYHU\ /83 +< Lf RI &' < f M XQ GHU HYHU\ JHQHUDOL]HG (XFOLGHDQ HUURU ZUW D QQG 4 $VVXPH WKDW Ha KDV DQ HOOLSWLFDOO\ V\PPHWULF SGI JLYHQ E\ KrHr_$ Uf RF U $ I r7D $n9f f P US A ZKHUH DV DOUHDG\ GHILQHG Y H f DQG $ 'LDJ' f DQG WKH NQRZQ QRQQHJDWLYH IXQFWLRQ I LV VXFK WKDW

PAGE 93

T (nY L QW (_HL L On I UH rWDL HrMGH RR f ZKHUH Y 9 T f 7 DQG H L }f f f r 1 7 f7 :H ZLOO GHQRWH fWKLV GLVWULEXWLRQ E\ BS U A$f ZKHUH iBSL GHQRWHV WKH GLVWULEXWLRQ ZKRVH SGI LV JLYHQ E\ N W IL 4r Df RF Ar M 4 I e 74rBe rf f 3 ZKHUH W DQG >L DUH LQ 5 4 S[Sf LV SG DQG D 1RWH WKDW WKH QRUPDOLW\ RI Hr ZLWK PHDQ DQG YDULDQFHFRYDULDQFH PDWUL[ UA$ LV VXIILFLHQW EXW QRW QHFHVVDU\ IRU f DQG f WR KROG ,W IROORZV IURP f WKDW (Haf H[LVWV DQG IURP f (Hrf 1RWH WKDW H U ;$f Hr KDV D VSKHULFDOO\ V\PPHWULF GLVWULEXW LRQ (AH[S L XAH r r JI rf ZLWK FKDUDFWHULVWLF IXQFWLRQ FIf 7n FX Xf IRU VRPH IXQFWLRQ F VHH .HONHU f ZKHUH L A 8 8M XArf A DQG Tr 1 +HQFH Hr KDV FI JLYHQ E\ 7 T (AH[S L XAHrfA FU A$Xf f 1RZ ZULWH :r }=A\ M f ZI Z r7 7 KDV D FI JLYHQ E\ 7KHQ : (AH[S LXA:rfA@ FU AXrA6Xrf f

PAGE 94

ZKHUH ( ='=A &RPSDULQJ f DQG f RQH FDQ VHH IURP f fWKDW : KDV DOVR DQ HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQ eA U A(f ZLWK SGI JLYHQ E\ KZr_ ( Uf D U( L I A UZ r7YO f 7KHRUHP 8QGHU WKH PRGHO f f DQG f HJ)< `f XQLYHUVDOO\ GRPLQDWHV HYHU\ /83 f +< f B f f RI < A IRU HYHU\ SG 4 5HPDUN 7KHRUHP GRHV QRW FRQWDLQ 7KHRUHP VLQFH 7KHRUHP UHTXLUHV WKH HOOLSWLFDO V\PPHWU\ RI WKH GLVWULEXWLRQ RI :r ZKLOH WKH RWKHU GRHV QRW ,W VKRXOG EH QRWHG WKRXJK WKDW WKH PRGHO DVVXPSWLRQ PDGH LQ f LV QRW QHFHVVDULO\ VWURQJHU WKDQ WKH XVXDO DVVXPSWLRQ RI ILQLWHQHVV RI FHUWDLQ PRPHQWV 7KLV LV EHFDXVH WKH DVVXPSWLRQV RI 7KHRUHP KROG HYHQ LI D GLVWULEXWLRQ KDV LQILQLWH VHFRQG PRPHQW HJ IRU FHUWDLQ PXOWLYDULDWH Wf EXW WKH %/83 SURSHUW\ LV PHDQLQJOHVV LQ VXFK LQVWDQFH 1RZ ZH ZLOO VWDWH DQG SURYH D OHPPD 7KH SURRI RI 7KHRUHP UHVWV FUXFLDOO\ RQ WKLV OHPPD /HPPD ,I :1UA[Of KDV SGI KAZ A U M WKHQ IRU HYHU\ /X[1USf X 1US /: 7 f§ //AfA:8 ZKHUH : Q X OX 4f: ZKH UH 1A XfM LV D QXOO PDWUL[ DQG PHDQV HTXDO LQ GLVWULEXWLRQ

PAGE 95

3URRI RI /HPPD 7KH SURRI IROORZV fWKH DUJXPHQWV RI +ZDQJ f )URP f LW IROORZV WKDW : KDV FI (AH[S LWA:fFUWAWf +HQFH (AH[SLWn>/ZfFUBWn-n//7Wf f ZKHUH WA LV D X[O YHFWRU 1H[W XVLQJ f f7//7f+ f VR WKDW WKH OHPPD IROORZV IURP f DQG f ,W IROORZV DV D FRQVHTXHQFH RI /HPPD WKDW :7/7/: /:f7/:f 7 A//7f:XM A//7f:XM :X//7:X f :H VKDOO XVH f UHSHDWHGO\ IRU SURYLQJ 7KHRUHP 3URRI RI 7KHRUHP /HW +< A A EH D /83 RI 9ff 7KHQ XQGHU WKH OLQHDU PRGHO f DQG IURP WKH IDFW WKDW (f LW LV HDV\ WR VHH WKDW

PAGE 96

f B :ULWLQJ : ( :r DQG XVLQJ f DQG f RQH JHWV :r7>+$ B F@74>+$ F@Zr :7(>+$ B &@7Q>+$ F@HZ :A>+$ &@(>+$ &@7IL:X f 6LPL DUO\ f :ULWH 7 + $ &0 7KHQ UKV RI f UKV RI f Z;T Z X f VLQFH XVLQJ WKH GHILQLWLRQ RI 0 JLYHQ LQ f

PAGE 97

f7 7 f DQG XVLQJ f + $ &0f ; A A &;A 0;Af f 7KHRUHP IROORZV QRZ IURP f f $OVR VLQFH LV SRVLWLYH GHILQLWH LW IROORZV IURP f WKDW UKV RI f UKV RI f LI DQG RQO\ LI 7 WKDW LV + $ &0 FI +ZDQJ f %HVW 8QELDVHG 3UHGLFWLRQ DQG 6WRFKDVWLF 'RPLQDWLRQ LQ ,QILQLWH 3RSXODWLRQ ,Q WKLV VHFWLRQ ZH ZLOO EULHIO\ FRQVLGHU D IHZ RSWLPDO SURSHUWLHV RI HJM
PAGE 98

'HILQLWLRQ $ SUHGLFWRU \\7 \ f \ \ f@ FR LV %83 IRU J:f LI DQG RQO\ LI &RYA>8APAP
PAGE 99

(>^VELgf &E rW 7KLV LV SURYHG VLPLODU WR f (B
PAGE 100

1RZ ZH ZLOO VKRZ WKDW "JM
PAGE 101

f DQG f i%O
PAGE 106

XQGHU WKH ORVV IRU HYHU\ QQG 4 &RQYHUVHO\ LI rRr < A A LV WKH EHVW HTXLYDULDQW SUHGLFWRU RI XQGHU WKH ORVV IRU HYHU\ QQG 4 WKHQ LW LV VR XQGHU /T :H VWDWH DQG SURYH 7KHRUHP ZKLFK HVWDEOLVKHV WKH RSWLPDOLW\ RI ZLWKLQ WKH FODVV RI DOO HTXLYDULDQW SUHGLFWRUV RI eA< A ?
/RLVf f +HQFH E\ &DXFK\6FKZDU] LQHTXDOLW\ ( Lm.g N\ ILQLWH 8QGHU WKH PDWUL[ ORVV /T

PAGE 107

(/ e ef a /L 6%)f_ (A A mN\ N\ f ( H 6%) f 9! fA,WN\ ff ( H mN\ f %Iff f 7 f 1RZ ZH ZLOO VKRZ WKDW WKH ODVW WZR SURGXFW WHUPV DUH QXOO PDWULFHV 8VLQJ WKH GHILQLWLRQV RI H %)
PAGE 108

( H r /R& f f /Te %) ( eN\ Lf fmWN\ f 4 f ZLWK HTXDOLW\ LI DQG RQO\ LI 7KH SURRI RI 7KHRUHP LV FRPSOHWH 1H[W LQVWHDG RI ZH FRQVLGHU WKH JURXS i WUDQVIRUPDWLRQV JLYHQ E\ ^I""Gf AfU3f G! G\f G\ ;AM $ SUHGLFWRU J\ RI A

PAGE 109

G\ Lf ; f G\ f D[ f H[ f f IRU DOO \ Lf c DQG G 1RWH WKDW iJS < A  LV DJDLQ DQ HTXLYDULDQW SUHGLFWRU RI < f \f QGHU fWKH DERYH FULWHULRQ $OVR QRWH WKDW LI LV HTXLYDULDQW XQGHU WDNLQJ G XQGHU LQ f LW IROORZV WKDW LW LV HTXLYDULDQW 6R WKH FODVV RI HTXLYDULDQW SUHGLFWRUV XQGHU 4 LV VPDOOHU WKDQ WKDW XQGHU %XW ZH ZLOO SURYH WKDW fEI7 LV EHVW LQVLGH WKLV VPDOOHU FODVV XQGHU D ODUJHU IDPLO\ RI GLVWULEXWLRQV RI < ZKLFK LQFOXGHV WKH QRUPDO GLVWULEXWLRQ DV D VSHFLDO FDVH :H ZLOO DSSO\ WKH JURXS RI WUDQVIRUPDWLRQV A RQ fWKH IDPLO\ RI HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV ZLWK SGI JLYHQ E\ I"\f D ( I\ ;Ef"A ;EfUf f ZKHUH I LV D NQRZQ QRQQHJDWLYH IXQFWLRQ VDWLVI\LQJ \ ;Ef7"\ ;Ef@ I \ ;Ef"AA ;Ef M7 MG\ 22 f 4 1RWH WKDW < a 1;E D (f LV D VSHFLDO FDVH VDWLVI\LQJ f DQG f ,W LV HDV\ WR VHH WKDW WKH DERYH JURXS RI WUDQVIRUPDWLRQV A RQ 1USG L PHQV L RQDO (XF LGHDQ VSDFH IRU < LQGXFHV D JURXS RI WUDQVIRUPDWLRQV

PAGE 110

^LG! L H 5S G J G f§ + X f 7 7[7 GD GEf ,f` f RQ WKH SDUDPHWHU VSDFH JLYHQ LQ f $V EHIRUH D ORVV IXQFWLRQ /e 6nf IRU SUHGLFWLQJ f < f? / FfOf? f E\ 6? LV LQYDULDQW XQGHU WKH JURXS RI WUDQVIRUPDW LRQV LI / G f f? L9Of \ D[ F[ f fL! I"GAf G\ f f rr [! F[A / 8! f f} } \ Lf f IRU DOO \AO? \A? 3 G !f DQG :H VKDOO FRQVLGHU WKH VSHFLDO ORVVHV /e f /i fD f DQG /e J Jf /MLL Lf$ f %RWK WKHVH ORVVHV VDWLVI\ f 7R ILQG WKH EHVW HTXLYDULDQW SUHGLFWRU RI f \f LQ WKLV VHW XS ZH ZLOO SUHVHQW WZR OHPPDV ZKLFK SURYLGH D XVHIXO FKDUDFWHUL]DWLRQ RI WKH FODVV RI HTXLYDULDQW SUHGLFWRUV $ SURRI RI /HPPD LV RPLWWHG /HPPD f Zf /HW \f, EH DQ HTXLYDULDQW SUHGLFWRU IRU < 7KHQ D QHFHVVDU\ DQG VXIILFLHQW FRQGLWLRQ

PAGE 111

IRU D SUHGLFWRU A< f? [f RI < WR EH HTXLYDULDQW LV WKDW \ff AR\ff Lf f IRU DOO \ A ZKHUH KG\ r1 [f GK\ f f IRU DOO \ f $ DQG G :H ZLOO IROORZ WKH DUJXPHQWV RI /HPPDV DQG RI 'DWWD DQG *KRVK f WR ILQG D UHSUHVHQWDWLRQ RI WKH IXQFWLRQ K LQ f 'HILQH L.\ Lf .\f\ '?\fA L >\f7.\f!R@ f 1RWH WKDW VLQFH .(A. VR \ f7fB f Of979 ff .\ .\AUA.\Af DQG W LV LQGHHG D IXQFWLRQ RI .\ Lf ,W FDQ EH VKRZQ WKDW W.\ f? LV D PD[LPDO LQYDULDQW XQGHU WKH JURXS RI WUDQVn IRUPDW L RQV  ^V}G!  5 3 G 6LG\ Lf G\A ;f f L QGXFHG E\ i Q f WKH \ VSDFH 7KH IROORZLQJ OHPPD FKDUDFWHUL]HV WKH FODVV RI IXQFWLRQV KA\AA VDWLVI\LQJ f :H ZLOO XVH WKLV OHPPD WR FKDUDFWHUL]H WKH FODVV RI HTXLYDULDQW SUHGLFWRUV

PAGE 112

/FPQLF/ f f f $ IXQFWLRQ \ff X[Lf VDW LVILHV f LI DQG RQO\ LI K KDV WKH UHSUHVHQWDWLRQ Lf 'WN '?f ZKHUH V X[Of LV DQ DUELWUDU\ IXQFWLRQ RI WA.\ f 3URRI RI /HPPD $VVXPH K KDV WKH UHSUHVHQWDWLRQ JLYHQ E\ f 1RZ VLQFH G\ f ; .G\ A [[! G\!7.\f DQG VLQFH W L"\ Lf D PD[LPDO LQYDULDQW XQGHU 4T VR W A [AA DQG FRQVHTXHQWO\ W.\ f EG\ ;' G \ f7f fQ_ .\ V W f GK\ f +HQFH f LV VDWLVILHG 2Q \ LI 6LQFH K VDWLVILHV f IRU DOO \ Lf DQG G WDNLQJ G ZH VHH WKDW K PXVW VDWLVI\ Lf [ f \ A A IRU DOO \ DQG 7KLV LPSOLHV WKDW K PXVW EH LQYDULDQW XQGHU DQG KHQFH PXVW EH D IXQFWLRQ RI .\Af 6R K\ 2K V.\ ZKHUH V X[Of LV DQ DUELWUDU\ IXQFWLRQ LQJ GV i.\ Of L H VAG.\ 2K GVA2f i.\ f

PAGE 113

IRU DOO G 1RZ WDNLQJ G A\ A A.\ A A r IRU \f7MA\Of A T ZH KDYH IURP f E\ff I.\ff W\ f7.\fUL.\ff\f7.\ffL 6W.\ff9\f.\f 1RZ IRU \A7.\nnf LI ZH WDNH K WKHQ f LV VDWLVILHG DQG ZH FDQ UHSUHVHQW K E\ f 6LQFH A A LV DQ HTXLYDULDQW SUHGLFWRU RI f ff? < LW IROORZV IURP /HPPD DQG /HPPD WKDW A< A A LV DQ HTXLYDULDQW SUHGLFWRU RI eA/A f K8 4f@ LV Q f Q f G f 5HPDUN 1RWH WKDW LI T
PAGE 114

A f? :H QRZ HVWDEOLVK WKDW HQQ < LV EHVW HTXLYDULDQW f f SUHGLFWRU RI Y n < WKHRUHP LV SURYHG a%) XQGHU WKH ORVV / r 7KH IROORZLQJ 7KHRUHP 8QGHU WKH PRGHO JLYHQ LQ f DQG WKH IDPLO\ RI HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQV DV JLYHQ E\ f DQG f WKH JURXS RI WUDQVIRUPDWLRQV 4 JLYHQ LQ f DQG WKH ORVV / JLYHQ LQ f WKH EHVW HTXLYDULDQW SUHGLFWRU RI H6)
PAGE 115

,W LV HQRXJK WR SURYH WKDW WKH ODVW WZR SURGXFW WHUPV LQ f DUH QXOO PDWULFHV 3URFHHGLQJ DV LQ WKH SURRI RI 7KHRUHP ZH KDYH f 1RZ IURP WKH SURSHUW\ RI HOOLSWLFDOO\ V\PPHWULF GLVWULEXWLRQ LW IROORZV WKDW ;!E 6OL < f DQG XVLQJ WKLV ZH KDYH IURP f WKDW f Yf9Yf7H['A f7 L f LOO ; f79B f f§ f§ f 7KHQ WKH VHFRQG WHUP RI WKH UKV RI f f 9f99f7 O[Off [ ( f 7LUY f?AY f 7AL f .< :[ 6LL< ; f76O`[!Ef [ V7OW.
PAGE 116

1RZ ZH ZLOO VKRZ fWKH H[SHFWDWLRQ RI WKH UKV RI f LV D QXOO PDWUL[ 1RWH WKDW VLQFH ( '7N\2f RR DQG ( Lf ; f E E\ &DXFK\6FKZDU] LQHTXDOLW\ [ ;A7(\nnf KDV ILQLWH H[SHFWDWLRQ 1RZ XVLQJ WKH LQGHSHQGHQFH RI f7 W.
PAGE 117

\AAN\A 9[A7e\A [fEf £ \f7N\ffA[f9L\f [fEf f 1RZ WDNLQJ H[SHFWDWLRQV RQ ERWK VLGHV ZKLFK H[LVW ILQLWHO\ GXH WR WKH SUHYLRXV REVHUYDWLRQ RQH JHWV f +HQFH IURP f f f DQG f ZH KDYH V W .< ZKLFK LV QQG IRU DOO DQG WKH WZR ULVN PDWULFHV DUH HTXDO LI DQG RQO\ LI V W .< IRU DOO 3 L H IRU DOO

PAGE 118

5HPDUN VXIILFLHQW DQG $ K W.< f7f9 f? YOf7ABLYOf .< nf r (7 < V f? LV DQFLOODU\ %DVX"V 7KHRUHP f FDQ QRW EH DSSOLHG VLQFH WKH VXIILFLHQW VWDWLVWLF LV QRW FRPSHWH %HVW (TXLYDULDQW 3UHGLFWLRQ LQ ,QILQLWH 3RSXODWLRQ ,Q WKLV VHFWLRQ ZH FRQFHQWUDWH RQ WKH HTXLYDULDQW SUHGLFWLRQ RI eE
PAGE 119

IRU DOO \ \ DQG 1RWH WKDW WKH PDWUL[ ORVV /T&! 6f 6 f f DQG WKH TXDGUDWLF ORVV /A& 6f &fA46 &f DUH LQYDULDQW 1RZ ZH KDYH WKH IROORZLQJ OHPPD VLPLODU WR /HPPD ZKLFK FKDUDFWHUL]HV WKH FODVV RI HTXLYDULDQW SUHGLFWRUV RI eE
PAGE 120

,OO 6LQFH M < ;"f HJL
PAGE 121

WUDQV RUPDW L RQV DQG WKH PDWUL[ ORVV /T WKH EHVW HTXLYDULDQW SUHGLFWRU RI &2"r
PAGE 122

7R DYRLG WKH UHSHWLWLRQ RI WKH DUJXPHQWV D GHWDLOHG SURRI RI WKLV WKHRUHP LV RPLWWHG 3URFHHGLQJ H[DFWO\ DV LQ 7KHRUHP DQG 7KHRUHP ZH FDQ SURYH WKLV WKHRUHP

PAGE 123

&+$37(5 )285 $6<03727,& 237,0$/,7< 2) +,(5$5&+,&$/ %$<(6 35(',&7256 )25 0($16 ,QWURGXFWLRQ :H KDYH LQWURGXFHG LQ WKH SUHYLRXV FKDSWHUV VHYHUDO +% SUHGLFWRUV ERWK LQ WKH FRQWH[W RI ILQLWH DQG LQILQLWH SRSXODWLRQ VDPSOLQJ XQGHU JHQHUDO OLQHDU PRGHOV 7KHUH ZH KDYH XVHG XQLIRUP SULRU RYHU WKH DSSURSULDWH (XFOLGHDQ VSDFHf IRU WKH IL[HG HIIHFWV DQG LQYHUVH JDPPD SRVVLEO\ LPSURSHUf IRU WKH YDULDQFH FRPSRQHQWV WR SUHGLFW 7 A Pf WKH YHFWRU RI ILQLWH SRSXODWLRQ PHDQV $ QDWXUDO TXHVWLRQ WR DVN LV WKDW LI LQGHHG WKHUH LV D fWUXHf RU fHOLFLWHGnf SULRU ZKHWKHU WKH +% SUHGLFWRU RI LV LQ VRPH VHQVH FORVH WR WKH fWUXHf %D\HV SUHGLFWRU ZKLFK ZH VKDOO UHIHU WR DV VXEMHFWLYH %D\HV SUHGLFWRU 6XFK D FRPSDULVRQ FDQ EH PDGH FRQYHQLHQWO\ LQ WHUPV RI %D\HV ULVNV RI WKHVH SUHGLFWRUV FRPSXWHG XQGHU WKH HOLFLWHG SULRU )ROORZLQJ 5REELQV f ZH VKDOO FDOO D SUHGLFWRU RI fDV\PSWRWLFDOO\ RSWLPDOf $f LI WKH GLIIHUHQFH LQ WKH %D\HV ULVNV RI WKH SUHGLFWRU DQG WKH VXEMHFWLYH %D\HV SUHGLFWRU FRQYHUJHV WR ]HUR DV P f§ 7KH $4 SURSHUW\ RI FHUWDLQ (% SUHGLFWRUV DULVLQJ QDWXUDOO\ LQ WKH FRQWH[W

PAGE 124

RI ILQLWH SRSXODWLRQ VDPSOLQJ ZDV SURYHG LQ *KRVK DQG 0HHGHQ f *KRVK DQG /DKLUL Df DQG *KRVK /DKLUL DQG 7LZDUL f ,Q WKLV FKDSWHU ZH SURYH WKH $2 SURSHUW\ RI VHYHUDO +% SUHGLFWRUV XQGHU DYHUDJH VTXDUHG HUURU ORVV 7R RXU NQRZOHGJH VXFK DQ DWWHPSW LV WKH ILUVW RI LWV NLQG ,Q 6HFWLRQ ZH VWDUW ZLWK WKH QRUPDO OLQHDU PRGHO f ZKHQ $ WKH YHFWRU RI WKH UDWLRV RI YDULDQFH FRPSRQHQWV LV NQRZQ :H VSHFLI\ RXU ORVV IXQFWLRQ WKH HOLFLWHG SULRU GLVWULEXWLRQ DQG GHULYH WKH VXEMHFWLYH %D\HV SUHGLFWRU RI ,Q 6XEVHFWLRQ ZH GHULYH D JHQHUDO H[SUHVVLRQ IRU WKH GLIIHUHQFH RI %D\HV ULVNV RI RXU VXEMHFWLYH %D\HV SUHGLFWRU DQG DQ\ SUHGLFWRU RI ,Q SDUWLFXODU ZH FRQVLGHU WKH VDPSOH PHDQ YHFWRU DQG WKH +% SUHGLFWRU RI ZKLFK FDQ EH GHULYHG IURP 6HFWLRQ ,Q 6XEVHFWLRQ ZH FRQVLGHU WKH UDQGRP UHJUHVVLRQ FRHIILFLHQWV PRGHO RI 'HPSVWHU HW DO f DQG LQ 6XEVHFWLRQ WKH QHVWHG HUURU UHJUHVVLRQ PRGHO RI %DWWHVH HW DO f DV VSHFLDO FDVHV RI WKH JHQHUDO PRGHO RI 6XEVHFWLRQ :H KDYH VKRZQ WKDW WKH +% SUHGLFWRU LV DV\PSWRWLFDOO\ RSWLPDO ZKHUHDV WKH WUDGLWLRQDO VDPSOH PHDQ YHFWRU LV QRQRSWLPDO ,Q 6HFWLRQ ZH FRQVLGHU WKH )D\+HUULRW PRGHO RI 6HFWLRQ ZLWK DOO WKH VDPSOLQJ YDULDQFHV 5 VD\f NQRZQ :H DVVXPH IRU WKH VDPSOH VL]HV QM 9 AQ A DUH DOO HTXDO :H ILUVW VKRZ WKH $2 SURSHUW\ RI WKH +% SUHGLFWRU

PAGE 125

RI P[Of ZKHUH LV DV JLYHQ LQ 6HFWLRQ 7KH UHVXOW LV WKHQ XVHG WR SURYH WKH $ SURSHUW\ RI WKH +% SUHGLFWRU RI 7KH +% SUHGLFWRU RI LV WKH VKULQNDJH HVWLPDWRU REWDLQHG E\ VKULQNLQJ WKH PD[LPXP OLNHOLKRRG HVWLPDWRU RI WRZDUGV D UHJUHVVLRQ VXUIDFH 7KH PRGHO ZH ZLOO FRQVLGHU LQ 6HFWLRQ LV D VOLJKW JHQHUDOL]DWLRQ RI WKH RQHV JLYHQ LQ 0RUULV f DQG *KRVK f DQG LQFOXGHV DOVR DV VSHFLDO FDVHV WKH RQHV FRQVLGHUHG LQ 6WUDZGHUPDQ f DQG )DLWK f 7KH DVVXPSWLRQ RI NQRZQ ILUVW VWDJH YDULDQFH FRPSRQHQW RI 6HFWLRQ LV GLVSHQVHG ZLWK LQ 6HFWLRQ DQG D SULRU GLVWULEXWLRQ SURSHU RU LPSURSHUf LV DVVLJQHG WR WKH ILUVW VWDJH YDULDQFH FRPSRQHQW 7KLV VLWXDWLRQ LV QRZ D VSHFLDO FDVH RI WKH QHVWHG HUURU UHJUHVVLRQ PRGHO ZKHUH WKH FRYDULDWH YHFWRUV DVVRFLDWHG ZLWK HDFK XQLW ZLWKLQ D VWUDWXP DUH WKH VDPH ,Q 6WURXG f VHH DOVR *KRVK DQG /DKLUL LQ SUHVVf RQH FDQ ILQG VXFK DQ H[DPSOH 2QFH DJDLQ WKH $2 SURSHUW\ RI +% SUHGLFWRU RI LV SURYHG 7KH UHVXOW LV WKHQ XVHG WR SURYH WKH $2 SURSHUW\ RI WKH +% SUHGLFWRU RI ,Q WKH UHPDLQGHU RI WKLV VHFWLRQ ZH ZLOO LQWURGXFH D IHZ QRWDWLRQV )RU D VHTXHQFH RI D[E PDWULFHV )Pf ILMQfff` DQG IRU DQ D[E PDWUL[ ) A Mff f ZH LP )Pf Pf§ARRa Y H[LVWV DQG ZULWH LP )Pf Pf§r!RRa Y )r LI LP I Pf Pf§!RR Y I IRU L fff A r M f f f E )RU DQ D[D

PAGE 126

PDWUL[ 7 ZH GHQRWH LWV ODUJHVW VPDOOHVWf HLJHQ YDOXH E\ FK/7f FKJ7ff 0RGHO /RVV 3ULRU DQG 3UHGLFWRUV :H VWDUW ZLWK WKH OLQHDU PRGHO f QDPHO\ < < f f [f? ]f [9a f f f ZKHUH DOO WKH TXDQWLWLHV DSSHDULQJ LQ f DUH WKH VDPH DV LQ f :H ZDQW WR VWXG\ $2 SURSHUW\ RI SUHGLFWRUV RI WKH ILQLWH SRSXODWLRQ PHDQ YHFWRU /HW $ Q L O9 OL7 f L OQLf DQG & BQ f ZKH L On 1L QLn Q M DQG 1 M WKH EHIRUH 7KHQ $< A A &< A 6LPLODUO\ ZULWH WKH VDPSOH PHDQ YHFWRU < < 9 L a Vf 9fOVfff fPVf ?7
PAGE 127

ZKLFK < a 1A;E4 UA(A ZKHUH !T S[Of 7T !f DQG ( g ='$Tf=A DUH NQRZQ 1RZ fWKH VXEMHFWLYH %D\HV SUHGLFWRU RI XQGHU WKH ORVV f LV JLYHQ E\ H6%[fE Lc\f ;fEf f ZKHUH ( L M DUH SDUWLWLRQHG PDWULFHV RI ( )RU NQRZQ $ WKH +% SUHGLFWRU RI ZLWK LQGHSHQGHQW XQLIRUP53f SULRU IRU % DQG JDPPDAD4 ARf 3UARU ARU 5 AQG WKH ORVV f IROORZV IURP 6HFWLRQ DQG LV JLYHQ E\ A%)
PAGE 128

*HQHUDO ([SUHVVLRQV IRU %D\HV 5LVNV 'LIIHUHQFH :H ZLOO QRZ GHULYH WKH GLIIHUHQFH LQ WKH %D\HV ULVNV RI DQ DUELWUDU\ SUHGLFWRU VDPA@ VWDQGDUG L DQ FDOFXODWLRQV JLYH 4 rR! ef UQ P X :P rf H 6% f )RU P\ 6% f
PAGE 129

8VLQJ WKLV IURP f DQG f ZH REWDLQ 4P n2f EIf U 4 P n2f 6 6% B B U4 P WU 4P&[ f f f 9}ff ; f 6L6LL[ff9 f )RU < Vf f ZH KDYH IURP f U4P7f Vff U4P7f a6Ef WU P PL (W< Vf i6E2A Af(nRaVf a 6%A9 A f WU U P 4PA/ $ mO6Of6OOW $ 7 WU P 4P/; f WYOf r1f?O OB7U Y f WYOf fYf? $; F[ fEL"RW; } : 7n $; &; WU U P 4P/O A f rrr &(fK(fHOO / $ &(A(7Lf rrr 7 f :H ZLOO XVH f DQG f UHSHDWHGO\ LQ WKH IROORZLQJ VXEVHFWLRQV 5DQGRP 5HJUHVVLRQ &RHIILFLHQWV 0RGHO :H ZLOO H[DPLQH WKH EHKDYLRU RI WKH ULVN GLIIHUHQFH JLYHQ E\ f DQG f LQ WKLV VSHFLDO FDVH )RU VLPSOLFLW\ D UDQGRP UHJUHVVLRQ FRHIILFLHQWV PRGHO DV

PAGE 130

DSSHDUHG LQ 3UDVDG DQG 5DR f ZLOO EH FRQVLGHUHG 7KH PRGHO FDQ EH GHVFULEHG DV IROORZV Lf &RQGLWLRQDO RQ 5 U % E DQG % M f§ } P < a 1EL;L U $f M E L 1L" L P LQGHSHQGHQWO\ LLf FRQGLWLRQDO RQ 5 U DQG % E % a 1E U$f P LQGHSHQGHQWO\ LLLf% LV XQLIRUP RR RRf DQG 5 a JDPPDA4 LQGHSHQGHQW \ :H FDQ ZULWH Lf DQG LLf DV \X E[ Y [ f L f f A f f m L M LMf M 1 L Pf ZKHUH H DQG Y DUH PXWXDOO\ LQGHSHQGHQW ZLWK Y 1 U$R! DQG f 1 +H P ; ;L f [L1f7} A f f f P A = H[ nS L L La a a7 e1f§ f $f P DQG ( B P W 17 $R f :H ZLOO VKRZ WKDW XQGHU DSSURSULDWH FRQGLWLRQV WKH ULVN GLIIHUHQFH JLYHQ E\ f JRHV WR ]HUR ZKHUHDV WKH ULVN GLIIHUHQFH JLYHQ E\ f GRHV QRW JR WR ]HUR DV P RR 7R WKLV HQG ZH ZLOO SURYH WZR WKHRUHPV 7KH ILUVW WKHRUHP ZLOO SURYH WKH $* SURSHUW\ RI WKH +% SUHGLFWRU ZKHUHDV WKH VHFRQG WKHRUHP SURYHV WKH QRQRSWLPDOLW\ RI WKH VDPSOH PHDQ YHFWRU 7R SURYH 7K QHHG WKH IROORZLQJ FRQGLWLRQV $VVXPH

PAGE 131

P -On6RFK / P f ) VD\f r f DQG IRU WZR SRVLWLYH QXPEHUV SA DQG Z L WK 3L 3 3L ;M f S IRU M 1 P f %HIRUH JHWWLQJ LQWR WKH WKHRUHPV ZH ZLOO PDNH D UHPDUN RQ WKH FRQGLWLRQ f ZKLFK SHUWDLQV WR DOO WKH WKHRUHPV LQ WKLV FKDSWHU 5HPDUN 1RWH WKDW IRU DYHUDJH VTXDUHG HUURU ORVV 4P ,P DQG WKH FRQGLWLRQ f WULYLDOO\ KROGV 7KHRUHP 1 [E ZKH 6XSSRVH XQGHU WKH SULRU 7T < E4 LV D VFDODU $VVXPH WKH FRQGLWLRQV f DQG f KROG 7KHQ IRU WKH UDQGRP UHJUHVVLRQ FRHIILFLHQWV PRGHO WKH +% SUHGLFWRU A LQ f RI LV DV\PSWRWLFDOO\ RSWLPDO XQGHU WKH ORVV f 7R SURYH WKH WKHRUHP ZH QHHG WKH IROORZLQJ OHPPD :H ZLOO XVH WKLV OHPPD UHSHDWHGO\ LQ WKLV FKDSWHU $OVR LQ RXU DSSOLFDWLRQV WKH PDWUL[ 3 RI WKH OHPPD KDSSHQV WR EH Q Q G /HPPD /HW 3S[Sf EH D V\PPHWULF PDWUL[ DQG 8T[Sf EH DQ DUELWUDU\ PDWUL[ 7KHQ FKV3fWU\7Xf WUS87\f FK/ 3f WU\7Xf

PAGE 132

3URRI RI /HPPD 8VH WKH VSHFWUDO GHFRPSRVLWLRQ S S 3 ( ZKHUH Uc DUH WKH HLJHQ YDOXHV RI 3 DQG e DUH L O Oa a WKH FRUUHVSRQGLQJ RUWKRQRUPDO HLJHQ YHFWRUV 7KHQ WU Z&OLLL'X 7 RQH 7KH OHPPD IROORZV 3URRI RI 7KHRUHP ,Q /HPPD WDNLQJ 3 4P DQG 8 ""A A A f LW IROORZV IURP f WKDW A f

PAGE 133

f 7 f 7? 7 3DUWLWLRQLQJ [A A[ ;M f L ;M f [L ff} ; f7 LQ f mL f ;LQc 7 f [ f c\M M L P ZH FDQ ZULWH (MA L O,QL $[>nf[Snf7f DQG ( $ 6 [>![3!7 L aL L 'HILQH Z Q Q ( [" Z LfI L[L Xf@ ZKHUH [ L Xf 1L QLf 1 L ( [ Q f DQG W U f LLL[ff[f 6,[ff F 7 _I0Xf : Lf P R  ; [ cWn !Xf f Of7fLYOf < n f n B 9 ‘/< HQL P f7W BL f Of7[ f  6L OLQL nU ; mcAQ mL e>}^ L O/ f7 f f7 fDf7 f YLOf7 f [O -[O M [OO<[O 6L rL rL OrL 6L fI $ P $ e ZL f )URP f f ZH KDYH

PAGE 134

UAFK/4PfP 2 -r"-YUf f P R 1RZ E\ f P ( [A [ r Y n L O LXf LQILQLW\ DV P JRHV WR LQILQLW\ f DQG P ( Z JRHV WR 7KHQ E\ f WKH UKV RI f JRHV WR ]HUR DV P RR DQG KHQFH IURP f L PO Pf§RR >UJf7R eEIf U4Pff 6Ef@ r f %XW +4PLn2r A%If U4P >rR! VEf A IRU DO P DQG KHQFH LP P f§ RR_ >UJPnR AEIf U4Prf a6Ef@ r f &RPELQLQJ f DQG f RQH JHWV WKH WKHRUHP 1RZ ZH ZLOO FRQVLGHU WKH VDPSOH PHDQ YHFWRU 7R SURYH LWV DV\PSWRWLF QRQRSWLPDOLW\ ZH QHHG WKH IROORZLQJ FRQGLWLRQV LQ DGGLWLRQ WR f 6XS Q VD\f RR f L!L 1L QL IRU L f DQG OMMQ FK4 4Pf X f Pf§ARR A a 1RZ ZH VWDWH DQG SURYH WKH IROORZLQJ WKHRUHP

PAGE 135

7KHRUHP 6XSSRVH XQGHU WKH SULRU WWT < a 1A;E4 $VVXPH WKDW WKH FRQGLWLRQV JLYHQ E\ f DQG f f DUH WUXH 7KHQ IRU WKH UDQGRP UHJUHVVLRQ FRHIILFLHQWV PRGHO WKH VDPSOH PHDQ YHFWRU LV DV\PSWRWLFDOO\ QRQRSWLPDO SUHGLFWRU RI 3URRI f§ GP f L P G P f§ RI 7KHRUHP /HW 6LQFH GP IRU DOO P UPIf a V!f 66%f LW LV HQRXJK WR SURYH WKDW P )URP f XVLQJ /HPPD LW IROORZV WKDW FKV4Pf UOWUW f 1RZ f ZKHUH P $ VR &(A( AA 7KHQ

PAGE 136

$ &(A(f"O / $ VVLVLLfnU P RI B [ LpOIL/QL $ [LXfB[LVf f 9 ZLf[LXf@ DQG KHQFH $ A L f6QW $ &"L"L`f 7 L L / $ $ [ f ; [ [f Y OXf LVfn & ZL![LXf@ eL>Q, $OZL[LXf / $Q [ f ?; f ? I $Q Xf LVf O[L6f@ P R ( I" L O $OZL[L Xf ZLO[LVff A& : r ;f[" ; n OVf P ( L O Q $OZ L[LXf Z LO[LVff P Q ( I" L O QL ( [LM [ M OY Q QL( ;LMf $ M L Z I[ ? f§ Z f ; ? @ L? LXf LVf I"$OZL[L Xf Z ,OVLVff f )URP f f GP FKJ4PfP B ( I L L U$f OZL[LXf Z ,O[LVff ER[ L Xf [ L Vf L Vff FKJ4Pf. Lfr1 L O UR$Rf OZ L[LXf Z ,O[LVff ER[L Xf ; L Vff f

PAGE 137

/HW DQG OLP DQG [ [ LI Z0[ a LXf f ? LL 2f OLP DQG ZLAf[L Vf 7KHQ BZ $A[" O 2 L ( LHV E$[ E2$2[L $U.SLf ;$ .SOf 6 [LVf E$.SAf e [ Vf E\ f f Of6 DQG f G = [ Vf GSP f L O A n E\ f ZKHUH )URP f DQG f ZH KDYH FKJ4PfP A. f AGSAP DQG E\ f

PAGE 138

LP GP Z. f AGS" SURYLGHG EQ A Pf§ RR X 7KLV FRPSOHWHV WKH SURRI RI 7KHRUHP 5HPDUN 1RWH WKDW IRU DYHUDJH VTXDUHG HUURU ORVV 4P ,P! DQG f DQG f DUH VDWLVILHG 7KHQ IURP WKH SUHFHGLQJ WZR WKHRUHPV LW IROORZV WKDW ZKLOH WKH VDPSOH PHDQ YHFWRU LV DV\PSWRWLFDOO\ QRQRSWLPDO WKH +% SUHGLFWRU LV DV\PSWRWLFDOO\ RSWLPDO 1HVWHG (UURU 5HJUHVVLRQ 0RGHO ,Q WKLV VXEVHFWLRQ ZH ZLOO H[DPLQH WKH DV\PSWRWLF EHKDYLRU RI WKH ULVN GLIIHUHQFHV JLYHQ LQ f DQG f XQGHU WKH QHVWHG HUURU UHJUHVVLRQ PRGHO 7KH FRUUHVSRQGLQJ OLQHDU PRGHO LV JLYHQ E\ WKH HTXDWLRQ f +HUH r ,1 $4 $4 '$4f $,P =f ,Q f§ OMX DQG e p LY AQAA1 $OVR ; A L O L O ,1LQL L On 1L X 7 f FRO FRO [>f DQG ; FRO FRO OLP OMQA a a L P QAOM :H ZLOO SURYH KHUH WKDW XQGHU VXLWDEOH FRQGLWLRQV JLYHQ EHORZ WKH ULVN GLIIHUHQFH EHWZHHQ WKH +% DQG WKH VXEMHFWLYH %D\HV SUHGLFWRU JRHV WR ]HUR DV P f§ RR ZKHUHDV WKH SRVLWLYH ULVN GLIIHUHQFH EHWZHHQ WKH VDPSOH PHDQ YHFWRU DQG WKH VXEMHFWLYH %D\HV SUHGLFWRU UHPDLQV ERXQGHG DZD\ IURP ]HUR 7KHVH WZR UHVXOWV ZLOO EH SURYLGHG LQ WKH IRUP RI WKHRUHPV 7KH ILUVW WKHRUHP LQ WKLV VXEVHFWLRQ SURYHV WKH $ SURSHUW\ RI WKH +% SUHGLFWRU ZKHUHDV WKH VHFRQG

PAGE 139

WKHRUHP SURYHV D QHJDWLYH UHVXOW DERXW WKH VDPSOH PHDQ YHFWRU ZKLFK LV GHVLJQ XQELDVHG IRU WKH ILQLWH SRSXODWLRQ PHDQ YHFWRU XQGHU VLPSOH UDQGRP VDPSOLQJ 7R SURYH 7KHRUHP ZH QHHG WR DVVXPH WKDW P [ L U 7 [ L VfL Vf LV SG 0RUHRYHU RR Pf DQG f f f ZKHUH [ L V L f§ A f P f 5HPDUN 1RWH WKDW LW WRU VRPH SG PDWUL[ 4 -7 [ 4 WKHQ FRQGLWLRQV f DQG f KROG $OVR WR SURYH 7KHRUHP EHORZ ZH QHHG DQ LQHTXDOLW\ LQYROYLQJ WKH HLJHQ YDOXHV RI PDWULFHV 7R WKLV HQG ZH VWDWH DQG SURYH WKH IROORZLQJ OHPPD /HPPD )RU D V\PPHWULF QQG PDWUL[ 3S[Sf DQG D V\PPHWULF PDWUL[ 7S[Sf

PAGE 140

Lf FK/,f FK/3 7f LLf FKJ7f FKV3 7f 3URRI RI /HPPD )URP S RI 5DR f ZH KDYH L P? DA7D DA3 7f D B[ FKL 7f VXS a UU" a VXS APf§f§ FKL 3 7f f DA D D D A D D 6LPLODUO\ LLf FDQ EH SURYHG 7KHRUHP 6XSSRVH WKH SULRU UQ LV JLYHQ E\ < 1 [E! DQG WKH FRQGLWLRQV JLYHQ E\ f f DQG f f DUH WUXH 7KHQ a%) Lf WKH +% SUHGLFWRU RI XQGHU WKH QHVWHG HUURU UHJUHVVLRQ PRGHO LV DV\PSWRWLFDOO\ RSWLPDO XQGHU WKH ORVV f 3URRI RI 7KHRUHP /HW D P 4 P 7f EIf U4 P n2f V 6% 1RWH WKDW DP IRU DOO P 7R SURYH WKH WKHRUHP LW LV HQRXJK WR VKRZ LP DP Pf§! 1RZ IURP f WDNLQJ 3 4P DQG 8 Lf7 LfUOYf rr IRRZV WKDW LQ /HPPD LW DP FKI 4Pf WU f ( (aA; A A [ f 6L67[fn, &7 f

PAGE 141

1RZ WDNLQJ 3 ;nn a (7L; A A Of7\BL DQG 8 7 ; f "OeOOr Y ff9 LQ WKH /HPPD ZH KDYH IURP f WKDW Of7ALYOfM [ W U F[ f 6L6LL[f9 ; f L,L;ff77 f +HUH (}( n P U k Q f L O_B 9 Q ( (A ; A A A f FR f f QLQL DTf O ?W f [ DQG ; f ( (BA; A A &2 L P FRO [7f Q A M 1 a QM&QM $f LL n a1Qf f [QLVf ) URP ZH KDYH WU f OYOf9Yf 6L6cLff & 7 P LOIL"LXf QLQL$Rfn6LVff7LXf QLQL$RfiLVff P L V"$6LXf n QLQL$Rfa L L 2f
PAGE 142

)URP WKLV DQG f ZH JHW WKDW $ R. $! A 7 ; ?; ; LAOaVfaVf LV QQG DQG KHQFH IURP /HPPD WDNLQJ 3 6Of96Of $ $Qf A 6n [ [[7 r L LfpfaVf DQG 7 $. f P ;Qf ZH KDYH FKRO $[. $Tf A 9r [ Z V[ [ [ FK 6 P UU\ \r [ Q[ [ L La V [A76LL;Af L H FK / ; f7 .$4fFK/ P m7 L"LaLVfaLVf f )URP f f ZH KDYH DP U B K/4PfO .$RfFK/AaLVf7Vf P F P US P US 9 [ [; [ U [r L fXfnXf f6LVf 74 .$RfFK/PfFKOL rLVf7Vf Le[7Xf;LXf P r,VfrLVf L O 2f P f 7KH UKV RI f JRHV WR ]HUR DV P JRHV WR LQILQLW\ E\ FRQGLWLRQV f DQG f f +HQFH IURP f OLPD P RR P

PAGE 143

%XW DP IRU DOO P &RQVHTXHQWO\ L PBDP DQG Pf§RR WKH SURRI RI 7KHRUHP LV FRPSOHWH 5HPDUN 1RWH WKDW 7KHRUHP UHPDLQV WUXH LI P ZH DVVXPH 7 A A ; f ?; f ? L OaXfaXf Q P nS R PAf DQG A [ L LaA LI ZH P US SODFH f E\ e [I L La nf VfaLVf VfLVf PHf RPAf RU DQG P 7 f E\ <@ [ L ![ ’ P IH f IRU DQ\ H Q D VSHFLDO FDVH RI WKH QHVWHG HUURU UHJUHVVLRQ PRGHO WR EH FRQVLGHUHG LQ 6HFWLRQ f [ M M [L" 1 L P LQ WK LV FDVH FRQGLWLRQV f DQG f DUH LGHQWLFDO $OVR LQ WKLV FDVH P L"LaLVfaLVf P Y 7 / [ f ; f ,6 L O SG VLQFH LW LV DVVXPHG WKDW UDQNO; f 3 1RZ ZH ZLOO EULHIO\ FRQVLGHU WKH LQILQLWH SRSXODWLRQ VHW XS IRU WKLV PRGHO $Q DOWHUQDWLYH SUHGLFWRU IRU M LQ WKH QHVWHG HUURU UHJUHVVLRQ PRGHO DV ZH KDYH VHH Q LQ 6HFWLRQ LV WKH SUHGLFWRU RI ;E \ L VD\f WKH FRQGLWLRQDO PHDQ YHFWRU JLYHQ WKH YDOXHV RI WKH FRYDULDWHV DQG WKH UHDOL]HG YDOXH RI WKH UDQGRP YHFWRU Y RI WKH VWUDWXP HIIHFW ZKHUH ; 1 FRO I[7 A L Q$a Sf DQG [ Q Sf L 17 ( [ )URP 6HFWLRQ RQH FDQ VKRZ WKDW WKH +% M USUHGLFWRU VD\f RI L XQGHU WKH ORVV f LV JLYHQ E\

PAGE 144

FR L P ; }f9;f FRO L P QLQL $f 1RZ DVVXPH WKDW WKH SULRU LU WKDW A L f 091 ZLWK (7R
PAGE 145

ZH ZLOO HVWDEOLVK WKDW WKH VDPSOH PHDQ YHFWRU LV DV\PSWRWLFDOO\ QRQRSWLPDO XQGHU FHUWDLQ FRQGLWLRQV 7KHRUHP 8QGHU WKH SULRU WWT JLYHQ E\ < A 1 ;IHR UAf(A DQG WKH FRQGLWLRQV f f IRU WKH QHVWHG HUURU UHJUHVVLRQ PRGHO WKH VDPSOH PHDQ YHFWRU LV DV\PSWRWLFDOO\ QRQRSWLPDO 3URRI RI 7KHRUHP $V LQ 7KHRUHP LW LV HQRXJK WR VKRZ WKDW IRU G P 4Pnf a V U 4 P rf 6 6% LP G Pf§P[f P )URP f DQG /HPPD ZH KDYH GP UnP F KJ 4Pf $ 4" f Oa a a eOO-An f ,Q WKLV FDVH / $ &(L"LL P B L B L QU $L"ILA$ QLf QL OQL DQG KHQFH O D &6DL6L\6QLIF D FHAHannW $'LDJIQ$R Q[fB IPQP$f QPfB f ZH KDYH 7KHQ IURP AP UAP OFKVPf$ A ILQL$$ QLf L O P B U$$ .f B.B OfFK44Pf J Y ZP E\ f DQG f +HQFH E\ f LP GP :O P f§ [f P Q $ .f . fA DQG WKH WKHRUHP IROORZV

PAGE 146

$V\PSWRWLF 2SWLPDOLW\ ZLWK .QRZQ )LUVW 6WDJH 9DULDQFH &RPSRQHQW )D\+HUULRW 0RGHO &RQVLGHU WKH IROORZLQJ KLHUDUFKLFDO PRGHO &RQGLWLRQDO RQ A } % E DQG $ 6 f
PAGE 147

,r FRQGLWLRQDO RQ % E DQG $ • RR@f f ZLWK D DQG UHDOf ,Q RUGHU WR HQVXUH D SURSHU SRVWHULRU GLVWULEXWLRQ IRU $ ZH VKDOO LPSRVH VRPH UHVWULFWLRQ RQ ODWHU 6RPH DXWKRUV FRQVLGHUHG PRGHOV ZKLFK DUH HLWKHU VSHFLDO FDVHV RU JHQHUDOL]DWLRQV LQ VRPH VHQVH RI WKH PRGHO JLYHQ E\ f§ ZLWK < UHSODFLQJ
PAGE 148

6WUDZGHUPDQ f FRQVLGHUHG fWKH PRGHO ZLWK NQRZQ E DQG D "! ZKLOH )DLWK f FRQVLGHUHG WKH FDVH RI NQRZQ E EXW DVVLJQHG D JHQHUDO FODVV RI SULRUV WR $ LQFOXGLQJ EXW QRW OLPLWHG WR WKH 7\SH,, EHWD GHQVLW\ DV JLYHQ LQ f ZLWK D DQG 7KURXJKRXW ZH DVVXPH ZLWKRXW PHQWLRQ WKDW P LV VR ODUJH WKDW LP Sf ) $OJHEUDLF PDQLSXODWLRQV OHDG WR WKH IROORZLQJ IDFWV IURP ,,,, Lf FRQGLWLRQDO RQ < Vf DQG $ 1+\Vf 9+f f [r[O[ ZKHUH + 9 [ 9 [ P 96n" a[M! "; ; ;7 LLf FRQGLWLRQDO RQ < [ Y n Vf \ LVf $ KDV SGI I V L \ Vf RF H[S c9 P ‘ H[f" Vf ; 9 6f 39B9 f RR f /HW 8 99 I $f DQG X 99 f 7KHQ LW IROORZV IURP f W K DW f $ VR ZULWLQJ VP LW IROORZV IURP

PAGE 149

f WKDW WKH FRQGLWLRQDO SGI RI 8 JLYHQ < m H[S_AXVP_XA A XfDB >8@r f $FFRUGLQJO\ XQGHU WKH TXDGUDWLF ORVV / Df ZLWK / JLYHQ LQ f WKH +% HVWLPDWRU RI LV JLYHQ E\ (f!LLVffH[-Vf f ZKHUH XIDXO"VffGX X .PSf Xf D AX6P ,GX 8PSf"O X Xf D aX6P MGX f DQG ,Q WKH DERYH A GHQRWHV WKH H[SHFWDWLRQ ZUW WKH SGI I D" JLYHQ LQ f :H VKDOO QRZ HYDOXDWH WKH SHUIRUPDQFH RI WKH +% HVWLPDWRU HAM RI IRU P XQGHU WKH ORVV f DQG WKH 1 f,Pf SULRU IRU 7KH VXEMHFWLYH %D\HV

PAGE 150

HVWLPDWRU RI XQGHU WKLV SULRU VD\ 7T DQG fWKH ORVV f LV JLYHQ E\ f ZKHUH X 99 f ,Q WKLV VHW XS IRU DQ HVWLPDWRU EDVHG RQ < RI aVf ZH GHQRWH LWV %D\HV ULVN XQGHU WKH SULRU 7T DQG WKH ORVV / LQ f E\ f ZKHUH WKH H[SHFWDWLRQ (A LQVLGH WKH VTXDUH EUDFNHW LV ZUW WKH FRQGLWLRQDO GLVWULEXWLRQ a 1 9,Pf DQG WKH RXWHU H[SHFWDWLRQ (A LV ZUW WKH SULRU GLVWULEXWLRQ 7T $V LQ f ZH KDYH U4P Wf 2 a U4P n2f 6 VEf PB( 7 P Vf f ZKHUH WKH H[SHFWDWLRQ (r LV ZUW WKH PDUJLQDO

PAGE 151

GLVWULEXWLRQ < 0Vf 1 ;rE 9 f,Pf REWDLQHG E\ DYHUDJLQJ WKH FRQGLWLRQDOf GLVWULEXWLRQ R a 2f JLYHQ LQ ZUW WKH SULRU U :H ILUVW VKRZ WKDW 4 P rf 6 %, 4 P n2f J VEf DV P RR 7R DFKLHYH WKLV ZH QHHG WR SURYH D VHULHV RI OHPPDV 'HQRWH WKH H[SHFWDWLRQ ( 2f JLYHQ LQ f E\ 7PA< r! "f WKDW L V 7P< 4! "f X iPS f" Xf 2I H[SfX6PfGX .PSf" X Xf +X 6 P f :H VXEVFULSWHG LW E\ P WR VKRZ LWV GHSHQGHQFH RQ P 1RWH WKDW 7P\ D "f LV D VWDWLVWLF VLQFH R DQG S DUH NQRZQ SULRU SDUDPHWHUV DV DSSHDUHG LQ ,,, 7KH ILUVW OHPPD RI WKLV VHFWLRQ VKRZV WKDW IRU R 7P< Vf f F/ f 6 f X T ZKHQ KDV WKH SULRU UQ LH PDUJLQDOO\ < 2f 1 [rE 9 fLPf /HPPD 6XSSRVH a Vf 1 [rER! \ 7KHQ 7PVf f f FRQYHUJHV DV DV P 22f WR 8T 3URRI RI /HPPD )RU R IURP f ZH KDYH

PAGE 152

PSf X LPSf"O f )LUVW LQWHJUDWLQJ E\ SDUWV WKH QXPHUDWRU LQ f LW IROORZV IURP f WKDW S f6P LPSfO f $VVXPH ILUVW WKDW P S LV DQ HYHQ LQWHJHU DQG c LV DQ 8VLQJ WKH V\PERO Gf GG f G U f IRU G U VXFFHVVLYH LQWHJUDWLRQ E\ SDUWV OHDGV WR PSf LP Sf LPSfO U LP Sf f $FFRUGLQJO\

PAGE 153

LPSf"O U f 1H[W ZULWLQJ 6 P 0Vf ;rERf7M < P H[M\ 2f ;rEfY DQG QRWLQJ WKH LGHPSRWHQF\ RI ,P 3 ; LW IROORZV WKDW 6 P f§ 8T [PSf 1RZ XVLQJ 0DUNRYfV LQHTXDOLW\ IRU HYHU\ H! f $SSOLFDWLRQ RI %RUH &DQWH L /HPPD QRZ OHDGV WR 6PXP 3 A A D V DV P f§rRR +HQFH ( APSf"O U LP Sf U LV AP ( LUQBSfAB U LP Sf OP 3f V P P Sff U ( U LP Sf P 3f 6PP Sff U f

PAGE 154

&KRRVH P ODUJH WKDW OfP Sf J4 ZKHUH J4 8T +HQFH IRU P LT UKV RI f err JfUVPP U X Y 3ff U 6/ f V f (fXR JRff RR f $ VR HW ePf F AP M PSf "f NP Sf H[S_ LV P +HQFH ORJ ePf LVP ( PSf M ORJ M LP Sf f ORJ LVP L6P LPSf"O ORJ [ G[ sP Sf ORJO 6 P $VP LP Sf @ ORJ_ eP Sf a J2 Sf"OfO AP Sf OrJ_ J6P 6P P Sf ORJ f P S OP Sf f LP Sf ORJ T AP P Sf f 6LQFH 6PP Sf f 8TA LW IROORZV IURP f WKDW LP P Pf§RR ORJPf AXT ORJ 8Tf D V f

PAGE 155

+HQFH IURP f DQG f f RQH JHWV WKDW KV RI f RR D V P FR 2QFH DJDLQ UHFDOOLQJ WKDW 6PP Sf r X A P FR LW IROORZV IURP f DQG f WKDW 7P< "f D V X ,I P S LV RGG DQG LV DQ LQWHJHU XVH WKH P RR LQHTXDOLW\ 8P Sf"O X AX6P OGX ? PSf X S> JX6P _GX f DQG SURFHHG VLPLODUO\ DV EHIRUH WR FRQFOXGH WKDW 7P< Vff ? F/ f 6 f f f§ X P 7KXV IRU LQWHJHU + 7P"@ GHQRWHV WKH LQWHJHU QRW H[FHHGLQJ ( 08O"Vff (!,9ff (OQIOOXOrVff L H 7PA< } >"@f 7PA< A 7 P "@L f

PAGE 156

6LQFH 7 P P Vf >"@f AXR DQG 7 L T? } f 6 f L SM f§‘ X f P Vf >"@Of D X FR VR 7 P
PAGE 157

HL!_\ 2f DV 8 Lf P RR f IURP /HPPD 6LQFH 6PP Sf P RR LW IROORZV IURP f f f DQG f WKDW 7PL< D "f ( Vf 8T D V f P RR IRU D 7KXV IRU D OLP 7PA< Y D "f Pf§!RR 8T DV 1RZ IRU OD XVLQJ WKH 0/5 SURSHUW\ RQFH DJDLQ H!X_" mff (mA8,"6ff (O8OVff f 6LQFH ERWK ( 7 $a  Vff a nP?a Vf L "f DQG (A8_< Vf 7}"fc If IURP f WKDW DV WR X P RR LW I R RZV 7PVf ( 2f ? DV f X P IRU OD 7KXV 7P
PAGE 158

7P
PAGE 159

+RZHYHU RQH KDV WR MXVWLI\ WKH FKDQJH LQ WKH RUGHU RI WKHVH LPLWV :H QRZ WXUQ WR WKH PDLQ WKHRUHP RI WKLV VHFWLRQ 7KHRUHP EHORZ SURYHV WKH $2 SURSHUW\ RI HAM GHILQHG LQ f 7KHRUHP 6XSSRVH
PAGE 160

7KH ODVW HTXDOLW\ IROORZV VLQFH (DA8OVff LV D IXQFWLRQ RI <7 I VfI ,n ([f" (D"8, Vf "fO ([ mf D IXQFWLRQ RI P 3[-
PAGE 161

1RZ ZH ZLOO UHWXUQ WR WKH SUHGLFWLRQ RI ILQLWH SRSXODWLRQ PHDQ YHFWRU )URP LW IROORZV WKDW WKH +% SUHGLFWRU RI LV JLYHQ E\ %I
PAGE 162

VLQFH I" IRU DOO L ` P f )URP WKLV ZH KDYH DV LQ 7KHRUHP WKDW Uf LU H)f UJ r J6%f 4 P RR $V\PSWRWLF 2SWLPDOLW\ ZLWK 8QNQRZQ 9DULDQFH &RPSRQHQWV ,Q WKLV VHFWLRQ ZH GLVSHQVH ZLWK WKH DVVXPSWLRQ RI D NQRZQ YDULDQFH FRPSRQHQW DW WKH ILUVW VWDJH DQG FRQVLGHU D KLHUDUFKLFDO PRGHO VLPLODU WR WKH RQH FRQVLGHUHG LQ 6HFWLRQ ,, &RQGLWLRQDO RQ J % E 5 U DQG $ $ < 1 nc}1 L a1f f P LQGHSHQGHQWO\ FRQGLWLRQDO RQ % E 5 U DQG $ $ 1 [rE $Uf -PM ZKHUH ;rP[Sf LV DVVXPHG WR KDYH UDQN S P ,,, % 5 DQG H $5 DUH PXWXDOO\ LQGHSHQGHQWO\ 3 GLVWULEXWHG ZLWK % a XQLIRUP5 f H KDV SGI IHf RF fDO D LPSfA DQG 5 a JDPPDAD4 WKDW LV 5 KDV SGI KUf RF H[SADJUf U AR ZKH D4 A DQG J4 UHDOf VDWLVILHV Q OfP I J4 ZKHUH WKH VDPSOH VL]H IURP HDFK VWUDWXP 7KXV H KDV Q LV LPSURSHU JDPPD SGI ZKLOH 5 KDV D SURSHU RU DQ LPSURSHU JDPPD SGI ZLWK LWV SDUDPHWHUV VDWLVI\LQJ FHUWDLQ FRQGLWLRQV +HUH DOVR ZH DUH LQWHUHVWHG LQ SURYLQJ WKH $4 SURSHUW\ RI

PAGE 163

‘WKH +% SUHGLFWRU RI :H ZLOO DWWDLQ WKLV E\ SURYLQJ WKH DV\PSWRWLF RSWLPDOLW\ RI WKH +% SUHGLFWRU RI $VVXPH WKDW ZH KDYH D VDPSOH Q$f Q\Vf$3[!W\Vf QUf Q$fB P $([-@ f LLf FRQGLWLRQDO RQ ?I6 V DQG $ $ a Vf LVf 5 *DPPD YD Q$ 7 Q $a V fGP 3[Af\Vf f L QPS L L Lf FRQG LW LRQDO RQ DQG 6 V $ KDV SGI a 2f rVf L$_\Vf Vf RF $LQ$ffA 3A$ D [ D Q$ 7 W BS _-QQSmJf Q $ Vf9aP H;rM\Vff

PAGE 164

/HW 8 $Q $f 7KHQ IURP f LW IROORZV WKDW WKH FRQGLWLRQDO SGI RI 8 JLYHQ AVf DQFr A V LV X LPSDfO XffO X)f LQPSDJf f X ZKHUH ) Q\AJA,P "[rf\ VfD Vf 1RWH WKDW LI D4 ) LV D PXOWLSOH RI D XVXDO ) VWDWLVWLF $OVR IURP f DQG f RQH JHWV WKH +% HVWLPDWRU RI XQGHU WKH ORVV f JLYHQ E\ 6KEg Vf ZKHUH f (f8_<6f Vf I OPSDf OQPSDJf X Xf X)f GX I OPSDfO D A .QP3DJf U X Xf X)f GX f

PAGE 165

:H QRZ H[DPLQH WKH DV\PSWRWLF DV P RRf EHKDYLRU RI iIM% DV DQ HVWLPDWRU RI XQGHU WKH VXEMHFWLYH SULRU 7T ZKLFK VSHFLILHV WKH YDOXH UQ IRU U DQG WKH GLVWULEXWLRQ a 1 URf Pf f DQG XQGHU WKH ORVV / Df JLYHQ LQ f 7KHQ PDUJLQDOO\ [ DQG 6 DUH PXWXDOO\ LQGHSHQGHQW ZLWK
PAGE 166

2f 6 Of $ W X)f WOWf GX 2 X WL X)f WAWf GX 2 ) U W X) X)ff OO X)ff Wa I X)ff GX 2 W L X)O X)ff OO X)ff Wa I X)ff GX 2 )O)f ) W Y Yf )O)f U WL W4O Y Yf ] GY GY 2 2 f ,QWHJUDWLRQ E\ SDUWV JLYHV )O)f Y )O)f A 9 f &RPELQLQJ f DQG f RQH OUAO WLW f)f W fB) $ )f )O)f WL Y W4O Yf = GY WWLf f

PAGE 167

1RWH WKDW VLQFH 6 a U2A;AQOfPf AAQ OfPf Da U U P RR $JDLQ XVLQJ WKH LGHPSRWHQF\ RI ,P 3< LW [r B W IROORZV fWKDW Q< A 2f P ([f" Vf Vf rRf7P "[M Vf a ;AE4f a U X ;Pf3 W +HQFH Q< f 2f P (;f"f DV U X P 7KXV WW f)f ^P S fQ fP J4 f`I n Q fB [ Q OfU UROXR P RR +HQFH IRU SURYLQJ /HPPD LW VXIILFHV WR VKRZ WKDW WKH VHFRQG WHUP LQ WKH UKV RI f FRQYHUJHV WR ]HUR DV DV P RR XQGHU WKH GLVWULEXWLRQ RI DQG 6 JLYHQ LQ WKH OHPPD :LWK WKLV HQG DVVXPH ILUVW WKDW WA LV D SRVLWLYH LQWHJHU DQG XVH VXFFHVVLYH LQWHJUDWLRQ E\ SDUWV WR REWDLQ )O)f W Y WO Yf GY )f WLWLf L WMMO W ff W -f W[ LfUWfUW L Wf f ZKHUH [f ‘ [fM [ [ f [ M } f IRU M [ DQG If LV WKH XVXDO JDPPD IXQFWLRQ +HQFH IURP f

PAGE 168

W 1 ))f W Ir W UMf§ Wf I W f )f ) n 9 WO Yf = GY r n WO WOfM) M WWf WMf )f WOWB [ ) A W OfWA f IrfIAL Af W a W L L WOffWWO f( n ) )f M W M [ ) rf r AfrL LfUAf)r rf W 29(" ^r M R r‘ f$` I )f WOWf [ ) r ‘rffr f f Urf)rO rAn f 1RWH WKDW WKH ILUVW WHUP LQ WKH UKV RI f FRQYHUJHV DV WR (MB4Q f AQ fXRfM aArrB48 B a XRfB DV P f§ $OVR XVLQJ 6WLUOLQJfV ERXQGV IRU JDPPD IXQFWLRQV VHFRQG WHUP LQ WKH UKV RI f WWO WAOf L )f ) UfW f

PAGE 169

WWOf )f ) WWf W f 5HFDO WKDW WA ) r Q fBA8TA DV P f§ RR :ULWH KPf IRU WKH ORJDULWKP RI WKH UKV RI f 7KHQ P S f W LQ fP J4f DQG LP P RR P AK Pf LQ fRJQ f Q ORJ Q f XVLQJ [fQ Q[ IRU [ +HQFH KPf f§rf RR DV DV P f§ RR DQG LW IROORZV IURP f DQG f

PAGE 170

WKDW WKH VHFRQG WHUP LQ WKH UKV RI f FRQYHUJHV WR ]HUR DV DV P f§ RR ,I WA LV QRW DQ LQWHJHU ZULWH W KAB Xf X X)f WLWf W L I WQO WWf I X X)f GX WnOWO 7KHQ IRU WA W A KAB Xf RF AXO X)fM 8VLQJ WKLV 0/5 SURSHUW\ DQG ZULWLQJ :PA< [ 6 A LQ X H 8_9 Vf Vf L LW IROORZV WKDW (eA8_<6?! 6f L V W L Q W ZKHUH (A GHQRWHV WKH H[SHFWDWLRQ ZUW KA Xf +HQFH LI >WA@ GHQRWHV WKH LQWHJHU SDUW RI W >W[@ L >W ( Q8,< [ 6f (r 6 (r>WO@LX,WA@ DV WR X P RR LW IROORZV WKDW (A 8O"Vf! X_< 6f FRQYHUJH 2f Vf D X P RR 7KH SURRI RI /HPPD LV FRPSOHWH /HPPD &RQVLGHU WKH VHW XS RI /HPPD ZLWK D 7KHQ :PA
PAGE 171

:P< XWO Â’ffn X)fnWWfGX XW XfDO X)f WWfGX f &RQVLGHU ILUVW WKH FDVH ZKHQ D ,QWHJUDWLRQ E\ SDUWV XWO Xfr}O X)fWWfGX a WJ WA f [ X)f r GX _D OfWJ W [ fDWO XfO X)fWWfGX WW W f`) X XfD ; ; X)f A A AGX ^AA A 9O ffm X)fnWW!GX

PAGE 172

^D OfWJ W f`IB XWO XfmO X)fWWf f &RPELQLQJ f DQG f RQH :PVff 6 ff WW G-I ^D OfW OA) [ (D>8O 8)f 8f ,< Vf r 6M f ZKHUH UHFDOO fWKDW (D LV WKH H[SHFWDWLRQ ZUW JLYHQ LQ f +HQFH IURP f : P Vf 6 Df WW ) D 8T DV P RR $OVR ZULWLQJ I DAX < I XfIWWXf D D Vf Vf DV I4Xf IRU D D W OD f§D )f XfL Q X +HQFH XO Xf LV W LQ X IRU D f V L QHH X)f WWf

PAGE 173

7 XW ; XfO X)f WWfGX X)fnWWfGX WR WQWRf U X Xf X)f GX 0 8}Vf 6f@ n r XRfB P f§r RR f )URP f DQG f f f A Vff V`@ F/ f 6 f XA X P RR f ,W IROORZV IURP f DQG f WKDW IRU D :P
PAGE 174

8QGHU WKH VXEMHFWLYH SULRU WWT ZKLFK VSHFLILHV U 7T $ $T DQG Q[r!T URARfBaPf WKH VXE-HFLYH %D\HV HVWLPDWRU RI LV JLYHQ E\ X f < 2f XTAr%T f 7KH IROORZLQJ WKHRUHP SURYHV WKH $2 SURSHUW\ RI f 7KHRUHP $VVXPH WKH FRQGLWLRQ f KROGV 7KHQ IRU WKH SULRU 7T JLYHQ DW WKH EHJLQQLQJ RI WKH VHFWLRQ DQG IRU D 7T-WWT i+Jf HJJf DV P RR 3URRI RI 7KHRUHP /HW (r GHQRWH WKH H[SHFWDWLRQ WDNHQ ZUW WKH MRLQW GLVWULEXWLRQ RI
PAGE 175

1RWH WKDW &RYWW> ,P 3[rf
PAGE 176

VL QHH WU &f WU H[UDQN;rf 3 $V LQ WKH SURRI RI OUr 7KHRUHP ZH FDQ SURYH P (D 8 < A 6f 2f [B XRf r 7 Vf [ P f [A Vf P RR +HQFH IURP f f DQG f ZH KDYH P LP UQ U r‘ FMF\PY f a+% HPf U 4 P 7f 6 6% 7 LP ( f§ Pf§rRR (f8_<6f 6f Xf"7 P H[-< A rF 2f +HQFH UQ [4 +%f rr4 r HJ%f 4 P RR P :H PD\ QRZ DSSO\ WKLV WKHRUHP IRU +% SUHGLFWLRQ RI )URP ,,, LW IROORZV WKDW WKH +% SUHGLFWRU RI LV JLYHQ E\ RVf 'LDJ I [ OIPf< %)O IV9 rfPQVf 'LDJI LPf iKEA Vf f 6f f ZKHUH I Q1M L P )URP LW IROORZV WKDW WKH VXEMHFWLYH %D\HV SUHGLFWRU RI XQGHU WKH SULRU 7T LV JLYHQ E\ V6EgVff 6f 'LDJ&OIM OIPf< Vf 'LDJIA LPfL6%
PAGE 177

1RWH WKDW A%IgVff 6f 6EAVff 6f 'LDJI WIPfA+%AVff 6f f A6EAVff JLYHV R _VEI
PAGE 178

&+$37(5 ),9( 6,08/7$1(286 %$<(6,$1 (67,0$7,21 2) 60$// $5($ 9$5,$1&(6 ,QWURGXFWLRQ ,Q fWKLV FKDSWHU ZH DUH LQWHUHVWHG LQ VLPXOWDQHRXV +% HVWLPDWLRQ RI YDULDQFHV IURP VHYHUDO VPDOO DUHDV ZKHUH HDFK VPDOO DUHD KDV D ILQLWH QXPEHU RI XQLWV 7KLV FKDSWHU LV GHYHORSHG IROORZLQJ LQ SDUW WKH RXWOLQHV RI WKH SUHFHGLQJ WKUHH FKDSWHUV :H ZLOO GHYHORS WKH +% HVWLPDWRU RI WKH ILQLWH SRSXODWLRQ YDULDQFH YHFWRU DVVXPLQJ DQ XQGHUO\LQJ QRUPDO OLQHDU PRGHO IRU WKH VXSHUSRSXODWLRQ XQGHU D TXDGUDWLF ORVV :H XVH WKH QRWDWLRQV LQWURGXFHG HDUOLHU /HW QQ SPf GHQRWH WKH ILQLWH SRSXODWLRQ YDULDQFH YHFWRU ZKHUH S 1 f L 1 ( 9 M O9 8 < Lf f f P :H ZDQW WR ILQG WKH +% HVWLPDWRU RI S DQG VWXG\ LWV DV\PSWRWLF RSWLPDO SURSHUW\ XQGHU WKH ORVV f ZKHUH 4P P[Pf LV D NQRZQ QQG DQG QRQQXOO PDWUL[ 7KLV ORVV KDV EHHQ XVHG LQ WKH SUHYLRXV FKDSWHU

PAGE 179

%D\HVLDQ HVWLPDWLRQ RI YDULDQFHV IURP VWUDWLILHG VDPSOHV ZHUH FRQVLGHUHG HDUOLHU E\ *KRVK DQG /DKLUL Ef DQG /DKLUL DQG 7LZDUL LQ SUHVVf ZLWKRXW LQFRUSRUDWLQJ DQ\ DX[LOLDU\ LQIRUPDWLRQ +DUWOH\ DQG 5DR f LQWURGXFHG DX[LOLDU\ LQIRUPDWLRQ LQ WKH HVWLPDWLRQ RI YDULDQFHV IRU VWUDWLILHG VDPSOHV RQO\ LQ D YHU\ VSHFLDO VHW XS 7KH SUHVHQW FKDSWHU WUHDWV WKH YDULDQFH HVWLPDWLRQ SUREOHP LQ WKH JHQHUDO IUDPHZRUN RI &KDSWHU 7ZR ,Q 6HFWLRQ ZH KDYH GHYHORSHG XQGHU WKH VHW XS RI 6HFWLRQ LH ZLWK NQRZQ UDWLRV RI YDULDQFH FRPSRQHQWVf WKH +% HVWLPDWRU RI D TXDGUDWLF IRUP :H KDYH XVHG WKLV UHVXOW WR GHULYH H[SOLFLWO\ WKH +% HVWLPDWRU RI 3 7KLV HVWLPDWRU LV FRQVLGHUHG LQ JUHDWHU GHWDLOV LQ 6HFWLRQ LQ WKH VSHFLDO FDVH RI QHVWHG HUURU UHJUHVVLRQ PRGHO f ZLWK [ A f [ f 1 L P $V LQ &KDSWHU )RXU ZH VWXG\ WKH DV\PSWRWLF RSWLPDOLW\ RI WKH SURSRVHG HVWLPDWRU 6LQFH WKLV SUREOHP LV PXFK PR DOJHEUDLFDOO\ LQYROYHG WKDQ WKH RQH ZH FRQVLGHUHG IRU WKH PHDQV LQ &KDSWHU )RXU ZH UHVWULFW RXUVHOYHV WR WKLV VSHFLDO FDVH :KLOH ZH SURYH WKDW WKLV +% HVWLPDWRU LV DV\PSWRWLFDOO\ RSWLPDO XQGHU WKH ORVV f DQG WKH VXEMHFWLYH SULRU 7T RI 6HFWLRQ ZH HVWDEOLVK WKDW WKH XVXDO VDPSOH YDULDQFH YHFWRU WXUQV RXW WR EH QRQRSWLPDO 7R SURYH WKHVH UHVXOWV ZH GR QRW QHHG WKH Q WR EH HTXDO

PAGE 180

6R IDU ZH KDYH DVVXPHG fWKDW WKH UDWLR RI YDULDQFH FRPSRQHQWV LV NQRZQ ,Q 6HFWLRQ IRU WKLV VSHFLDO FDVH DJDLQ ZH GHULYH WKH +% SUHGLFWRU IROORZLQJ WKH KLHUDUFKLFDO %D\HV VHW XS RI 6HFWLRQ ZKHQ ERWK YDULDQFH FRPSRQHQWV DUH XQNQRZQ DQG DUH DVVLJQHG JDPPD SULRUV SURSHU RU LPSURSHUf )ROORZLQJ WKH DUJXPHQWV RI 6HFWLRQ ZH KDYH EHHQ DEOH WR SURYH WKH DV\PSWRWLF RSWLPDOLW\ RI RXU +% SUHGLFWRU XQGHU WKH DGGLWLRQDO DVVXPSWLRQ WKDW DOO WKH QA DUH HTXDO *KRVK DQG /DKLUL Ef DQG /DKLUL DQG 7LZDUL LQ SUHVVf KDYH DOVR SURYHG WKH DV\PSWRWLF RSWLPDOLW\ RI WKHLU (% HVWLPDWRUV XQGHU DYHUDJH VTXDUHG HUURU ORVV ZLWKRXW UHTXLULQJ WKH Q WR EH DOO HTXDO %XW DV SRLQWHG RXW HDUOLHU WKH\ KDYH QRW XVHG DQ\ DX[LOLDU\ LQIRUPDWLRQ %D\HV (VWLPDWLRQ RI D 4XDGUDWLF )RUP ZKHQ 5DWLRV RI 9DULDQFH &RPSRQHQWV .QRZQ :H ZLOO FRQVLGHU WKH VHW XS GHVFULEHG LQ 6HFWLRQ 7 :H DUH LQWHUHVWHG LQ HVWLPDWLQJ D TXDGUDWLF IRUP < )< ZKHUH ) 1US[1USf LV D NQRZQ V\PPHWULF PDWUL[ :ULWLQJ DQG ) 1M f QUSf [ 1I f Q7Af ZH FDQ EUHDN XS
PAGE 181

\WI\
PAGE 182

)URP f ZH ZLOO GHULYH WKH +% HVWLPDWRU RI S 1 f 1L enf ,
PAGE 183

, ,, &RQGLWLRQDO RQ % E DQG 5 1 0Q L ,Qf r P LQGHSHQGHQWO\ FRQGLWLRQDO RQ % E DQG 5 U H a 1[rE $Uf sP ZKHUH ;r FR f LQ$ n LV DVVXPHG WR KDYH IXOO FROXPQ UDQN S P ,,, % DQG 5 DUH PXWXDOO\ LQGHSHQGHQWO\ GLVWULEXWHG ZLWK % a XQLIRUP5Af DQG 5 a JDPPDAD4 ZKHUH DA DQG J4 f VXFK WKDW QU\ J4 S :H FDQ ZULWH DQG ,, DV D OLQHDU PRGHO
PAGE 184

f§ QUM J 4 < 9XLr7 (& 2f§ f 8VLQJ f DQG f ZH KDYH IURP f DIWHU VRPH VLPSOLILFDW LRQV I 1 f L U\ r ZKHUH D Xf [r L P :H ZLOO QRZ VKRZ fWKDW L !! UQ LV DV\PSWRWLFDOO\ RSWLPDO XQGHU WKH ORVV f DQG XQGHU WKH VXEMHFWLYH SULRU 7T ZKLFK VSHFLILHV WKDW
PAGE 185

XLf f ZKHUH GHQRWHV WKH H[SHFWDWLRQ WDNHQ XQGHU WKH SULRU GLVWULEXWLRQ 7T ; ; /HW 3E A% AP%f DQG 6% A6% f AP6Ef r 8VLQJ WKH QRWDWLRQV RI &KDSWHU )RXU WR SURYH WKH $ IRU 3J ZH QHHA 7R VKRZ Ef UTP7} 6Ef f§ DV P r RR :H HVWDEOLVK LW LQ WKH IROORZLQJ WKHRUHP DVVXPLQJ WKH FRQGLWLRQ f 7KHRUHP $VVXPH WKDW WKH FRQGLWLRQ f KROGV 7KHQ IRU WKH SULRU 7T JLYHQ DERYH WKH PRGHO f DQG WKH ORVV f SJ WKH +% SUHGLFWRU RI S LV DV\PSWRWLFDOO\ RSWLPDO 3URRI RI 7KHRUHP 8VLQJ VWDQGDUG %D\HVLDQ FDOFXODWLRQV ZH JHW DP VD\f

PAGE 186

6LQFH DP 2 9 P LW LV HQRXJK WR SURYH %\ /HPPD DP A DQG KHQFH E\ FRQGLWLRQ f 7 LP P f§ P f§RR 6R LW LV HQRXJK WR VKRZ WKDW 1RZ f r Q7 JR S f DQG

PAGE 187

Vf r7tf < L Vf rr"Rf VD\f >[7E Ef@ [7E Ef"c Vf rf 1RWH WKDW XQGHU WWT < L Vf US a [ME DQG E DUH LQGHSHQGHQWO\ GLVWULEXWHG ZLWK < L Vf [ME a 12 UQ7 8 $T[7 P ( D D Oa a DQG E a 1 E4 $ 72 f 7KHQ DIWHU VRPH VLPSOLILFDWLRQV (URW Lf 9!UQf9, i 7 2/f§ rL f DQG 9AWL` R$Rf0 ILH D 7 rL X7; [7 Df§ 7 rL U$f[D Df§ 7 ;M8 f )URP f DQG f 1L A L % A L6% I QLXLWL (URnWL K_ >D
PAGE 188

79 P X f e M 6 DDDD 9r 7? mL D@< 6 DD-f YD O n Dr DQG WKH PDWUL[ 3 US P 7 B!L ?? D DDDf Dff LV V\PPHWULF LGHPSRWHQW a D O a a 7KHUHIRUH D7 ( DDDMf D r D a L 8VLQJ WKLV ZH JHW KM1M Lf 1M f AI1 8M X@ VLQFH ZH ZLOO DVVXPH WKDW QM )URP DOO WKHVH REVHUYDWLRQV LW IROORZV WKDW SL6% D4 7 < fQ7 J4 S f +HQFH f )URP f DQG f ZH KDYH U$f ([ < Q7 J S f f f

PAGE 189

1RZ VLQFH $T $T Qf DQG DA f 5H FDOO W K DW A f§ k fAQ k r Q m, r ; f§ FRO A MM [ f A DQG a L O/ L X a LM OLPna OA ^ (A; A A[ A -[ A A 6LQFH XQGHU 7T a 1A;AAE4 LW IROORZV IURP 5DR VHH 5HVXOW YLLf RQ S f WKDW a URA$QASr 7KHQ LW LV URXWLQH WR FKHFN WKDW LP (LU P f§ RR D a_ +2@ f Q 7 RR DV P &RPELQLQJ f f f IROORZV 7K FRPSOHWHV WKH SURRI RI 7KHRUHP /HW X LV IHO !fffr V VP 7 ZKHUH 2L Lf nLMOLL
PAGE 190

WKHRUHP WKDW XQGHU VRPH FRQGLWLRQV WKLV HVWLPDWRU LV QRW DV\PSWRWLFDOO\ RSWLPDO 7KHRUHP $VVXPH WKH FRQGLWLRQV f f KROG 7KHQ 7RU WKH SULRU WWT JLYHQ DERYH WKH PRGHO f DQG WKH ORVV f SMM WKH WUDGLWLRQDO HVWLPDWRU RI S LV DV\PSWRWLFDOO\ QRQRSWLPDO Xf U4P 7KHQ 3URRI RI 7KHRUHP /HW G P G P P HrR8 eV%f7PX ‘ 6Ef FK J4PfP 3 6Eff E\ /HPPD DQG KHQFH LP GP OR LP P P f§ RR P f§RR P R 9 (WW V f A RY L A6EfA (\ rf ,W LV HQRXJK WR SURYH WKDW ‘$0/n(nRV" "VEf $IWHU VRPH VLPSOLILFDWLRQV SL6% L >L f1 f !" U U QL1L < L Vf r7"Rf UROQLOX, 8QGHU 7T R B V f DQG < [ Vf DUH LQGHSHQGHQWO\ GLVWULEXWHG IRU P 0RUHRYHU ( ‘RWIf Un UR f 7KHQ

PAGE 191

(r 2 AVE@ 1 L QLf1L f@b"f QI1L [ (AM 2 < L Vf VOtRf UROQLOXaL U4I"QL f UAO QM&QM ff QL a f E\ f +HQHH WT. f. fB E\ f LP P P P e HW L O 2 A6E@ U R f. Of R 7KH SURRI RI 7KHRUHP LV FRPSOHWH $V\PSWRWLF 2SWLPDOLW\ LQ 1HVWHG (UURU 5HJUHVVLRQ 0RGHO ZLWK 8QNQRZQ 9DULDQFH &RPSRQHQWV ,Q WKH SUHYLRXV VHFWLRQ ZH DVVXPHG WKH UDWLR RI YDULDQFH FRPSRQHQWV ZDV NQRZQ +HUH ZH ZLOO GLVSHQVH ZLWK WKLV DVVXPSWLRQ DQG XVH WKH KLHUDUFKLFDO PRGHO GHVFULEHG LQ 6HFWLRQ 7KLV LV DQ H[WHQVLRQ RI WKH PRGHO LQ 6HFWLRQ ZKHUH $ $Tf LV DVVXPHG WR EH NQRZQ +RZHYHU XQOLNH LQ 6HFWLRQ ZH ZLOO DVVXPH DOO WKH Q DUH HTXDO WR Q :H ZLOO ILUVW GHULYH WKH +% SUHGLFWRU 3SMJ RI S WKDW WKH +% SUHGLFWRU 3MMMJ RI S A LV JLYHQ E\ 1RWH

PAGE 192

A L+% ( 3L ,< Of ( ( SL $ < f < f f 7R IDFLOLWDWH WKH GHULYDWLRQ RI ( SL $ < &ff f L Q f ZH ZLOO ZULWH GRZQ WKH H[SUHVVLRQV RI WKH ILUVW WZR PRPHQWV RI WKH GLVWULEXWLRQ RI JLYHQ DQG $ ) URP ,,, RI 6HFWLRQ LW HDVLO\ IROORZV WKDW ,9 FRQGLWLRQDO RQ $ $ % DQG 5 DUH PXWXDOO\ LQGHSHQGHQWO\ GLVWULEXWHG ZLWK % a XQLIRUP5Af DQG 5 a JDPPDOD OJ4Dff :ULWH 8 $Q $f DQG X $Q $f )URP ,, DQG ,9 ZH KDYH DV LQ f DQG f ZLWK DSSURSULDWH PRGLILFDWLRQV WKDW ( f_D \>f 2 8f
PAGE 193

$V LQ 6HFWLRQ ZH JHW IURP f f WKDW ( A $ < f 1M f
PAGE 194

VWDWH DQG SURYH WKH IROORZLQJ WKHRUHP 1RWH WKDW D DSSHDULQJ LQ WKH WKHRUHP LV D SULRU SDUDPHWHU 7KHRUHP $VVXPH WKH FRQGLWLRQ f KROGV 7KHQ IRU WKH SULRU 7T JLYHQ DERYH DQG IRU D A+% rV DV\PSWRWLFDOO\ RSWLPDO IRU WKH EDODQFHG QHVWHG HUURU UHJUHVVLRQ PRGHO 3URRI RI 7KHRUHP $V LQ 7KHRUHP OHW (r GHQRWH WKH H[SHFWDWLRQ WDNHQ ZUW WKH MRLQW GLVWULEXWLRQ RI DQG 6 VSHFLILHG E\ WKH SULRU 7T 7KHQ IROORZLQJ WKH SURRI RI 7KHRUHP LW LV HQRXJK WR VKRZ WKDW P P e (rSL+% 3 L 6E@ P RR /HW (DX < Vf f 6f YP\ 2f 6 Df 9P DQG X4 $4 Q $Tf 7KHQ +% 3 VEf 1Of ILQ "LV f}7Ef 9P
PAGE 195

r 1 OffQ
PAGE 196

?U Q XJ9r LW IROORZV IURP D FDOFXODWLRQ LQ WKH SURRI RI 7KHRUHP WKDW OLP Pfr 9 (rWAf Pf§rRR ? f§ ? 9 A 6LQFH 1M IRU DOO L ZH KDYH P A A WR  DA 6 665:Pf L O / BQ U QP J4 D S f U4 M f f 6/ V f 5HFDO WKDW :P A 8T $OVR LW FDQ EH VKRZQ WKDW XQGHU 7T 6PQ ff U DQG 665P Sf f WA8T E\ XVLQJ 0DUNRYfV LQHTXDOLW\ DQG %RUH &DQWH L OHPPD DV LQ 6HFWLRQ )URP WKHVH LW IROORZV WKDW $OVR LW FDQ EH VKRZQ XQLIRUPO\ LQWHJUDEOH (rAD4 6 665:Pf QP J4 D S f P f§ RR 7KHUHIRUH IURP f +HQFH DV f ZH KDYH W L

PAGE 197

:H FDQ DOVR VKRZ DV EHIRUH : P D4 6f 6659 P QP J D S f UOXR@ Dr6 DQG XQLIRUPO\ LQWHJUDEOH )URP WKLV ZH KDYH L P P Pf§RR ,6 A ( (r W L &RPELQLQJ f f ZH JHW LP P P f§!RR P e (rSL+% A L 6%M DQG WKH WKHRUHP IROORZV f f

PAGE 198

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f WR EH WKH %83 Ef WR XQLYHUVDOO\ RU VWRFKDVWLFDOO\

PAGE 199

GRPLQDWH DOO OLQHDU XQELDVHG SUHGLFWRUV DQG Ff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

PAGE 200

WR WDNH LQWR DFFRXQW ERWK WKH ZLWKLQ XQLW DQG EHWZHHQ XQLW YDULDWLRQ 7KH SUHVHQW GLVVHUWDWLRQ GLG QRW DGGUHVV WKH UREXVWQHVV LVVXHn H[FHSW 7RU WKH VWXG\ RI XQLYHUVDO DQG VWRFKDVWLF GRPLQDWLRQ ZKLFK LQ VRPH VHQVH SURYLGHV UREXVWQHVV IRU WKH ERZOVKDSHG ORVV IXQFWLRQV 2QH LPSRUWDQW LVVXH LV WKH SULRU UREXVWQHVV )RU WKH YHFWRU RI IL[HG HIIHFWV ZH KDYH XVHG XQLIRUP SULRU ZKLFK FDQ EH YLHZHG DV D GLIIXVHG PXOWLYDULDWH QRUPDO SULRU 7KH SHUIRUPDQFH RI WKH SURSRVHG +% SUHGLFWRU IRU RWKHU SULRUV ZKRVH WDLOV DUH KHDYLHU WKDQ WKH QRUPDO HJ PXOWLYDULDWH Wf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

PAGE 201

PHDQV RU YDULDQFHV LQ SUHVHQFH RI UHVSRQVH ELDV PHDVXUHPHQW HUURU FDQ DOVR EH H[SORUHG RU

PAGE 202

$33(1',&(6

PAGE 203

$33(1',; $ 3522) 2) 7+(25(0 8QGHU WKH DVVXPSWLRQV R WKH 7KHRUHP WKH MRLQW SGI RI < % 5 DQG $ LV JLYHQ E\ I\ E U $f $ f f 1RZ \ [Ef7"B\ ;Ef E Ef7;7e;E Ef \7T\ $ f ZKHUH E [7(B [fB ;7(< DQG 4 (B ( [[7( [fa ;7(a )URP $Of DQG $f RQH JHWV LQWHJUDWLQJ E RXW WKH MRLQW SGI RI < 5 DQG $ LV JLYHQ E\

PAGE 204

I \ $f RF [WH[ M1US ( JASf U $f 1RZ LQWHJUDWLQJ Z U W U RQH ILQGV fWKH MRLQW SGI RI < DQG $ LV JLYHQ E\ L I\ $f [ ; 7(n; W [ D R ;rL[L \74\ 17L,6L 3f W LII LI L O $f 6LPLODUO\ VWDUWLQJ ZLWK WKH MRLQW SGI RI < Lf % 5 DQG $ RQH JHWV WKH MRLQW SGI RI < Lf DQG $ LV JLYHQ E\ KI\AA $M RF 9 ; f7 L f R W f 7 Gnf D4 e DL$L \n .\ A n L O LQ7_RJLSf [ IW $A L O $f 6R WKH FRQGLWLRQDO SGI RI < f JLYHQ $ $ DQG < AA \AA LV JLYHQ E\

PAGE 205

RF _e_e [WHB[ [AAHM-[AA W [ D e DL$L fIF \Of7. Lf9Q7L RVc Sf [ ODJ W < DLAL L O \74\ M 1W AT6L3f $f 1RZ ; n7Y ; ; Of7BBOY f Yf ( f 9a < OL &6f Y YH A f? f ;7(; ; f7 OYOf Yf 9 LY ( HL[< f [ ( OO &6f Y \L\ Hff f ; HLf7YBL Lf L L L[fHM[ff f 9
PAGE 206

XVLQJ WKH VDPH H[HUFLVH ZH KDYH A $f 8VLQJ WKH GHILQLWLRQ RI IURP f DQG $f DQG $f ZH KDYH [f7VM`[f ,6 [fW(,`[f $f 1RZ XVLQJ D UHVXOW R 6DOODV DQG +DUYLOOH f ZH KDYH 4 LP ( ( F;; [[7f LP % f f§ 22 a VD\f $f :H SDUWLWLRQ %Hf LQWR $f ZKHUH %MHf (M H; A A A L M f 1H[W XVLQJ D VWDQGDUG IRUPXOD IRU SDUWLWLRQHG PDWULFHV HJ S RI 6HDUH f

PAGE 207

\7%B2\ \f7iL2\f \f %L%c-Hf\ff7 [ AO\f AOIf%O\nnff $f ZKHUH i L f # % A Hf % A f 8VLQJ D IRUPXOD VLPLODU WR $f DQG UHFDOOLQJ WKH GHILQLWLRQ RI LQ f ZH KDYH LP % ` f f§ 22 a 6LO[f[f7O-[ff [nf7"M$f 1H[W REVHUYH WKDW pOHf If ff f;f7 [ HL[ I[9!7f $JDLQ E\ 6DOODV DQG +DUYLOOH f Vf7VLLrfff7VL )URP $f $f DQG f LW IROORZV WKDW F8VIH&2%-Hf (. ;A[A7(L-[Af;A7(M0 $f $f $f

PAGE 208

1RZ VRPH ) URP QRWH WKDW AAAA2L fYf7?YfYf7 (mK% H"Oa ;9an; H; :[:I9X H;A9Af YfYf7 OfYOf7[O f9f7 W;nf;nnn + H;AA;AAf ;f§; $f ;AOf LV RI IXOO FROXPQ UDQN WKHUH 1RZ PDWUL[ ) VW ; A A ); A ? 7KHQ H; !; f7"; ; I;8;ALLf ;9n; OfYOf7[OYOfYf7 H) )H; f;f7 f"Q O:Of7[O fYf7 f r ; H);f;f7 """ H;AO\O;AL rf 9nnIm fYf7?Yf_Yf7 e[9fW (6Q86Q r fff7Q f f7 ; f (; $f $f O DQG $f LW IROORZV WKDW f§ iAf f a HfaO a

PAGE 209

^OO I[![A7f;f[f7 6LD!m6&!7} [![f7f` )6OOM6OO rf7ff9f`Vf7 A $f 1RZ XVLQJ $ f DQG $ f ZH KDYH IURP $ f WKDW LP % f f§ 22 ( AU OY OfY OfAQOY Of?aAYf7 6OL; [ [ fYf7fBYfQYf7WB "LLrY r \a a LYLfYLf9LYLfUYf7 e""n 6LL; [ ( "OOOA LYLfYLf9LYLfUL "LL[ [ [ f7YU! [ r LYL9YLf9LYLfUYf7 "M[ } [ 9YOf7fBOYOfUOYOf77B Y L < HQG ; A [fI[f76,O[!U[!7 ; f ); f f OYOf9YOf9OYOfUO ",L; f f Y \ < aaa f?7 DIWHU VRPH UHDUUDQJHPHQWV XVLQJ f $f 1RWH WKDW LV SG 1RZ IURP $f $f $f DQG $f RQH JHWV

PAGE 210

\74\ \ f7 f $f .\ 0\ Of?7ROf 0\ f $f 1RZ IURP $f $f DQG $f LW IROORZV WKDW Ir fX \f W f 7 RF _*_ D ( DL$L \Y n n.\ L f?a A7Q7A [ f 0\ Of?7fOf f f6 0\ W W R A erLrL D U? I \f7.\f a ? A7A AT La3f $f :U L W L QJ W ( iL L S D ( DMAL L L \f7.\f,* 4 LW IROORZV IURP $f WKDW I\ f8 \r 2& QUS W ( iL L S f W f 0\ f Q7A4tLa3L1L7fQ7f $f )URP $f DQG f LW IROORZV WKDW WKH FRQGLWLRQDO GLVWULEXWLRQ RI A JLYHQ

PAGE 211

W ( mL L SDUDPHWHU 4 $VR VLQHH S ORFDWLRQ SDUDPHWHU 0\ Lf DQG VFDOH $f LW I ROORZV IURP $f WKDW f7 f nOOr

PAGE 212

$33(1',; % $1 ,1'(3(1'(1&( 5(68/7 ,1 $ )$0,/< 2) (//,37,&$//< 6<00(75,& ',675,%87,216 /HPPD % &RQVLGHU D )DPLO\ RI GLVWULEXWLRQV n3 ^JI ; D (f eH53 D ZKHUH BS f ff LV DV GHILQHG E\ f ; Q[Sf LV D NQRZQ PDWUL[ RI UDQN S DQG ( Q[Qf LV D NQRZQ SG PDWUL[ 7KHQ IRU D UDQGRP YDULDEOH < Q[Of ZLWK GLVWULEXWLRQ BS;" D (f V
PAGE 213

5HFDOO IURP f f WKDW I LV DVVXPHG WR EH NQRZQ )URP %Of DQG %f LW IROORZV WKDW a LV ; ( < VXIILFLHQW IRU A %XW
PAGE 214

)RU \ DFWV RQ IURP WKH OHIW E\ J\f G\ ; % 1RWH WKDW WKH UHVXOW RI DFWLQJ RQ IURP WKH OHIW LV HTXLYDOHQW WR WKH JURXS RI WUDQVIRUPDWLRQV i ^J Gf eJU3f G J U X a \f G\ ;"` JLYHQ E\ f &RQVLGHU DQRWKHU JURXS 4 rf ZKHUH ^W G? R Q3 5 [ 5 DQG r LV D ELQDU\ RSHUDWLRQ ZKLFK ZH GHILQH IRU GL L L L E\ r HDV\ WR YHULI\ WKDW 4 LV D JURXS ZLWK GOG GA LO f ,W L WKH LGHQWLW\ HOHPHQW DQG f G f DV WKH LQYHUVH aG HOHPHQW RI * 1RZ ZH ZLOO VKRZ WKDW LV D KRPRPRUSKLF LPDJH RI 'HILQH WKH IXQFWLRQ U IURP RQWR E\ rOVf G f ZKHUH ["f * ,W LV HDV\ WR YHULI\ WKDW IRU T DQG 4UA * ADARDAf U@TAf r AAf a DQG UfD OVff 7KHQ WKH IXQFWLRQ L V D KRPRPRUSKLVP IURP RQWR DQG LV D KRPRPRUSKLF LPDJH RI )RU V LI DFWV RQ ,I IURP WKH OHIW E\

PAGE 215

6RPHWLPHV ZH ZLOO VD\ D JURXS ZLWKRXW H[SOLFLWO\ UHIHUULQJ WR WKH JURXS RSHUDWLRQ 1RWH WKDW DFWV PHDVXUDEO\ RQ L F[f DQG LV D ORFDOO\ FRPSDFW DQG MFRPSDFW WRSRORJLFDO JURXS ZLWK WKH XVXDO WRSRORJ\ LQ ZLWK DOJHEUD &A $OVR WKH PDSSLQJ Vf 7Vf GV GV" IURP *[ WR &Af LV PHDVXUDEOH +H &L;&L LV D SURGXFW DOJHEUD RI & ZLWK LWVHOI 6R WKH FRQGLWLRQV RI 3URSRVLWLRQ RI (DWRQ f DUH VDWLVILHG IRU DQG WKH PHDVXUDEOH VSDFH }f &Mf $OVR QRWH WKDW DFWV WUDQVLWLYHO\ RQ LI 6LQFH VG< ;"f G6L
PAGE 216

%,%/,2*5$3+< %DVX f 2Q VWDWLVWLFV LQGHSHQGHQW RI D FRPSOHWH VXIILFLHQW VWDWLVWLF 6DQNK\£ B %DWWHVH HUURU *( +DUWHU 50 DQG )XOOHU :$ FRPSRQHQWV PRGHO IRU SUHGLFWLRQ XVLQJ VXUYH\ DQG VDWHOOLWH GDWD RI L VW $VVRF f $Q FRXQW\ FURS $PHU %HUJHU -2 f 6WDWLVWLFDO 'HFLVLRQ 7KHRU\ DQG %D\HVLDQ $QDO\VLV QG (GLWLRQ 6SULQJHU9HUODJ 1HZ
PAGE 217

'HPSVWHU $3 5XELQ '% DQG 7VXWDNDZD 5. f (VWLPDWLRQ LQ FRYDULDQFH FRPSRQHQWV PRGHO $PHU 6WDWLVW $VVRF (DWRQ 0/ f 0XOWLYDULDWH $QDO\VLV $ 9HFWRU 6SDFH $SSURDFK :LOH\ 1HZ
PAGE 218

*KRVK 0 DQG /DKLUL 3 f %D\HV DQG HPSLULFDO %D\HV DQDO\VLV LQ PXOWLVWDJH VDPSOLQJ 6WDWLVWLFDO 'HFLVLRQ 7KHRU\ DQG 5HODWHG 7RSLFV 9 (GV 66 *XSWD DQG - %HUJHU 6SULQJHU9HUODJ 1HZ
PAGE 219

+DUYLOOH '$ LQ SUHVVf %/83 DQG EH\RQG $GYDQFHV LQ 6WDWLVWLFDO 0HWKRGV IRU *HQHWLF ,PSURYHPHQW RI /LYHVWRFN (GV *LDQROD DQG +DPPRQG 6SULQJHU9HUODJ 1HZ
PAGE 220

.DFNDU 51 DQG +DUYLOOH '$ f $SSUR[LPDWLRQV IRU VWDQGDUG HUURUV RI HVWLPDWRUV RI IL[HG DQG UDQGRP HIIHFWV LQ PL[HG OLQHDU PRGHOV $PHU 6WDWLVW $VVRF .HONHU f 'LVWULEXWLRQ WKHRU\ IRU VSKHULFDO GLVWULEXWLRQ DQG D ORFDWLRQVFDOH SDUDPHWHU JHQHUDOL]DWLRQ 6DQNK\D $ /DKLUL 3 DQG 5DR -1. f 5REXVW HVWLPDWLRQ RI PHDQ VTXDUH HUURU RI HPSLULFDO EHVW OLQHDU XQELDVHG SUHGLFWRUV 3UHSULQW /DKLUL 3 DQG 7LZDUL 5& LQ SUHVVf 1RQSDUDPHWULF %D\HV DQG HPSLULFDO %D\HV HVWLPDWLRQ RI YDULDQFHV IURP VWUDWLILHG VDPSOHV 6DQNK\D /HKPDQQ (/ f 7KHRU\ RI 3RLQW (VWLPDWLRQ 1HZ
PAGE 221

3UHVV 6f $SSOLHG 0XOWLYDULDWH $QDO\VLV +ROW 5LQHKDUW DQG :LQVWRQ 1HZ
PAGE 222

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

PAGE 223

+H JDLQHG VRPH YDOXDEOH H[SHULHQFH LQ FRPSXWLQJ ZKLOH ZRUNLQJ ZLWK DFWXDO GDWD LQ WKH 8QLWHG 6WDWHV &HQVXV %XUHDX ZKHUH KH DFFRPSDQLHG 'U 0DOD\ *KRVK DV DQ $6$16)&HQVXV 5HVHDUFK $VVRFLDWH EHIRUH KLV JUDGXDWLRQ 8SRQ JUDGXDWLRQ KH ZLOO MRLQ WKH 'HSDUWPHQW RI 6WDWLVWLFV 8QLYHUVLW\ RI *HRUJLD $WKHQV

PAGE 224

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n U KWX 0LFKDHO $ 'H/RUHQ $VVRFLDWH 3URIHVVRU RI 'DLU\ 6F L HQFH 7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH 'HSDUWPHQW RI 6WDWLVWLFV LQ WKH &ROOHJH RI /LEHUDO $UWV DQG 6FLHQFHV DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ $XJXVW 'HDQ *UDGXDWH 6FKRRO

PAGE 225

81,9(56,7< 2) )/25,'$