A few days ago we were asked to analyse the relationship between Interest Rates, Non Performing Assets (NPAs) and Credit Growth in the Economy for our Commercial Banking Course. A part of the analysis involved the use of computation of the NPA Ratio which is defined as the ratio between the Net NPA and Net Advances or Gross NPA by Gross Advances. After the computation of the ratios they were to be classified based on the nature of ownership of the banks. This is where I realized a first level analysis using SPSS could have been used. It would have given us a fair idea about the various categories we were analyzing, namely:
· Public Sector Banks
· Private Sector Banks
· Foreign Banks
A bivariate analysis would also have been possible to study the relationship between say, the nature of the scheduled bank and the capital adequacy ratios (CAR) which is the most important ratio for a bank.
Clustering is the process of dividing data into meaningful groups so that they can be used for analysis. Rather than looking at the theoretical aspects of the topic I would again like to highlight the use of a basic clustering technique for financial analysis. Being a student of Finance the easiest correlation between whatever we learn in class and the practical world are the assignments which we are given. So would again state an example of an assignment which we will be doing for our Investment Analysis and Portfolio Management. For the course, we have to make a list of 200 stocks with prices and their book values over a period of five years. These companies would then be “clustered” into groups based on price to book value ratios into various portfolios each year. A further step using clustering could then be used to classify these companies into Growth and Value stocks. To make this decision a number of ratios would be taken into account. Some of them include the growth rates, price to book value ratios, price to earnings ratio etc. All of these factors would make a cluster based on which the decision can be taken and a portfolio created.
Another use of cluster analysis could be for stock picking where various factors would be made into clusters like Macro economic factors, Growth Prospects, Firm Specific Risk, Foreign Institutional Flows risk etc. Each cluster mentioned above would consist of individual factors which would together make up the cluster. For example, the Macroeconomic Cluster would consist of factors like:
· Wholesale Price Index
· Index of Industrial Production Figures
· GDP Rate
· Tax Collection Rate
· Domestic Savings Rate
· Capital Formation Rate
· Fiscal Debt Rates
This is the sort of application that I could think of after the first day of the Business Analytics course. These are based on the First Level Analysis and Basics of Cluster Analysis discussed in class today.
Author: Rishi Sonthalia (13039)
Group: Finance 3