Unbalanced datasets

How would you proceed if you the default rate was 1% instead of 9% as in the the case shown before. Would you need to add some pre-processing to the dataset?

Also, it seems you used one line for each loan. How would you proceed if you had quarterly financial information on the borrower (such as corporate borrowers). Would you have a line per financial information and the default or not if the default occurred a year after the financial data? How would you proceed?