International Journal of Emerging Markets

Market structure, efficiency and profitability of insurance companies in Ghana

Abdul Latif Alhassan George Addison Michael Effah Asamoah
Article information:
To cite this document:
Abdul Latif Alhassan George Addison Michael Effah Asamoah , (2015),"Market structure, efficiency and profitability of
insurance companies in Ghana", International Journal of Emerging Markets, Vol. 10 Iss 4 pp. -
Permanent link to this document:
http://dx.doi.org/10.1108/IJoEM-06-2014-0173
Downloaded on: 16 August 2015, At: 12:08 (PT)
References: this document contains references to 0 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 26 times since 2015*
Users who downloaded this article also downloaded:
Solomon W. Giorgis Sahile, Daniel Kipkirong Tarus, Thomas Cheruiyot, (2015),"Market structure-performance hypothesis in
Kenyan banking industry", International Journal of Emerging Markets, Vol. 10 Iss 4 pp. -
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Access to this document was granted through an Emerald subscription provided by emerald-srm:463575 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

 Market Structure, Efficiency and Profitability of Insurance
Companies in Ghana
1.0 Introduction
Following the separation of life insurance business from non-life insurance business in Ghana with
the passing of the Insurance Act 724 in 2006, most of the insurance firms maintained their non-life
businesses in the original companies to set up new companies to manage the life businesses. The
insurance market has since been opened to keen competition among players with the enactment of
the law and attracted new entrants bringing the total number of industry players to forty-three (43)
from twenty-five (25) licensed firms in 2006. Five out of the eighteen (18) life insurance companies
control about 75% of gross premiums over the last six years up to the period ending 2011. In the
non-life insurance market, concentration is still among five (5) out of the twenty-five (25) firms. It
is, however, interesting to note that the other thirteen (13) companies in the life insurance and
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

twenty (20) in the non-life insurance business that control at most 25% of the market have been
gaining some points in their share of the market. The industry has also been experiencing growth in
premium mobilization above the growth of the non-life insurance business (FINSSP II, 2012).
These changes would have an impact on the structure of the industry and reflect a growing
insurance industry which has the benefit of accumulating long term funds for access by financial
agents for investment purpose to propel economic growth.

The pricing policy of insurance companies reflects assumed risks with high risk clients charged high
premiums. Low risk clients being charged low premiums to reflect the risk they bring into an
insurance pool. However, the ability of the insurers to correctly classify their clients depends on the
factors that interplay in the marketplace with the impact of market structure on firm pricing
behaviour having received considerable attention in empirical literature studies from an industrial
organization perspective. The traditional structure conduct hypothesis (SCP) hypothesis of Bain
(1951) and made popular by Baumol (1982) posits that a firm’s pricing and conduct are determined
by the structural features of the market within which it operates and that a highly concentrated
industry is characterized by collusion among the few large firms in setting prices to achieve
abnormal profits. This theoretical view of the market structure hypothesis was subsequently
challenged by the efficient structure (ES) hypothesis of Demsetz (1973) and Peltzman (1977).
According to the ES hypothesis, efficient producing firms generate higher sales through lower
pricing. This would result in higher market share for efficient firms leading to concentration. An
understanding of the interrelationships between market structure, efficiency and profitability is
therefore necessary to inform both competition and regulatory policies.

1.1 Motivation for the Study

Since effective regulation requires realistic understanding of consumer and firm behaviour, there is a
need for the insurance regulator to better understand the dynamics of pricing behaviour among
firms in the industry to formulate policies for a stable industry. With different empirical studies
having found mixed results regarding the relationship between market structure and the pricing
behaviour of firms, this study seeks to extend regulators understanding of firm behaviour in
emerging markets. The coming into effect of the Insurance Act 724 (2006) coincided with the
dwindling in market share of the top five insurers in both life and non-life market. This study seeks
to find answers to the following questions, “How has the structure of the new markets (life and
 non-life markets) impacted on firms pricing behaviour? While several studies have examined
insurance markets in Europe and America,1 very few studies have been carried out in African
insurance markets2 with only three empirical studies3 having been carried out on the insurance
market in Ghana. While Ansah-Adu et al. (2012) focused on cost efficiencies of both life and non-
life insurance companies, Akotey et al. (2013) examined the financial performance of life insurers in
Ghana. The obvious omission from these two studies is the link between insurer’s efficiency and
profitability. Additionally, the effect of the structure of the emerging Ghanaian insurance market on
profitability has not been empirically investigated. This study enhances the understanding of the
industrial economics of the Ghanaian insurance industry by investigating the impact of market
structure and efficiency on profitability. The main contribution of this study is the empirical
application of the classical industrial economics theory on SCP and ES paradigms to an emerging
insurance market in Africa. This study thus extends the work of Akotey et al. (2013) by examining
the effect of market structure on profitability of both life and non-life insurers. With regards to
Ansah-Adu et al. (2012), we further explore the relationship between efficiency and profitability.
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

The rest of the paper is organized into section 2 on the overview of both life and non-life insurance
markets in Ghana, Section 3 on related studies on the SCP and ES paradigms, section 4 on
methodology and section 5 empirical findings. Section 6 concludes the study and discusses the
policy implication of the findings.

2.0 Overview of Insurance Industry in Ghana

Insurance, as a form of business, first began in the Gold Coast way before attainment of
independence. The premier insurance company, the Royal Guardian Enterprise (now operates
under the name Enterprise Insurance Company Limited) was established in 1924. Other insurance
companies that were later established included the Gold Coast Insurance Company, General
Insurance Company, and Cooperative Insurance Society in 1955, 1957 and 1958 respectively. Within
five years of independence, the first national insurance company, the State Insurance Company
(SIC), came to existence in 1962 through a merger of Gold Coast Insurance Company and
Cooperative Insurance Society. As at the end of the year 2013, the insurance industry was made up
of 25 Non-Life companies, 18 Life companies, 2 Reinsurance companies, 54 Broking companies, 1
Loss Adjuster, 1 Reinsurance Broker, 1 Oil and Gas company, and about 4000 insurance agents
(www.nicgh.org) regulated by the by the National Insurance commission. The National Insurance
commission came to operation when the then Insurance Law 1989 (PNDC Law 227) was
promulgated, however, the industry is now regulated by the Insurance Act, 2006 (Act 724). The Act
724 mandates the National Insurance Commission with the supervising, regulating and controlling
the business of Insurance in Ghana. The minimum start-up capital of US$1million for an insurance
company was reviewed to US$5million or in its cedi equivalent by the Insurance Act, 2006 (Act
724). This was done to improve upon the financial capacity of insurance companies to enable them
underwrite more risky businesses especially with the discovery of oil in commercial quantities (NIC,
2010).

1 Chidambaran et al. (1997), Choi and Weiss (2005), Jedlicka and Jumah (2006), Berry-Stölzle et al.(2011)
2 Jones et al. (1999), Liebenberg (2000), Theron (2001) and Liebenberg and Kamerschen (2008) have all examined the
South African Insurance market
3 Ansah-Adu et al. (2012), Akotey et al. (2013) and Akotey and Abor (2013)
 Table 2.1: Growth in Total Gross Premium Income (Non-Life and Life) (GHS)
Non-Life Insurance Life Insurance
2007 142,020,077 67,534,641
2008 187,010,274 91,245,062
2009 220,704,263 122,269,456
2010 270,773,967 187,343,779
2011 358,352,702 270,176,073
Source: National Insurance Commission (NIC) Report, 2010; 2011

Table 2.1 presents the gross premiums written for both the non-life and the life markets from 2007
to 2011. It is observed that the industry has experienced consistent growth in premiums throughout
the period. The gross premiums written for the non-life and the life businesses increased to
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

GH¢358,352,702 and GH¢270,176,073 in 2011 from GH¢142,020,077 and GH¢67,534,641 in 2007

respectively. The yearly growth rates of the premiums in both markets are presented in Figure 2.1. It
is evident that the premium growth rate of the life markets has been consistently greater than the
non-life market. This shows that life insurance businesses has made it the leading contributor to the
industry’s total premium income than the motor business which was hitherto the driving force
behind the insurance business in Ghana (NIC, 2010; NIC, 2011). It is worth noting that life
insurance in Ghana hitherto was equivalent to insurance against death. However, this phenomenon
has changed, especially when the life insurance was separated from general business. Life insurance
now includes endowment policies, funeral and others. (http://www.nicgh.org)

Figure 2.1: Growth Rate in Premium Income

60.00%
53.20%
50.00% 44.20%

40.00% 36.10% 35.10% 34% 32.30%

31.60%
30.00% Non-Life
23.90% 22.60%
18% Life
20.00%

10.00%

0.00%
2006 - 2007 2007 - 2008 2008 - 2009 2009 - 2010 2010 - 2011

Source: NIC, 2006 to 2011

2.1 Life Insurance Market

Life insurance is mostly deemed a contract entered into between a policy holder and an insurer in
which the insurer is expected to pay a pre-determined sum of money upon the death of the insured
person to a person so designated by the policy holder in the contract. Other forms could include
pension, savings and health insurance. In cases of pension and some life insurance, the policy holder
may be the same as the beneficiary. The birth of the life insurance business in Ghana came at the
 passage of the Insurance Act (2006), Act 724. Up until the passage of Act 724, there was no clear
distinction between life and non-life insurance businesses as some companies were known to use
premiums from life to support non-life operations. However, with the passage of Act 724, a
separation of the two was a must for every insurance company. The business of insurance before
the Act was mainly the type of policy under taken. These were mainly life, fire, marine and
miscellaneous. Aside the aforementioned powers of Act 724, the act also forbids composite
insurance companies. Thus by December, 2008, all composite insurance companies were expected
to have separated their life and non-life operations into different companies. This distinction was
eminent to meet the fast growing changes in Insurance regulation and standards and to conform to
global methodologies as adopted by The International Association of Insurance Supervisors (IAIS),
the global head of insurance business. For several years, there have been a phenomenal growth in
life business when compared to non-life business. This is likely to continue given the rate of product
innovation and development. Market share seems to be concentrated in the life industry. In the
sector, close to 78% of the total life insurance premium goes to only five out of the 18 life
companies that existed as of 2011. The five are State Insurance Company (SIC life), Glico life,
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Enterprise life, Star life and Metropolitan life insurance. Between these five companies, they
accounted for 77.7% of total premiums in 2007, 76.9% in 2008, 76.4% in 2009, 73.7% in 2010 and
77.5% in 2011. Figure 2.2 below illustrates the market share among these top five life insurance
companies.

Figure 2.2: Premium Distribution in Life Insurance (Top 5 firms)

35.0

30.0

25.0
SIC
Percentages

20.0 Glico

15.0 ELAC
Star
10.0
MET Life
5.0

0.0
2007 2008 2009 2010 2011

Source: NIC, 2006 to 2011

2.2 Non-Life Insurance Market
According to NIC (2011), non-life insurance has always contributed a higher premium to total
insurance premium than that of life insurance. The growth in non-life premiums is largely driven by
premiums form motor business, followed by accidents, then fire, marine and the others. For motor,
annual gross premium recorded was GH¢123.9 million for 2010 from a previous year’s figure of
GH¢105.6 million and landed at GH¢162.4 representing 25.8% of the entire premiums for the year
2011. Premiums from accident business for 2011 was GH¢78.9 million denoting a market share of
12.6%. This increment was on the back of GH¢71.8 million for 2010. The figure for 2010 was an
increase of nearly GH¢23 million above the 2009 figure. Fire insurance made an annual income of
GH¢77.7 million and market share of 12.4% for the period ending 2011. It is believed that the
 compulsory insurance of all commercial buildings which was to start on August 1, 2011 will help
boost the income for this line of business (NIC, 2011; www.ghanabusinessnews.com). Most of the
market premiums are underwritten by State Insurance Company Limited (SIC), Enterprise
Insurance Company Limited (EIC), Metropolitan Insurance Company Limited (MET), Vanguard
Assurance Company Limited and Star Assurance Company limited; this accounted for 72.69% of
gross premiums in 2007, 68.03% in 2008, 66.25% in 2009, 63.13% in 20010 and 59.98% in 2011.
The entrant of six new firms between 2006 and 2009 led to a decline in the market share of the top
5 insurers (NIC, 2011).

Figure 2.3: Premium Distribution in Non-Life Insurance (Top 5 firms)

40.00

35.00

30.00
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

SIC
Percentages

25.00
EIC
20.00
MET
15.00 Vanguard
10.00 Star

5.00

0.00
2007 2008 2009 2010 2011

Source: NIC, 2006 to 2011

3.0 Empirical Studies on Market Power and Efficient Structure Hypotheses

Most empirical literature on market power hypothesis have mainly focused on banking institutions
in both developed and developing countries, majority of the studies have been on USA and
European banks with incongruous results. Most recent of these include Tregenna (2009), Maudos
and Fernandez de Guevara (2007). Studies across other regions include Pakistan (Farooq, 2003),
Australia (Sathye, 2005), Bangladesh (Samad, 2008) and Korea (Park and Weber, 2006), UAE (Al-
Muharrami and Matthews, 2009; Alzaidanin, 2003). In Africa, Mensi and Zouari (2010) focused on
the Tunisian banking system. Other industries examined under SCP include agriculture (Marion,
1976), pulp and paper (Li et al., 2004), steel (Purkayastha, 2005), tourism (Tung et al., 2009) as well
as inter-industry (Lehmann et al, 2004).

Fewer studies have examined the insurance industry. One of the first researchers in the insurance
industry that studied S-C-P hypothesis was Joskow (1973). In examining the U.S. non-life insurance
industry competitive structure, he found that despite competitive market structure, insurers set
prices through cartel-like rating bureaus. The author concludes that the collusive behaviour among
insurers lead to the cuts in supply, inefficient sales systems and over-capitalized industry.

Chidambaran et al. (1997) empirically analysed the economic performance across 18 different lines
of the U.S. property and liability insurance industry covering a 10-year period from 1984 to1993.
 The main was that profitability was driven by market concentration to the SCP hypothesis.
Bajtelsmit and Bouzouita (1998) also analysed the structure-conduct-performance and efficient
structure hypotheses in the automobile market in US from 1984 to 1992, the authors found no
evidence in support of the efficient structure hypothesis. In extending the work of Chidambaran et
al. (1997) using a different dataset on the US property and liability insurance market, Choi and Weiss
(2005) examined the market structure, efficiency, and performance over the period 1992–1998 using
data at the company. By estimating both revenue and cost efficiencies, they found evidence in
support of the ES hypothesis to indicate that cost-efficient firms charge lower prices and earn
higher profits. Taking motivation of regulatory differences in different state in the US, Weiss and
Choi (2008) examined the structure–conduct–performance (SCP), relative market power (RMP),
and efficient structure (ES) hypotheses. They find evidence to that market power is mostly exercised
by insurers in non-stringently regulated but competitive markets.

Meanwhile, Jedlicka and Jumah (2006) tested the SCP hypothesis on Austrian insurance industry by
considering a sample of 52 insurers between 2002 and 2003. Although they find evidence to suggest
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

the insurance market is concentrated, the SCP hypothesis of collusive firm’s behvaiour was rejected.
Pope and Ma (2008) examined the SCP hypothesis on life insurance market from an international
perspective by employing a multiple regression analysis on a panel of 23 countries from 1996 to
2003 and found evidence in support of SCP hypothesis although they concluded that the effect of
market concentration on performance depended on the level of market liberalisation.

By employing data from 1980 to 2000, Liebenberg and Kamerschen (2008) examined South African
auto insurance market and found no support for S-C-P hypothesis. In an examination of the effect
of liberalisation on market structure and performance in the non-life insurance industry in Eastern
European countries, Njegomir and Stojic (2011) employed a panel data from 2004 to 2008 to
provide evidence in support of the SCP. With motivation from restructuring and consolidation of
the European insurance markets, Berry-Stölzle et al. (2011) studied the SCP, RMP and ES
hypotheses in the property and liability insurance industry in 12 European countries from 2003 to
2007. They find strong support for the efficient structure hypothesis while the SCP was rejected.
The findings of the above studies provided strong evidence in support of the ES hypothesis. This
could be explained in terms of anti-competitive legislation that promotes easy entry and exits which
serves as a check on the collusive behaviour in concentrated markets in these developed countries as
per Baumol (1982) contestable market theory. However in less than developed financial markets in
developing economies like Ghana, the market imperfections and high levels of concentrations
makes collusive behaviour very likely. To this end, we seek to provide empirical evidence on the
SCP and ES hypotheses in less developing insurance market in Africa.

In examining profitability determinants of life insurers in Ghana, Akotey et al. (2013) employed a
panel of ten life insurers over a 10 year period to identify gross premium written, insurers size,
reinsurance, claims, management expenses and interest rate as the significant determinants of life
insurer’s profitability in Ghana. The authors however, did not consider the effect of the structure of
the life insurance industry on profitability. Based on the mixed results for the SCP hypothesis in
empirical studies reviewed, this study seeks to bring new evidence on the SCP hypothesis from a
developing insurance market undergoing structural changes. In this study, we employ the DEA
technique of Charnes et al. (1974) and Banker et al. (1984) to estimate both technical and pure
technical efficiency scores for both non-life and life insurers while using the structural measures of
market structure in the form of the Herfindahl index and 4-firm concentration ratio. Unlike
Liebenberg and Kamerschen (2008), we bring robust evidence to our findings by considering other
 known correlates of profitability in leverage, size, underwriting risk, GDP growth and inflation.
Additionally, this study employs two different estimations in the panel corrected standard errors of
Beck and Katz (1995) and the random effects estimation to provide robust evidence on the
relationship between profitability and market structure and efficiency of both life and non-life
insurance market in Ghana.

4.0 Methodology
This section describes the empirical strategy employed to examine the relationship between market
structure, efficiency and profitability of Ghanaian insurance companies. We first describe our choice
of market structure proxies and how they are measured. The data envelopment analysis (DEA)
technique is employed to measure insurance efficiency. This study employed a panel dataset which
covers 14 life and 22 non-life insurance companies in Ghana covering the period 2007 to 2011. This
sample was taken from the 18 life and 24 non-life insurance firms in operation during the study
period. The 14 life insurers account for about 95% of premium market share while the 22 non-life
insurers also account for about 93% of premium market share. The data is contained in annual
financial report of the sample as obtained from the National Insurance Commission (NIC), the
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

regulator of the insurance industry.

4.1 Measures of Market Structure

Several metric measures have been provided by literature on how to measure the market structure of
an industry. Among these measures are the Herfindahl Hirschman Index (henceforth HHI), the
Lerner index and the concentration ratios. For the purpose of the study, both the HHI and
concentration ratio (4-firm CR) were used to proxy the structure of the Ghanaian life insurance
market. The HHI is measured as the sum of squares of the market share of firms with an industry.

This was computed by the formulae below;

 =  


Where  is the market share of insurer i. The market share is proxied as the ratio an insurer’s
premium to gross industry premium. The concentration ratios, which is common indicator for the
market structure measures the degree to which few dominant firms within an industry account for
greater portions of economic activities within the market. This study employs the 4-firm
concentration ratio (CR4) as it has been found to be the most preferred concentration ratios
(Scherer and Ross, 1990).

4.2 Estimating Efficiency: Data Envelopment Analysis

The data envelopment analysis (DEA) technique was employed in estimating efficiency. The DEA
technique measures the relative performance of firms by comparing multiple inputs and outputs.
The efficiency score is estimated as the ratio of the weighted sum of outputs to weighted sum of
inputs. Assuming n number of decision making units (DMU’s), with m inputs and s outputs, we
obtain the relative efficiency score of a test DMU p through the estimation of a model proposed by
Farrell (1957) and made popular Charnes et al. (1978) described below;
 

max ℎ =   


 :   



 

   −    ≤ 0,  = 1,2, … ( (1)
 

 ,  ≥ 0 , = 1,2, …  - = 1,2, … 

where xij is the quantity of input i used by bank j,  is the quantity of output r produced by bank j,
and  and  refer to weights chosen for output r and input i respectively

min ℎ = 0
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

 : 0  −  1  ≥ 0



∑3 1   ≥ 4

1 >0

The input oriented model above seeks to minimize cost in achieving a desired level of output. An
efficient DMU has an efficiency score 0, of one (1) and serves as the bench mark for the DMU’s
within in an industry and employing the same technology. The modeled linear programming
assumes constant returns to scale as per Charnes et al. (1974) which implies that each DMU operate
at their optimal scale and that any increase in inputs will result in proportional increase in outputs.
The efficiency scores estimated under the assumption of constant returns to scale is called technical
efficiency (TE) which denotes the ability of firms to employ technology to maximize out. However,
when inputs changes results in disproportional changes in the output variables, the DMU’s are said
to be operating at variable returns to scale, which Banker et al. (1984) describes as pure technical
efficiency (PTE).

4.2.1 Input and Output Variables

Based on three principal services provided by insurers in the form of real services, risk pooling and
risk bearing and intermediation functions, there is a general consensus on what constitute input
variables in efficiency studies. The input variables are classified into labour input and capital inputs.
While labour inputs are classified into business services input and labour cost, equity capital and
debt capital make up capital inputs. Our choices of input and outputs variables employed were
influenced by data availability. In line with Ansah-Adu et al. (2010) and Al-Amri et al. (2012), the
input variables considered were total operating expenditure and equity capital. Total operating
expenditure proxy for both labour and business services input. Although the issue of insurance
output remains contentions in literature, we follow the arguments of Leverty et al. (2004) and Al-
Amri et al. (2012) to employ net premiums instead of claims incurred as our choice of output. Since
outputs have to be desirable, no insurer seeks to maximize incurred losses. We also employ net
 income after tax as the other output measure. The efficiency scores are estimated under the out-
orientation based on the notion that insurers seek to maximize their earned premiums and profits to
be able to adequately cater for cover for any incurred losses. The descriptive statistics of the input
and output variables are presented in Table 4.14.

Table 4.1: Summary Statistics of Input and Output Variables

PREMIUMS NET PROFIT EXPENDITURE EQUITY
NON-LIFE INSURANCE
Mean 8,302,143 475,539 924,6518 7,675,981
Std. Dev. 11,100,000 14,900,000 20400000 16,300,000
Min 1,682,000 115,000 10,300 199,822
Max 61,600,000 28,700,000 158,000,000 86,400,000
LIFE INSURANCE
Mean 11,000,000 2,625,075 833,8457 14,300,000
Std. Dev. 14,100,000 5,544,575 923,2396 1,7200,000
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Min 29,005 532,2000 18,893 204,457

Max 71,500,000 27,000,000 44,500,000 10,1000,000
Notes: All values in Ghana Cedis

4.3 Empirical Model

The estimated market structure based on the Herfindahl index (HHI) and four firm concentration
ratio (CR-4) in section 4.1 and the efficiency scores under both constant returns to scale (TE) and
variable returns to scale (PTE) are used to model the SCP and efficient structure hypotheses.
Profitability is measured as the ratio of profit after tax to total assets. While the market structure
variables are proxied by the HHI and CR, the efficiency scores estimated in section is used to proxy
for the efficient structure hypothesis. The empirical model to test the hypotheses is presented
below;

Υ, = 6 + 8 9: + 8 ;:
, + 8< =, + >, .......... (2)

Where subscripts i and t denotes insurer and years respectively, Υ is the profitability of insurers
proxied as the return on assets, 9: represents market structure proxied by the Herfindahl index and
4-firm concentration ratio, ;: represents efficiency scores estimated under DEA technique namely
technical efficiency (TE) and pure technical efficiency (PTE), while = denotes list of both firm
specific and macroeconomic determinants of insurers profitability. >, is the disturbance term.

4.3.1 Control Variables

We control for three firm specific variables in size, leverage and risk. Inflation and GDP growth are
also controlled to account for the effect of the macroeconomic environment of insurer’s
profitability. These variables are selected based on the studies of Akotey et al. (2013) in examining
profitability determinants in the Ghanaian life insurance industry. A brief description of the
variables are given below;

Size: This variable is measured as the natural logarithm of total assets. The relationship between size
and profitability is ambiguous. A positive relationship indicates the benefits of economies of scale
advantages while a negative relationship is attributed to diseconomies of scale.

4 Firms with negative values were excluded from the efficiency analysis.
 Risk: Insurers risk is proxied by the ratio of incurred losses to earned premiums. This measure
captures the uncertainty that premiums earned may not be enough to cover for losses incurred
under the policies underwritten. We hypothesize that insurers with high underwriting risk would be
less profitable, hence a negative relationship with profitability is expected.

Leverage: Insurer generate leverage from unearned premiums from unexpired policies and any
outstanding claim amount. The quality of investments of from the leverage assumed determines its
impact on profitability. Hence, the relationship between leverage and insurance becomes positive
through sound investment choices that generate revenues above expected losses. Bad investments
decisions leads to a negative leverage-profitability relationship. While Studies like Adams (1996),
Adams and Buckle (2003) and Akotey et al. (2013) have found evidence in support of a positive
relationship, Malik (2011) and Ahmed et al. (2011) find an inverse leverage-profitability relationship.

GDP Growth: The confidence of every business environment in reflected in the growth in the real
economy as captured by the gross domestic product. Increasing growth stimulates demand for
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

services such insurance cover by businesses to provide cover their expanding businesses. The
growth would also lead to increasing demand by consumers for goods and services. In this line, we
expect GDP growth to have a positive relationship with profitability through growth in premium
income.

Inflation: Increases in interest rate arising from high inflationary pressures means that returns on
investments also increases. Hence, we expect inflation to have a positive effect on insurer’s
profitability due to high investment yields. However, a negative relationship can be experienced
when incurred losses exceeds any gains in investment returns. Additionally, the negative effect of
inflation on real incomes means that increasing inflation would result in reduced sale by insurance
companies.

Equation 3 is thus expanded to form equation 3 as described below;

?@A, = 6 + 8 9B + 8 ;
, + 8< :C;, + 8D E;F, + 8G ?:H, + 8I JKB + 8L MN +

 + , .......... (3)

Where ?@A, is the return on assets for insurer i in year t. Return on assets is defined as the ratio of
profit after tax to total assets.  denotes the firm specific-fixed effects whiles , represents the
firm specific unobservable effects which vary over time. The measurements of the independent
variables are presented Table 4.2.

Table 4.2: Variables Description

Variable Measurement Expected Signs
Return on Assets Profit after Tax to Total Assets Ratio
Independent variables
Market Power HHI/CR4 +/-
Efficiency Estimated using the DEA (TE, PTE) +
Size Natural logarithm of Total Assets +/-
Leverage Liabilities to Assets ratio +/-
Risk Incurred Losses to Premiums ratio -
GDP GDP growth rate +
Inflation Inflation Rate -
 4.4 Estimation Procedure

The panel data methodology employed in this study enables a better identification and measurement
of phenomenon that are not captured in either cross-sectional or time series models. This allows for
the modelling and testing of complex behavioural models (Baltagi, 2001). The structure of the error
terms in panel data modeling makes the ordinary least estimations an inefficient and biased
estimator of panel models. The time-series increases the likelihood of correlation between the error
terms of different periods. Additionally, the assumption of homoskedasticity does not always hold
for panel data models. In estimating panel’s models therefore, Beck and Katz (1995, 2011) argued
that by considering the presence of non-spherical error terms (the presence of autocorrelation or
heteroskedasticity), the OLS can be an efficient estimation technique for estimation of panel data
models. Accordingly, the authors proposed the ordinary least squares panel-corrected standard
errors (OLS-PCSE) to replace the OLS standard errors in an OLS estimation, making the OLS
robust to the presence of non-spherical error terms. This study therefore employs the OLS-PCSE
after the providing evidence of heteroskedasticity and autocorrelation. Panel data models are
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

traditionally estimated using the fixed or random effects. The choice of either is influenced by the
Hausman (1978) test under the null hypothesis of random effects model. We test the robustness of
the OLS-PCSE estimation by employing the Hausman test to select between the fixed effects (FE)
and random effects (RE) estimations.

5.0 Results and Discussion of Findings

Presented in Table 5.1 is the indicators for the structure of both non-life and life markets in the CR
and the HHI. Consistently across all the study period, the life markets is highly concentrated than
the life markets. This probably reflected by the high number of firms in the non-life market as
opposed to fewer firms in the life market.
Table 5.1: Indicators of Market Structure
Non-Life Life
Years CR4 HHI CR4 HHI
2007 0.6700 0.1723 0.6956 0.1554
2008 0.5800 0.1306 0.6766 0.1478
2009 0.5900 0.1126 0.6665 0.1377
2010 0.5400 0.1054 0.6538 0.1386
2011 0.5100 0.0985 0.7086 0.1570
Source: Authors’ Estimation from NIC Data, 2007 to 2011

The results of the efficiency scores estimated using the Data envelopment analysis technique under
intermediation approach assuming both constant returns to scale (Technical efficiency) of Charnes
et al. (1978) and variable returns to scale (Pure Technical efficiency) of Banker et al. (1984) is
presented in Table 5.2. Across both technical and pure technical efficiency scores, life insurers have
exhibited higher efficiency scores compared to non-life insurers implying that life insurers in Ghana
are more efficient in resource utilization compared to their non-life counterparts. This findings
support the findings of Ansah-Adu et al. (2012) who examined the cost efficiency of insurance
companies in Ghana.
 Table 5.2 Efficiency of Ghanaian Insurers (2007 to 2011)
NON-LIFE LIFE
TE PTE TE PTE
2007 0.290 0.570 0.7337 0.8780
2008 0.778 0.885 0.8051 0.8908
2009 0.773 0.857 0.7381 0.8795
2010 0.617 0.779 0.6948 0.8458
2011 0.692 0.759 0.7582 0.8453
Source: Authors’ Computation from EMS

As shown in Table 5.2, obvious differences exist among the efficiency scores for both life and non-
life insurers. Table 5.3 presents the results of both Kruskal-Walis and Man Whitney U tests for
differences mean values for both technical and pure technical efficiencies. As indicated, significant
differences was found between the technical efficiency scores for life and non-life insurers at 10%
but no significant differences was found for pure technical efficiency. This indicates that life
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

insurance firms in Ghana have the ability to better maximize their outputs from the production
inputs compared to insurers in the non-life market.

Table 5.3: Test for Mean Differences

TE PTE
LIFE 0.752 0.871
NON-LIFE 0.651 0.782
Kruskal-Walis O 2.995* 1.686
Man-Whitney U test: z 1.736* 1.34
* indicates significance at 10%

5.1 Descriptive Statistics

Table 5.4 presents a summary statistics of the regression model variables. Insurer’s profitability,
earnings before interest and taxation is on average 6% of total assets for life insurers compared to
3.3% for non-life insurers over the study period. Over the period, the concentration ratios for the
four largest insurers (CR4) were 57.8% and 67% for non-life and life insurers respectively indicating
a degree of market concentration. The Herfindahl Index as a measure of market structure confirms
that the life insurance market is more concentrated compared to the non-life market. On the
efficiency scores, both pure technical efficiency (PTE) and technical efficiency (TE) for life insurers
were higher than that of non-life insurers implying that Ghanaian life insurers are more efficient
compared to non-life insurers. Insurer’s size, leverage ratio and underwriting risk were 15.977%,
60.2%, 18.1% respectively for non-life insurers with the corresponding ratios for life insurers of
16.2%, 33% and 36% respectively. On the macroeconomic variables, inflation rate and GDP growth
rate averaged 13.188% and 8.256% respectively over the study period.
 5.2 Correlation Matrix
In examining the possibility of high correlation among the independent variables, the Pearson
correlation coefficient was estimated and the results is presented in Table 5.5. There is strong and
significant correlation between both market structure measures and efficiency scores for both life
and non-life insurers. With regard to the correlations between the independent variables to test for
the presence of multicollinearity, correlation coefficients of less than 0.5 was found among pairs of
the independent variables giving an indication that potential multicollinearity problems will not be
encountered in using all the independent variables in the regression equation.

Table 5.4: Summary Statistics

NON-LIFE Mean Std Dev. Min Max
ROA 0.033 0.0847 -0.199 0.177
HHI 0.124 0.027 0.098 0.172
CR4 0.578 0.054 0.510 0.670
PTE 0.782 0.248 0.173 1.000
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

TE 0.651 0.281 0.117 1.000

SIZE 15.977 1.085 13.951 18.822
LEV 0.765 0.214 0.216 0.982
RISK 0.181 0.279 -0.163 2.261
GDP 8.256 3.454 3.990 14.390
INFL 13.188 4.015 8.730 19.250
LIFE INSURANCE
ROA 0.06 0.16 -0.43 0.47
HHI 0.14 0.01 0.10 0.16
CR4 0.67 0.04 0.51 0.71
PTE 0.87 0.15 0.51 1.00
TE 0.75 0.18 0.38 1.00
SIZE 16.28 1.16 13.30 18.71
LEV 0.67 0.22 0.23 0.77
RISK 0.36 0.18 0.02 0.77
Note: ROA=Return on Assets, HHI=Herfindahl Hirschman Index, CR4=4 firm Concentration ratio, PTE= Pure Technical
Efficiency, TE=Technical efficiency, SIZE=Insurers Size, Risk=Underwriting Risk, LEV= Leverage, GDP= Economic
Growth Rate, INF=Inflation Rate
 Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Table 5.5: Correlation Matrix

ROA HHI CR4 PTE TE SIZE LEV RISK GDP INFL
NON-LIFE INSURANCE
ROA 1
HHI 0.002 1
CR4 0.07 -0.85*** 1
PTE 0.28*** -0.22** 0.07 1
TE 0.40*** -0.39*** 0.24** 0.72*** 1
SIZE -0.08 -0.23** 0.25** -0.01 0.001 1
LEV 0.54*** 0.22** -0.09 0.04 0.11 -0.11 1
RISK -0.46*** -0.14 0.16 -0.11 -0.19* 0.28** -0.01 1
GDP 0.14 -0.36*** 0.20*** -0.12 -0.02 0.20* 0.08 0.14 1
INFL -0.245** 0.08 -0.40*** 0.27** 0.30*** -0.15 -0.16 -0.04 -0.393*** 1
LIFE INSURANCE
ROA 1
HHI -0.09 1
CR4 -0.15 0.96*** 1
PTE 0.57*** -0.17 -0.2 1
TE 0.70*** -0.15 -0.19 0.63*** 1
SIZE 0.36*** 0.17 0.19 -0.07 0.19 1
LEV 0.01 -0.15 -0.17 0.29** -0.03 -0.2 1
RISK -0.42*** 0.03 0.06 -0.30** -0.25** 0.22* -0.32***
GDP 0.18 0.34*** 0.42*** -0.1 0.01 0.22* -0.1 -0.12 1
INFL -0.25** -0.26** -0.08 0.07 0.02 -0.15 0.05 0.09 -0.41*** 1
Note: ROA=Return on Assets, HHI=Herfindahl Hirschman Index, CR4=4 firm Concentration ratio, PTE= Pure Technical Efficiency, TE=Technical efficiency,
SIZE=Insurers Size, Risk=Underwriting Risk, LEV= Leverage, GDP= Economic Growth Rate, INF=Inflation Rate. ‘***’, ‘**’, and ‘*’ indicate significance at the 1%,
5%, and 10% levels respectively.
 5.3 Regression Results
As discussed earlier, the OLS-PCSE estimation provides consistent regression estimates after
correcting for either heteroskedasticity or autocorrelation or both. The Breusch and Pagan (1979)
test for heteroskedasticity and Wooldridge (2002) test for autocorrelation were employed. In
providing justification for the use of PCSE, evidence was provided for the presence of
heteroskedasticity for both non-life and life samples but autocorrelation was only present in the life
sample. The reliability and validity of the regression results are indicated by the Wald O which tests
the joint significance of the independent variables in explaining the variations in profitability. With
all the probabilities values of the O test (Prob >O ) less than 0.05, we conclude that the model
variables significantly explain profitability of insurance companies in Ghana. Additionally, the R-
squared indicates that on the average, about 60% of variations in profitability of non-life insurers are
explained by the model whereas for the life insurance market, about 70% variations in profitability
are explained by the model variables. The results of these tests are provided in Table 5.6.
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

From Table 5.6, the SCP hypothesis was rejected in all estimations using the two different measures
of market structure. In the non-life insurance model, inconsistent evidence was found for the SCP
hypothesis. Using the HHI as a proxy for market power, a significant positive relationship was
found at 10% whereas a negative and significant relationship at 10% when CR4 proxied for
insurance market structure. This results clearly does not provide strong evidence for the SCP
hypothesis in the non-life insurance market in Ghana. The tendency for firms to follow market
leaders in their pricing conduct could explain the conflicting evidence for the SCP; such behaviour
characterises oligopolistic markets where market leaders compete against each other. The proxies
for efficient structure (PTE and TE) exhibited significant positive relationships with return on assets
at 1% to provide support for the efficient structure hypothesis. This implies that efficient non-life
insurance firms produce at a lower prices to drive up sales and market share, hence higher
profitability. This results is consistent with the findings of Liebenberg and Kamerschen (2008) who
found mixed results5 for the SCP hypothesis in the South African Auto insurance market.

With regard to the life model, the coefficients of market power (HHI and CR4) are negative and
statistically significant at between 5% and 1%. This indicates that a competitive life insurance
industry leads to higher profitability implying a rejection of the SCP hypothesis in the Ghanaian life
insurance industry and denotes the absence of collusive behaviour in the non-life market. This result
is also consistent with the findings of Choi and Weiss (2005), Jedlicka and Adusei (2006) and Stölzle
et al. (2011). Evidence was also found for the Efficient Structure (ES) hypothesis as indicated by the
positive significant coefficient of technical efficiency and pure technical efficiency at 1%. This
implies that profitability of life insurers in Ghana thus results from efficient firms being able to
lower per unit cost to sell more products to capture larger market.

For the control variables, size, underwriting risk, leverage, and inflation exhibited significant
relationships with profitability of non-life insurers in Ghana whereas underwriting risk, leverage, and
inflation rate exhibited significant relationship with life insurer’s profitability. For insurer’s size,
larger life insurers exhibited significant positive relationship with profitability which could imply the
benefits of economies of scale enjoyed by larger firms. This results is consistent with Akotey et al.
(2013) on Ghanaian life insurers. This relationship was significant at 1%. However, larger non-life

5 Generally, the SCP hypothesis was rejected for most indicators of market structure except 10 firm concentration ratio.
 insurers were found not able to leverage on their size to earn higher profitability as indicated by
insignificant relationship between size and profitability in the non-life model.

Underwriting risk which indicates the riskiness of the insurance business exhibited significant
negative relationship with return on assets in both non-life and life models at significance of 1% and
5% respectively. As insurers underwrite high risky policies, the likelihood of high claims payout
lowering underwriting profits, hence reduced return on the insurer’s assets all other things being
equal. This effect of risk on profitability for the non-life insurance market was greater than that of
the life insurance market. This indicates that selling more risky insurance business leads to high
claims payment, hence reduced profitability. The non-life market would therefore benefit stringent
regulatory policies that improve underwriting practices. These results are consistent with Akotey et
al. (2013).

Leverage exhibited mixed relationships with profitability for non-life and life models. For non-life
models, leverage exhibited significant positive relationship at 1% implying that highly levered non-
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

life insurers are profitable. For the life model however, a significant negative relationship was
exhibited between leverage and profitability. On the macroeconomic variables, only inflation rate
exhibited significant relationship with profitability for both non-life and life insurers. The
relationship between inflation rate and insurers profitability was negative at a 1% significance level
confirming the findings of Akotey et al. (2013) who reported similar relationship for the life
insurance industry in Ghana.

5.4 Robust Analysis; Random Effects Estimation

In examining the robustness of the results presented in Table 5.7, both the random and fixed effects
estimations were considered. In choosing between the most efficient estimators, the Hausman
(1978) specification test was employed and the results as shown in Table 5.6 favoured the random
effects estimation. With heteroskedasticity and autocorrelations established through the Breusch and
Pagan (1979) test and Wooldridge (2002) respectively, the random effects estimation was carried in
the presence of heteroskedasticity and serial correlations. Consistent with the OLS-PCSE
estimations, the results for the SCP hypothesis was mixed for the non-life sample by exhibiting
significant positive relationship with profitability using HHI as a proxy market structure whilst
exhibiting negative relationship with profitability using concentration ratio to proxy market
structure. These relationships were significant at 10%. For the life model, the SCP hypothesis was
rejected in all four models with the market structure (HHI and CR4) either exhibiting significant or
insignificant negative relationship with profitability. Consistent with the basic estimation, the ES
hypothesis was accepted for both life and non-life samples in the random effects estimations.
 Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Table 5.6: OLS Panel-Corrected Standard Errors Estimation

DEPENDENT VARIABLE: RETURN ON ASSETS
NON-LIFE INSURANCE LIFE INSURANCE
Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4
Constant -0.23 -0.27 -0.04 1.29 -0.85 *** -0.732*** -0.788*** -0.692***
HHI 0.22 2.99 * -1.59 *** -1.131**
CR4 -0.35 -2.50 * -0.489*** -0.318**
PTE 0.45 *** 0.46 *** 0.454*** 0.443***
TE 0.70 *** 0.71 *** 0.392*** 0.384***
SIZE 0.02 0.02 0.02 0.02 0.06*** 0.058*** 0.065*** 0.059***
RISK -0.70 *** -0.57 *** -0.70 *** -0.57 *** -0.42*** -0.366*** -0.431*** -0.372***
LEV 0.40 *** 0.32 *** 0.39 *** 0.32 *** -0.08 ** 0.0001 -0.091** -0.002
GDP 0.00 -0.01 0.002 0.01 -0.0002 -0.003 -0.001 -0.004
INF -0.03 *** -0.05 *** -0.03*** -0.05 *** -0.008 *** -0.009*** -0.007*** -0.009***
R-squared 0.6131 0.6741 0.6133 0.6764 0.7252 0.7126 0.7285 0.7018
Adj. R-squared 0.575 0.642 0.5752 0.6445 0.6926 0.6785 0.6963 0.6664
Wald O (7) 82.37 128.94 82 131.22 100.88 147.81 109.67 137.72
Prob > O 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Hettest 5.28 7.45 5.36 7.39 8.5 15.11 8.71 14.82
Prob > O 0.0215 0.0064 0.0206 0.0066 0.0036 0.0001 0.0032 0.0001
AR(1): 0.0096 0.611 0.075 1.074 10.828 47.081 10.465 47.256
Prob > F 0.7618 0.4475 0.7886 0.3176 0.0059 0.0000 0.0065 0.0000
Insurers 22 22 22 22 14 14 14 14
Observations 79 79 79 79 67 67 67 67
Note: HHI=Herfindahl Hirschman Index, CR4=4 firm Concentration ratio, PTE= Pure Technical Efficiency, TE=Technical efficiency, SIZE=Insurers Size, Risk=Underwriting Risk, LEV= Leverage,
GDP= Economic Growth Rate, INF=Inflation Rate. Hettest=Heteroskedasticity Test, AR (1) =Autocorrelation test, Numbers in parentheses are z-statistics. ‘***’, ‘**’, and ‘*’ indicate significance at the
1%, 5%, and 10% levels respectively. Source: Authors Estimation in STATA12
 Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Table 5.7: Random Effects Estimations with Heteroskedastic and Autocorrelated Error Terms
DEPENDENT VARIABLE: RETURN ON ASSETS
NON-LFE INSURANCE LIFE INSURANCE
Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4
Constant 0.205 0.174 0.254 1.379 * -0.809*** -0.601*** -0.753** -0.569 **
HHI 0.006 2.323* -1.732 * -1.114
CR4 -0.210 -2.019 * -0.560 ** 0.332
PTE 0.367** 0.379 ** 0.434 *** 0.428 ***
TE 0.566*** 0.579 *** 0.370 *** 0.362 ***
SIZE 0.001 -0.003 0.004 0.001 0.063 *** 0.052 *** 0.067 *** 0.054 ***
RISK -0.549*** -0.497*** -0.547*** -0.501*** -0.451*** -0.396 *** -0.453*** -0.400 ***
LEV 0.362 *** 0.304*** 0.360*** 0.301 *** -0.054 -0.011 -0.057 -0.013
GDP -0.004 -0.011 -0.001 0.010 -0.001 -0.004 -0.001 -0.004
INF -0.031*** -0.045*** -0.031*** -0.043 *** -0.008*** -0.010 *** -0.007 *** -0.009 ***

Wald O 2833.29 3103.66 2773.27 3070.12 103.64 185.42 99.1 174.32

Prob > O 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
R-squared 0.608 0.6693 0.6084 0.672 0.668 0.6937 0.6736 0.6923
Adj. R-squared 0.5694 0.6367 0.5698 0.6397 0.6286 0.6574 0.6349 0.6558
Hausman 7.8 6.84 3.10 5.17 1.74 1.32 1.72 1.39
Prob > O 0.1676 0.336 0.3771 0.3956 0.9729 0.9878 0.9738 0.9867
Insurers 22 22 22 22 14 14 14 14
Observations 79 79 79 79 67 67 67 67
Note: HHI=Herfindahl Hirschman Index, CR4=4 firm Concentration ratio, PTE= Pure Technical Efficiency, TE=Technical efficiency, SIZE=Insurers Size, Risk=Underwriting Risk, LEV= Leverage,
GDP= Economic Growth Rate, INF=Inflation Rate. Numbers in parentheses are z- statistics. ‘***’, ‘**’, and ‘*’ indicate significance at the 1%, 5%, and 10% levels respectively. Source: Authors Estimation
in STATA12
 6.0 Conclusions and Recommendations
This paper sought to examine how the structure of both non-life and life insurance markets in
Ghana has impacted on the behaviour of firms with regards to their pricing policies after the
enactment of the Insurance Act 724 (2006) which led to the separation of both life and non-life
businesses. By empirically testing the structure-conduct- performance (SCP) and efficient structure
(ES) hypotheses on the insurance industry in Ghana, the study sought to provide an understanding
of firm behaviour in the new markets created to inform regulatory and competition policies.

Based on our structural measures of market structure, both markets have levels of concentration
with the premiums concentrated among the four biggest insurers. However, the life market is highly
concentrated than the non-life market. From the data envelopment analysis, we also conclude that
the life insurers are more efficient compared to non-life insurers. In the non-life market, we find
inconclusive evidence in support of the SCP hypothesis but strong evidence is alluded for the ES
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

hypothesis that profitability is driven by efficient producing insurers. For the life market, we find
also find strong evidence for the ES hypothesis while rejecting the SCP hypothesis. We also
identified underwriting risk, leverage and inflation as the other significant determinants of
profitability in both markets. Most importantly, the effect of underwriting risk on profitability is
more pronounced in the non-life market compared to the life market. Larger life insurers were also
found to be more profitable than small life insurers. In conclusion, we do not find any evidence of
collusive behaviour in the concentrated insurance markets in Ghana.

The results of this study have implications for the regulation of insurance markets in Ghana and
other developing countries. The rejection of the SCP hypothesis coupled with the high levels of
concentration in both markets indicates that competition policies would not only lead to increased
profitability but also improve consumer welfare. Hence, we recommend that efforts aimed at
promoting competitive industry should be accelerated within a conducive framework to promote
healthy competition among industry players. Management of insurance companies should also aim
to develop manpower and invest in new technology to improve efficiency. This could result in new
product development and effective delivery systems to maximize resource usage. Adherence to the
best underwriting practices should be pursued at both the regulatory level and firm level to reduce
the risk inherent in premium underwriting. This recommendation therefore re-enforces the decision
by the NIC in the adoption of the risk based supervision framework and also in line with the
solution proffered by Akotey et al. (2013). Other significant determinants of profitability identified
could serve a benchmark for insurance companies in improving profitability.

It is the recommendation of this study that future researchers could examine the effect of
competitive insurance market on efficiency of insurers. Other forms of efficiency such as profit and
revenue efficiency as well as different forms of insurance profitability could also be examined in
both markets. Data limitations made it impossible for this current study. Productivity analysis of the
Ghanaian insurance market could also be an area of interest to for further studies. From a
methodological perspective, a dynamic panel estimation can employed to deal with a potential of
endogeneity common with panel data models as well as examine the persistence of insurer’s
profitability in Ghana.
 Acknowledgements

The authors are grateful to the Editor and Area Editor Prof. Ronald Schramm for their comments.
Comments from two anonymous reviewers which greatly improved the earlier draft of the paper is
appreciated. We also appreciated the help of Cosmos Owusu of the National Insurance
Commission in us with the data. This paper was prepared while the corresponding author was a
Teaching Assistant at the Department of Finance at the University of Ghana Business School. An
earlier draft of this paper was presented at the 1st University of Ghana Business School Conference
on Business and Development, April 8-9, Accra, Ghana.

References

Ahmed, N., Ahmed, Z. and Usman, A. (2011), “Determinants of performance: a case of life
insurance sector of Pakistan”, International Research Journal of Finance and Economics, Vol. 61, pp.
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

123-128.
Akotey, J.O. and Sackey, F.G., Amoah, L. and Manso, R.F. (2013) The financial performance of life
insurance companies in Ghana, The Journal Risk Finance, Vol. 14 No. 3, 2013 pp. 286-302.
Akotey, J.O. and Abor, J. (2013), Risk management in the Ghanaian insurance industry, Qualitative
Research in Financial Markets, Vol. 5 No. 1, 2013, pp. 26-42
Al-Muharrami, S. & Matthews, K. (2009). “Market power versus efficient-structure in Arab GCC
banking” Applied Financial Economics, Vol. 19, No. 18, pp. 1487 – 1496.
Alzaidanin, J.S. (2003). “An Investigation of Bank Profitability and Market Concentration in the
United Arab Emirates Financial System”, Bangor Business School Staff Publications &
Working Papers.
Ansah-Adu, K., Andoh, C. and Abor, J. (2012), “Evaluating the cost efficiency of insurance
companies in Ghana”, The Journal of Risk Finance Vol. 13 No. 1, pp. 61-76
Banker, R., Charnes, A. and Cooper, W.W. (1984), “Some models for estimating technical and scale
inefficiencies in data envelopment analysis”, Management Science, Vol. 30, pp. 1078-92.
Baltagi, B.H. (2001). Econometric Analysis of Panel Data, 2nd Ed. Chichester: John
Wiley & Sons.
Bain, J.S. (1951) “Relation of Profit-Rate to Industry Concentration: American Manufacturing”,
1936-1940, Quarterly Journal of Economics, 65: 293-324.
Bajtelsmit, V.L. and Bouzouita, R. (1998), “Market Structure and Performance in Private Passenger
Automobile Insurance”, Journal of Risk and Insurance, Vol. 65, No. 3.
Baumol, W.J. (1982), “Contestable Markets: An Uprising in the Theory of Industry Structure”, The
American Economic Review, Vol. 72, No. 1, pp. 1-15
Beck, N. and Katz, J.N. (1995), “What To Do (and Not To Do) with Time Series Cross-Section
Data”, American Political Science Review 89:634–47
Berger, A.N. and Hannan, T.H. (1998), “The Efficiency Cost of Market Power in the Banking
Industry: A Test of the “Quiet Life” and Related Hypotheses”, Review of Economics and
Statistics, 80: 454-465.
Berry-Stölzlea, T.R., Weiss, M.A. and Wende, S. (2011), Market Structure, Efficiency, and
Performance in the European Property-Liability Insurance Industry, Accessed: 9th August,
2013. American Risk and Insurance Association Meeting, 2011
 Available at:
http://www.aria.org/meetings/2011%20papers/Market_Structure_in_EU_insurance_mark
ets%20072311.pdf
Breusch, T. and Pagan, A. (1979). A Simple Test for Heteroskedasticity and Random Coefficient
Variation. Econometrica, 47, 1287-1294
Charnes, A., Cooper, W.W. and Rhodes, E. (1978), “Measuring the efficiency of decision-making
units”, European Journal of Operations Research, Vol. 2, pp. 429-44.
Choi, B.P. and Weiss, M.A. (2005), “An Empirical Investigation of Market Structure, Efficiency, and
Performance in Property-Liability Insurance”, Journal of Risk and Insurance, 72: 635-673.
Demsetz, H. (1973) Industry Structure, Market Rivalry, and Public Policy, Journal of Law and
Economics, 16: 1-9.
Evanoff, D.D. and Fortier, D.L. (1988), Re-evaluation of the Structure-Conduct-Performance
Paradigm in Banking, Journal of Financial Services Research, 1: 277-294.
Farooq, A.M. (2003). Structure and Performance of Commercial Banks in Pakistan, State Bank of
Pakistan, Munich Personal RePEc Archive Paper No. 4983
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

Hausman, J. A. (1978), “Specification tests in econometrics”, Econometrica 46:6, pages

1251-1271
Jedlicka, L. and Adusei, J. (2006) “The Austrian Insurance Industry: A Structure, Conduct and
Performance Analysis,” Economics Series 189, Institute for Advanced Studies.
Lehman, E., Warning, S. and Weigand, J. (2004) “Governance Structures, Multidimensional
Efficiency and Firm Profitability”. Journal of Management and Governance. Vol. 8, Issue 3, pp
279-304
Leverty, T., Lin, Y. and Zhou, H. (2004), ‘‘Firm performance in the Chinese insurance industry’’,
available at: http://rmictr.gsu.edu/Papers/WP04_10.pdf
Lloyd-Williams, D.M., Molyneux, P. and Thornton, J. (1994), Market Structure and Performance
in Spanish Banking, Journal of Banking and Finance, 18: 433-444.
Liebenberg, A.P. and Kamerschen, D.R. (2008), “Structure, Conduct and Performance Analysis of
the South African Auto Insurance Market: 1980-2000”, South African Journal of Economics,
Vol. 76, No. 2.
Li, H., McCarthy, P. and Urmanbetova, A. (2004), “Industry Consolidation and Price-Cost Margins
- Evidence from the Pulp and Paper Industry”, available from:
http://www.cpbis.gatech.edu/research/findings/Industry
%20Consolidation%20and%20Price-Cost%20Margins.pdf
Malik, H. (2011), “Determinants of insurance companies profitability: an analysis of insurance sector
of Pakistan”, Academic Research International, Vol. 1 No. 3, pp. 315-321
Marion, B. (1976), “Application of the Structure, Conduct, Performance Paradigm to Subsector
Analysis”, N.C. Project 117 Working Paper No. 7.
Maudos, J. and De Guevara, J.F (2007). The Cost of Market Power in Banking: Social Welfare Loss
vs. Cost Inefficiency, Journal of Banking and Finance, Vol. 31, pp. 2103-2125.
Mensi, S. and Zouari, A. (2010), “Efficient Structure versus Market Power: Theories and Empirical
Evidence”, International Journal of Economics and Finance, Vol. 2, No. 4; pages 151-166
Molyneux, P. and Forbes, W. (1995), Market Structure and Performance in European Banking,
Applied Economics, 27: 155-159.
Njegomir, V. and Stojic, D. (2011) Liberalisation and Market Concentration Impact on Performance
of the Non-Life Insurance Industry: The Evidence from Eastern Europe. The Geneva
Papers on Risk and Insurance: Issues and Practice 35, 94-106.
 NIC (2010), National Insurance Commission Annual Report, (Accessed: 10th March, 2013,
Available at:
http://www.nicgh.org/live/images/photos/downloads/NIC_Annual_Report_2010.pdf)

NIC (2011), National Insurance Commission Annual Report, (Accessed: 15th July, 2013. Available
at: http://www.nicgh.org/live/images/photos/downloads/NIC_Annual_Report_2011.pdf)

Park, K.H., and Weber, W. (2006). Profitability of Korean Banks: Test of Market Structure versus
Efficient Structure, Journal of Economics and Business, Vol. 58, pp.222-239.
Peltzman, S. (1977), “The Gains and Losses from Industrial Concentration”, Journal of Law and
Economics 20(2), 229–263.
Purkayastha, S. (2005), “Structure Conduct Performance (SCP) Paradigm and the Indian Steel
Industry: An Analysis”, The Icfai University Journal of Industrial Economics, Vol. 2, No. 3
Pope, N. and Ma, Y-L (2008). The Market Structure–Performance Relationship in
the International Insurance Sector, Journal of Risk and Insurance, Volume 75, Issue 4, Pages
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

947–966,
Rhoades, S.A. (1985), “Market Share as a Source of Market Power: Implications and Some
Evidence”, Journal of Economics and Business, 37: 343-363.
Samad, A. (2008). Market Structure, Conduct and Performance: Evidence from the Bangladesh
Banking Industry, Journal of Asian Economics, Vol.19, pp.181-193.
Sathye, M. (2005). Market Structure and Performance in Australian Banking, Review of Accounting and
Finance, Vol. 4, No2, pp.107-122.
Shepherd, W.G. (1986), On the Core Concepts of Industrial Economics, in: H.W. de Jong, and W.
G. Shepherd eds., Mainstreams in Industrial Organization (Boston: Kluwer Academic).
Tregenna, F. (2009). The fat years: the structure and profitability of the US banking sector in the
pre-crisis period. Cambridge Journal of Economics, 33, pp. 609–632
Tung, G., Lin, C. and Wang, C. (2009), “The Market Structure, Conduct, Performance Paradigm
Reapplied to the International Tourist Hotel Industry”, paper presented at the
International Symposium on Finance and Accounting in Asia, Kuala Lumpur.
Weiss, M.A. and Choi, B.P. (2008), “State regulation and the structure, conduct, efficiency and
performance of US auto insurers”, Journal of Banking and Finance, Vol., 32, Issue 1, Pg 134–156
Wooldridge, J. (2002). Econometric Analysis of Cross-Section and Panel Data. MIT Press, 130, 279,
420-449.
 About the Authors:

Abdul Latif Alhassan is currently pursuing his PhD in Business Administration (Finance) at the
Graduate School of Business, University of Cape Town, South Africa. His research
interests are in Insurance economics, efficiency and productivity analysis, development
finance, econometric modelling and industrial organization theory. He has undertaken ad-
hoc refereeing for the Journal of Risk Finance, International Journal of Emerging Markets
and The Journal of International Trade and Economic Development. He can be contacted
at lateef85@yahoo.com; alhabd004@myuct.ac.za

George Kojo Addisson is the Managing Director of StarLife Insurance Company in Ghana with
over ten years’ experience in the Ghanaian insurance industry. He has an MBA in Finance
at the University of Ghana Business School. He can be contacted at
Downloaded by UNIVERSITY OF EXETER At 12:08 16 August 2015 (PT)

gaddison@starlife.com.gh.

Michael Effah Asamoah is Lecturer at the Central University College in Ghana. He is also a
Chartered Accountant and has an M.Phil in Finance from the University of Ghana
Business School. His research interests are in Foreign Direct Investment and
Macroeconomic uncertainties. He can be contacted at iameffah@gmail.com.