Quant Shit

Question 1

Conditional heteroskedasticity is

Answer 1

residual variance related to level of X’s

Question 2

Serial correlation is

Answer 2

correlated residuals

Question 3

Multicollinearity is

Answer 3

two or more X’s are correlated

Question 4

Effect of conditional heteroskedasticity

Answer 4

Type I errors

high t stat, caused by low std errors

Question 5

Effect of serial correlation

Answer 5

Type I errors

positive correlation

Question 6

Effect of multicollinearity

Answer 6

type II errors

Question 7

Detection of conditional heteroskedasticity

Answer 7

Breusch-Pagan Test

Chi-Square Test

Question 8

Detection of serial correlation

Answer 8

Durbin-Watson test

Question 9

Detection of multicollinearity

Answer 9

Conflicting t and F stats

Correlations among ind variables if k=2

Question 10

Correcting conditional skedasticity

Answer 10

white-correct std errors

Question 11

Correction serial correlation

Answer 11

Hansen method

Question 12

Correcting multicollinearity

Answer 12

Drop a correlated variable

Question 13

Functional Form Misspecifications

Answer 13

• important variables omitted
• variables not transformed properly
• data pooled improperly

Question 14

Time-Series Misspecification

Answer 14

• X is lagged Y with serial correlation present
• Forecasting the past
• Measurement error

Question 15

Probit model

Answer 15

estimates probability of default given values of X based on normal dist

Question 16

Logit Models

Answer 16

estimates probability of default given values of X based on logistic dist (computationally easier than normal dist).

Logistic dist NOT logarthimic

Question 17

Discriminant models

Answer 17

produces a score or rank used to classify into categories

ex- bankrupt, not bankrupt

Question 18

Economic Significance

Answer 18

not significant just because of statistical significance

-commissions, taxes, risk, etc.

Question 19

If a time series is mean reverting

Answer 19

the value of the dependent variable tends to fall when above its mean; and rise when below its mean

Question 20

Mean Reverting Level Formula

Answer 20

b0/ (1 - b1)

Question 21

Forecasting Accuracy of ARCH measured by

Answer 21

root of mean squared error.

Use model with lowest RMSE based on out-of-sample forecasting

Question 22

Without a mean reverting level, the time series is

Answer 22

non-stationary

Question 23

Dickey-Fuller Tests for

Answer 23

unit root

Question 24

Dickey Fuller Test method

Answer 24

subtract x(t-1) from both sides; first differencing
where g1 = (b1 - 1)

If there is a unit root in AR(1) model , g1 will be 0.