Uncovering Key Performance Indicators for Private Sector Banks in Pakistan: An Application of Exploratory Factor Analysis

Efficiency of banking system leads to efficient allocation of scarce resources by the financial system of a country. Bank’s efficiency is ascertained through different mechanisms. One of the mechanisms uses the aspect of Key Performance Indicators (KPIs). Considerable literature exists on KPIs for non-financial organizations. However, for the financial sector, deliberation on KPIs is uncommon. This research paper endeavors to fill this gap especially for Pakistani Commercial Banks belonging to the private sector. A total of 25 Pakistani commercial banks are listed with the State Bank of Pakistan, out which five are owned and controlled by the Government of Pakistan. The remaining 20 belong to the private sector. Out of these 20, secondary data of top 10 private sector banks selected on the basis of asset size was analyzed for the five year period from 2011 to 2015. Statistical technique of Exploratory Factor Analysis (EFA) was applied on 28 different financial ratios to uncover four categories of possible KPIs. Citation: Saeed MA, Siddiqui FA (2017) Uncovering Key Performance Indicators for Private Sector Banks in Pakistan: An Application of Exploratory Factor Analysis. J Bus Fin Aff 6: 248. doi: 10.4172/2167-0234.1000248


Introduction
Performance evaluation of banks in academic literature devotes considerable focus on competition, concentration, efficiency, productivity, and profitability. In context of social welfare, competition and efficiency are vital as they promote low prices, high quality, and innovation. Financial resources thus become readily accessible and affordable. Further, the speed and strength of monetary policy transmission becomes stronger because of better competitive environment. Techniques for the direct measurement of bank's efficiency are problematic. One of the obstacle is the non-availability of data on individual banks' output prices (or credit rates) and figures on cost of different individual banking products [1].
This paper attempts to identify the KPIs for private sector commercial banks in Pakistan through application of EFA technique. The use of KPIs in performance measurement and evaluation is wide spread amongst business firms. KPIs are popularly used along with balanced score card to outline the strategy map of an organization.

Literature Review
Efficiency investigations and study of performance indicators in context of commercial banks is of interest for academics, bank management, central banks, and for the financial markets. Performance measurement models include; analysis of financial ratios, production analysis, Delphi analysis, analytic hierarchical process (AHP), data envelopment analysis (DEA) and balanced scorecard (BSC).
A comprehensive theoretical and empirical frame work for evaluation of bank performance has been propounded by Biker and Bos [1]. Their work in the shape of a book outlines the theoretical framework of profit maximizing bank, its basic and more advanced models, their assumptions and the empirical results. The authors finally conclude by developing a balanced scorecard which they recommend for ultimate adoption by bank performance evaluators.
KPIs for private sector banks were investigated by Nimalathasan for Srilanka by using Exploratory Factor Analysis [2]. The study used primary data collected from banking executives through a questionaire containing 21 variables (both financial as well as non financial) and based on five points Likert scale, from strongly disagree (1) to strongly reserves for loan losses, and the size of the bank were important factors affecting the profitability of the banks in both countries [5].
In Summary it can be concluded that ROA and ROE have been widely used as measures of banks' performance. Regarding factors affecting bank performance, different factors have been used by researchers such as; shareholders' equity to total assets, liquid assets to assets, total loans to total deposits, fixed assets to total assets, total borrowed funds to total assets, reserves for loans to total assets, market concentration, market size, labor productivity, bank portfolio composition, capital productivity, bank capitalization, financial ratios, level of capitalization, age of the bank, per capita GDP, cost to income ratio and customer satisfaction.

Objectives
The objectives of this research paper are as under; 1. To uncover the relevant performance indicators of private sector commercial banks in Pakistan.

2.
To extract the key indicators of performance of private sector commercial banks in Pakistan.

Research Design
Sampling procedure A total of 25 commercial banks are listed with the State Bank in Pakistan as on 30 th September, 2016. Out of these, five are owned and controlled by the Government of Pakistan [6]. The remaining 20 are owned and controlled by the private sector. This study focuses on 10 largest private sector banks selected on the basis of asset size. All of the 10 banks selected are listed on the Pakistan Stock Exchange. As such these commercial banks have partial general public ownership as well.

Data source
The study has been compiled on the basis of secondary data extracted from the published annual reports of the ten selected banks [7]. The sample period is from 2011 to 2015.

Statistical tool used
EFA has been employed to uncover the KPIs from a large list of possible performance indicators. The well-known software package SPSS version 23 was used to analyze and apply EFA technique [8].

Variables and factors analyzed
The following Table 1 contains the list of different variables, for which data was analyzed and the pertinent factors were extracted [9]. It may be noted that all of these variables are quantitative and are of ratio level. The code used in the SPSS software for the variables is also mentioned in the table.

Extraction of factors
Statistical technique of Exploratory Factor Analysis was applied on the 28 variables obtained from the financial ratios of the 10 selected commercial banks. The relevant statistics of the extracted factors which are given below indicates that the four factors (components) impound nearly 74% of the total variance explained. The extracted components have eigenvalues greater than one which is a common criterion for a factor to be useful. When eigen value is less than one, this means that less information is explained by the particular factor and there would be less justification to keep the factor. Since four factors were specified to the software, as many factors were extracted [10]. If small number of factors were not specified, then the software would have extracted the best factors, which in this case would have been seven. The extracted factors were also rotated. The rotation method employed was orthogonal (varimax). The results are shown on the following pages in Tables 2 and 3.

Grouping of factors
As per the rotated component matrix, four categories have been developed which impound the large number of 28 variables. The four categories have been assigned names on the basis of the common characteristics of the different variables. The four newly named groups or indicators are as under:

Interest coverage (IC)
This category or indicator represents seven variables with factor loadings ranging from 0.953 to 0.653. The list of seven variables include; Spread, Interest income/interest Income, Net Interest Margin, Net Interest Income/Total Assets, Capital Ratio, Total Liabilities/Total Assets, and Return on Assets. This indicator accounted for 35.859% of the rated variance.

Assets coverage (AC)
This indicator also represents seven variables. Factor loadings range from 0.567 to 0.809 out of 1. These seven variables include; Administrative Expenses/Non-Interest Income, Non-interest Income/ Total Assets, Non-Performing Loans/Gross Advances, Provisions Non-Performing Loans/Non-Performing Loans, Total Deposits/Equity, Provisions Non-Performing Loans/Gross Advances, and Deposits/ Total Assets. Variance of 20.652% was explained by this indicator [11].

Deposits efficiency (DE)
Six variables are included in this indicator. Factor loadings of these variables range from 0.619 to 0.872 out of 1. The list of variables include; Advances/Total Assets, Gross Advances/Borrowings+Deposits, Investments/Total Assets, Gross Advances/Deposits, Administrative Expenses/Profit Before Tax, and Non-Interest Expense/Total Income. Proportion of Variance explained by this factor is 10.39.  Rotation converged in 10 iterations.