Database Systems Homework #5 Solution

HW5: (the biggest one)




 
For the group of companies you discovered in HW4, we want to look at their fundamentals in the year that they produced great returns.




Identify one or two factors in their fundamentals that most of the high-performers have in common. We want factors which you believe could have contributed to why these companies

performed well.




To help identify why these companies had good returns, answer the following questions:

(turn in SQL queries that answer the following questions).




Do the high-performers have any of the following in common?




High net worth - (total assets - total expenses)

Or: (even better) high net worth growth year-over-year.

High net income growth year-over-year?

High revenue growth year-over-year?

High earnings-per-share?

High earnings-per-share (eps) growth?

Low price-to-earnings ratio ?

(this is stock price relative to eps - i.e.: pe ratio = share price / eps )

Amount of liquid cash in the bank vs. total liabilities?




Hints: In class, we will go over how to get year-over-year changes using LAG() in a window function, using EXTRACT() to get years from dates, making temporary tables, performing JOINs ON more than one column at a time, and other things that may help.




ALSO: you may find that we do not have enough data in this small data set to answer all your questions. If you run into difficulties here - try going back to the end of HW4, and "widening the funnel at the top". I.e., allowing more candidates into your high-performers group, to increase the chances that we have enough data on the high-performers to make a conclusion.




If you don't like these questions, you may look for any characteristic that you like, but clearly identify what you have chosen and why.




If you choose to answer different questions than those above, or want to answer some combination of the above and some of your own, then turn in at least six queries of equivalent or greater difficulty as those listed above.




It's ok to ask questions and get negative answers - i.e., conclude that some characteristic did not contribute to your high-performing results (because the high-performers do not share this in common). But you need to find at least one or two characteristics that the high-performers do have in common.




 
Once you have identified some factors that you believe contributed to high performance, go back to the original list of all the companies (not just the highest returns),




and find the top 25-30 companies that have the most similar fundamental factors to your high-performers for the year 2016.




These will be your potential candidates for investment.




From this list of potential candidates, select ten companies to invest in:

 
Make sure there are no more than 2 companies in the same sector.

We want diversification, and selecting ten companies in the same sector fails to achieve that.




- Try to base your final selection on some reason you find in the data.

Identify why you made your final selections (sector diversification, profits, etc.).




Extra Credit:

From your potential candidates list (from Q #2, above), check the correlation between

each of your potential candidates, and try to choose companies from the list that have

the lowest correlation with each other.




Note: this is not the minimum correlation value - as two stocks may be highly negatively correlated. We want the correlation values which are closest to zero.




Submit one SQL script which includes the queries that you used to make your selections. Submit a list of your 10 investment choices, and a (very) breif explanation

of why you chose them (data-driven reasons).