Does build model has beta
No
Grinold kronar formula?
[Inflation + growth + change in pe+ div yield ]- rfr
Do not forgo the rfr
Aswath damodaram
ERP= ERP developed + [ LAMDA*Crp]
Crp= sovereign spread * ( sd eq developing / sd deb developing]
Matrix pricing
Old YTM + ( DIFFERENCE OF YTM ) * adjust for years ( big/ less)
Combing a portolio x with a benchmark pf , calculate sr
Portfolio SR =[ SR²B + IR²]½
Closet fund has ______ ir
Negative , if a manager has added any value that is representative in IR, then it is not a closet fund
Heterodasticity error test resolve
Type 1 error, Bp chi sq test , introduce robust std errors
Serial correlation error , test , resolution
Type 1 error , durbin watson ( DW TEST) , BG test robust standard errors
Multicollinearty error resolution
Type 2 error , drop one of the correlated varb.
Excluding one or more Regression varb
Using different proxies
Increasing the sample size
R ²?
sum of explained varb squares
/ sum of total varb
Adj R²
1 -([ 1-R²] [n-1/[n-k-1])
AIC ?
USED FOR BETTER FORECAST
Aic = n ln ( sse/n) +2 (k+1)
BIC
Bettet goodness of fit
n ln ( sse /n)+ln n (k+1)
Logit model
Ln ( p/1-p)
If say dw test have values 0.10 and 0.25 , CV 0.75 is scor present?
Yes , CV> T STAT
How is df conduct?
One must sub the id from xt-1 , from both sides of ar 1 model. This results in the change of dependent varb and change in regression coeff from b1 to b1-1
Dw test is invalid when?
Variables include past value of dependent varb.
What what we test to find out serial correlation
Correlation of coefficient and residuals ; via BG and Dw test , dw is irrelevant in case variable include past values of dependent varb
To perform non stationary test we?
Subtraction of independent varb from both sides of eqn
To test seasonality we?
Add an lag to the eqn
Do we ever change the independent varb in regression
NO , we can jus add sub both sides in case of stationary test
With serial correlation present in data which model we can use & which can’t
Use ar model., log ar
Not linear , log linear
Bp test?
n*r²
If less than cv , we reject that hetero is present
AIC AND BIC _____ the better
Lower
