• Choose a publicly traded security for which you can find a series of historical values and make a conjecture about related data (at least two data series) that might be used as predictors for this series of values
  • Find online data sources to get current data for both the predictors and the values of the security (up to 12/31/19 or more recent, if possible)
  • Download this data and copy it into Excel
  • Create graphs of the data
• Use Excel to conduct a regression of the values of the security against the predictors and verify the validity of underlying assumptions
  • Check for homoscedasticity and serial correlation
  • If necessary, rerun the regression using robust standard errors
  • Look for evidence of multicollinearity and eliminate redundant predictors if necessary
• Submit written report & Excel spreadsheet
  • Report can be .doc or .pdf
  • Document regression inputs with text comments
• Report should:
  • Clearly state the conjecture being investigated
  • Clearly describe the data sources used & data handling
    • Include links to data if possible
  • Clearly describe the analysis performed
  • Clearly state the conclusions you have drawn from this analysis