Reference no: EM132855039

M31222 Business Analytics

Assignment brief: You are acting as an analytics consultant to fast-growing SMEs that wish to incorporate Business Intelligence, Analytics, and elements typically found in emerging technology frameworks to improve the business performance of the companies, or generally improve the decisions taken by these firms. Using data sets that the SMEs are providing you with and materials found on Moodle or further suggested readings, make suggestions and advise your clients based on the questions found in each exercise below. Remember that your clients are not familiar with statistical analysis so your presentation should reflect this and use simple language to explain your results.

Task: Create a brief presentation that answers the questions found within each of the case studies below. The total length of your presentation should not exceed 15 minutes (it can be shorter than this) and should include the graphs you produce. At the end of the slide pack accompanying your presentation, you should include slides containing the commands you executed in R to get your results and/or the screenshots of your results. You do not need to include these slides in your presentation; they are simply there to check that you have run the R code correctly. You should submit an mp4 file of your presentation and a copy of the slides used in your presentation (including the end slides containing the commands).

Case studies
Case study 1

You are acting as a consultant to an SME that would like to choose a mix of strategies in how it manages its finances to improve its profitability. Your client has complete lack of knowledge in financial management and tries to run her business based on her previous experience. Having access to balance sheets and income statements of firms in the vicinity, you are able to help the manager find patterns in the data of how other firms manage their profitability, and disentangle the relationships between profitability (as proxied by return on assets: ROA), liquidity (as proxied by Current Ratio), revenue efficiency (as proxied by Operating Margin) and debt ratio (as proxied by long-term leverage).

Instructions

• Download the file "SME Exercise1.xls" from Moodle. Use Microsoft Excel or other visualisation software to produce suitable charts illustrating the distribution of ROA, and the relationship of ROA with the other three measures. Include these in your presentation, along with descriptive statistics for ROA and a reflections on the shape of the distribution and the strength and nature of any linear relationships with the other three measures.

• Load dataset "SME Exercise1.xls" into RStudio

• Fit the following log-log model:

log(ROAi) = a + b1log(currentRatio) + b2log(Operating Margin) + b3log(Leverage)

• Ensure that your presentation includes discussion of the following:
a) Discuss how the regression model can provide valid insights to the firm.
b) State if the coefficients are statistically significant and what this means.
c) Having a transformed (i.e. log-log) model, explain to the firm manager what do the b1...b3 coefficients denote in this case?
d) Discuss their interpretation in this particular example.
e) The firm in question had losses last year and wants to use cash to cover those. Explain to the manager, what would you expect to happen to his firm's profitability (ROA) if Liquidity (Current Ratio) was to decrease by 5%?
f) So far, you've provided good insights to the manager, but what is the goodness of fit of this model? What does it show? Explain to the manager in simple terms.

Case study 2

You are working as a consultant for a credit underwriting company (companies that decide whether to lend to higher risk clients). The company has observed that it loses too much time in processing each application individually, and it has instead asked you to help automate its process on whether it should accept or reject a client based on the criteria it normally looks at when its analysts make a decision. This could help the firm potentially reduce its staff count or put it to better and more efficient use.

Instructions
• Upload the historical data of the company's clients (file "SME Exercise2.xlsx"). Note that the data set describes:
o Whether an offer was made by the firm in the past (0/1): "offer"
o The interest rate of the loan: "int.rate"
o The monthly instalments owed by the borrower if the loan is funded: "installment"
o The natural logarithm of the borrower's income: "logincome"
o The debt-to-income ratio of the borrower: "dti"
o The borrower's credit score: "creditscore"
o The number of days the borrower has had a credit line of any kind: "dayswithcreditline"
o The number of times the borrower has been late in repayments by more than 30 days: "timeslaterpayment"

• Apply a Logistic (Logit) Regression Model, where "offer" is your dependent variable, and all remaining variables (in quotes, above) are the independent ones.

• Ensure that your presentation includes discussion of the following:

a) Discuss if there is evidence that analysts of this company were looking at all above factors when they were making an offer.

b) Develop the logistic regression model containing all factors that analysts were looking at. This will serve as the model on which firm can be based to automate its procedure in the future.

c) How certain are you about the model you've developed1? Explain to the manager what this means in simple terms about the accuracy of the built model and the insights made from it.

You will need to load the ‘pscl' library. To do this, execute the following commands: install.packages("pscl")

Attachment:- Business Analytics.rar