CS代考 RT 27 forneale to get – cscodehelp代写
Gunny Rit stock i return
Mit af mean forstock
qeiialo.gl
multiple stock return
Copyright By cscodehelp代写 加微信 cscodehelp
fluctuation
covariance
independent oftime Contents
on period t
measure of uncertainty in return G
Best guess
Return E ELRit
for allstocks ten N
linear associationbeteen stock returns
My Cig or Coo
We don’t know
or Gig or Cor Rig
Mi fr stocki or SD Roe
RANDOMVARABLES
parameters
it RÉ i Variable
Suppose R T R
estimator for the men a
1 RT 27 forneale to get
stocks return 51 251 401 51
substitute
for the mean a
Suppose the last 5months of a 101
estimator for men I Rtkthfutt substitute whees I
estimator for a
if compute Wh
valuer 7 it
estimators for
Ritkejthari
it u estimator
Bias of an
We would like to have
it a E it ee
ieal R BiasECRn gym
0 Suppose it Rather
Rthe Rn a FLEEELI ELIJ
gosayaa ae
inspect the average Robof Heeds
arose p.pt
examriesgfjt.IT
SAMPLE MEAN Egret
mil Rie di unbiased
estimator puevalue
www.gesmhf
lij going not
A Reise estimator is me whose venaility
Bias I E Cui
Squared Emr ii my MeanSquared Error
ie ut MSE ie m
MSECal u Precise Estimator E
Varlet Bias se
NSE is low
BiasCal al
jÉfrt frarfrittke est
unbiased estimator
94.1 int are independent
AsTinereesee MSE get smaller it gets more precise
tr 11ft True variance
Root Mean Squared Error
RouseCol m
Alternet meme Estimator
in sample ́ samplemaned
standard error or standards
T is higher Replace O
estimator is mole precise
Pijslample comelehin
as T o un ee then it is
happers to our estimator If
consistent estimator
Does Mee it I
we’re working Normattioned
disbehishin
Igad uncurl
or this is true for all the estimators
N N M É If
is not Normally Distrinuted NORMALLY DISTRIBUTED
NORMALLY DISTRIBUTED For
the coin T tries
Result of coin toss E Xe
w Mh P w prob i p
thbability of Heads
Pepe ep var flattie
fluffy tree
VarXe EFFEt s p p pere
Construct objective
a new random variable Z Bjp
carry of simulehin woo times a plot hatpin A
II.fi gf distributed
f fh a whiny man masks with meena
se f sample mean
sample mean as Tt
and variance
Also written as
I n N Mx I as T
Iiin N lij III as to
n’ÉmhT MEE
ie 1.96Ey e m e
95CI i ngi
in none 1R
o 9T or 951 metro.sn
E 1.96 Prob iguIfIdme
The window
it 1967,1 Igf
doesn’t artein u
if igift It i asFy contains a with a rub8951
1.2 3 6 we dont
either contains u know 6
Workaround Try using A
instead of 6
If T lo ie
tg dishbuhin 226to I t 226ft
WORKAROUND FOR
large enough T
It 0 025,9 9110975,9
2.26 t 2.26
he 1.9 É CI for remaining estimators using normed approx
Then the 951 981
i.ae EFei set the
is commonly used here r
Ej i.gg TTfj I 1.96 ÉdI
I’t 196 last
lij ij cat
Notts SEC I has the same unit as the estoneetor itself
SE f is smell precise estimate
SEL is daye imprecise estimate
SE ai depends on G stockswith higher 6
will have worse estimates for u
is swathe for layer sample sizes
If we want a more precise estimate get more data
historied returns maynot follow the
TO STANDARD Using it as
distribution as decreases as
recent returns 1
ERROR CONFIDENCE Intervals an example
T resamples f the date
I1it edz e The Jackknife estimate of bias
eatable the same man Resample R R2 Rte
calculate the sample mean
For each resouple calculate Rz I
calculate the sample men Resample R Rs Ru RT
leave one out estimators
biasjackla T1 sackknife I
For means selfie a Sejade lil
THEBuoTIRA
start with original sample Create B bootstrap samples
standard error
barton Isaak mean
og sample 1 GI
1st Bootstrap sample and Bootstrap seance
simple ist Bother source
and Bertha sank
Rutstge seance
Bootstrap estimator for bias of M
Use Bootstrap to estimate 951 Confidence Intervals NORMAL Approximation
Bootstrap estimator for se f
sebout lil SD Mit Mit if
o 587 Create new quantity
estate 10.574
vector f Afar it qf.org
meanfrefums
154 1900.491 I rift lay u491
untapeI 1.96Sebootd
original senile mean
Fall to reject Ho
HYPOTHESIS TESTING
HYPOTHESES TO BE TESTED
Null Hypothesis ftp AlternateHypothein
directional
SIGN FANCE
A Test Statistic
OF TEST formate
null hypothesis
USE TEST STATISTIC
C Reject the
a FALSE No Error
N error Significance level x
Probability of making Type I
Pr Rejecting Ito f
Power Of A TEST Pr Reject a It is False
EI s Low L t High T
Error it is True
Use 3 Suppose
0,0 102 is it Mto05
Rit Mi t it e i min
us AI no1010
E 1,2 italo ri
a specific value
Test whether mean I stuck i s retain equals Ho ie a n ve H M no
Significance Level Estimator for u
of null hypothesis is free
to make a decision
tf z staring
is me N NCoil
hypothesis
o or HiiHo as vto
asdontrefute
don’treject’t
3 score IIs
hypothesis
196Ey en Ft
on Ho n o05
196 Ift mo
ughhhh prom 0 050 En
perches If I if as 5f E pvalue 521
p value to significancelevelCa then don’t reject null hyphens
Reject null hypothesis if pvalue is less then a
9t 0.025T 1 2
i gtfo975T
Eftleast 30
truevalue standard error estimated for
Mall Hypothesis Alt Hypothesis
Nool 6010us HI64010
III Rtdoraqt Ho
calculate sample sd
calculate standard em t I i self fat
not sufficient evidence to reject n_
conclude Not sufficient evidence to reject null hypothesis I’WkfPtesnfn
X significance level 12 4 5
Calculate threshold valuesfor 3 score based on a 2 2
null hypothesis
HYPOTHESIS TESTING B W 2
Test Ho M Mz us He de M
GHo agrois
MSFT I SBVX
varia in EÉ
wyd tf 2M www.sina.in
estimate them
using 0.27
or 5 yeas I reorthby returns Tome
81 a threshold valuesfor 3 score from 0.04 e gnome 0.96 1.75 41.75
191 C 1.75 as we have sufficient evidence to reject Ho
程序代写 CS代考 加微信: cscodehelp QQ: 2235208643 Email: kyit630461@163.com
