- CFA Exams
- CFA Level I Exam
- Study Session 2. Quantitative Methods (1)
- Reading 5. Multiple Regression
- Subject 4. The Standard Error of Estimate in Multiple Linear Regression Model
CFA Practice Question
Suppose you want to know whether Fidelity Select Technology Fund (FSTF) behaves more like a large-cap growth fund or a large-cap value fund. You want to estimate the regression Y-hatt = b0 + b1X1t + b2X2t + et, where Yt is the monthly return to the FSTF, X1t is the monthly return to the S&P 500 Growth Index, and X2t is the monthly return to the S&P 500 Value index. The table below shows the results of a multiple linear regression using monthly data from December 1994 through March 2009.
Calculate the standard error of estimate (SEE) for the regression in the table above.
Correct Answer: From the table we can calculate the standard error of estimate as (0.3682/169)1/2 = 0.0467. Thus the residual standard error is 4.67% a month.
User Contributed Comments 6
| User | Comment |
|---|---|
| hrai123 | SEE is really square root of Mean square of residuals in the Anova table |
| katybo | That is right: MSE = SSE / n-k-1 |
| MasterD | Thus SEE = (MSE)^(1/2) = (SSE/(n-k-1))^(1/2) |
| mikeb119 | And according to the previous question: "The standard error of estimate is equal to the standard error of the residual." So you can just pull the 0.0467 from the table. |
| Nightsurfer | mikeb119, I was going to point out the same thing. |
| quanttrader | SEE=sqrt(MSE), MSE provided but can also be calculated as SSE/n-k-1 |