COUNTRY RISK ANALYSIS



This report concentrates on performing analysis for particular factors across various countries, evaluating their influence on a country’s risk dynamics. Country risk is associated with those factors which determine or affect a country’s ability and willingness to pay on schedule interest and amortization on its external debt. More specifically, it is the credit risk of borrowers in a country as a whole viewed from a specific country perspective. Thus country risk analysis consists mainly of the assessment of the political and economic factors of a borrowing country which may interrupt timely repayment of principal and interest.

Country risk could be attributed to the following:

● A country’s position in the economic growth life cycle.

● The corruption rate of the country.

● The legal system.

● The degree to which non-violent conflict resolution is actually practiced within a society.

The country risk analysis results are used as pre-lending as well as post-lending decision tools. Prior to lending, decisions such as whether or not to lend, how much to lend, and how much risk premium it should charge, are based on the measured risk. After lending, periodic country risk analysis serves as a monitoring device, providing a pre-warning system. To accomplish this, we are utilizing the Country Risk 2019 dataset accessible from the website (link: www-2.rotman.utoronto.ca/~hull). The CSV file contains six primary attributes: Country, ABV, Corruption, Legal, GDP growth, and Peace. The dataset comprises 242 rows and 6 columns, predominantly containing numerical data, with the exception of the categorical “Country” and “ABV” columns.




The objective of this project is to employ the K-means clustering technique to categorize countries based on their similarity in terms of risk levels, distinguishing between those with high, medium, and low country risks. Additionally, the project seeks to ascertain the extent to which these factors contribute to or impact a country's risk level. We begin by loading the essential python libraries and reading the dataset. This is followed by visualizing the data using appropriate charts and graphs in order to gain insights.

First, we plot a correlation matrix in order to study the relationships between the features.

From the diagram above we can see that there is a high correlation between Corruption and Legal risk whereas there is no correlation between GDP growth and Peace.


I also plotted a pair plot in order to perform multivariate analysis on the variable.







 


From the above graph, we can note the following:
  1. The diagonal cells show the histograms of each feature. We can see that GDP growth is skewed to the left.
  2. There is a negative relationship between Corruption and Peace.
  3. There is a positive relationship between Corruption and Legal risk.
  4. There is a negative relationship between Peace and Legal risk.


  1. An increase in Peace leads to a decrease in Corruption rate and vise verser
  2. An increase in the corruption rate in a country is positively related to an increase in legal risk.
  3. An increase in the rate of legal risk leads to a decrease in the rate of peace in a country.
Moreover, the data is preprocessed, and in this case this involves normalization in order to ensure that variables with different scales do not dominate the clustering process.










I then performed the Elbow method to determine the ideal number of clusters for the dataset. The marginal gain of adding one cluster dropped quite a bit from k=3 to k=4. I chose k=3 (not a clear cut though). Below are the visualizations of the result;
A 3D plot.


Corruption vs Peace

Corruption vs Legal risk

Peace vs Legal risk


From the analysis, I find out that the countries with a high country risk also had low peace rate, high corruption rate and high legal risk rate. Conversely, the countries with low country risk were associated with low legal risk, low corruption rate and a high peace rate. Therefore , they are more likely to attract investors.

This is a link to the summary report: LINK




Comments

Popular Posts