support@w4writers.com +44 7743 307695
Mar 12, 2024

Assignment Task

Distribution Project Statement

Wholesale distribution companies typically purchase products from manufacturers suppliers and then sell them to retail stores making them available for consumers. Typically wholesale distributors deal in large quantities of goods and are set up to have warehouses distribution centres and logistic functions to manage and deliver inventory to retail stores. We are interested in better understanding the profitability of wholesale distribution companies.

Looking at the profitability of wholesale distribution companies globally over the past five years (PwC to provide excel containing raw data), is there a correlation (positive or negative) between their profitability and their local jurisdiction’s GDP and other key economic metrics or events (e.g., the COVID-19 pandemic). If so, what may

In addition, with a straightforward business model, wholesale distributors aren’t involved in other key business functions such as manufacturing, R&D, retail trade etc. Are researchers able to review the publicly available information of key global distribution companies and corroborate their key functions, assets, and risks across various jurisdictions (e.g., comparing the activities performed, assets held, and risks borne by wholesale distributors based in the US vs China) to determine the other drivers of profitability that may exist? Please also provide any supporting analysis for these additional considerations.

Context

This is a real business question that PwC is investigating. The role you are to play is one of a consultant contracted by PwC to assist with the analysis of data using the COMM5000 data analysis tools, which include descriptive and inferential statistics.

The work will be scaffolded into two milestones M1 (20%) & M2 (20%)) and a final project report (60%). Every milestone will require you to use what you have learned to address specific aspects of the data. Generally, M1 consists of an exploratory data analysis, whereas M2 is concerned with identifying hypotheses and formulating key inferential questions. In the final project report, all the insights gathered from M1 and M2 are used to model the data to answer the project questions.

PwC schedule of engagement

It is very important that you attend these sessions where delegates from PwC will hold live synchronous sessions to provide information about the importance of analysing the factors affecting the profitability of distribution companies to their operations. During these sessions, you are free to ask questions and discuss any aspects of the project.

Description of Assessment Task

In this final report, you will use the information you have collected in your analysis in M1 and M2 to build a linear regression model. This model will capture the key objectives we have been pursuing since M1.

(1) We want to investigate what company characteristics are determinant factors of its profitability.

(2) We want to know if the country of jurisdiction has a premium or effect on a company`s profitability.

(3) We want to address the question of whether COVID-19 has affected the profitability of wholesale companies.

In M2, you have performed a series of hypothesis testing to see whether there is evidence that profits differ between countries (at least for some industries) and whether there was a significant change from 2019 to post Covid19 years. Hypothesis testing is very informative but does not provide us with a means to control for confounding effects. For example, suppose you conclude that you reject a null of equal population means of operating revenues growth rates between the UK and Australia. In that case, you can’t conclude that this difference is due solely to the country`s effect. Other factors may affect revenues even if you control for the country of jurisdiction. Meaning that rejecting the null hypothesis here is driven by factors other than differences in the country.

To provide a ceteris paribus analysis, we need to use a regression model to identify and estimate the country effect (if any) while controlling for other factors. Other factors are defined by the data we are given. We will use the information on the company characteristics to control for the factors that may drive profitability and estimate the effect of the country of jurisdictio

DATA Considerations

(i) Profitability Variables For the dependent variable in your regression model, choose one of the variables you have already analyzed in M2. i. Operating Revenue ($’000)

ii. EBITDA: earnings before interest, taxes, depreciation, and amortization

(ii) Countries for analysis You need to have three countries to analyse the country effect. You can use the countries allocated to you in M1 and add to that to make three. You need to have Australia on your list. If you want to change countries, you can choose any combination of three from the list Australia, the UK, NZ, and the US.

(iii) Company Classification For the report rubric, there is no requirement to use any classification. You may keep the pool of companies as is to estimate the regression model. But some of you may find it easier to focus on samples of a small group of companies belonging to a class/ industry and do the regression analysis.

Task

(A): Within-Country Analysis for Australia

(1) We investigate what company characteristics determine factors for the selected profits variable Y! in Australia. First, consider the year 2019 and estimate a regression model for Y.

How do you select what X variables go into the regression? One simple way is to start by putting all the variables in the Australian dataset in the regression. Then perform a process of eliminating those that are not statistically significant.

Model_Australia2019: Consider all the 2019 company variables and the past years` variables as regressors, including the Y variable`s past values.

(i) Estimate the model and use the estimation results to remove all the variables that are not statistically significant.

(ii) Repeat the regression with these significant variables and check if all are statistically significant.

(iii) Repeat this process until you get a model where all regressors are significant.

(B): Between-Country Analysis

Consider now the three countries` data. We will use the determinants (regressors) we have selected in Task(A) for Australia and use them as a starting set of regressors for the between-country models. For illustration in this document, assume that the other countries are NZ and the UK.

(1) Consider the year 2019, estimate a regression model of Y on the selected regressors in (A) for both the UK and NZ

(2) Country effect: To estimate the country effect, we need to stack the data from the three countries in one file. We use a dummy variable to capture the regional effect or effect of the country of jurisdiction. There are two ways to add a dummy variable with multiple categories

Ethics Considerations

Considering what you have learnt in Week 8, are there any ethical considerations around data collection, data analysis and use/implementation of the report’s recommendations? Refer to the PwC code of Ethics:

Executive Report

The final report unifies the insights you have gathered in M1 and M2 and this final analysis. It showcases all the statistical techniques you have learnt in COMM5000. This report is a final take on this problem and should include the key results you wish the PwC team to get from the data. Your report will consist of the following core components. You can restructure the section as you see fit for your report.

Executive summary

This is the punch line of this case study. This should tell PwC what you have found in clear and precise language. This shouldn’t be technical. It should refer to PwC description of the problem and their expectations.

Introduction

The introduction should present the problem analysed in this case study. The background and the expected outcomes or target questions of the analysis. Your introduction should include some key conclusions from M2 (and M1) and how these conclusions provide some basis for developing the model for wholesale company profit characteristics. This is not a copy/paste of all your reports for the first two milestones. It should report the take-home message(s) from the detailed analysis you have completed no need to re-report the tables and graphs from those milestones.

Section

1: Data considerations, including what countries you are analysing, any major limitations in the data like a significant % of missing points for some variables/countries. This will inform the reader that you will restrict the analysis according to data availability.

2: Analysis of the determinants of profitability (growth rate of operating revenues) in Australia, including your analysis of the covid19 effect. Is there a Covid19 premium? Was the move from pre-Covid19 to Post Covid19 costly for the company/or industry in your analysis? Can you express it in the per cent change of growth in the profit variable?

3: Analysis of the countries` differences in profitability determinants. How important is the country`s effect? Is it the same size across the three countries you have analysed? Is there a premium or penalty to a company located in a specific country compared to Australia? In this section, you can put your news commentator hat on and add a comment on how you believe the country effect may have be driven by the.

4: Robustness analysis and model limitations. Any statistical analysis is based on assumptions that ensure that the inference you perform and the estimation of the model parameters is also correct. Some points to discuss in your analysis are the key assumptions:

  • Zero conditional means: does your model satisfies the exogeneity condition? This will ensure that no confounding factors will bias the estimates of your key target parameters in the model. A brief explanation of what this assumption implies for the context of the model(s) you are estimating. you believe it holds and why? What can be done to ensure it is satisfied if you had more data and more time to develop the model? If it isn’t satisfied, what are the implications on the model estimation and inference?
  • Multicollinearity: With the selected set of regressors in your models, check whether the regressors satisfy the assumption of no perfect collinearity. Explain what it means in your context and the implications of the regressors failing the condition on the validity of the inference.
  • Concluding remarks. In this section, you should summarise the data`s key take-home messages. Your view on ethics considerations and how this impacts the collection of the data and the use of your results can be discussed in the conclusion. 
Recent Post

Order this Assignment now

Total: GBP120

fables template