BASICS
PetDt
RawGrossReturn Pe i Rt
Average Rpt FE ftp.RK N aRi
Pt P dins
Is
Raw Net Return re
p
Variance
Rp wi O two 0221 2W We0,02pm
Log Return log Re rt footfolio
dork
Cord2 EAR EIRD R2 E R2
Cover.az
NR Ra O or
NOT SO BASICS
Y Yi Yi row vector Yi YT are random variables not scalars
1 x T
to
f
t Yoke
Tx
Avg Y It To sum obs in a row vector multiplyby Ii
To sum obs in a colourvector multiply It by thevector
Var ly Create 2 Y Aug I't lsince avg is 1 number youmultiply by 1 to
method 1 subtract from
average eachobs matchproportions of X
Var ly
f ZZ
rsmatipY Alwaysavarkou
matrix
gesturing.ve
a
µ.gg
Farias
squared
avgsquared II
Aught Aug Y
VfAug Vt Ii Vlf Eye 1Eye
5 ELAug INAug 0 if ii'd
if i id Coto and Vie o htt
so
to't f o's
if Nyt o ft but Cov o
f
, to TTT HCov
EEE E'Iii
Rule 1 El'll Ely Ely I
Rule 2 EIA Y A ELY DOESN'T APPLYTo VH1
Rukh Sandwich rule
VLA YI A WH A
get.ae
YI
VAFA A Vh1 A
Univariate regression in normal stats
A dtpry.tn
Cov IR Ra EIR R2 EIR Elka
VIR ECR R ETR EC2D EfRf
EfR.Yvar.isjustcov.oftvwbyitsd
playcgvaIIg .p in normalstats
If A LEG A Ide ta be A A Inversematrix
a o.o II
X
Tx Isamedimentionas It
g
for YitwaslY.INbecause Ywasrowvector
Aught f Il't X
vwkt fki.tl
fkkl
I lwas4
wasYY IT
forrowvector forrowvector
covlx.wt ffxwl ffh.in
ffleiw
uedorEfx.w
justarother
Efx1 EW
ofsainedianentions
ask
X Cti Z Newvector X differentfrom before
4
f
, Ii
Y Xp tu Univariate regression ineconometrics
I ftp.z coTnesfromFoc
derivations
justtake given
B ex xI H'Ht H t.IXBtuh.fr xx X.pItfxtXI xi.n
T T
since Btf I X'up
wehaveinversionmatrix we candivideby To not it'sthesame
be T
T TITI's f ft L
ypg y
Bi 132
Assumptions for sampleregression function Yi a Bet
1 Echt 0 errorshave
too
2 Vliet o ft so Vfu GI errorshave var 0 f Homoskedasticity
3 Cov
lui ujl o all errors arestatistically independent
ofeachother
4 Corfu Xi 0 allerrors arestatistically independent
oftheircorresponding Xvariable
V Ix X X hu Nx X5
tfptfx.XI.hu
since
B p J I T Ihriassspohondodfuetbeinga
aesstimfaorg.ffbieavwan.mu
A Y symetricinatrieitstransposeisthesame
dPiso
sincebyassumption
Vinko't o2X X x X o4x X
This is avarkovmatrix f
iii cities
SEIB Vote X
SEIB Vote X 2 fit's the
or
to same
If we have Vllin combo
from.aecovqatrir
I 22
PetDt
RawGrossReturn Pe i Rt
Average Rpt FE ftp.RK N aRi
Pt P dins
Is
Raw Net Return re
p
Variance
Rp wi O two 0221 2W We0,02pm
Log Return log Re rt footfolio
dork
Cord2 EAR EIRD R2 E R2
Cover.az
NR Ra O or
NOT SO BASICS
Y Yi Yi row vector Yi YT are random variables not scalars
1 x T
to
f
t Yoke
Tx
Avg Y It To sum obs in a row vector multiplyby Ii
To sum obs in a colourvector multiply It by thevector
Var ly Create 2 Y Aug I't lsince avg is 1 number youmultiply by 1 to
method 1 subtract from
average eachobs matchproportions of X
Var ly
f ZZ
rsmatipY Alwaysavarkou
matrix
gesturing.ve
a
µ.gg
Farias
squared
avgsquared II
Aught Aug Y
VfAug Vt Ii Vlf Eye 1Eye
5 ELAug INAug 0 if ii'd
if i id Coto and Vie o htt
so
to't f o's
if Nyt o ft but Cov o
f
, to TTT HCov
EEE E'Iii
Rule 1 El'll Ely Ely I
Rule 2 EIA Y A ELY DOESN'T APPLYTo VH1
Rukh Sandwich rule
VLA YI A WH A
get.ae
YI
VAFA A Vh1 A
Univariate regression in normal stats
A dtpry.tn
Cov IR Ra EIR R2 EIR Elka
VIR ECR R ETR EC2D EfRf
EfR.Yvar.isjustcov.oftvwbyitsd
playcgvaIIg .p in normalstats
If A LEG A Ide ta be A A Inversematrix
a o.o II
X
Tx Isamedimentionas It
g
for YitwaslY.INbecause Ywasrowvector
Aught f Il't X
vwkt fki.tl
fkkl
I lwas4
wasYY IT
forrowvector forrowvector
covlx.wt ffxwl ffh.in
ffleiw
uedorEfx.w
justarother
Efx1 EW
ofsainedianentions
ask
X Cti Z Newvector X differentfrom before
4
f
, Ii
Y Xp tu Univariate regression ineconometrics
I ftp.z coTnesfromFoc
derivations
justtake given
B ex xI H'Ht H t.IXBtuh.fr xx X.pItfxtXI xi.n
T T
since Btf I X'up
wehaveinversionmatrix we candivideby To not it'sthesame
be T
T TITI's f ft L
ypg y
Bi 132
Assumptions for sampleregression function Yi a Bet
1 Echt 0 errorshave
too
2 Vliet o ft so Vfu GI errorshave var 0 f Homoskedasticity
3 Cov
lui ujl o all errors arestatistically independent
ofeachother
4 Corfu Xi 0 allerrors arestatistically independent
oftheircorresponding Xvariable
V Ix X X hu Nx X5
tfptfx.XI.hu
since
B p J I T Ihriassspohondodfuetbeinga
aesstimfaorg.ffbieavwan.mu
A Y symetricinatrieitstransposeisthesame
dPiso
sincebyassumption
Vinko't o2X X x X o4x X
This is avarkovmatrix f
iii cities
SEIB Vote X
SEIB Vote X 2 fit's the
or
to same
If we have Vllin combo
from.aecovqatrir
I 22
