Find solutions for your homework
Find solutions for your homework
businessfinancefinance questions and answerswhile discussing correlation and regression models in chapters 8 and 9 of quantitative investment analysis, the authors note that the coefficient estimates for a particular model may be sensitive to a few outliers or too small changes in the data set. outliers can also influence sample averages, simple linear (slope-intercept) regression models, and
This question hasn't been solved yet
Ask an expert
Question: While Discussing Correlation And Regression Models In Chapters 8 And 9 Of Quantitative Investment Analysis, The Authors Note That The Coefficient Estimates For A Particular Model May Be Sensitive To A Few Outliers Or Too Small Changes In The Data Set. Outliers Can Also Influence Sample Averages, Simple Linear (Slope-Intercept) Regression Models, And
While discussing correlation and regression models in Chapters 8 and 9 of Quantitative Investment Analysis, the authors note that the coefficient estimates for a particular model may be sensitive to a few outliers or too small changes in the data set. Outliers can also influence sample averages, simple linear (slope-intercept) regression models, and multiple linear regression models.
For the sample average, the impact of outliers may be reduced by excluding the most extreme observations in the data set. For example, an "Olympic average" is a particular type of trimmed mean in which the sample average is computed after the largest and smallest observations are discarded. Trimmed means and other tools are less sensitive to outliers, but critics of this approach argue that these tools make inefficient use of the sample information.
Suppose you read a financial prospectus that reports the expected returns for an investment opportunity based on the Olympic average of annual returns from the past 10 years. Would you trust the results? Would you like to know the values of the excluded outliers before making a decision? Would you prefer to see the average return computed from all of the data?
