While talking to Mackintosh about the results of the regression, Wallace made the following comment: Comment: “The interpretation of the slope coefficient of annual % increase in GDP is that for every 1-unit increase in GDP, we would expect returns to increase by 0.667 units. The coefficient value of 0.667 will remain constant even if we remove the second independent variable.” When Mackintosh inquired about the effect of changes in market return on the return to his Brazilian investment, Wallace stated that if we compare the returns to the investment
Level II of CFA Program Mock Exam 1 – Questions (PM) FinQuiz.com © 2019 - All rights reserved. at the beginning of a year with those at the end of the year during which time market return has increased by 1%, we would expect the returns to the investment to increase by 2.245%. However, he stated that to ensure that this holds true, we would have to calculate the F-statistic using data of the sum of squared residuals and the regression sum of squares. After his meeting with Mackintosh, Wallace met Colin Edwards, a statistician at QIA’s headquarters. Edwards has been working with Wallace toward the management of the U.S. stock portfolio of an institutional fund worth $25 million. In applying regression analysis to financial data, Edwards specified a multiple regression model to determine whether an increase in the U.S. literacy rate and a change in the technology industry’s competitive structure have any effect on technology stocks’ P/E multiples. While discussing the violations of regression assumptions with Wallace, Edwards stated that although heteroskedasticity does not affect the consistency of regression parameter estimates, it does however lead to mistakes in inference. Wallace stated that serial correlation, however, did affect the consistency of parameter estimates, but only if at least one of the independent variables is a lagged value of the dependent variable. Wallace questioned Edwards about machine learning (ML) and its major forms. Edwards replied with the following statements. i)Machine learning involves several approaches by which computer programs are used to learn from experience to improve performance. ii)Supervised learning and unsupervised learning are two distinct classes of ML. The distinction between these categories is whether the technique makes or does not make use of labelled training data. iii)Unsupervised learning analyzes X variable and there is no Y target variable set. 13.Which of the following about the regression on Brazilian stock returns is mostaccurate using a critical value of 2.00 for the t-test statistics? A.The Brazilian stock returns are very closely related to the annual return to the Brazilian market index. B.The Brazilian stock returns are closely related to the annual % increase in GDP. C.The Brazilian stock returns are unrelated to the annual return to the Brazilian market index and to the annual % increase in GDP.
Level II of CFA Program Mock Exam 1 – Questions (PM) FinQuiz.com © 2019 - All rights reserved.
You've reached the end of your free preview.
Want to read all 34 pages?
- Spring '16
- The Land, Financial services, Private equity fund, Chartered Financial Analyst