The study examined critical factors that discriminate between Nonperforming loans and advances and performing ones in commercial Banks. Non-performing credits has been a major cankerworm that continuously affects the Nigerian Banking System. A linear discriminant function was developed after considering the eight factors responsible for discriminating between Performing credits and Non-performing credits. The study revealed that only two factors; years of experience and the tenor of the credit facility were the most important discriminatory factors that successfully discriminate between Performing and Non-performing credits. The model developed for the analysis is Y = 0.282 X4 – 0.234X8 . This model was evaluated using F-value, Chi-square, Eigenvalue, Canonical correlation and Wilk’s Lambda, and it was confirmed to be significant in discriminating between the two credit groups.


Certification Page i
Dedication ii
Acknowledgement iii
List of Tables iv
Abstract v
1.0 Introduction 1
1.1 Background of the Study 1
1.2 Discriminant Function Analysis as a Multivariate
Technique 4
1.3 Statement of Problem 6
1.4 Limitation and Scope 6
1.5 Aims and Objectives 7
1.6 Significance of the Study 8

1.7 Purpose of Discriminant Analysis 8
1.8 Classification Problems Examples 9
1.9 Definitions of Terms 10
2.0 Review of Related Literature 15
2.1 Discriminant Function Analysis 19
2.2 Problems In Discriminant Analysis 21
2.3 Commercial Loans and Advances 24
3.0 Research Methodology 30
3.1 Sources of Data 30
3.2 Sampling Procedure 31
3.3 Method of Data Analysis 31
3.4 Problems Encountered During Data Collection. 31
3.5 Discriminant Analysis 32
3.5.1 The Form of Data for A Discriminant Analysis 33
3.5.2 Fisher’s Linear Discriminant Function 34
3.6 Stepwise Discriminant Function Analysis 38

3.6.1 Wilk’s Lambda Distribution 39
3.6.2 Near-Exact and Asymptotic Distributions
for the Generalized Wilks  Statistic 42
3.6.3 The Theory of Canonical Correlation in Discriminant
Function Analysis 45
3.7 Probability Of Misclassification 47
3.8 Cutting Scores 49
3.8.1 Relative Discriminatory Power of the Variables 49
3.8.2 Computer Programme Used In the Analysis 50
3.9 Evaluation of Performance of the Model 51

4.1 Data Description. 52
4.2 Data Presentation 52
4.3.1 Application Of Stepwise Discriminant Method 58
5.1 Summary of Findings 66
5.2 Conclusion 67

5.3 Recommendations 69
References 71
Appendices 76


Non-performing loans and advances have been a major problem
affecting the banking industry. It creates problem of liquidity into
the system and also is a sign that a bank is unhealthy.
As at the end of March, 2004, the CBN’s ratings of all the banks
classified 62 as sound/satisfactory, 14 as marginal and 11 as
unsound, while 2 of the banks did not render any returns during
the period. The weakness of some ailing banks are manifested by
their overdrawn positions with the CBN, high incidence of non
performing loans, capital deficiencies, weak management and non
corporate governance.
A further analysis of the returns of the marginal and unsound
banks reveals that they account for 19.2 percent of total assets of
the banking system, 17.2 percent of total deposit liabilities while
the industry non-performing assets account for 19.5 percent.
Although below the trigger points for declaring the system as

distressed, they are nevertheless of major supervisory concern
(Soludo 2004).
Banks consolidation and 25bn recapitalization in Nigeria reduces
the number of deposit money banks from 89 as at end-June 2004,
to 25 banks as at January 2006. The reform is designed to ensure
a diversified strong and reliable banking sector which will ensure
the safety of depositors’ money, play active developmental roles in
Nigerian players in the African Regional and Global Financial
In the United States of America, there had been over 7000 cases of
bank mergers since 1980. In Korea, for example, the system was
left with only 8 commercial banks with about 4,500 branches after
consolidation. A bank in South Africa – Amalgamated Banks of
South Africa (ABSA) has asset base larger than all of the Nigerian
commercial banks put together (Soludo 2004)
The Central Bank of Nigeria removed 5 Banks’ Chief Executive
Directors (CEO) and their Executive Directors on 14th of August,
2009 for excessive high level of non-performing loans in five banks
which was attributable to poor corporate governance practices, lax

credit, administration processes and absence or non-adherence to
banks credit risk management practices. Thus, the percentage of
non-performing loans to total loans ranged from 19% to 48%
(Vanguard Online, 14 August, 2009).
Furthermore, additional three Banks’ Managing Directors and
Executive Directors were also fired on 2nd October, 2009 for the
same offences. The huge provisioning for the non-performing loans
have virtually eroded the shareholders fund. Thus, the banks are
under-capitalized for their current levels of operations and are
required to increase their provisions for loan losses, which impacted
negatively on their capital (Sanusi 2009).
In other words, these banks were unable to meet their maturing
obligation as they fall due without resorting to the CBN or Inter
Bank market. Their liquidity ratios ranged from 17.65% to 24% as
at May 31, 2009 (Regulatory minimum is 25%). Hence, the need to
identify the factors contributing to non-performing credits in Nigeria
Banking Industries.
In furtherance of the efforts of the Central Bank of Nigeria (CBN) to
assist the banks affected by the outcome of the recent CBN/NDIC

Special Examination, published the list of non-performing loans of
N100m and above for Bank PHB, Spring Bank, Unity Bank, Wema
Bank and Equitorial Trust Bank on The Nation Newspaper. The
number of the non-performing loans of N100m and above for the
banks stated above respectively are 149, 221, 120, 79 and 45 (The
Nation, October 14, 2009)
The banking sector reform is valid for now.
Multivariate analysis can be referred to as all statistical methods
that simultaneously analyze multiple measurements on each
individual or object under investigation. Any simultaneous analysis
of more than two variables can be loosely considered as multivariate
analysis. As such, multivariate techniques are extensions of
univariate analysis (analysis of single-variable distributions) and
bivariate analysis (cross-classification correlation).

However, to be considered truly multivariate all of the variables
must be random variables that are interrelated in such ways that
their different effects cannot be meaningful interpreted separately.
Discriminant analysis is a multivariate technique concerned with
separating distinct set of objects or observations and with allocating
new objects to previously defined groups.
It can be referred to as a statistical technique by which we can
make decisions to categorise groups or classify individuals (objects)
into their respective groups usually on the basis of some
measurement observed on the individuals (objects). This implies
that the basic problem of discriminant analysis is to assign an
observation, X, of unknown origin to one of two (or more) distinct
groups on the basis of the value of the observation.
In some problems, fairly complete information is available about the
distribution of X in the two groups. In this case we may use this
information and treat the problem as if the distribution are known.


The Central Bank of Nigeria is saddled with the responsibility to act
and protect all depositors and creditors and ensure that no one
loses money due to bank failure. The commercial banks give out
depositors money as loan and advances which should be paid back
at expiration of such facilities.
The basic question is what factors contribute or identifies non
performing loans or performing loans. What factors or indices
should be looked into, to avoid non-performing loans which can
eventually make a bank to have liquidity problem which can
culminate to classify a bank as distressed and depositors losing
their hard-earned money. Hence, the research is focused on
Discriminant Function analysis of performing loans/advances and
non-performing loans/advances.
This project work is limited to commercial loans and advances
granted by commercial banks in Nigeria between 2006 and 2008,

both year inclusive. Consumer loans are not included in this
The study is aimed at performing a discriminant analysis on
performing loans/advances and non-performing loans/advances.
Using data compiled from schedules received from 5 randomly
selected commercial Banks (Skye Bank, Intercontinental Bank,
Union Bank, UBA Bank, and Bank PHB) through their
Account/Credit officers. The project seeks to achieve the following
sets of objectives:
1. To identify the socio-economic characteristics that
discriminate between performing loans/advances and
nonperforming loans/advances;
2. To be able to predict the likelihood that loans/advances given
out by a bank will belong to a group of performing or not
based on the rule to be derived with Fisher’s Discriminant


Non-performing loans/advances cause liquidity problems in the
banking industry whereby banks were unable to meet up with their
obligations as they fall due. This study is justified in that it will add
to the body of knowledge pertaining to factors contributing to non
performing loans/advances.
The inferences from the research will enable the management of
various banks to restructure their loans and advances to address
the socio-economic characteristics that contributed to non
performing loans.

 To classify cases into groups using a discriminant prediction
 To investigate independent variable mean difference between
groups formed by the dependent variable.
 To assess the relative importance of the variable in classifying
the dependent variable.

 To discard variables which are of little discriminating power to
the group distinctions.
Instances of classification problems can be applied in many field of
endeavour such as the following:
1. A medical practitioner who intends to classify new born babies
into different categories of blood groups, based on measurements
obtained from the blood samples of the babies.
2. A geologist can as well wish to classify fossils into their respective
categories of fossils-groups on the basis of measurements on the
ages, size and shapes of the fossils.
3. A guidance and counselling consultant might desire to categorise
students with different course in a university, for which they are
best-suited based on measured scores of the students in related
subjects in the J.M.E results.
4. A biochemist might desire to classify foods into the distinct
categories of food nutrients as protein, fat and oil, vitamin,

carbohydrates, minerals and water based on measurement of the
comparative amount of different nutrients in the food.
5. An agronomist can as well be faced with the problem to classify a
particular breed of animal or plant into its proper class.
6. An automobile engineer might as well decide to classify an engine
into one of the several categories of engine on the basis of
measurements of its power output, shape and size.
All the above related problems of classification can be effectively
solved by discriminant analysis.
a) Performing credit: A credit facility is deemed to be performing
if payments of both principal and interest are up-to-date in
accordance with agreed terms. The borrower must effect payment
such that outstanding unpaid interest must not exceed 90 days.
b) Distress Institution: is a financial institution with several
financial operational and management weakness which have
rendered it difficult to meet its obligation to its customers, owners
and economy as at when due.

c) Non-performing credit: a credit facility should be deemed as
non-performing when any of the following conditions exists:
(i) interest or principal is due and unpaid for 90 days or
(ii) interest payments equal to 90 days interest or more have
been capitalised, rescheduled or rolled over into a new loan
(except where facilities have been reclassified as specified in
(iii) below);
(iii) the practice whereby some licensed banks merely
renew, reschedule or roll-over non-performing credit
facilities without taking into consideration the repayment
capacity of the borrower is objectionable and unacceptable.
Consequently, before a credit facility already classified as
“non-performing” can be reclassified as “performing” the
borrower must effect cash payment such that outstanding
unpaid interest does not exceed 90 days.
Non-performing credit facilities should be classified into
three categories namely, sub-standard, doubtful or lost on
the basis of criteria below:


(d) Sub-Standard : The following objective and subjective criteria
should be used to identify sub-standard credit facilities:
( i) Objective Criteria: facilities as defined in c(ii) on which
unpaid principal and/or interest remain outstanding for
more than 90 days but less than 180 days;
(ii) Subjective Criteria: credit facilities which display well
defined weaknesses which could affect the ability of
borrowers to repay such as inadequate cash flow to service
debt, under-capitalisation or insufficient working capital,
absence of adequate financial information or collateral
documentation, irregular payment of principal and/or
interest, and inactive accounts where withdrawals exceed
repayments or where repayments can hardly cover interest
(e) Doubtful : The following objective and subjective criteria
should be used to identify doubtful credit facilities:
(i) Objective Criteria: facilities on which unpaid principal
and/or interest remain outstanding for at least 180 days but
less than 360 days and are not secured by legal title to

leased assets or perfected realisable collateral in the process
of collection or realisation.
(ii) Subjective Criteria: facilities which, in addition to the
weaknesses associated with sub-standard credit facilities,
reflect that full repayment of the debt is not certain or that
realisable collateral values will be insufficient to cover bank’s

(f) Lost Credit Facilities: The following objective and subjective
criteria should be used to identify lost credit facilities:
(i) Objective Criteria: facilities on which unpaid principal
and/or interest remain outstanding for 360 days or more
and are not secured by legal title to leased assets or
perfected realisable collateral in the course of collection or
(ii) Subjective Criteria: facilities which in addition to the
weaknesses associated with doubtful credit facilities, are
considered uncollectible and are of such little value that
continuation as a bankable asset is unrealistic such as
facilities that have been abandoned, facilities secured with

unmarketable and unrealisable securities and facilities
extended to judgment debtors with no means or foreclosable
collateral to settle debts.
Provision should be made for non-performing credit facilities
as follows:
(i) Sub-Standard Credit Facilities: 10% of the outstanding
(ii) Doubtful Credit Facilities: 50% of the outstanding
(iii) Lost Credit Facilities: 100% of the outstanding