Theoretical and practical issues in Solvency II

Introduction
Since the 70’s the role of the insurance industry has grown in importance
and various normative were released.
Solvency II is the last growing project on this subject. It is the new frame-
work for the regulation of European insurance undertakings, that is cur-
rently discussed in order to provide an appropriate response to the changing
needs of insurance regulation.
The basic ideas of Solvency II are the same of Basel II, however, while
the objective of Basel II is to reinforce the soundness and stability of the
international banking system, in the insurance industry, the main driver of
regulation is consumer protection.
The customer protection will be guaranteed by the new Risk Manage-
ment techniques adopted by the normative. The main ideas regarding this
is the adoption of two capital requirements (the SCR and the MCR) and the
possibility of calculating these with two approaches (the Standard Model
Approach and the Internal Model Approach)
Moreover, Solvency II will equally protect all the European consumers,
guaranteeing the respect of the same rules through the EU.
The objective of this thesis is to analyze the specificities of the Solvency
II project.
The analysis can be split in three parts, that correspond to the three
following Chapters.
In the first part, we will analyze the developments faced by the insurance
industry since the 70’s, we will describe the Solvency II project, its objec-
tives, the concept of Capital Requirement, and the phases of the project. We
will also hint the problems related to pension funds.
In the second part we will analyze Solvency II using a ”quantitative ap-
4
proach”. Firstly, we will familiarize with two important concepts: the ruin
probability, and the concept of ”coherent measure”. The next step is to un-
derstand how to model the SCR and the problems in its construction. Thus,
we will be able to comprehend the basic ideas for constructing, at EU level,
the standard approach. Finally we will discuss the creation of the internal
models.
In the last part we will speak of the actual context and of the future ex-
pectancies at EU level, also summarizing the results of different researches.
In Chapter 3 we will also analyze the Italian context, watching at the results
from the QIS 4, and the RAS case.
5
Chapter 1
SOLVENCY II: HARD WORK FOR
THE INSURANCE COMPANIES
1.1 Historical reviews
The aim of EU solvency rules is to ensure that insurance undertakings are
financially sound and can withstand adverse events, in order to protect pol-
icyholders and the stability of the financial system as a whole.
However, the current EU solvency system is over 30 years old and fi-
nancial markets have developed dramatically in recent years, leading to a
large discrepancy between the reality of the insurance business today and
its regulation.
1.1.1 Before Solvency I: the origin of solvency margin
The first European directives concerning the solvency margin were pub-
lished in 1973 (first non-life directive), and in 1979 (first life directive). These
marked the first step toward the establishment of a free insurance market
through the European Community.
The non-life directive was accepted in Italy in 1978, while the first life
directive was accepted in 1986. Before the acceptance of the directives, it
was requested only a social capital that was only dependent on the branches
of the company and independent on the number of sinister. So, all the
companies had the same capital requirements.
6
Figure 1.1: The evolution of the solvency rules through the time
Source: our elaborations
The works of Campagne were the main base of the directives. With his
works Campagne tried to find the size of the security margin in terms of
percentage of premiums, and with a certain probability, necessary to avoid
all the default possibilities. He did not claim that the model should give any
information about the solvency position of the company, but only provide an
early warning system.
Campagne proposed that the ruin probability
1
should be taken approxi-
mately as 0.03% in one year. From this and from the model used, it was
recommended that a solvency margin of 25% of the retained premiums was
enough to meet the requirement of avoiding ruin. Finally, it was suggested
that an additional 2.5% of ceded reinsurance premiums should be added to
cover against the reinsurance failure.
Campagne’s non-life approach is simple in its nature. Let the net re-
tained premium be 100%, we have to deduct a constant fraction equal to
the average expense ratio of each country from that. The remaining part is
what remains for claims payment. Next we calculate the VaRLR (Value At
Risk of the Loss Distribution) in each country and add this to the difference
between 100% and the expense ratio. The part that is above 100 would cre-
ate a solvency margin expressed in percent of the net premium minimum
income according to this approach.
Campagne used the same approach also for the life insurances. Let us
1
See chapter 2
7
Figure 1.2: Illustration of Campagne’s non-life approach
From: A. Sandström; Solvency: models, assessment and regulation, 2006
say that the risk on investments is the most important factor for life insur-
ance companies, and that the technical provisions are the most important
amount. Starting from here Campagne considered a Minimum Solvency Mar-
gin(MSM) as give as a percentage of the latter.
One main objection to the approach upper hinted is that the more pru-
dence there is in the technical provisions, the higher the MSM will be. In
other terms, a more prudent company pays more than a less prudent one
for solvency.
In the sixties, the Conference of EEC Insurance Supervisors Authorities
(the equivalent of the modern CEIOPS) began to discuss the steps toward
a free insurance market. From the dawning they discussed the technical
reserves, asset backing these reserves, and control over the assets.
2
It was
set up a Study Commission that, starting from the approach of Campagne
developed it and proposed new criteria for the MSM based on three ratios:
• Free assets premium received during the last year;
2
Source: Solvency: models, assessment and regulation, A. Sandström, 2006
8
• Free assets to average incurred claims over the last three year;
• Free assets to technical reserves.
The study produced the following results as standards of solvency margin:
• 24% of gross premium written;
• 34% of incurred claims;
• 19% of technical reserves.
The third method supposes that technical reserves would be set up in a
uniform manner in different countries.
However this approach was not used further. As claims payment fluc-
tuates between years, it was proposed that an average over the last three
financial years should be used. Moreover the EEC member states did not
have the same opinion about the indexes, for someone they were too high,
for others they are not sufficient.
In the following years were issued some directives about insurance regu-
lation.
The earliest was the first non-life directive (1973), that has the objective
to remove restrictions on the opening of branches and agencies by insurance
undertakings in other member states. The first non-life directive did not give
a definition of insurance (and this produced difficulties later) but clearly
define what is an insurance undertaking. That was essential to eliminate
differences through the different legislations and to coordinate the concept
of financial guarantee.
Indeed, article 16 of the first non-life directive start his definition of in-
surance undertaking saying: ”Each Member State shall require every under-
taking whose head office is situated in its territory to establish an adequate
solvency margin in respect of its entire business. The solvency margin shall
correspond to the assets of undertaking, free of all foreseeable liabilities, less
any intangible items.”
3
Article 17 defines two concepts of guarantee funds: the Minimum Guar-
antee Fund that is one third of the MSM, and the Absolute Guarantee Fund
defined as a fixed amount categorized according to branches of insurance.
3
Source: First non-life insurance directive, European Community, 1973
9
In 1975 the commission proposed to implement the normative with a
second directive, and an early expression of the principle of home country
control with mutual recognition of standards was proposed too. This means
that insurers established in one country, e.g. in Italy, but wishing to cover
risks in another one, e.g. in Spain, have to apply to the Italian supervisor to
have the permission.
In 1987 the council working party for economic questions discussed the
proposed second non-life directive. This led to the adoption of that on June,
1988. The solvency rules defined in the first directive were not changed.
This was an important step toward the internal market in insurance,
but the commission was not satisfied with it, as it did not make the internal
market complete for all insurance risks. In 1989 the commission announced
that the new policy was to achieve freedom of services with home country
control and using a single-license concept. This means that only one au-
thorization is needed to distribute insurances through the member states.
Third non-life directive was set up in 1992.
The life directives follow roughly the same structure of the non-life ones,
and reflects the same approach to the problems that could born. Moreover,
also for those the Campagne approach was the start point for the proposals.
The first two articles of the first life directive (1979) defines life insurance
activities as ”the type of insurance existing in Ireland and in the United King-
dom and known as permanent health insurance not subject to cancellation.”
4
Article 19 defines the MSM as a early warning signal, the guarantee fund,
defined in Article 20, has to be the MSM’s coadjutor giving the so-called
wind-up barrier.
In the second life directive (1988) a separation between home country
control and destination state control was set up.
As for the non-life business, this did not satisfy the commission, and in
1992 a third life directive was issued to include the single-license principle
without changing the rules defined in the two first directives.
Annex 1, at the end of the chapter, shows the procedure to calculate the
SM for non-life insurance business according to those directives.
4
Source: First life insurance directive, European Community, 1979
10
1.1.2 Solvency I
5
During the process with the third directives, European Council discussed
the possibility of analyse the previsions concerning the solvency margin.
But in order not to postpone the completion of the insurance single market,
it was decided to do so later.
At the Insurance Committes’s meeting in April 1994, the question about
a solvency review was raised. The IC agreed to ask the European supervisory
authorities to create a working group to look into solvency issues in a wide
sense. Helmut Müller, from the German insurance supervisory authority,
was the chairman of the group. The report that the working group presented
in 1997 will be named The Müller report.
A questionnaire was sent to its members by the Müller group, and similar
questionnaire was sent out by the commission to three European organisa-
tions: Group Consultatif (GC); Comité Européen des Assurances (CEA) and
Association des Assureurs Coopératifs et Mutualistes Européens (ACME).
The commission’s work on solvency evolved in two directions:
Solvency I: a review of the current regime, based on the Müller report;
Solvency II: a fundamental new approach.
In the Müller report the current solvency margin requirement was consid-
ered satisfactory. However, the insurance industry and the EU supervisory
authorities were in favor of a simplification in the calculation for the pre-
mium index and claims index for the non-life insurance business.
The Müller report pointed out some specific causes of deficiencies that
could have been avoided by more accurate solvency margin regime and con-
cludes that ”it was found that even if the solvency rules had been applied and
observed more strictly, and even if they had contained stricter requirements
than they do at present, a number of the economic collapses that happened
could not have been prevented. The solvency margin as a rule fulfils its warn-
ing and safety function but it does not at all replace an effective company
5
Source: Solvency: models, assessment and regulation, A. Sandström, 2006;
MARKT/2535/02: Considerations on the design of a future prudential supervisory system.
Paper for the solvency subcommittee, European Commission, November 2002
11
analysis and even less a prudent establishment and coverage of the technical
provisions.”
At least the report of the Müller Working Party suggested to keep in ac-
count the solvency rules applied in others contexts, e.g. the US’s Return
Based Capital system and the system used by European banks.
The commission considered that the principles governing the operation
of the present regime should be maintained but with the view that further
work should be done to improve the solvency margin regime and harmonize
the provision. The commission proposed to set up a new working group
chaired by the commission’s services and consisting of government experts.
For non-life insurance the group proposed the use of at least three in-
dices: the premium index, the claims index, and a provision index in order
to keep in account the long-tail businesses.
The latter should be applied alternatively or additively and the index re-
sulting in the highest margin should be the decisive one.
A fourth index could be an investment index that could be applied addi-
tively, and the yardstick for this index should be the weighted assets of the
insurance companies.
For life insurance, the group propose that the solvency margin should
not be increased. The technical risk should, as it is in the first life directive,
be taken into account by 0.3% of the CaR (Capital at Risk, valued by the VaR
method).
It was also suggested that the new regulations had to refer not only to
the solvency margin, but also to the composition of the margin and the
guarantee fund. The admissible own fund were listed in the catalog of the
directive.
The working group identified three groups of risk with 20 risk categories.
The three groups were classified as:
• Technical risks, divided in:
 current risks
 special risks
• Investment risks
12
• Non-technical risks.
Another council, the GC, was summoned to give opinions about the future
Solvency directives. The Groupe Consultatif states that from an actuarial
point of view the current solvency regime is valuable and is becoming gen-
erally accepted.
It also states that the EU directives do not govern properly the risks of
the non-life technical provisions. So, The GC list those risks dividing them
in:
• Quantifiable risks
• Risks hard to quantify.
GC wanted to avoid double counting, so in the solvency margin should ap-
pear only risks that are not well provisioned in the balance sheet.
GC also preferred to have the requirement differentiated according to
classes of insurance and distinction made between short-term and long-
term insurance.
Following the Müller report and the commission’s report there were four
expert meetings during 1997 and 1998. In the first meeting it was decided
to make simulations in order to analyze the financial impact of
1. the third index in non-life insurance based on the technical provisions
2. an increase in Minimum Guarantee Fund(MGF).
The efficiency of the provision index was doubtful, as it did not seem to
affect the targeted insurance undertakings. With regard to the MGF, very
large increases were estimated for small undertakings.
The working documents conclude that the future EU legislation should be
more flexible in order to incorporate developments in the financial services
industry faster.
The working document sets out draft proposals for an improvement of
the system and also some sketches on further developments that could be
considered (a future Solvency II system). Indeed, the report says that ”at
some future date it may be desirable to undertake a wider review of the EU
13