Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
cover.indd 1-2 2/9/09 15:27:02
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
The nature and history of insurance ...........................................................1
Role of government in insurance ................................................................3
Profile of the modern insurance sector ......................................................4
Developm
ent paths for the insurance sector .............................................5
Insurance and development .......................................................................10
Policy issues in insurance market developm
ent ......................................11Industrial structure and m
inimum
capital ..........................................11Financial insurance ................................................................................12Liberalization .........................................................................................13Financial supervisor cooperation or integration .................................14Catastrophe risk .....................................................................................14Accounting and inform
ation systems ...................................................14
Insurance taxation .................................................................................15D
istribution and intermediaries ...........................................................16
References .....................................................................................................17
Annex I: Bank deposits versus insurance and pension assets,
OECD
survey 2001 ...................................................................................19
Annex II: N
on-life insurance and private sector credit growth ............20
Box 1. The global m
etrics of insurance markets .................................4
Table 1. Geographical em
phasis of transnational insurers ................5Table 2. Sequencing of classes of insurance .........................................6
Figure 1. Non-life insurance elasticity ..................................................7
Figure 2. Life insurance elasticity versus GD
P growth ......................8
Figure 3. Minim
um capital requirem
ents for non-life insurers w
orldwide ............................................................................................12
Insurance in its pure form is a social good and in a num
ber of cases can be classified as a public good (that is, it generates desirable externali-ties). Insurance com
panies, mutuals and cooperatives enable individ-
uals and firms to protect them
selves against infrequent but extreme
losses at a cost which is sm
all compared to the feared loss. Th
ey do this through the w
orkings of the law of large num
bers and the central limit
theorem w
hich ensure that a sufficiently large num
ber of reasonably hom
ogenous risks will produce w
ell behaved and highly predictable aggregate results follow
ing a roughly Gaussian loss distribution.
Life insurance contracts can be for short periods (for example,
accidental death) or very long periods (for example, w
hole of life). In consequence life insurance can intrinsically include a savings elem
ent, and in m
any late transition and industrial countries this component
dominates funds flow
s in the sector. This flow
of funds effect can be exaggerated by the fact that in recent decades the life insurance sector has begun to com
pete directly with m
utual funds and unit trusts through unit linked contracts offering a life insurance ‘tax w
rapper’. N
on-life insurance contracts, which insure m
aterial and finan-cial risks, typically run for one year and are renew
ed on the basis of updated risk inform
ation. G
iven its fundamental role in spreading risk it is not surprising
that references to insurance can be found in antiquity. A form
of risk sharing for m
arine ventures known as bottom
ry (not unlike modern
catastrophe bonds in concept) was in existence m
ore than two
millennia ago. N
on-life insurance in its modern form
(also known
as Property and Casualty or P&
C, and general insurance) became
established in Italian cities such as Genoa in the fourteenth century
to support their active marine based trading activities, w
ith the first regulation appearing in 1336. Th
e concept spread rapidly to other parts of Europe and eventually to the N
ew W
orld. Life insurance goes back to at least Rom
an times w
hen funeral societies became popular.
Various forms of insurance have been banned from
time to tim
e for religious reasons (for exam
ple fire insurance was shunned in Southern
Germ
any for a period because it was seen as prevented G
od from
exercising his displeasure) and because of fraudulent or morally ques-
tionable schemes such as tontines and pyram
id arrangements. Th
e grow
th of the sector was m
ost rapid in the United Kingdom
, largely because of a long-standing liberal approach to m
arkets, and the U.K.
remains the m
ost ‘insured’ country in the world H
owever the m
odern approach to the prudential regulation of the insurance sector w
as largely pioneered in the U
nited States in the nineteenth century, and in M
assachusetts in particular.Insurance by its nature is an intangible good, involving paym
ent in advance for an unknow
able quality of delivery in the future. Thus trust
is a critical element, and public good classes such as health, disability
and work place injury or illness have to date often been delivered
through state entities. How
ever most classes of insurance are usually
delivered through private markets and insurance regulation tends
to reflect solvency concerns and information asym
metry betw
een suppliers and policyholders: m
any countries have explicit reference to insurance contracts in their civil codes and specialized law
s including specific provisions for retail (B to C) m
arkets. As fiscal pressures
mount there is an increasing trend to entrust m
ore social good classes, such as w
orkman’s com
pensation and annuities, to the private sector and this adds to the pressure for effective m
arket conduct and pruden-tial regim
es.G
overnment intervention to enforce m
andatory insurance (most
comm
only motor third party, w
orkman’s com
pensation and construc-tors all risk) aside, private insurance m
arkets form naturally w
hen risk aversion ensures that individuals w
ill pay more than the expected loss
in order to hedge against the possibility of a loss that is large relative to available resources. Th
e reason that the market prem
ium is greater than
the expected loss is that the expenses of running the business need to be factored in, together w
ith the cost of the capital set aside to under-w
rite the risks involved. Typically the market prem
ium is not greatly
in excess of the expected loss for most com
mon classes such as m
otor collision insurance, w
here claims are relatively frequent (around 11 to
15 percent annually) and the average claim size is sm
all, thus gener-ating statistically credible databases. H
owever the cost of capital can
dominate w
here large elements of uncertainty exist and system
ic events can affect m
any risks at once, such as credit insurance and natural disasters coverage.
A m
ajor industry subset, the global reinsurance sector, has largely arisen from
its ability to modify the frequency and claim
s profile of a direct insurer at a price that is effi
cient for both parties (that is, rein-surers are insurers of insurers). For a typical m
ature non-life insurer in an industrial country the net cost of reinsurance in a norm
al year accounts for approxim
ately 8 percent of gross premium
revenues. H
owever reinsurers also effectively ‘lend’ capital to rapidly grow
ing insurers (both life and non-life and som
etimes know
n as surplus relief reinsurance) and the proportion reinsured can be m
uch larger in em
erging markets. Th
e reinsurance sector has also been the core source of technology transfer betw
een developed and developing markets and
is the subject of Module 2 of this series.
Given the balance sheet risks involved, the capital m
anage-m
ent of insurers bears some resem
blance to that of banks. In both cases it ideally reflects a desire to set som
e upper limit on the prob-
ability of failure within a defined period of tim
e. Hence elem
ents such as m
arket, credit and operational risk are increasingly reflected in solvency regim
es as various countries introduce risk based supervi-sory regim
es. How
ever insurance capital determination is consider-
ably more com
plex as both sides of the balance sheet are stochastic in nature (liabilities are the present values of uncertain future obligations) and the correlations betw
een liability and asset risks can be much less
obvious than for banks.
Governm
ent has three roles in insurance. The first is to ensure that
those who are granted licenses are com
petent to enter the business and w
ill have sufficient scale. Th
e second is to ensure that there are suffi-
cient competitors to prevent cartels from
developing, while lim
iting num
bers to a level that prevents pyramid structures (know
n as cash flow
underwriting) from
emerging. Insurers are prone to pyram
id structures because prem
iums are paid w
ell in advance of claims and
the ultimate cost of claim
s is uncertain. The m
ain tools to achieve these potentially conflicting objectives are m
inimum
capital require-m
ents (see figure 3), licensing rules and centralized claims data collec-
tion so that indicative pure premium
s (that is, expected losses per unit
of coverage) can be published. Market prem
iums are pure prem
iums
loaded for expenses and cost of supporting capital.Th
e third role is to protect the public. This is not the sam
e as preventing insurer insolvency (Stew
art Economics, Inc. 2003). In fact
allowing an insurer in trouble to delay insolvency (know
n as regulatory forbearance) can be inim
ical to the public interest and involve substan-tial unexpected and unnecessary public expenditures. Th
is recogni-tion has contributed to the developm
ent of risk based supervisory approaches, including corrective actions and enforcem
ent based on the ratio of actual capital to statutory m
easures of risk based capital.
Today the insurance sector is a major global industry covering a
huge range of risks ranging from natural disasters and environm
ental hazard, through life and disability and standard property risks (fire, explosion, burglary, and so forth) to various types of liability under tort and civil codes to protecting the balance sheets of credit granting institutions. In the latter case the sector has developed overlaps w
ith, and becom
e the backstop for significant sections of the banking and shadow
banking sectors. It is also a significant source of investment
funds (box 1).In 2003 seven insurers w
ere in the top fifty corporations in the w
orld in terms of revenues and insurers accounted for tw
o of the top four global financial institutions by m
arket capitalization. The industry
is global by nature as it is in the risk spreading business, but in prac-tice less than fifteen insurers can be called genuine global players
(and these are mostly European), w
ith the balance having more of a
regional approach or being confined to their home countries. In term
s of geographical involvem
ent it is clear that colonial pasts and cultural affi
liations have played a large part in foreign investment, in addition to
proximity (table 1). H
owever other factors are also im
portant.O
f particular note is the fact that, the global success of the sector notw
ithstanding, there is a massive dichotom
y between developed
and developing markets (see box 1). Th
e density of insurance (that is, prem
ium per capita) in an industrial country like France at U
S$4,075 is m
any multiples of the U
S$38.4 reported by a poor country like India (Sw
iss Re 2007). In parts of Africa the sector is effectively non existent,
despite its fundamental social and econom
ic role. The insurance devel-
opment path is discussed in detail below
but can in part be explained by the sequencing of the introduction of the various classes of insur-ance (table 2), w
hich in turn follow developm
ents in the real sector and the establishm
ent of property rights.Th
e major challenge for developm
ent institutions is to work to
change the traditional development patterns to support real sector
development and hence generate m
ore rapid sectorial development in
a virtuous cycle.
The grow
th of insurance markets has been different to that of banking
sectors, with the insurance sector being m
ore sensitive to fundamental
real sector development, and in particular to average incom
e levels (A
nnex I). This partly reflects the fact that insurers are usually not
allowed to borrow
or take on net foreign exchange exposures. 1 Their
balance sheet leverage (typically around 10 to 1 for life and 5 to 1 or less for non-life) is naturally generated by the need to hold reserves against claim
s incurred but not yet settled and future obligations. Thus
insurers are more dependent on local financial resource m
obilization and in turn on grow
th in domestic econom
ic activity. Form
al life insurance markets norm
ally develop well after non-life
insurance markets. Th
is reflects the former’s dependence on the em
er-gence of a m
iddle class and, ideally, the development of reasonably
deep and liquid capital markets. H
owever culture can play a part, w
ith life m
arkets appearing sooner where strong savings ethics apply, such
as in North A
sia and former British colonies.
If non-life premium
s/capita (premium
density) is graphed against G
DP/capita it can be seen that econom
ic growth is the m
ain driver (figure 1), w
ith a global elasticity between prem
ium/capita and G
DP
capita of approximately 1.3. 2 H
owever densities vary considerably
between countries at sim
ilar stages of development.
The m
echanisms of non-life insurance grow
th have varied both over tim
e and according to country context. Originally m
ost non marine
insurers started as mutual organizations set up to m
eet a comm
on need
1. Som
e major m
arkets (for example, the European U
nion and the United States) do in fact
allow insurers to borrow
to support solvency, but only on terms that strictly lim
it the investors’ capacity to redeem
the relevant notes.2.
That is, a 1 percent increase in G
DP per capita corresponds to a 1.3 percent increase in
premium
per capita.
or out of natural linkages. State Farm, now
one of the largest insurers in the w
orld, began life as a mutual to insure farm
ers’ vehicles. Fire insur-ance w
as founded in France by the water com
pany, which organized
the first fire brigade service. The life insurance industry grew
out of w
idows’ funds and arrangem
ents set up by various religious groups, and subsequently by the fraternal m
ovements (supported by industrial-
ization and increasing population concentration). Many large and long
established life insurers in industrial countries had mutual structures
until relatively recently, when com
petition and the obvious existence of lazy capital (and the chance for current m
anagement to utilize the
profits created over many decades of operations) forced a m
ove to shareholder structures.
The current pattern in developing m
arkets tends to be different. Transport (M
AT), industrial and comm
ercial non-life lines first appear, driven by the requirem
ents of trading partners and foreign investors/ partners and the arrival of international insurance brokers. Even the m
ost extreme centrally planned econom
ies require marine and trans-
port insurance. In addition certain economic sectors, such as agri-
culture, often see the early emergence of insurance m
echanisms for
consumption sm
oothing purposes: such classes are often subsidized by the state. H
owever the general public usually rem
ains unexposed to insurance until m
otor car fleets begin to build, often with a sudden
increase in traffic related deaths and injuries, at w
hich point compul-
sory third party liability insurance is almost inevitably introduced by
law. Because of the strong cash flows generated by this class it often
attracts questionable entrepreneurs and the insurance sector can rapidly lose the public’s confidence. Th
ereafter the main driver tends
to be the development of form
al credit markets w
ith attendant needs to cover loss of a w
age earner’s life or loss of collateral. This helps to
explain the strong link to GD
P growth and an observed nexus w
ith the developm
ent of credit markets (see A
nnex II). Life insurance based savings develop along a som
ewhat different
path. Sales people offering hard currency contracts arrive on the scene as soon as a m
iddle class with surplus incom
e begins to appear. Th
e insurers are often from other countries and the transactions
are carried out in hotel rooms. Local life insurers usually cannot be
sustained until a reasonable level of income is attained by a signifi-
cant number of fam
ilies (typically corresponding to a GD
P/ capita of betw
een US$5,000 and U
S$8,000 in 2007 dollars), the local regulatory system
has been strengthened and there is at least some local currency
capital market developm
ent. Thereafter it tends to grow
rapidly, with
a premium
density elasticity to GD
P/capita growth of approxim
ately 2 tim
es. Often the first stage w
ill involve some variant of m
ulti level selling (that is, using social contacts) and professional sales channels follow
some tim
e later. How
ever banc-assurance, in one of its various form
s, is increasingly being tried at the early stages in emerging
markets, w
ith varying degrees of success (see discussion below). Th
us, w
hile life insurance development is also prim
arily driven by GD
P/
capita, there are strong threshold effects, and other factors such as incom
e distribution, the role of social transfers and religious mores
can have a significant impact 3 (figure 2). Recent history also plays a
role with a w
ell known international brand being seen as a positive in
some countries, particularly those of the form
er Soviet Union w
hich saw
savings destroyed by hyperinflation and other countries where
pyramid schem
es have collapsed.A
side from these factors the m
ajor determinant of life insurance
development is the level of involvem
ent the sector has in managing
funded pension arrangements. Th
ere is evidence for a substitution effect w
hen mandatory second pillar pensions are introduced in m
ore developed countries, although forced savings in poorer countries does appear to lead to a net increase in overall savings.
Little work has been done on the tradeoffs betw
een establishing local insurers and introducing foreign players, although U
NCTA
D
has encouraged the development of locally ow
ned sectors. The m
ain offi
cial drivers of this policy stance usually involve retention of foreign exchange, creation of a com
petitive domestic m
arket and local control of long term
funds. The econom
ic justifications behind the second and third drivers can be questioned. FD
I has no apparent relation-ship to m
arket size but it does appear to be related to non-life under-w
riting capacity and product/distribution innovation, and foreign insurers tend to be less driven by local bank funding needs and issues of governm
ent paper. Th
e retention of foreign exchange argument m
ay be more sustain-
able provided the mechanism
adopted efficiently uses local capacity.
For example a m
ore efficient use of local financial and hum
an resource capacity is som
etimes achieved through m
andatory reinsurance through a local or regional reinsurer. Such entities enable a m
ore attractive and larger portfolio to be offered to international risk transfer m
arkets and, through a critical mass of technical staff, can offer advice
to small insurers.
Most insurance m
arkets are now opening up as the G
ATS negotia-tions proceed although these liberalization efforts are being im
ple-m
ented with highly varying level of skill and som
etimes in the absence
of necessary institutional settings (see policy issues below). Like
banks, state controlled insurers created to provide mandatory insur-
ances and to insure state owned enterprises have been m
isused under directed lending and investm
ent programs. Even in som
e late transi-tion countries insurers have been required to direct funds to socially preferred sectors such as housing (for exam
ple, New
Zealand in the
3. Takaful insurance is now
showing signs of grow
ing rapidly in Muslim
comm
unities.
1970s). This and the strong linkages that typically exist betw
een the insurance sector and the influential classes in developing m
arkets has been a m
ajor source of resistance to effective market liberalization in
some jurisdictions.
A new
role outside the normal developm
ent path is now em
erging for the insurance sector. Th
is involves protecting the working poor from
falling into poverty due to catastrophic idiosyncratic events such as death or illness of the w
age earner or systemic events such as drought
and is somew
hat similar to the social role played by friendly societies
and industrial insurers in the U.K. and Europe in the late nineteenth and early twentieth centuries. It com
es under the general heading of micro-
insurance and emerged out of the m
icro finance industry in the late twentieth century as a m
eans to raise collateral against micro-lending
and to develop alternative sources of contribution to overheads. The
official funds flow
and GD
P numbers likely to be generated by this sub-
sector are small but the social and poverty im
pacts can be considerable, w
ith more than 100 m
illion poor people already covered and the number
increasing rapidly. This series discusses m
icro-insurance in Module 3.
There has been virtually no academ
ic research on the economic
role of insurance: risk considerations appear to be absent from m
ost econom
ics courses. How
ever a body of research, largely supported by the W
orld Bank and UN
CTAD
, is now em
erging that demon-
strates growth related causalities from
insurance sector development to
economic developm
ent. These appear to apply for non-life insurance
at all stages of development (no country can trade w
ithout insurance) and for life insurance at the later stages of developm
ent (Arena 2006).
Anecdotal evidence w
ould point to the process as being iterative. Research has show
n the central role of insurance in supporting credit creation (A
nnex II) and trade finance, and the development of
equity and long term debt m
arkets (Impavido 2003). Recent financial
crises have also pointed to the unreliability of bank lending (particu-larly cross border lending) under stress and the im
portance of devel-oping m
ore reliable long term sources of funding. Finally it is now
becom
ing clear that risk markets can play a key role in protecting
developing countries from fiscal and liquidity stresses follow
ing major
natural catastrophes.Th
ere are very few exam
ples of insurance sector failure adversely system
ically affecting the economy. In m
ost cases where this has
occurred it has arisen from links w
ith the banking sector: the Jamaican
meltdow
n in the late 1990s is perhaps the best known exam
ple. Only
one recent case of a direct impact on the econom
y is known. Th
is was
the HIH
collapse in Australia, which threatened to bring the construc-
tion industry to a halt because of lack of required insurance coverage. H
owever this w
as rectified quickly. Similarly only one recent case of a
significant liquidity sourced insurer failure is known. Th
is was C
onfed-eration Life in Canada, w
hich was siphoning its liquid assets to leasing
and property subsidiaries. Most insurers fail because of insolvency,
although reduced cash significant flows are often a sign that problem
s are brew
ing for non-life insurers.M
icroinsurance for the poor and informal sectors is still at an early
stage of development but has considerable potential, both in the credit
related sense through MFIs and for social safety net purposes through
mutuals and com
munity based organizations.
A rough rule of thum
b is that the minim
um sam
ple size (in terms of
number of claim
s) to develop a pure premium
for a given rating cell (that is, a collective satisfying a given set of risk characteristics such a m
otor vehicle power, age of driver, and so forth) is 1,000. A
ssuming
a claims frequency of 10 percent this m
eans one rating cell needs 10,000 policyholders. Th
us many insurers in developing, transition
and industrial countries are too small to be statistically effi
cient. They
are able to operate because reinsurance is able to remove the poten-
tial extreme results of their risk portfolios. Essentially these sm
all insurers are ‘fronts’ for international reinsurers and the European reinsurers in particular have a long history of providing technical support to set up such entities. O
perating scale and scope can also be im
portant but studies carried out to date indicate that there is an optim
um level beyond w
hich complexity and loss of local know
ledge begin to reduce effi
ciency. From
a supervisory point of view a large num
ber of insurers is undesirable given the lim
ited resources typically allocated by govern-m
ents to the non bank sectors and the possibility of excess competition
arising from a breakdow
n in signaling mechanism
s. This is relevant to
the insurance sector because of the opacity of pricing and a consequent inherent tendency tow
ards pyramidal structures.
Several significant jurisdictions (Nigeria and Russia are tw
o recent exam
ples) have increased minim
um required capital to force consoli-
dation of overly fragmented insurance sectors and to help develop local
champions. Som
e major countries opening up to foreign players (for
example, China and India) have set very high initial capital require-
ments to filter out all but the m
ajor global insurers and to support heavy initial capital needs (figure 3). It is now
almost universally
required that life and non-life insurance be written through separately
licensed companies and as a rule the m
inimum
capital requirements
are greater for life insurers (and for reinsurers).
Financial insurance includes such lines as mortgage lenders’ insur-
ance, which typically covers default risk on higher levels of LTV
ratios (particularly if secondary m
ortgage markets have form
ed), debenture guarantee insurance (w
hich supports debenture credit ratings), project finance insurance (covering large long term
projects dependent on supplier or end user financing), retail consum
er credit insurance and fiduciary insurance (w
hich covers the dishonest behavior of company
officers handling financial transactions and is know
n as bankers blanket bonds for banks). It can be argued that, like reinsurers, credit insurers are centers of specialized underw
riting and risk managem
ent expertise, able to support a w
ider range of originators. How
ever it also needs to be recognized that credit related risks have parallels w
ith catastrophe reinsurance as the law
of large numbers tends to be irrel-
evant within a given jurisdiction w
hen such insurance is really needed (that is, arising from
some com
bination of systemic events such as high
interest rates and high unemploym
ent). In industrial and some transi-
tion countries credit risks typically have to be written through special-
ized ‘insurers’ called monolines w
ith heavy capitalizations reflecting the 400 to 600 percent loss ratios that they can experience every decade or so. W
here countries already have financial insurers or are contem-
plating them the norm
al advice is to have a separate specialized section of the law, to require that they are segregated from
other lines of busi-ness (preferably in a separate licensed entity) and are appropriately capitalized w
ith adequate reinsurance. Aside from
the highly differen-tiated capital needs the argum
ent for segregating monolines revolves
around their unique nature—effectively a lim
itless downside and an
upside constrained by competition and regulation and, possibly, a tight
linkage to the rating of capital market instrum
ents. In these situations a clean structure able to raise capital quickly and be assessed by the credit rating agencies in a w
ay that is meaningful for investors is essential.
There is strong pressure on transition countries to open up their local
insurance sectors to international competition, even if this only m
eans the ability to establish a local subsidiary or joint venture. Th
is pressure com
es from various sources, although the W
TO (G
ATS) negotiations tend to be the m
ost comm
on factor mentioned. A
number of countries
have agreed to open up their insurance sectors, sometim
es as a trade off for enhanced physical goods access to developed m
arkets or to deflect attention from
the banking sector. Unfortunately because the insurance
sector is small the liberalization is often hurried and poorly thought
through, with foreign insurers cherry picking the m
iddle class (if there is one) and little or no technology transfer occurring. Critical steps required before an insurance m
arket is opened up include strength-ening the regulatory regim
e and raising the capacity of the supervi-sory entity, developing necessary local skills and standards (accounting, auditing actuarial) and privatizing, or at least corporatizing any govern-m
ent owned or controlled insurers. It m
ay also be necessary to first put m
andatory insurance classes such as motor third party liability onto a
sound financial and operational basis.
While it is tem
pting to apply a one size fits all approach to supervi-sory coordination, experience to date show
s that every country is different and a degree of caution is needed before m
aking any substan-tive recom
mendations. In sm
aller countries there has often been a bias tow
ards moving all financial sector supervision under the central
bank—and w
here this has happened results have generally been encouraging. H
owever central banks often decline the opportunity and
at this point it is better not to be specific about options. While the need
for greater coordination should naturally emerge from
any reasonably thorough assessm
ent using some or all of the IA
IS ICPs the safest route is to form
a joint working group to look at needs and options and to
then develop a detailed plan—possibly w
ith technical support.
Many developing and transition countries are subject to severe natural
catastrophe risk. How
ever the poor development of local insurance
markets can lim
it opportunities to transfer risk because of limited or
even negligible penetration into the household and SME sectors. In
addition there is a certain size of direct insurer below w
hich interna-tional insurers cannot justify any support for transaction cost reasons. A
number of structures and products have been developed by the
World Bank and other developm
ent organizations to fill this lacuna.
Insurance accounting needs to be on an accrual basis of meaningful
information is to be generated. Th
is means that the unearned portion
of premium
s as at the accounting date need to be reserved together w
ith the present value of future claims obligations. Th
is latter amount
can be simply an addition of claim
s outstanding in claims files, w
ith a m
echanistic adjustment for claim
s incurred but not reported, or very com
plex calculations based on models in the case of long term
life insurance and classes of non-life insurance w
here claims take a
long time to stabilize and settle (such as for exam
ple a quadriplegia w
orkman’s com
pensation claim). Revenues need to reflect changes in
unearned premium
s, and claims costs changes in claim
s reserves and provisions. A
module on insurance accrual accounting appears later in
this series.
Historically the reserving process has been controlled by actuaries,
who tend to build in substantial safety m
argins. In addition assets were
held at the lower of book or m
arket, building in further safety margins.
How
ever accounting doctrine now requires that reported profits are a
good representation of reality and that surplus distribution to policy-holders and ow
ners is equitable over time. Th
is has led to a require-m
ent that assets are held at ‘fair’ value (that is, market for listed assets)
under international accounting standards. The valuation of liabilities
is also conceptually to be at ‘fair value’ but as no effective market in
insurance liabilities exists the meaning of this concept rem
ains unclear. Th
e core insurance accounting standard (IFRS4 – Insurance contracts) is currently in an half w
ay house and its final version will probably
remain unresolved until 2012 at earliest. In the interim
Solvency II, the EU
approach to prudential controls of the insurance sector, will require
that reserves and provisions have some resilience but that m
ost of the balance sheet risk is carried by capital.
Insurance information system
s are critical for both general accounting and statutory reporting purposes, and for effective m
anagement of insurers (and pricing in particular). In recent decades
there has been a move tow
ards relational databases so that analysis of em
erging experience can be carried out along a number of vectors.
The use of increasingly sophisticated analytical techniques such as
generalized linear models has also put increasing pressure on data
granularity and accessibility. Countering this has been a dependence
on legacy systems, particularly by life (long-term
) insurers that have sold a w
ide range of product types over the years. The role of legacy
systems has been a constraint on rationalization of life insurance
sectors, although some acquirers sim
ply put the relevant business into run off. Run off has in fact becom
e a profitable business in some juris-
dictions with supporting legislation.
Many IS firm
s now offer insurance system
s and this has become less
of a competitive factor over tim
e.
Insurance companies can conceptually be taxed on the sam
e basis as any other com
mercial enterprise. Th
at is on declared profits and at the corporate tax rate. For non-life insurance this is in fact the norm
al approach, although a num
ber of countries will only allow
for case esti-m
ates in setting claims provisions, rather than also allow
ing for incurred but not reported claim
s. This reflects a desire to be able to audit the tax
accounts and to minim
ize scope for creative tax accounting.
For life insurance the position is much m
ore complex as reserving
offers substantial scope for creativity and there is the additional compli-
cation of how the policyholders’ equity in em
erging surplus (profits) should be allocated and taxed. As a rule of thum
b policyholders are enti-tled to participate in surplus receive 90 percent of the profits em
erging from
the business lines they are supporting. It is not uncomm
on for taxation regim
es in developing markets to tax investm
ent income less
the expenses incurred in generating that income (called the I-E m
ethod) rather than the declared surplus. Th
e investment incom
e arising from
pension related business may be exem
pted from tax. Th
is approach has the advantage of being relatively objective but can result in an unprofit-able insurer being taxed and further weakened.
Insurance intermediation com
es under a range of headings:
Tied agents, w
ho normally represent only one insurer for each
class of insurance.
General agents, w
hich may have agency arrangem
ents with a
number of insurers for a given class, and w
hich employ sub
agents. This category is banned in m
any countries because of the confusion it can cause for consum
ers.
Brokers, who represent the insured and w
ho can deal with any
insurer. Brokers can accept comm
ission from the insurer in m
ost countries and ideally they should be required to disclose this to their client.
Insurers ow
n staff selling from the insurers ow
n premises.
Bancassurance, w
here a bank either acts directly for an insurer or provides space for an insurer’s representative in its retail outlets. Th
is channel has become very successful in such countries as
France and Italy and is being implem
ented in many countries
around the world as high initial costs can be offset by low
er distribution costs in the long run.
Call centers have had m
ixed success but can be very effective for highly specialized m
arkets and when linked w
ith radio and internet advertising. Like bancassurance initial costs can be very high.
O
ther businesses where insurance is ancillary to a m
ain business, such as travel agents.
For non-life insurers distribution and marketing costs typically
range between 10 and 40 percent, depending on the pow
er of the inter-m
ediary networks and the class of business. M
andatory lines typically have a cap placed on com
mission levels, although this has proved to be
difficult to enforce in som
e countries.For life insurance, m
arketing and distribution is typically 70 percent of the total annual costs. For agency based system
s the costs of putting a new
policy on the books is typically more than the prem
iums
received in the first year (the so called new business strain) and is
recovered from subsequent years’ prem
ia.In m
ost transition and industrial countries insurance agents need to be at least registered and there is an increasing trend to require a level of training com
mensurate w
ith the complexity of the product being
sold. The European U
nion now has a directive covering insurance
intermediaries. Th
is topic is covered in greater detail in a later module.
Arena, M
arco. 2006. “Does Insurance M
arket Activity Promote
Economic G
rowth?” W
orld Bank Policy Research Working
Paper, No. 4098, W
orld Bank, Washington. D
.C.
Brainard, Lael. 2008. “What is the role of insurance in econom
ic devel-opm
ent?” Zurich: Zurich Insurance Group.
Impavido, G
regorio, Alberto R. M
usalem, and Th
ierry Tressel. 2003. “Th
e Impact of C
ontractual Savings Institutions.” World Bank
Policy Research Working Paper, N
o. 2948, World Bank, W
ash-ington. D
.C.
Liedtke, Patrick M. 2007. “W
hat’s Insurance to a Modern Econom
y?” Th
e Geneva Papers on Risk and Insurance Issues and Practice 32 (2): 211-221.
OECD
(Organisation for Econom
ic Co-operation and D
evelopment).
2003. Institutional Investors—Statistical Yearbook, 1992–2001.
Paris: OECD
.
Stewart Econom
ics, Inc. 2003. Managing Insurery Insolvency
2003—Updating the 1988 Report. W
ashington, D.C.: N
ational A
ssociation of Insurance Brokers.
Swiss Re. 2007. Sigm
a No 4/2007.
UN
CTAD
(United N
ations Conference on Trade and D
evelopment).
2005. Trade and Developm
ent Aspects of Insurance Services and Regulatory Fram
eworks: UN
CTAD/D
ITC/TNCD
/2005/7. G
eneva: UN
CTAD
.
Webb, Ian. 2006. Assessm
ent on how strengthening the insurance industry in developing countries contributes to econom
ic growth.
Washington, D
.C.: USA
ID.