OLS and linear model

CFA Level II
#1

What is Ordinary Least Square method? And what is the difference between OLS and linear model?

#2

Ordinary least squares is minimizing Σ(Yi − Yi*)², where

  • Yi is the actual value of the dependent variable
  • Yi* is the estimated value of the dependent variable from the model

A linear model is one of the form:

Yi* = _b_0 + b_1_X_1_i + b_2_X_2_i + . . . + bnXni

You can use:

  • ordinary least squares with a linear model
  • generalized least squares with a linear model
  • ordinary least squares with a nonlinear model
  • generalized least squares with a nonlinear model

OLS and linear are independent.

#3

In schweser , in time series topic OLS is related to the value of t, what does schweser mean by that?

#4

To estimate the values of the parameters in a time series model, you minimize Σ(Xt − Xt*)², where

  • Xt is the actual value of the dependent variable at time t
  • Xt* is the estimated value of the dependent variable from the model at time t

Same as before.

#5

Could you please explain the difference between OLS and GLS and when to use which one? To my understanding, we usually use OLS, however to correct for conditional heteroscedasticy we could use GLS as one remedy. Other examples?

#6

The best person here to answer questions about GLS is likely tickersu. He’s had a lot more experience with statistical theory than most people (maybe everyone else) here; certainly more than I have had.