Docstoc

Solvency II Revealed Reinsurance Thought Leadership Aon

Document Sample
Solvency II Revealed Reinsurance Thought Leadership Aon Powered By Docstoc
					Solvency II Revealed
October 2011
Contents
4    An Optimal Insurer in a              27   Boosting Knowledge of Life
     Post-Solvency II World                    Catastrophe Risk

10   Changing the Landscape of            32   Rating Agencies and Solvency II
     Insurance Asset Strategy
                                          35   Reinsurance Assets: Aggregate or
16   Capital Relief Through Reinsurance        Individual?

21   Natural Catastrophe Capital          41   Risk and Capital Modelling for
     Requirement Under Solvency II             Solvency II: A Pillar of Strength
                                                                                                                     Aon Benfield




Solvency II Revealed
The insurance industry has seen an extraordinary rise to prominence for the proposed Solvency II regulation
which will bring fundamental reform of insurance supervision across Europe. As the deadline for implementation
approaches, and the economic environment continues to present significant challenges, the key is in
understanding how and where to prioritise resource to not only achieve compliance but also make the most of
the business opportunities that Solvency II offers.

Aon Benfield, in collaboration with its clients, has made enormous progress in understanding the practical implications
of the new regulatory landscape. Solvency II Revealed explores new ways of thinking about the regulatory challenges
and practically addresses these through a series of in-depth articles and case studies. Aon Benfield identifies where
Solvency II will have most impact on the industry, with advice on how best to plan ahead for all roles involved in
managing the regulation from CFO and CRO to actuaries and catastrophe modellers.

Seven key themes are explored in the report:

                                                      After Solvency II implementation, what will a capital-efficient insurer
  look like? Taking a long-term approach, the article predicts how an insurer would structure its business to maximise
  capital efficiency under the Solvency II rules, considering both the asset and liability sides of the balance sheet.

                                                             The insurance industry faces significant challenges
  transitioning to an economic framework for investments. The article reveals how Solvency II will impact insurance
  asset strategy and identifies the key considerations for the CFO and CIO in repositioning their portfolio to achieve
  capital efficiency and sidestep possible dislocations in the financial markets.

                                         This case study examines the potential impact of a non-proportional
  retention protection reinsurance on the non-life solvency capital requirements (SCR) for a notional company under
  the Standard Formula. The study demonstrates how such a contract can substantially reduce capital requirements as
  an additional benefit of the reinsurance protection.

                                                                     Internal models, despite requiring a significant
  investment, result in more accurate reinsurance recoveries and, consequently, net capital requirements, than the
  Standard Formula. The article delves into the alternatives to calculate insurers’ natural catastrophe SCR.

                                                 Assessment of terrorism and pandemic risks has been taken to a
  new level through innovative developments in partial internal models. Case studies are used to illustrate the
  risk-mitigating effect of reinsurance.

                                     Not only are regulators focused on Solvency II, but to no surprise, so are the
  rating agencies. Reviewing recent feedback from rating agencies, the article helps reinsurers prepare for the
  potential impacts of rating agencies’ calculations of capital requirements.

                                                        Calculating the fair value of a reinsurer’s share of technical non-life
  liabilities could be a challenging task if the reinsurance programme has changed in recent years. This article examines
  Solvency II’s framework directive requirements and presents two different approaches from a practical perspective.

                                                    a pillar of strength: Companies using internal models need to ensure
  they satisfy each pillar of Solvency II. The article highlights how an internal model can deliver tangible benefits
  when completing the Own Risk and Solvency Assessment (ORSA) and help companies prove to regulators that risk is
  being effectively managed.

Aon Benfield is helping insurers prepare for all pillars under Solvency II by identifying cost effective means of
improving capital efficiency, by assisting with modelling of asset and underwriting risks and by validating
(partial) internal models and ESGs. The firm offers expertise on both sides of the balance sheet and is advising
clients on designing optimal insurance and asset strategies under Solvency II. Our Solvency II capabilities
comprise asset management, reinsurance and capital market solutions covering life, non-life and health insurance.

As the industry rapidly approaches implementation, Solvency II Revealed aims to provide insurers with a fresh view
of Solvency II and empower firms to achieve both regulatory compliance and a level of capital efficiency that
exceeds investors’ expectations.
                                                                                                                                  3
Solvency II Revealed




An Optimal Insurer in a
Post-Solvency II World

Uniting the management of insurance and asset risk provides a valuable opportunity for
insurers to implement better management practices by viewing risk and capital
holistically. This approach targets the overall balance sheet risk rather than insurance or
investment risk as a silo. By leveraging the internal model framework, insurers can
optimise business strategy across insurance and investment to improve both shareholder
return and economic efficiency.


Solvency II is changing the way regulatory capital is         In practice, the two sides of the balance sheet have
assessed for European insurers. The results of the latest     been managed by separate business functions and
impact assessment study, QIS 5, suggest that the              strategies are set without a full understanding of the
average solvency ratio for non-life European insurers         impact on the overall level risk and capital. For
will drop from over 200% to around 165% (overall ratio        example, credit insurance losses are highly correlated to
since non-life ratio is not available separately).            economic risks and setting asset strategy without
Additionally, unlike the existing Solvency I regime,          consideration of insurance risks may result in a strategy
Solvency II uses a risk-based approach to set the level of    that actually increases overall risk for the firm.
each individual insurer’s solvency capital — thus
requiring more capital to be held for riskier insurance       New approach for optimisation
and investment activities. This means that insurers that      Aon Benfield has developed an optimisation process for
take a higher level of risk, as measured by Solvency II,      setting consistent strategy across asset and liability
will suffer a far greater fall in solvency ratio than those   risks, recognising all relevant economic and capital
with less risky portfolios, whose solvency ratio may          constraints. This process will support insurers to better
even improve.                                                 manage their risk and capital under Solvency II. In a
Despite presenting clear challenges, Solvency II also         post-Solvency II world, those insurers that can
offers insurers the opportunity to improve their              transform their business to maximise economic and
business strategy through better allocation of risk and       capital efficiency will enjoy competitive advantages and
capital to target opportunities that provide the highest      improved shareholder returns.
rate of return per unit of risk. Solvency II encourages       The process is outlined below, with sample exhibits for
firms to view risk, capital and value from a top-down          a hypothetical insurer “Multi-line Plc”. Figure 1
perspective, rather than from a silo based approach.          illustrates the process for optimising the firm’s overall
Insurers must set strategy in accordance with two sets        business strategy across insurance and asset risk.
of constraints simultaneously: the capital constraints        The initial strategy of Multi-line Plc is derived from the
imposed by regulators, and the economic constraints           insurance and asset strategy of the average non-life
imposed by stakeholders, including shareholders,              company in Europe. The risk assessment of the existing
policyholders and management. To maximise                     business strategy has been carried out using the Aon
performance, insurers must pursue a combined strategy         Benfield Insurance Risk Study assumptions for 2011 for
for both sides of the balance sheet — a strategy that         underwriting volatility and correlations and Aon Hewitt
comprehends the potential dependence between                  Capital Market Assumptions 2011 for asset risk.
insurance and asset risk behaviour.                           Additional assumptions for reserve risk and underwriting
To date, very few organisations have optimised their          performance have been assessed using industry data.
allocation of risk and capital across both insurance and      The Aon Benfield business strategy optimisation process
asset risk under a consistent measurement framework.          for Multi-line Plc is illustrated on the next page:


4
                                                                                                                                            Aon Benfield




Figure 1: Aon Benfield Process for Insurance and Asset Strategy Optimisation
                                                         Insurance Asset Strategy Optimisation




   1          Articulate the firm’s overall risk
              appetite, capital target and driver
              of shareholder value                    2           Identify optimal allocation of
                                                                  insurance risk under selected risk
                                                                  and capital measure                     3           Utilise remaining risk and capital
                                                                                                                      budget to develop optimal
                                                                                                                      investment strategy

 Risk                                               Insurance                                          Asset
                                                    Classes         risks                              Asset &
                                                                                                       Classes          firm
 Risk                                               Insurance                                          Classes &
                                                                                                       Constraints
                                                    Classes                                            Constraints
                                                                    characteristics                                     constraints

 Capital                                            Insurance                                          Insurance
                e.g. 150% Solvency II ratio         Constraints
                                                    Insurance       current strategy                   Insurance
                                                                                                       & Asset Risk     asset risk
 Capital                                                                                               & Asset Risk
                                                                                                       Model
                                                    Constraints
                                                                                                       Model
                model

 Value                                              Optimise                                           Optimise
                earnings volatility combined        Optimise
                                                    Insurance                                          Optimise
                                                                                                       Asset            to identify portfolios that
 Value                                              Insurance
                                                    Strategy                                           Asset
                                                                                                       Strategy
                                                    Strategy                                           Strategy
                                                                                                                        budget and capital budget

Source: Aon Benfield




1        Risk Appetite, Capital Target
         and Drivers of Value
                                                                               2       Identify Optimal Allocation of Insurance Risk
                                                                                       The firm wants to improve its insurance strategy to
        Overall risk for the firm will be quantified as the                              generate improved returns for shareholders and
        volatility of surplus, i.e. assets less liabilities, from                      achieve better risk characteristics. Shareholders of
        all sources of insurance and asset risks. An internal                          non-life insurers typically seek firms that offer
        model of the full balance sheet will be utilised to                            exposure to a carefully selected portfolio of
        measure surplus volatility and to consistently                                 insurance risks, with an asset strategy that supports
        allocate risk between insurance and asset risks.                               their liabilities and enhances risk-adjusted return
        Aon Benfield’s price-to-book regression study                                   within the overall risk and capital budget.
        points to a volatility measure of risk as best                                 Therefore, when optimising the insurance and
        capturing investor risk tolerances.                                            asset strategy of a non-life insurer, the first stage is
                                                                                       to optimise the insurance portfolio. Once the
        The binding capital metric for Multi-line Plc is the
                                                                                       optimal allocation of insurance risks has been
        Solvency II capital requirement (SCR) under the
                                                                                       selected, the remaining risk and capital budget can
        Standard Formula. Capital utilisation for insurance
                                                                                       be deployed to enhance shareholder returns
        and asset risks will be measured by the
                                                                                       through the asset strategy.
        contribution to the overall SCR from non-life
        insurance and market risk. This assumes that                                   Multi-line Plc has defined upper and lower
        shareholders are motivated by optimal exposure to                              bounds for premium volume by class of business
        insurance risks, careful management of balance                                 and agreed that the total premium volume can
        sheet volatility and attractive returns on equity.                             vary between 85% and 100% of the current level
                                                                                       (€100m): by optimising the risk allocation it may
         Following consultation with Multi-line Plc’s
                                                                                       be possible to achieve higher profitability at a
         management, the overall risk appetite for the
                                                                                       lower premium volume.
         company has been articulated as 10.0% surplus
         volatility across both insurance and asset risk,
         hence maintaining its existing level of overall
         balance sheet risk. Its existing Solvency II ratio of
         165% has been judged an appropriate long term
         position and the strategy review should maintain
         this level of capital adequacy.
                                                                                                                                                           5
Solvency II Revealed




Table 1: Insurance Allocation Constraints
                          LOB                                                                  Allocation

                                                               Initial                            Min                           Max

                                                               33%                               28.0%                       38.0%

                                                               18%                               15.5%                       21.0%

                                                                4%                               2.5%                         4.5%

                                                               30%                               25.5%                       34.5%

 General Liability                                             12%                               8.5%                        14.5%

                                                                4%                               2.5%                         4.5%

Source: Aon Benfield




An internal model of the insurer is created and, using         Figure 2: Superimposition of Economic
Aon Benfield’s proprietary optimisation framework,              and Capital Efficient Frontiers
determines the economic and capital efficient frontiers                 4.1            Economic           Capital

of the insurance portfolios that provide the maximum                   3.9                                                          B
expected profit for a specified level of economic                        3.7

volatility or Solvency II capital utilisation, respectively.           3.5
                                                               Profit




                                                                       3.3
The Solvency II capital efficient frontier differs from the                                                                Initial Portfolio
                                                                       3.1        A
economic frontier, and capital allocations that are
efficient under the proposed Standard Formula can be                    2.9

suboptimal from an economic perspective. This is                       2.7

because the proposed Standard Formula assesses                         2.5
                                                                        8.0%   8.1%     8.2%      8.3%      8.4%   8.5%   8.6%       8.7%     8.8%
capital based on prescribed volatilities and correlations
for non-catastrophe underwriting risk and prescribed                                              Economic Volatility

events for natural and man-made catastrophes — these            Source: Aon Benfield
prescribed factors are not based on economic best
estimates and are often conservative.                          In Figure 3, the insurance portfolio composition is
                                                               shown along the economic efficient frontier and two
The goal of optimisation is to identify portfolios of          performance metrics: the economic Sharpe ratio and
insurance risk that are efficient from both the economic        return on capital. In comparing options A and B, the
and capital perspective. The identification of jointly          highest ratio of profit to risk and capital is sought: this
efficient portfolios is achieved by comparing the               will achieve the optimal allocation of insurance risk.
economic and capital efficient frontier and seeking             The selected portfolio is at the left most point of the
portfolios that lie on the intersection of the two             intersection of the economic and capital efficient
frontiers. Figure 2 plots the economic and capital             frontier in region B, where both the economic Sharpe
efficient frontiers on a single graph in volatility-return      ratio and return on capital is higher than at region A,
space. The economic and capital efficient frontiers             leading to greater profit per unit of risk and capital. This
coincide in locations A and B, which provide jointly           portfolio provides the best combination of economic
optimal allocations. In order to decide which mix of           and capital efficiency.
insurance risk is preferable, it is necessary to consider
performance metrics at the two candidate portfolios.




6
                                                                                                                                                                          Aon Benfield




Figure 2: Optimisation of Insurance Risk Under Economic Risk Measure

             100%                                                                                                                                                               40%

             90%                                                                                                                                                                35%
             80%
                                                                                                                                                                                30%
             70%
                                                                                                                                                                                25%

                                                                                                                                                     B
             60%

                                          A
Allocation




             50%                                                                                                                                                                20%

             40%                                                                                                                                                                15%
             30%
                                                                                                                                                                                10%
             20%
                                                                                                                                                                                5%
             10%

              0%                                                                                                                                                                0%
                    8.04%    8.10%         8.20%          8.30%      8.35%       8.40%       8.45%        8.50%      8.55%          8.60%       8.65%       8.70%      8.75%

                                                                                     Economic Volatility


                            Credit and Suretyship                 Marine, Aviation and Transport            General Liability               Fire and Other Damage to Property
                            Motor and Vehicle Liability           Motor, Other Classes                      Sharpe Ratio                    Return on Capital



             4.5
                                                                                                                                   Optimal Portfolio
             4.0

             3.5

             3.0                                                                                                                Initial Portfolio
             2.5                                                                                                                                        B
Profit




             2.0

             1.5

             1.0

             0.5

             0.0
               8.0%                8.2%                   8.3%               8.4%                  8.5%               8.6%                   8.7%               8.8%                 8.9%

                                                                                         Economic Volatility




3              Optimisation of Asset Strategy
               Having selected the optimal insurance portfolio, the                                          does not exceed 10.0%. This will be computed as
               next stage is to investigate how the asset strategy                                           part of the optimisation as it is dependent on
               can be improved within the remaining risk and                                                 economic liability correlations
               capital budget for the firm. Overall, the level of
               insurance volatility has increased by 3 bps and the
                                                                                                             be such that the overall SCR remains at the current
               non-life Solvency II capital has increased by EUR0.1m.
                                                                                                             level of EUR69.56m: through calculation this
               As the total budget for risk is 10.0% and the firm is
                                                                                                             implies that the SCR_Mkt should be EUR29.86m
               targeting a 165% Solvency II ratio, this imposes two
               constraints on the asset portfolio:                                                         In assessing the overall level of surplus volatility for
                                                                                                           the optimised insurance strategy alongside candidate
                                                                                                           asset portfolios, it is important to consider the impact
                    volatility must be such that overall surplus volatility
                                                                                                           that liability volatility has upon economic risk.


                                                                                                                                                                                        7
Solvency II Revealed




Figure 2b: Optimisation of Insurance Risk Under Economic Risk Measure
                                                                                             Portfolio

                                                           Initial          1     2             3           4            5       6


                                                           8.62%       8.0%     8.2%          8.4%        8.5%       8.65%     8.77%

                  Profit                                     3.3         2.6      3.2           3.4         3.6          3.8      4.0
    Statistics




                                                            56.8        54.1    55.0           55.7       56.3       56.86      57.3

                                                           10.2%       22.3%    30.6%         31.1%       31.1%      31.0%     31.0%

                                                           5.9%        4.8%     5.9%          6.2%        6.5%          6.7%    6.9%

                                                           33.2%       28.0%    28.0%         28.0%       28.0%      28.0%     28.0%

                                                           18.0%       18.0%    15.5%         15.5%       15.5%      15.5%     15.5%

                                                           3.7%        2.5%     4.5%          4.5%        4.5%          4.5%    4.5%
    Allocation




                                                           30.1%       25.5%    25.5%         26.9%       29.0%      31.1%     33.0%

                  General Liability                        11.5%       8.5%     8.5%          10.5%       12.1%      13.7%     14.5%

                                                           3.5%        2.5%     4.2%          4.5%        4.5%          4.5%    4.5%

                  Total                                   100.0%       85.0%    97.3%         89.9%       93.6%      97.3%     100.0%

Source: Aon Benfield




                 In performing the asset strategy optimisation in the           not exceed EUR29.86m. The optimal asset portfolio
                 context of the overall insurance balance sheet, the            for the company is then determined by the portfolio
                 following key characteristics are incorporated into            lying on the efficient frontier that achieves an overall
                 the asset liability model:                                     surplus volatility of 10.0%.

                   Interaction of liability uncertainty with                    However, while this portfolio will provide optimal
                   economic risk: unavoidable market risk arises                return characteristics within the risk and capital
                   due to liability volatility interactions with interest       budget, it lacks a number of desirable qualitative
                   rate risk. For example, if the liabilities increase by       features. The asset strategy is refined by overlaying
                   50% then impact of interest movement will also               qualitative constraints for insurance:
                   increase by 50%.

                   Economic liability correlation: some insurance                     liquidity purposes (e.g. cat events)
                   risks, such as credit and surety, are highly
                   correlated to the economy. Ignoring economic
                                                                                      years at each key rate duration
                   liability, correlation understates true level of
                   overall risk, leading to incorrect allocations
                                                                                      generating assets of 10%
                 Using the asset liability model under our
                 optimisation framework, a constrained efficient                 This will help ensure that the asset portfolio is robust
                 frontier of asset portfolios is determined for which           during economic downturns and has good asset
                 the Solvency II market risk capital requirement does           liability characteristics for non-life insurance.




8
                                                                                                                                                                                             Aon Benfield




Figure 4: Optimal Asset Portfolio Composition
        100%                                                                              2.6%                                                                                         Portfolio
             90%
                                                                                          2.5%                                                                                  Initial           Optimal
             80%
             70%                                                                          2.4%                                                                                    -                20%




                                                                                                  Excess Return
Allocation




             60%                                                                                                                                                                  -               ±1 year
                                                                                          2.3%
             50%
                                                                                                                                                                                  -                10%
             40%                                                                          2.2%
                                                                                                                                                                                  -                10%
             30%                                                                          2.1%
             20%                                                                                                                                                                          10.0%
                                                                                          2.0%
             10%                                                                                                                                                                          29.86
              0%                                                                          1.9%                                                                                  2.12%             2.50%
                                 Initial                             Optimal
                                                                                                                                                                               3.54%              3.46%


                                                                                                                                   Statistics
                   Em Eq                   AA Credit 5          Equities         Gov Bonds 1
                   FoHF (Hedged)           Gov Bonds 3          A Credit 10      Excess Return
                                                                                                                                                                                30.00              29.86
                   Real Estate             Cash                 AA Credit 10                                                                                                   32.68%             43.06%
                   A Credit 5              Private Equity       Gov Bonds 5
                                                                                                                                                                               17.41%             20.02%
                                                                                                                                                                                0.96               0.95

Source: Aon Benfield

The optimal asset strategy shown in Figure 4 has provided a 38 bps increase in return compared to the initial portfolio,
while meeting quantitative constraints for risk and capital budgeting and insurance.


Conclusion                                                                                                        This portfolio provides the greatest profit per unit of risk
Multi-line Plc’s economic and financial characteristics                                                            and capital among all possible allocations of insurance
have been transformed under Aon Benfield’s proprietary                                                             risk within the specified constraints.
optimisation framework. Underwriting performance has                                                              After allocating risk and capital to the optimal insurance
been significantly enhanced by optimising premium                                                                  strategy, the remaining risk and capital budget was
volumes across each class of business, while applying                                                             allocated to an optimal asset strategy. The selected asset
realistic constraints to limit significant deviation from                                                          strategy fully utilises the remainder of the risk and capital
the initial underwriting strategy. The recommended                                                                budgets and provides optimal return while meeting
insurance strategy was selected as the portfolio of                                                               bespoke qualitative constraints specific to insurance.
insurance risks that:
                                                                                                                  As shown in Figure 5, the overall financial and economic
                                                                                                                  impact of the business strategy optimisation is an
             efficient frontier                                                                                    increase to expected profit of EUR1.4m, an improvement
                                                                                                                  of shareholder return from 13.3% to 14.5%. In addition,
             return on capital                                                                                    there has been no increase in volatility or capital
                                                                                                                  requirement under Solvency II.


Figure 5: Overall Comparison of Initial and Optimal Business Strategy
                                                                                                                                       17.00
                                                                                 Sharpe
                                             Profit           Vol.          SCR                    ROE
                                                                                  Ratio                                                            Optimised                  ROE of 14.45%

                           Initial           3.34           8.62%      64.37     24.45%          2.91%
                                                                                                                  Expected Profit




 Liabilities                                                                                                                           16.00
                                             3.81           8.65%      64.46     31.03%          3.32%

                           Initial          11.89           3.43%      30.00     32.17%        10.36%                                                      Initial            ROE of 13.27%
 Assets                                                                                                                                15.00
                                            12.77           3.46%      29.86     43.06%          11.13%

                           Initial          15.23                                29.15%          13.27%
                                                            9.97%      69.56                                                           14.00
                                            16.59                                36.07%          14.45%                                    8.00%   9.00%             10.00%            11.00%             12.00%


Source: Aon Benfield                                                                                                                                              Volatility


                                                                                                                                                                                                             9
Solvency II Revealed




Insurance Asset Strategy

Solvency II will change the investment behaviour of insurance companies. It introduces an
economic balance sheet and capital charges for assets that reflect the degree of asset risk and
asset liability matching. Under the Standard Formula, the calibration of some capital charges is
inconsistent with an economic view of risk. It is important to understand what potential market
dislocations could occur if a significant number of insurers choose to alter their investment
strategy accordingly.

Solvency II also encourages a holistic view of risk and capital across insurance and investment.
Allocating risk and capital across underwriting and investment more dynamically provides an
opportunity to deliver a more stable return to shareholders through the underwriting cycle.



Introduction                                                   The transition into Solvency II raises a number of
Solvency II is a major catalyst for insurance companies to     important questions for insurance companies and this
revisit their asset strategy, driven by capital requirements   article reveals how these can be addressed:
that reflect the riskiness of each asset class and how well
assets and liabilities are matched in a fair value               market risk relative to underwriting and reserve risk?
accounting world. This is in contrast to the current state
of play under Solvency I, where the same level of capital
is required whether assets are held in cash or private           strategy under Solvency II to achieve capital
equity and no consideration is given to the sensitivity of       efficiency whilst targeting attractive returns?
the firm’s valuation to movements in economic variables
such as interest rates and credit spreads.
                                                                 markets and what steps could be taken to avoid
The capital charge under Solvency II for asset and               potential market dislocations?
economic risks is called the Market Risk Solvency
Capital Requirement and represents the potential               Capital Allocation at the Enterprise Level
deterioration in the net asset value of the firm                Historically non-life insurance companies have tended
following a 1 in 200 year event over a one year time           to manage underwriting risk and investment risk as silo
horizon across all asset and economic risks. This              activities under the Chief Underwriting Officer and the
includes the potential loss in asset values and increase       Chief Investment Officer without joined up
in liabilities due to changes in the discount rate. The        measurement of risk and capital for the purposes of
market risk charge is decomposed into contributions            setting strategy. Solvency II changes the way insurers
from each underlying economic risk that drives                 think about risk, capital, volatility and value generation
changes in asset and liability valuation: this consists of     through unified risk management processes. Many
a number of sub-modules that are described in Table            companies are introducing the role of Chief Risk Officer
1. The capital charge for each sub-module is calibrated        who is responsible for managing the overall level of risk
to the 1 in 200 year return period over a 1 year time          and capital utilisation in the organisation across both
horizon. The overall market risk charge is computed            sides of the balance sheet.
by aggregating together each sub-module using a
prescribed correlation matrix to provide the 1 in 200          Having a more holistic view of risk and capital provides
level of loss across all sources of market risk.               the non-life insurance industry an opportunity to
                                                               achieve more consistent levels of return on equity



10
                                                                                                                                    Aon Benfield




throughout the underwriting cycle by dynamically                                       bearing capacity that can be redeployed to support
allocating capital between underwriting risk and                                       higher yielding investment strategies until the market
investment risk. As illustrated in Figure 1, by continually                            turns, at which point investments can be de-risked and
monitoring and forecasting the pricing cycle, business                                 a more aggressive underwriting strategy can be
plans can be adjusted to target business classes that                                  followed. From this perspective, the objective of
provide maximum profit per unit of risk and capital.                                    investment strategy for non-life insurance should be to
During soft markets, underwriting strategy can be                                      enhance the firm’s return on equity within the risk and
more cautious and premium volumes reduced                                              capital budget remaining after following the optimal
temporarily for less profitable lines. This will free-up risk                           underwriting strategy.


Table 1: Standard Formula: Overview of Market Risk Capital Changes

        Risk                                    Capital Change                                                       Implications


                                                                                               capital change




 Property

 Spread
                          income assets                                                        Solvency II


                                    Duration            Capital Charge by Duration
                      Rating
                                     Factor       1           3          5            10

                      AAA             0.9%       0.9%       2.7%       4.5%          9.0%      bonds may increase

                      AA              1.1%       1.1%       3.3%       5.5%          11.0%

                      A               1.4%       1.4%       4.2%       7.0%          14.0%

                      BBB             2.5%       2.5%       7.5%      12.5%          25.0%

                      BB              4.5%       4.5%      13.5%      22.5%          45.0%

                      B or lower      7.5%       7.5%      22.5%      37.5%          60.0%

                      Unrated         3.0%       3.0%       9.0%      15.0%          30.0%




 premium
                          markets



                                                                                               concentration threshold


                          assets for credit ratings A or above and 1.5% of total assets


Source: Aon Benfield




                                                                                                                                            11
Solvency II Revealed




Figure 1: Dynamic Risk and Capital Allocation




Source: Aon Benfield




Capital Efficiency of Investment Strategies                to Other Equity which includes a wide range of
The capital requirements for different asset classes      alternative assets. In the case of risky flavours of
under Solvency II vary considerably and are not always    private equity such as venture capital this is quite
set in line with economic principles. This creates        sensible but for a diversified fund of hedge funds, this
inconsistencies between optimal strategies as viewed      would be overly cautious: hedge funds have historical
from an economic risk measure and the Solvency II         levels of volatility significantly lower than listed equity.
Standard Formula capital requirements. It is important    In setting investment strategy it is instructive to
therefore to develop a framework for setting              understand the relative capital efficiency of different
investment strategy that can incorporate the              asset classes. One approach for comparing the capital
management’s own view of risk alongside the               efficiency is to consider the return on capital achieved
constraint of the Solvency II capital requirements:       for each investment under current market conditions
while achieving capital efficiency is important, it        from a silo perspective (i.e. ignoring its contribution
should not override the importance of careful risk        to diversification). This comparison can be helpful in
management. Where significant disparities exist            identifying whether the existing strategy is overweight
between the Standard Formula and economic                 in less capital efficient assets. In Figure 2 key asset
principles, one option is to develop a partial internal   classes’ return on capital under the Standard Formula
model covering market risk or specific asset classes       is compared to the economic view replicated using an
where greater risk granularity is desired. For example,   internal model view (based on current market
the Standard Formula assigns a capital charge of 49%      conditions).




12
                                                                                                                                                                                       Aon Benfield




Figure 2: Comparison of Return on Capital Across Asset Classes
      6.0%


      5.0%


      4.0%
ROC




      3.0%


      2.0%


      1.0%


      0.0%
              sh

                       s1

                              s3

                                      s5

                                             s7


                                                       0

                                                               3

                                                                      5

                                                                               7

                                                                                     10


                                                                                              3

                                                                                                       5

                                                                                                                7

                                                                                                                        10


                                                                                                                               3

                                                                                                                                        5

                                                                                                                                                 7

                                                                                                                                                         10


                                                                                                                                                                s

                                                                                                                                                                        te

                                                                                                                                                                               ty

                                                                                                                                                                                         d)

                                                                                                                                                                                               Eq
                                                                                                                                                              tie
                                                     s1




                                                                                                                                                                               ui
                                                                                                                                                                     ta
                                                            it

                                                                     it

                                                                            it




                                                                                             it

                                                                                                      it

                                                                                                               it




                                                                                                                              it

                                                                                                                                       it

                                                                                                                                                it




                                                                                                                                                                                       ge
             Ca

                    nd

                            nd

                                   nd

                                           nd




                                                                                                                                                                                              Em
                                                                                    it




                                                                                                                     it




                                                                                                                                                      it
                                                           ed

                                                                    ed

                                                                           ed




                                                                                            ed

                                                                                                     ed

                                                                                                              ed




                                                                                                                             ed

                                                                                                                                      ed

                                                                                                                                               ed




                                                                                                                                                              ui




                                                                                                                                                                             Eq
                                                                                                                                                                     Es
                                                  nd




                                                                                   ed




                                                                                                                    ed




                                                                                                                                                     ed




                                                                                                                                                                                    ed
                                                                                                                                                           Eq
                  Bo

                         Bo

                                 Bo

                                        Bo




                                                          Cr

                                                                 Cr

                                                                          Cr




                                                                                         Cr

                                                                                                  Cr

                                                                                                           Cr




                                                                                                                          Cr

                                                                                                                                   Cr

                                                                                                                                            Cr




                                                                                                                                                                   al
                                                Bo




                                                                                                                                                                          e
                                                                                Cr




                                                                                                                   Cr




                                                                                                                                                    Cr




                                                                                                                                                                                  (H
                                                                                                                                                                          at
                                                                                                                                                                Re
              ov

                       ov

                              ov

                                      ov




                                                     AA

                                                               AA

                                                                      AA




                                                                                        A

                                                                                                 A

                                                                                                          A




                                                                                                                         B

                                                                                                                                  B

                                                                                                                                           B




                                                                                                                                                                        iv
                                             ov




                                                                            AA




                                                                                                                A

                                                                                                                        BB

                                                                                                                              BB

                                                                                                                                       BB


                                                                                                                                                 B




                                                                                                                                                                               HF
             G

                     G

                            G

                                   G




                                                                                                                                               BB




                                                                                                                                                                     Pr
                                           G




                                                                                                                                                                             Fo
                                                                                                 Asset Class
                  ROC (Standard Forumla)                  ROC (Economic)


Source: Aon Benfield, Aon Hewitt




The Standard Formula significantly overstates the risk                                                  many non-life firms are focused on achieving positive
bearing capital required for longer duration credit and                                                investment return in their income statement, rather
hedge funds. It is also noteworthy that the riskiest                                                   than considering the asset return achieved relative to
asset classes provide the highest return on capital,                                                   the return of the liabilities. Solvency II is encouraging
despite having relatively high capital charges. Return                                                 insurers to think about the economic balance sheet
generating assets can still make an important                                                          and has significant capital charges for interest rate
contribution to return on equity despite the 49%                                                       duration mismatching. However, until IFRS 4 Phase 2
capital charge — the impact of the additional expected                                                 is implemented, the general accounting view will
yield is greater than the marginal increase in capital                                                 continue to be based on undiscounted liabilities. An
relative to less risky asset classes.                                                                  important consideration will therefore be the trade-off
                                                                                                       between capital efficiency and managing earnings
For non-life insurers an important consideration under
                                                                                                       volatility on current accounting principles: will non-life
Solvency II is the capital charges for interest rate risk.
                                                                                                       insurance investors understand that negative
Currently most categories of non-life insurance
                                                                                                       investment returns do not necessarily represent an
liabilities are accounted for on an undiscounted basis.
                                                                                                       economic loss when assets and liabilities are matched?
This means that the management and investors of




                                                                                                                                                                                                    13
Solvency II Revealed




Impact of Solvency II on the                                     2. Capital Constrained Insurers
Investment Market                                                   Under Solvency II, the capital constrained insurer’s
An important question is whether the new regulatory                 concern will be primarily to take steps to reduce
framework could itself have an impact on the                        the capital requirement. This means reducing
investment market through changing the investment                   exposure to return generating assets that attract
behaviour of the insurance industry. We have already                the 39% and 49% charges and any other assets
seen that in many cases there is a disconnection                    that have large capital charges. As illustrated in
between the basis on which the capital charges have                 Table 1, long duration corporate bonds are capital
been set under the Standard Formula and economic                    intensive and for credit rating BBB or lower are in
reality. The insurance industry plays a significant role             line with return generating assets. It is therefore
in institutional investment and is a major participant in           likely that insurance companies will reduce their
European bond markets.                                              exposure to equities and longer duration bonds
                                                                    rated BBB or lower. In addition, there is an
Changes in investment behaviour attributable to                     interesting secondary effect for life insurers.
Solvency II may originate from a number of sources                  Currently, life companies will invest in long
including:                                                          duration corporate bonds to match the duration of
1. Matching the components of the liability discount                their liabilities (typically 10 – 12 years). This works
   rate to reduce balance sheet volatility                          well as the strategy provides a good yield that
                                                                    enables competitive annuity pricing and the
2. Capital constrained insurers who need to improve                 liability discount rate is usually asset based so there
   their Solvency II ratio                                          is no additional balance sheet volatility. As noted
                                                                    previously, under Solvency II the liquidity premium
3. Insurers who target an investment strategy that
                                                                    component of the discount rate is based on a
   maximises their return on equity under the
                                                                    basket of corporate bonds, which supports
   Solvency II Standard Formula for market risk
                                                                    investment in matching bonds. However, under
                                                                    QIS 5, the calculation of the spread risk stress test
1. The Liability Discount Rate
                                                                    has been disconnected to the illiquidity premium
     Under the current proposals, liabilities are                   stress so assets and liability valuations are stressed
     discounted using a rate derived from the risk-free             separately. While an implicit link between spread
     rate plus an illiquidity premium. The risk-free rate is        risk and illiquidity risk has been maintained
     swap based with an adjustment for credit risk and              through a negative 50% correlation in the
     the illiquidity premium is variable depending on               aggregation calculation, this acts as a disincentive
     the level of illiquidity implicit in the liabilities: for      to match the spread duration of the liabilities. The
     annuities in payment which cannot be altered this              more capital efficient strategy is to invest in shorter
     will normally be 100% of the illiquidity premium               duration corporate bonds which have a lower
     and other more liquid liabilities will have a low              spread duration, and hence capital charge, and to
     percentage applied. The illiquidity premium itself is          utilise an interest rate swap to increase the rate
     based on the observed spread between a basket of               duration of the assets to that of the liabilities.
     corporate bonds (using the iBoxx index) and the
     swap rate that cannot be explained by credit risk.
     Some insurers will be motivated to invest the assets
     backing their technical provisions more closely to
     the liability discount rate under Solvency II as this
     will help to stabilise their Solvency II ratio.




14
                                                                                                           Aon Benfield




3. Maximising Return on Equity                               Conclusion
  As discussed earlier, the Standard Formula is not
                                                             Solvency II is driving non-life insurers to think
  consistent with an economic perspective which
                                                             holistically about risk, capital, volatility and value
  means that firms aiming to maximise return on
                                                             generation across insurance and investment. We
  equity may design an investment strategy that
                                                             believe that bringing together the management of
  differs substantially to their current asset allocation.
                                                             insurance and investment risk through the Chief Risk
  In particular, Figure 2 shows that long duration
                                                             Officer provides a valuable opportunity for insurers to:
  credit BBB or lower is less capital efficient and
  hedge funds also do not achieve a good return on
  equity under the Standard Formula. In general, the           risk and capital holistically across both insurance
  Standard Formula will encourage holding assets               and asset risk
  classes that provide maximum yield for the capital
  category they fall into: for example, within Other
                                                               underwriting and investment more dynamically
  Equity the most capital efficient assets will be risky
                                                               throughout the underwriting cycle to provide a
  forms of private equity investments. While many
                                                               more stable return to shareholders
  insurers are not expected to focus purely on capital
  efficiency, it is likely to be a consideration that will    There are many challenges for European insurers
  tilt the average insurance asset allocation away from      during the transitional period to Solvency II and
  less capital efficient asset classes such as high yield     beyond to the new international accounting
  debt that feature low quality credit exposures.            standards. To achieve attractive returns on equity
                                                             under the new capital regime for market risk,
                                                             significant changes to investment strategy will be
                                                             required to manage asset liability risk. Moving to an
                                                             economic view of the balance sheet has significant
                                                             implications for companies who report on an
                                                             undiscounted basis and careful communication with
                                                             senior management and investors is required to
                                                             carefully manage this transition.

                                                             Finally, in the current draft of the Standard Formula,
                                                             there are many areas of economic disconnect that
                                                             could have broader implications for the investment
                                                             market. Until the Standard Formula is finalised, it is
                                                             difficult to judge at what point insurers will start to
                                                             switch their portfolios, but it is important to be aware
                                                             of the potential market dislocations and consider how
                                                             to position your firm’s investment portfolio to
                                                             minimise the impact of the new regulatory framework.




                                                                                                                     15
Solvency II Revealed




Capital Relief Through Reinsurance


Insurers do not necessarily have to choose between a reinsurance programme which
makes business sense and one which reduces capital requirements — even if the
company appears thinly capitalised under QIS 5. Non-proportional reinsurance often
provides the best solution for the business by removing frequency risk and tail risks at a
cost that makes economic sense. Under the Standard Formula significant capital relief
can be obtained for non-proportional reinsurance, making it an attractive solution from
both a business and capital perspective.


Introduction                                                   The non-life underwriting risk module comprises three
Standard Formula risk mitigation techniques under
Solvency II have already become a hot topic for
actuaries, CROs and CFOs. For most companies,                  This case study considers the impact of the reinsurance
regulatory capital requirements have historically played       structure on both Premium Risk and Cat Risk which
a relatively small role in the decision to purchase a          must be calculated separately and then combined
particular reinsurance contract, where managing the            together using a prescribed correlation coefficient.
volatility of shareholder returns and economic and             Under the Standard Formula, the premium risk by class
rating agency capital requirements have typically had          of business is calculated as the product of a prescribed
the upper hand. However, QIS 5 indicated that                  underwriting volatility and the company’s premium
regulatory capital requirements under Solvency II are          volume. For proportional reinsurance, such as quota
likely to increase significantly for most non-life insurers.    share, the capital relief can be easily determined by
Although unlikely to become the dominant factor in a           multiplying by the ceded percentage. For non-
reinsurance purchasing decision, the impact of the             proportional reinsurance, the capital saving effect is less
reinsurance on Solvency II capital should be explored.         immediately apparent under Standard Formula.
In the Solvency II framework, an insurer can choose to         However, as this case study will demonstrate, non-
use the Standard Formula or its own internal model to          proportional reinsurance can offer significant capital
estimate its Solvency Capital Requirement (SCR). The           savings without requiring a partial or full internal model
Standard Formula is a non entity-specific risk-based            to be developed.
formula designed by EIOPA, the European insurance
regulator. Alternatively, the undertaking can build an         Notional Insurer and Reinsurance Structure
internal model and submit it for approval by the               The case study is based on a notional Swedish mono-
regulator to determine their SCR. For the vast majority        line company whose property portfolio has a premium
of companies the investment required to develop and            volume of EUR130m and is protected by existing risk
submit an internal model is too great and so a good
understanding of how the Standard Formula recognises           attachment of a 12 year return period. The focus of the
the risk mitigation effect of reinsurance is essential. This   study is the capital benefit of adding an aggregate
case study uses the QIS 5 version of the Standard              protection to the existing retention.
Formula to estimate the impact of a specific reinsurance
structure on a notional company’s non-life                     Since the company has existing risk and catastrophe

                                                               large individual and catastrophe claims, the additional




16
                                                                                                                     Aon Benfield




structure they are interested in purchasing is a risk and catastrophe aggregate reinsurance to provide more
sideways protection on their retention. With this new structure in place, the insurer is protected by the following
reinsurance contracts:




Table 1 shows the combined effect of these reinsurance protections for an example set of large individual claims:

Table 1: Combined effect of reinsurance protections

                                                                        Loss below    Presented to   Recovery from
 EURm                                 Gross Loss    Net of Risk XL                                                   Overall Net
                                                                         Retention     Aggregate       Aggregate


                                         30              10                10              9              0              10



                                         80              20                10              9              1.5           18.5



                                         40              10                10              9              9              1



                                          5               5                 5              4              4              1


Source: Aon Benfield



Undertaking Specific Parameters                                       The second method is known as Undertaking Specific
Under QIS 5, companies have the choice of two methods                Parameters (USPs). Non-life premium risk USPs allow
to estimate the impact of non-proportional reinsurance               insurers to determine the volatilities to apply in the
on their non-life insurance risk (on top of the effect of the        premium risk calculation using their own historical losses
reduced premium volume) — both involve customisation                 and one of three prescribed methods. The final premium
of the volatility factor.                                            risk USPs are weighted averages of the insurer’s
                                                                     calculation and the Standard Formula where the
The first approach is the Non-Proportional Adjustment                 credibility weights depend primarily on the number of
Factor for reinsurance. Most insurers failed to apply this           years of available data and the line of business. For
adjustment in their QIS5 submissions for a number of                 example, for property (fire) the weighting is 100% for 10
reasons, including: (1) it can only be applied for standard          years or more of data.

limits or deductibles are excluded, (2) the assumption               To apply one of the USP methods, the historical losses are
made by the adjustment calculation that individual large             first adjusted for elements such as inflation and then the
loss severities follow a lognormal distribution may be of
questionable appropriateness giving results which are                annual losses on an as-if basis. After all of these
hard to believe.                                                     adjustments have been made, the volatility to use for the
                                                                     premium risk calculation is derived. The appeal of this
                                                                     method is that the impact of any reinsurance structure
                                                                     can be taken into account. Also, in comparison with

                                                                     approval by the supervisors.




                                                                                                                                   17
Solvency II Revealed




Historical Loss Data                                                           Premium Risk Results
A credibility mechanism should be used when applying                           By applying the reinsurance programme to the 10 years
USPs. The credibility factors to be applied should be                          of historical data, the USP method can estimate both
chosen according to the length of historical loss data. In                     the gross and net volatilities as a percentage of gross
this case study of USP on the Premium Risk, USP Method                         and net premiums respectively. Figure 1 shows the
3 will be applied to 10 years of historical loss data from                     average large loss for each year before and after the
the notional Swedish company.                                                  reinsurance programme.

Figure 1: Effect of Reinsurance
            5.5

            5.0

            4.5

            4.0

                                                                                                                  B
Millions




            3.5

            3.0

            2.5

            2.0

            1.5

            1.0
                  2001    2002               2003         2004          2005         2006       2007       2008        2009       2010


                  Gross   Net of XL             Net of XL & AGG

Source: Aon Benfield

The volatility of the losses decreases significantly after the reinsurance programmes are applied.

Table 2 shows the premium volatility and premium SCR charges obtained from applying the USP method to the loss



Table 2: Premium Volatility and SCR Charges Using USP Method

                                 Net of XL            Net of XL & Agg


           USP                        9.1%                  8.5%



                                      32                     29


Source: Aon Benfield

Both of these volatilities are lower than the 10% prescribed for property (fire) premium risk under QIS 5. Since, in
this case, there is a property line with 10 years of data, a credibility weighting of 100% can be applied to the
insurer’s volatility calculation. The premium SCR decreases by approximately 9.5% due to the Aggregate Protection.




18
                                                                                                                            Aon Benfield




Cat Risk Model                                                      Table 3: Two Hypothetical Years for Cat Scenario
Property catastrophe exposure in Sweden is purely                    Cat Scenario (EURm)                 1                       2
Natural Catastrophe of which 100% is windstorm. In
this case study the Catastrophe SCR is estimated based                                                  72                       89
on real company data using Cat Method 1 of QIS 5 for
this property exposure in Sweden. Using the Cresta-
                                                                                                        36                       18
zone gross exposure data, the QIS 5 formula
determines the 1 in 200 year event loss for the peril.
To arrive at the 1 in 200 total peril loss (the CAT SCR)            Table 4: The Cat Reinsurance
(Table 5), the Standard Formula requires two
                                                                     Reinsurance (EURm)                 XL                   Agg
alternative hypothetical years to be created (Table 3).
This is first done on a gross basis and each is then
                                                                     Attachment                         15                   16.5
netted down for reinsurance, after which the maximum
net annual total of the two is taken (Table 4). For the
windstorm peril, the two hypothetical years are: 80%                 Limit                              75                       20


20% of the 1 in 200 year event.
                                                                    Table 5: The SCR Cat Results

                                                                                                     Before Agg           After Agg


                                                                                                        30                   18.5




Non-Life SCR Result
Figure 2 shows the result after the aggregation of the Cat risk and Premium risk by using the QIS 5 correlation matrix.

Figure 2: Aggregation of Cat and Premium Risks

           50                                                                                                49


           40                                                                                                                     38
                                                32
                      30                                       29
           30


           20
                                        19
Millions




           10


            0


           -10
                                                                             -9
                                                                                               -13
           -20
                              CAT SCR                Premium SCR                  Diversification                  Non Life SCR

                 Before AGG         After AGG

Source: Aon Benfield




                                                                                                                                       19
Solvency II Revealed




Since the aggregate contract protects both premium risk       Conclusion
and cat risk, some thought should be given to how the         This case study clearly demonstrates that, even for a
aggregate deductibles and limits are shared between the       reinsurance contract that is structured primarily to
two risk categories. By applying the aggregate                achieve very specific business benefits — such as the
conditions separately to the premium risk calculation         sideways retention protection — the risk-mitigation
and cat risk calculation, as has been done in this study, a
conservative assumption has been made due to the              significant even under the Standard Formula where it
aggregate deductible being imposed twice. Conversely,         may not be immediately apparent at first glance. As for
allowing for the full aggregate limit in both calculations    proportional reinsurance, non-proportional reinsurance
is an overly-generous assumption and therefore, a             programmes with tailored characteristics can also
further condition is imposed: the total reduction in SCR      significantly reduce the Solvency II Non-life
                                                              underwriting SCR.
contract, in this case EUR20m. A reduction of only
                                                              This is due to a combination of reduction effects
no such cap is required.                                      including the ability to fully recognise reinsurance in the
                                                              calculation of non-life catastrophe risk, as well as the
Therefore the aggregate reinsurance programme                 ability to capture the actual volatility of the company’s
decreases the capital charge for Cat Risk by 38% and          net premium risk using the USP method.
Premium Risk charge by 9% under the USP method.
After diversification, the total Non-life SCR decreases by     For companies for whom Solvency II capital is a key
22% from EUR49m to EUR38m.                                    constraint, the reinsurance programme could be
                                                              designed from more of a capital management
                                                              perspective. This first requires a company to define its
                                                              risk appetite for insurance underwriting risk, upon which
                                                              an optimal reinsurance programme can be structured to
                                                              help the company to meet these objectives, whilst
                                                              achieving other desirable outcomes such as reducing
                                                              ceded profit and retained volatility.

                                                              Reinsurance has always been a valuable risk-mitigation
                                                              instrument. Different reinsurance programmes provide
                                                              different business and capital benefits, and this case study
                                                              demonstrates that the two can go hand in hand under
                                                              Solvency II without necessarily using an internal model.




20
                                                                                                                 Aon Benfield




Natural Catastrophe Capital
Requirement Under Solvency II

A magnitude of difference can exist between the Standard Formula and an internal
model to calculate solvency capital for catastrophe risk. The discrepancies are revealed in
the case study which stresses the importance of making a strategic management


leading Cat risk mitigation tool and proves to be a cost effective source of capital, which
is now recognised by Solvency II.


Catastrophe risk is a key driver for capital under Solvency     Standard Formula
II, with the benchmark to withstand a 1-in-200 year event       Standardised scenarios are defined per European
for natural and man-made disasters. There is a basic            country and peril. The Standard Formula approach is
calculation method that insurers can use to determine           designed to be applicable to the majority of companies
their Solvency Capital Requirement. However the                 and will be a practical solution for smaller companies as
methodology for the standardised scenarios for natural          internal models can be costly and require a complex
catastrophe modelling overlooks key data features.              regulatory approval process.
As such, natural catastrophe (Nat Cat) calculations are         Standard Formula parameters, such as damage factors
ignoring 15 years of critical evolution under the               and correlations, as well as peril selection depending on a
currently proposed Solvency II Standard Formula,                country hazard profile, are based on the Catastrophe Task
which could lead to higher capital requirements for             Force (CTF) guidance. The CTF is a working group which
insurers when the regulation comes into force. Insurers         includes regulators, (re)insurance industry participants
need to choose between the Standard Formula and a               and catastrophe modelling agencies. The Nat Cat
partial internal model to assign a more appropriate             Standard Formula approach is currently under review after
capital charge. The article reveals the different               some criticism following the QIS 5 industry exercise.
outcomes through a detailed case study.
                                                                The factor-based method is used where standardised
Figure 1: Process Options to Calculate Nat Cat SCR              scenario is unavailable or non-applicable, including:

                                NatCat SCR




                                                (Partial)
                   Standard Formula
                                             Internal Model


                                                                Internal model
     Method 1:
                            Method 2:                Use of
    Standardised
                                                   Cat Models
                                                                Internal models based on catastrophe modelling software
      Scenarios            Factor based
                                                                output better reflect the risk profile of a company, which
Source: Aon Benfield                                             is particularly critical in producing results that reflect the
                                                                company’s potential exposure to Nat Cat risk.




                                                                                                                           21
Solvency II Revealed




Table 1 outlines the differences between the data requirements and therefore data quality impacting risk sensitivity of the
possible approaches used in QIS 5 for the Nat Cat SCR.

Table 1: Impact of the Different Data Requirements for Nat Cat SCR
                                                                                                        Internal model
 Parameters and metrics                Standardised scenario              Factor based
                                                                                                          (cat model)

 Perils

                                   Subsidence


                                                                Gross Written Premium




 Geographic resolution                                                                        All levels of resolution



 Property coverage
                                                                                              Interruption


 Line of business split




                                                                                              other secondary characteristics




 Loss scenario                                                  Single event



                                     and Subsidence


 Loss calibration


Source: Aon Benfield

Due to significant differences in data granularity between Standard Formula and partial internal model, the output
SCR will inevitably differ, with the former yielding a higher SCR in the majority of cases. Therefore companies will
base their choice of method on the approach which provides the more accurate representation of their risk in their
view. Significantly, an internal model offers several risk management applications in addition to the calculation of a
Solvency II SCR, and provides the opportunity to fully recognise the benefit of complex mitigation structures.



22
                                                                                                       Aon Benfield




Natural Catastrophe Reinsurance                            Case study
Under Solvency II                                          We examine below the differences between Standard
Both proportional and non-proportional reinsurance         Formula and internal model for a UK company writing
are reasonably taken into account as mitigation            property business.
methods under QIS 5.
                                                           The reinsurance protection in place (proportional
The QIS 5 technical specifications do not prescribe a
specific method to apply reinsurance to the proposed Nat
                                                           1) Surplus share one with occurrence limit GBP80m
Cat scenarios, because a single method is unlikely to be
appropriate for all reinsurance programmes. Instead
companies are asked to apply their reinsurance
programme using an appropriate methodology which is
then explained to the regulator. Since most internal
models fully capture the details of reinsurance
programmes, they should provide a more accurate picture       @ 100%
of the company’s net position at a 1 in 200 year level.

The Standard Formula suggests that companies assume
two events for windstorm, flood and hail, which are
composed in such a way to test the adequacy of
reinsurance protection: a combination of a large and
small event for vertical cover, and two smaller events
for horizontal protection. This allows the Standard

programmes with a reinstatement by providing capital
relief from the second limit. Internal models will of
course recognise all reinstatements.




                                                                                                               23
Solvency II Revealed




All types of reinsurance in this case are adequately recognised by both the Standard Formula and internal model.
The steps for calculating Nat Cat SCR are described in Table 2:

Table 2: Process for Applying Reinsurance

 Steps                                            Standardised scenario                       Internal model




                                                                          combined




                                 loss per peril                           Loss for all perils correlated




Source: Aon Benfield




24
                                                                                                                        Aon Benfield




Table 3 compares Nat Cat SCR using the Standard Formula and internal model for the case study example.

Table 3: Nat Cat SCR Comparisons Between the Standard Formula and Internal Model

 GBPm                                                  Standard formula                               Internal model


                                           Wind + Surge                                   Wind + Surge


                                             340.90                       185.41             243.40                    153.03

                                             409.08                       203.95             261.45                    164.51

                                                           500.65                                        304.78




 SS 1                                         107.02                      76.74               75.41                    54.86

 SS 2                                         58.11                       32.17              35.60                     29.74

                                               7.27                        3.37               4.09                      3.48

                                               7.10                        2.75               3.40                      2.30

                                              148.37                      54.09                          119.70

                                             327.87                       169.12                         328.59


                                              81.21                       34.83                           57.04

                                                            96.04                                         57.04

 adjustment **                                             +14.58                                         0.00
Source: Aon Benfield


** In the Standard Formula, there is a risk of double               This case study illustrates the magnitude of difference
                                                                    that can exist between solvency capital calculated using
                                                                    the Standard Formula and an internal model. If, as in this
being separate per peril and then correlating the net               case study, the company chooses the Standard Formula
results. Correlation mitigates double counting to some              route, it will need to maintain 94% more Nat Cat capital
                                                                    to meet the SII requirement than if it uses a partial
reinsurance recoveries, especially on lower layers in the           internal model. This difference originates from the gross
case of exposure to multiple perils. In our case the
mitigation effect of the bottom layer exceeded the total            calculation. This might hint at a calibration issue with the
limit available by GBP14.58m after combining perils                 Standard Formula. However, the Standard Formula gives
with the effect of correlation.                                     higher capital relief due to the two-event scenarios, and
                                                                    allows recoveries from reinstatements.




                                                                                                                                25
Solvency II Revealed




Figure 5: Comparison of Nat Cat SCR Using
Standard Formula and Internal Model                                             reinsurance is assessed using CatMetrica, a template
500                                                                             developed in ReMetrica. In addition to capital relief,
                                                                                CatMetrica provides measures of efficiency of
400
                                                                                reinsurance such as the Ceded RoE (calculated in this
300                                                                             example as the ratio of reinsurance margin to capital
200                                                                             relief before diversification). In Table 4, CatMetrica

100
                                                                                capital at a Ceded RoE of 4.94% under an internal
   0                                                                            model. Under the Standard Formula, the capital relief
-100                                                                            for this programme is 60% at a ceded RoE of 3.59%.

-200
                  Gross SCR                             Net SCR                 Conclusion
       Standard formula       Internal model     Difference
                                                                                As demonstrated, reinsurance is a competitively priced
                                                                                source of capital, which is expected to have full
Source: Aon Benfield                                                             recognition under the Solvency II regime.


Table 4: CatMetrica Reinsurance Evaluation
                                                                                              Probability of        Economic Capital      SII Standard
                          Premium
                                                                    Reinsurer                Attach/Exhaust             at 200 yr       Formula (QIS 5)
                                     Expected   Reinsurer St Dev of             Premium
                                                                    Margin to
                                     Recovery    Margin Recoveries              Multiple
                                                                    Std. Dev.                                      Capital    Ceded    Capital    Ceded
                           Total                                                           1st Limit   2nd Limit
                                                                                                                   Benefit      ROE     Benefit      ROE



                                        0.6                   11.7                         0.49%

                            2.3         0.6       1.8                                       1.37%       0.01%
                                                              5.8     30.53%     4.08                              46.6       3.78%     67.7     2.60%
 1@100%, 100% placed       3.33%       0.82%     2.51%                                      0.49%       0.00%

                            3.4         1.3       2.1                                       4.79%       0.07%
                                                              7.3     28.36%     2.55                               41.7      4.93%    50.9      4.04%
 1@100%, 100% placed       6.77%       2.65%     4.12%                                      1.39%       0.01%

                            4.3          2.3       2.1                                     15.90%       0.83%
                                                              6.7     31.31%     1.93                               31.4      6.67%    46.0      4.55%
 1@100%, 100% placed      17.39%       9.01%     8.37%                                      4.93%       0.13%

                                        10.7                  8.8

                                        15.4                  28.3                                                 176.7               275.2

                                                                                           15.90%       0.83%
                           10.1         4.2       5.9         16.8    35.15%     2.42                              119.7      4.94%    164.6     3.59%
                                                                                            0.49%       0.00%

                                        11.2                  15.2                                                  57.0               110.6



Source: Aon Benfield
*Expected reinsurance recoveries are based on catastrophe model output.




26
                                                                                                           Aon Benfield




Boosting Knowledge of



Pandemic and terrorism risks incur the largest catastrophe capital charges under

requirement of Solvency II, catastrophe models are evolving. Impact Forecasting’s UK
terrorism model incorporates input from counter terrorism experts on elements such as
credible attack types and damage profiles. In addition, the average cost of a catastrophe

mortality events other than pandemic.


Introduction                                               Pandemic risk
                                                           Pandemic risk has many different definitions but the
hold capital for the impact of 1 in 200 year extreme       common feature is the spread of infectious disease across
mortality events. To grasp what this definition means,      a large geographic region. The three major pandemic
in essence, this translates to not only considering one    influenzas in the last 100 years, namely the Spanish
specific scenario with the probability of 0.5%, but
thinking of holding enough capital to withstand the        caused over 50 million of deaths in total. The latest
cost of extreme events at the 99.5th percentile and        pandemic recognised by the WHO was the 2009 H1N1
therefore consider multiple tail events. Such events       which, although generating widespread public concern,
include, but are not limited to, pandemics, terrorist      was not as lethal as feared. In fact, with a reported death
attacks and natural catastrophes. The extent to which a    toll of less than 20,000 cases, it fell far short of WHO’s
company is exposed to catastrophe risk depends on          expectation of 250,000-500,000 annual deaths arising
various features of the underlying portfolio such as       from seasonal influenza.
demographic profile, geographical location and
                                                           The impact of Spanish flu, if the history was to repeat
product types.
                                                           itself on today’s insured portfolio, has the potential to
                                                           threaten the survival of many otherwise well capitalised
                                                           insurers. Many may argue that an exact repeat of a
followed by a much smaller share of terrorism risk.        Spanish flu is unlikely due to lessons learnt from the past
Impact Forecasting, Aon Benfield’s model development        such as the use of quarantine to contain the spread of
centre of excellence, has created models for pandemic,

however the article focuses on pandemic and terrorism
risks as these incur the largest catastrophe capital       facilitating rapid and effective communication of disease
charge of a combined 90%.
                                                           better access to medical assistance.
In the QIS5 technical specification, the Standard Formula
                                                           On the other hand, some may argue the prevalence of
additional deaths at 1.5 per thousand lives insured. A     international travelling, urbanisation and increased
case study will demonstrate how an internal model can      population density in city centres provide a favourable
                                                           environment for the spread of infectious disease. In
the standard formula.                                      addition to the changing landscape, we are constantly
                                                           under the threat of emerging pandemics with mutation
                                                           of pathogens or resurgence of past virus strains. The
                                                           outbreak of E-coli in Germany in 2011 had the world

                                                           situation was quickly contained. It is not a matter of if, but
                                                           when, the next major pandemic will strike.
                                                                                                                    27
Solvency II Revealed




Š                                                        Š                               Š Š                                                                                 Š
Figure 1: SIR Model

                                         Susceptible                           Latent                  Infectious                                           Hospitalized                              Dead




                                                                                                     Asymptomatic                                               Recovered



Source: Aon Benfield


In order to understand the impact of pandemic risk on                                                  Figure 3: U-shaped Excess Mortality
insurers’ portfolios, Aon Benfield has developed an                                                                                1000              1952-1956         1957
influenza pandemic model based on the Susceptible-
                                                                                                       (per 100,000 population)




Infectious-Recovered (SIR) methodology, commonly                                                                                   800

used by epidemiologists to model transmission of
                                                                                                             Flu Mortality




                                                                                                                                   600
infectious diseases. The model accounts for the
transition of healthy population (Susceptible) coming                                                                              400

into contact with the disease, developing symptoms,
                                                                                                                                   200
requiring hospitalisation, recovering from illness or
dying. In turn this enables monitoring of: the                                                                                       0
cumulative number of people becoming ill, admitted                                                                                       <1   1-4    5-14 15-24 25-34 35-44 45-54 55-64 65-74 75-84    >84

                                                                                                                                                                  Age (years)

with a given insured portfolio. As population density
                                                                                                       A probabilistic approach allows the financial impact to
transmission rate is therefore adjusted. For each                                                      be analysed by different return periods to form the
simulation, three different mortality profiles (W-shaped                                                basis of reinsurance optimisation strategy. Standard
in Figure 2, U-shaped in Figure 3 and uniform                                                          catastrophe cover is designed to remediate the tail risk
distribution) are considered, hence producing three                                                    of portfolio but does not provide appropriate
different estimates.                                                                                   protection against pandemic risk. This is mostly
                                                                                                       because the “hours clause” attached to a catastrophe
Figure 2: W-shaped Excess Mortality
                           3000              1913-1917        1918
                                                                                                       hours. However, based on what has been observed in
                           2500                                                                        the past, the full force of a pandemic can last for
(per 100,000 population)




                                                                                                       months if not years. For this reason, the industry has
                           2000
      Flu Mortality




                                                                                                       developed a range of specialist solutions including
                           1500                                                                        traditional stop loss cover, pandemic reinsurance and
                           1000                                                                        pandemic bond in the capital market.
                            500                                                                        At the height of H1N1 pandemic, some reinsurers
                              0                                                                        were quoting pandemic cover in excess of 10% rate
                                  <1   1-4    5-14 15-24 25-34 35-44 45-54 55-64 65-74 75-84   >84
                                                                                                       on line, which was prohibitively expensive. However,
                                                         Age (years)                                   in today’s market, without the over sensationalised
    Source:                                                                                            threat of pandemic, prices are more competitive and
                                                                                                       below the cost of internal capital for insurers. Similar

                                                                                                       excluded medical conditions, pandemic risk
                                                                                                       management is best considered at a time without
                                                                                                       immediate major threat.




28
                                                                                                         Aon Benfield




Terrorism risk                                            probabilistic terrorism model. Firstly, a list of potential
Terrorism has become an increasingly recognised risk      terrorism targets is identified, along with different types
factor over the last two decades. In the wake of the      of attacks ranging from nuclear devices to a shooting
                                                          rampage of a single gunman. For each type of attack,

                                                          hit zone is analysed. The probability of occurrence for
                                                          each attack or weapon type is also assigned.
Exhibit 2), insurers took a large hit on their balance    Another major piece of the puzzle is in determining
sheets but most survived by risk spreading through        the frequency of terrorist attacks. Historical records
                                                          have shown that attack frequency has increased over
have focused more on their management of
concentration risk.                                       upon in this rapidly changing environment with
Terrorism is a major catastrophe risk event for Group     reference to a range of frequencies indicated by the
business providers, especially if their portfolios have   Impact Forecasting database. As the model has been
concentrated exposure in high risk areas. Similar to      calibrated at postcode level, client exposure is
other catastrophe events, terrorism is difficult to        illustrated by postcode to evaluate terrorism risk at
predict when and where it will next strike. However,      various return periods. The terrorism model is helping
the act of terror is usually carried out to maximise      insurers to better understand their exposure of
                                                          terrorism risk and possible financial impact. In
why the World Trade Centre and the Pentagon were          addition it can be adopted b as an integral part of an
                                                          internal capital model and as an enterprise risk
                                                          management (ERM) tool to negotiate optimal
In determining the exposure to terrorism risk, Aon        reinsurance terms and demonstrate their ERM
Benfield has drawn on the knowledge and experience         strength to rating agencies.
of Aon’s counter terrorism experts to develop a

                                                                         Attack in the
Figure 4: Heat Map of Terrorism Risk Exposure                            City of London
                                                                                               Attack at
                                                                                               Canary Wharf

                                                                                               Concentric
                                                                                               Distance
                                                                                               from Impact



                                                                                          (exposure) for client data




                                                                                          overlaid with two
                                                                                          possible event attacks.

Source: Impact Forecasting


                                                                                                                    29
Solvency II Revealed




Case study                                                                     For this portfolio, the models suggest that the 1 in 200
To highlight the benefits of pandemic modelling to                              year influenza pandemic loss is GBP8.3m and terrorism
                                                                               cost is GBP6.2m. The pandemic and terrorism events are
                                                                               assumed independent of one another i.e. zero correlation.
Table 1:
                                                                               Technical Specifications, the combined 1 in 200 year
Table 1: Hypothetical LTA Portfolio                                            catastrophe loss will be GBP10.4m (=SQRT(8.3^2+6.2^2)),
                                                                               which is consistent with the capital requirement under the
       Age Group                      Lives insured    Sum Insured
                                                                               Standard Formula approach.
       35-44                                          £3,000,000,000
                                        20,000                                 Table 2: Modelled Pandemic Losses in London

                                                      £4,000,000,000                  Pandemic PML Points                  (£000)
       45-54                            20,000
                                                                                         Return Period                 Modelled Losses
       Total                            40,000        £7,000,000,000                             400                           10,742

Source: Aon Benfield                                                                              200                           8,316

                                                                                                 100                           5,643
Based on the Solvency II Standard Formula, the
                                                                                                 50                            1,980

the portfolio Sum at Risk (SAR). For the age profile of                                Average Annual Loss                       144
this particular portfolio, best estimate reserve is                            Source: Aon Benfield


capital requirement is therefore calculated at                                 Table 3: Modelled Terrorism Losses in London
GBP10.4million.                                                                       Terrorism PML Points                 (£000)

Assuming lives insured are evenly distributed across                                     Return Period                 Modelled Losses

cost is expected to be GBP10.3million for the next 12                                            400                           14,347

months (based on TMC00 table).                                                                   200                           6,193

                                                                                                 100                           4,774

                                                                                                 50                            2,016

                                                                                      Average Annual Loss                      1,131
                                                                               Source: Aon Benfield

Figure 5: PML of Pandemic losses in London
                       2.5%



                       2.0%
Exceeded Probability




                       1.5%

                                                                                                                  B
                       1.0%



                       0.5%



                       0.0%
                              2,000     4,000         6,000            8,000            10,000           12,000       14,000            16,000

                                                              Modelled Losses (£000)
 Source: Aon Benfield

30
                                                                                                                         Aon Benfield




A typical structure of pandemic cover is considered        Table 4: Summary of London and Edinburgh Portfolios
with the following features:
                                                                                                             London     Edinburgh

                                                                                                              10.4          7.7
  mortality loss
                                                                                                               6.3          1.6

                                                                                                               4.1          6.1
  mortality loss i.e. 90% excess 110%
                                                                                                              0.37         0.37

                                                                                                               9%           6%
                                                           Source: Aon Benfield
This results in a pandemic cover with an attachment
                                                           Figure 6: Illustration of Pandemic cover
                                                                                25
year for this example). The loss retained by the cedant                                                               200% of annual
before the pandemic cover is triggered is GBP1m so                              20
                                                                                                                      mortality loss
                                                           Losses (£millions)




                                                                                             Limit £9.27m
(=SQRT(1^2+6.2^2)) resulting in a capital saving                                15
                                                                                                                      Attachment point
ofGBP4.1m compared to no pandemic cover.                                                                              (110% of annual
                                                                                10
                                                                                                 Buffer
                                                                                                                      mortality loss)

                                                                                     Expected annual mortality loss
favourably to an average cost of capital of 10% to                               5
                                                                                               £10.3m
11%. Ceded RoE is the cost of reinsurance divided by
                                                                                 0
capital relief from buying reinsurance. Accretive risk
                                                               Source: Aon Benfield
transfer implies that ceded RoE is less than internal
cost of capital.
                                                           Terrorism risk can also be mitigated through
Assuming the same portfolio of lives insured is in
                                                           Benfield’s Global Death and Disability catastrophe
                                                           benchmarking study examines the catastrophe
                                                           reinsurance purchasing pattern between countries. The

arising from a terrorist attack is significantly less in

                                                           factors such as attachment point and the relative risk
If the Edinburgh portfolio has the same cost,
attachment point and limit for the pandemic cover as       a very cost effective method to mitigate extreme
                                                           mortality events other than pandemic. In addition,
becomes GBP1.6m (=SQRT(1^2+1.2^2)) with a capital          the cost of ceding the catastrophe risk to reinsurers
                                                           is well below the internal cost of capital of insurers.
it is clear, for this particular example, that pandemic
reinsurance is a cost effective risk mitigation solution   Conclusion
compared to holding capital.
                                                           Despite the relatively small capital requirement on
                                                           average for life cat risk in the overall SCR, reinsuring
                                                           life catastrophe risk with effective terms and conditions
                                                           is a very cost efficient tool to achieve a reduction.




                                                                                                                                    31
Solvency II Revealed




Rating Agencies and Solvency II



Ratings agencies are unlikely to change their core ratings processes, including capital
requirements and risk management expectations, as a result of Solvency II. However, the
new regulation will change how companies think about risk. Insurers that can demonstrate
an effective internal modelling process are likely to be at an advantage and may achieve
favourable capital adjustments under the rating agency models over time. This could
mean a reduction in the required capital needed to support their existing rating.


Rating agencies are just as focused on Solvency II as regulators and regulated companies. However, a rating looks
beyond a simple quantification of solvency (Table 1), considering ongoing financial performance, viability of
management and strategy, and other operating issues that support capital growth and sustainability over time.
Solvency II will significantly increase the prominence of regulatory capital but, ultimately, companies that wish to
maintain a strong rating and compete in the global marketplace will need to keep a focus on ratings and the
underlying capital considerations, beyond those of Solvency II.

Table 1: Drivers to Solvency II Versus Those of a Rating

                      Solvency II                                                                                    Standard & Poor’s*




                                                                                                                        Investments




Source: Aon Benfield

*In Standard & Poor’s ratings rationale an opinion of strength is provided for each of the first seven areas listed



Capital Implications                                                                 on the QIS 5 results, this would lead to a further 8% of
With the implementation date for Solvency II                                         participants unable to meet their solvency target. S&P
approaching and results from QIS 5 published, rating                                 stated that its overall concern is that approximately one
agencies believe Solvency II will have a significant capital                          quarter of European insurers will see their capital position
impact on the insurance industry.                                                    challenged under Solvency II.

Standard and Poor’s (S&P) report entitled Solvency II                                With the macro-economic uncertainty and softening
                                                                                     underwriting cycle, risk-mitigation has become
Uncertainty after Fifth Quantitative Impact Study                                    increasingly important. Insurers will need to rationalise
published in April 2011 states that insurers are likely to                           the link between risk tolerance, capacity and reward in
maintain a material buffer of 20% above the SCR. Based                               order to enhance business strategy.



32
                                                                                                                     Aon Benfield




Fitch states in its June 2011 report, Solvency II Set to Reshape Asset Allocation and Capital Markets, that insurers will
make significant changes to asset portfolios in order to enhance their capital position. Fitch anticipates a shift from
long-term to shorter-term debt and a migration towards higher-rated corporate debt and government bonds.
Ultimately, insurers would consider adopting lower risk investment policies in order to reduce their exposure to
volatile assets.

Table 2: Selected Differences in Capital Quantification

                               Solvency II — Standard Formula                                      S&P Risk Based Capital Model



                               Tier 1 and Tier 2 considerations                       Largely factor-based

                                    Largely factor-based                              Largely factor-based



                                    peril-based scenarios




Source: Aon Benfield



Transitional Arrangements                                         Future Rating Agency Capital
Proposed legislation project Omnibus II that will amend           Model Considerations
the original Solvency II directive will potentially reduce        S&P and A.M. Best are the two ratings agencies with
the risk of insurers being unable to meet capital                 established capital models. Neither agency has
requirements in the short term.                                   suggested that their respective models will be changed
                                                                  in light of Solvency II. However, both have stated that
Omnibus II outlines maximum transition periods of
                                                                  considerations for a company’s internal modelling
between 5 and 10 years for key aspects of Solvency II,
                                                                  process will be taken into account when assessing
such as meeting the full SCR and the treatment of
                                                                  capital management and risk management.
hybrid instruments in solvency measures.

In its March 2011 report entitled Weighing Solvency II’s
                                                                  ERM Analysis: Methodology for Assessing Insurers’
Impact on A.M. Best’s Ratings, the firm explains that
                                                                  Economic Capital Models, confirmed it was refining its
due to the transitional measures, the level of market
                                                                  methodology for assessing an insurer’s Economic
disruption will be much less than if Solvency II was
enforced in its entirety from the start date. Insurers are
                                                                  3 review, which is the next stage of its ERM focus. The
no longer at risk of having their market position
                                                                  review will concentrate on the quantitative and
abruptly challenged with fast-approaching deadlines,
                                                                  qualitative modelling considerations and specific risks
and transitional periods allow insurers a settling-in
                                                                  that an insurer embeds into its ECM framework.
period for when Solvency II is put into operation.

The proposed transitional periods may shift focus away
                                                                  meet certain criteria. An insurer must have an existing
from the quantitative capital impact that the regulation
                                                                  ECM which is incorporated into its decision making
brings to the detailed, non-capital related specifics of
                                                                  process and which is sufficiently documented, in order
the implementation. However, there is still debate on
                                                                  for the analysis to take place. It is also likely that only
the ultimate content and outcome of the final terms of
                                                                  insurers with ‘excellent’ or ‘strong’ ERM after
Omnibus II. Regulators must agree to the proposals
and, if there is significant deviation from these then
rating agencies may have new concerns for insurers
with which to contend.




                                                                                                                                  33
Solvency II Revealed




review to have a significant impact on ratings. However
the review can highlight risk management issues and
also demonstrate whether the ECM adequately
quantifies risks. This in turn can impact ERM and capital
conclusions, as well as the assessment of an insurer’s
management and strategy.


expected for those undergoing an internal model
approval for Solvency II, including the following:




  application




  modelling process


who can demonstrate a proficient Economic Capital
Modelling process, is at an advantage over those who
have not. Notably, those insurers whose internal capital
modelling processes are viewed by S&P to be credible
may achieve favourable capital adjustments within
S&P’s proprietary model up to one ratings category.
This could mean reduction in the required capital
needed to support their current rating. The question
remains whether rating agency capital will remain more
important than regulatory capital once Solvency II is
effective? This is likely, but time will tell.




34
                                                                                                               Aon Benfield




Reinsurance Assets: Individual vs.
Aggregate Valuation Methods

Solvency II offers opportunities to include the effect of risk mitigation techniques in
regulatory capital calculations and will likely generate new reinsurance techniques that fit
within the Standard Formula. However a key step is accounting for the impact of historical
risk mitigation and ensuring the true value of past investment is reflected in available
capital. Insurers should look deeper into the merits of calculating the fair value of their
reinsurance assets, as there may be an opportunity to unlock greater value than expected.


Introduction                                                   Case study
Reinsurance assets are one of the most complicated to          The case study presents an insurer’s balance sheet that
calculate within the Solvency II Fair Value Balance sheet.     is represented from three different angles.
Nevertheless, if the reinsurance bought is material,
                                                               The first approach (base scenario) is the current
correct calculations can have an important impact on
                                                               accounting view. Gross claim liabilities are valued on a
the Net Available Assets. The complexity of past and
                                                               case by case basis taking a conservative approach. The
present reinsurance programmes creates challenges to
                                                               reinsurance assets are valued on an individual claim
actuaries to deliver accurate estimates. The Solvency II
                                                               basis, based on the latest information available.
directive has recognised the importance of a correct
calculation of reinsurance assets by requiring a separate      Within Option 1, the actuary has valued the best
calculation for the gross best estimate claim provisions       estimate of the gross claim liabilities using actuarial
and the amounts recoverable from reinsurance                   techniques. Given the conservative approach each

                                                               gross technical claims. For valuing the reinsurance
In practice however, due to lack of time and knowledge
                                                               assets, the actuary is lacking individual data and
or data constraints, many companies calculate the
                                                               decides to use a net-to-gross ratio derived from
value of reinsurance assets using rules of thumb and
                                                               accounting data. The reinsurance assets are hence
relying on the gross results. Sometimes these give
remarkably good results, but sometimes they can be
                                                               valuation on both sides of the balance sheet, the Net
very misleading. This article aims to reveal more
                                                               Asset Value has changed from 20 to 35.3.
accurate approaches, along with possible pitfalls that
can easily be identified upfront, to indicate the
accuracy of rule of thumb approaches.

Table 1: Sample Balance Sheet in Different Approaches

        Assets            Liabilities                 Assets        Liabilities                 Assets       Liabilities

  110                                   20     110                                35,3    110                              40

   10                                   100     8,3                               83       13                              83

  120                                   120   118,3                               118,3   123                              123




Source: Aon Benfield




                                                                                                                            35
Solvency II Revealed




Figure 1: Incurred Position Towards
Its Last Known Incurred Value
250.00%


200.00%


150.00%

                                              99.64%       100.00%
                            82.04%      95.17%
100.00%
                      73.42%
                64.30%            88.49%
          35.42%
 50.00%


  0.00%
          1   2     3     4     5     6     7     8    9   10   11   12   13   14   15   16

Source: Aon Benfield

In Option 2, the actuary has analysed the large losses
individually and discovered a trend that is seen
amongst many insurance companies. Figure 1 explores
the development of the incurred position towards its
last known incurred value. The development periods
are represented in the x-axis. The y-axis shows (using a
box-plot) the spread of the observed ratios (current

example the claims were reported if their last known
incurred value exceeded EUR1.5m (note that the
example is based on a European country’s market
motor book in Euros).

After the first year of development, the median of all
claims in the database have reached only 42% of their
last known incurred value. This number increases over
time to top 90% after eight years of development. After
13 years of development, the uncertainty drops and
only a limited number of claims have an uncertain
ultimate value. Based on this, the actuary decides to do
a detailed analysis of the reinsurance assets. Applying a
net-to-gross ratio as in Option 1 will undervalue these
assets materially, hence leading to a calculated Net
Asset Value that is understating the true available
capital. The best estimate of the reinsurance assets is
now calculated at 13, which is 30% higher than the
current accounting value. The Net Asset Value increases
further to 40.




36
                                                                                                                                          Aon Benfield




Solvency II requirements
The Solvency II directive does not give any guidance on how these reinsurance assets need to be valued. They have
to follow the same principles as the gross claim liability which can be summarised and detailed in the below table:

Table 2: Guidance on the Calculation of Reinsurance Assets
 Directice                                                      Specific remarks for the calculation of Reinsurance Assets



 “The best estimate shall correspond to the probability
                                                                reinsurance program. Also the various clauses in the contract can increase the level
 weighted average of future cash-flows, taking into

                                                                can only be calculated from the probability weighted average.

 rate term structure.”

                                                                only be done once an accurate gross cash flow estimate
 deduction of the amounts recoverable from reinsurance
 contracts and special purpose vehicles. Those amounts
 shall be calculated separately, in accordance with
 Article 81”

 “The risk margin shall be such as to ensure that the value




 insurance and reinsurance obligations.”

                                                                be obtained if a separate calculation is done for both as the techniques, data
                                                                and contract issues are different: gross calculations should reflect the insurance

                                                                reinsurance policy conditions.



                                                                are protected by reinsurance, the risk mitigating effect of the reinsurance
                                                                programme should be included in the calculation of the risk margin. Also the
                                                                counterparty risk that arises should be included.



                                                                The timing of the payment of the reinsurance contract can deviate from the gross
 special purpose vehicles.

 “When calculating amounts recoverable from reinsurance
 contracts and special purpose vehicles, insurance and
                                                                no longer than three months.
 reinsurance undertakings shall take account of the time
                                                                The calculated recoverables should be reduced to reflect the credit position of

 “The result from that calculation shall be adjusted to
                                                                reduction of assets due to expected counterparty default

 counterparty. That adjustment shall be based on an
 assessment of the probability of default of the counterparty
 and the average loss resulting there from (loss-given-




Source: Aon Benfield

Within the scope of this document we will only focus on the calculation of the claim cash flows on a gross and net basis.




                                                                                                                                                       37
Solvency II Revealed




Aggregate methods                                            2. Change in reinsurance programme attachment
The most common approach to calculate the best                  and limits: It is assumed that the reinsurance
estimate for reinsurance assets is based on aggregate           programmes are stable and future trends can be
triangle techniques. Historical triangles of paid and           derived from past trends. Since claims triangles span
incurred claims data, gross of reinsurance, are modelled        many years, they can cover reinsurance cycles. These
and projected to ultimate by using a variety of different       cycles have an effect on the pricing and this is
mathematical curves. Assuming that past developments            compensated in the cedant book by retaining more
are a reasonable indicator of future developments the           or less risk by changing the priorities of the
Bornhuetter-Ferguson (BF) method incorporates prior             programme. Figure 2 details the payment pattern of
estimates of ultimate claims into the modelling process.        European motor losses that exceed a threshold value.
It is particularly appropriate for recent years of account      The higher the threshold value, the slower the
where the development factor modelling method may               payment pattern becomes.
produce unreliable results. Unlike development factor
                                                                Figure 2: Payment Pattern of
methods, the BF method can additionally take into               European Motor Losses
account collateral information, such as initial loss
                                                             100%               600 000            1 250 000     2 500 000

pricing models that may exist, and benchmark                 80%
development patterns. The prior estimates are adjusted
using development factor projection methods under a          60%
weighted average approach: the fewer the years of
development, the higher the weighting placed on the          40%

prior estimate.
                                                             20%
Net results are then obtained by netting down the
calculated gross results using net-to-gross ratios. These     0%
                                                                    1   2   3     4   5   6'   7   8   9 10 11 12 13 14 15 16 17 18 19 20
ratios are calculated from analysing the net incurred
                                                                Source: Aon Benfield
claims to the gross incurred claims or by a separate
estimate of the ultimate retained percentage after              Therefore, changing the reinsurance retentions not
examination of the reinsurance programme.                       only has an effect on the net-to-gross approach but
When using such a method, one implicitly makes many             also affects the payment pattern.
assumptions:                                                 3. Overall change in reinsurance program:
1. Stable incurred patterns: When bringing the                  Historically lines of business can be protected by
   incurred values to ultimate, the development ratios          many reinsurance programmes: individual line of
   show the release or increase of reserve surplus. In          business protection by means of proportional or
   practice (as shown higher) large losses tend to be           non-proportional treaties, portfolio protection by
   under-reserved whereas attritional losses carry the          means of aggregate covers protecting the overall
   bulk of reserve surplus. This means that within the          portfolio retention. In an incurred net triangle, these
   incurred triangle, two trends are aggregated.                various programmes can have a material impact on
                                                                the netting process and — probably even more
                                                                importantly — are a result of recovery allocation
                                                                rather than a true recovery of an individual loss.




38
                                                                                                           Aon Benfield




In many cases, when a closer examination of the              Current industry practices include both deterministic
(individual) claims data or the reinsurance policy           and stochastic approaches to quantify the reinsurance
wordings is made, even more practical issues arise:          asset. Due to the non-linear effects caused by
                                                             reinsurance, the stochastic approach is preferred. In the
1. Individual claims that have a substantial increase or
                                                             case study described below the technique of ’nearest
   decrease of reserves can have an important impact
                                                             neighbour’ has been applied to calculate the value of
   on the incurred development pattern. If the netting
                                                             reinsurance assets.
   is done on aggregated data, the effect of such a
   change in incurred can lead to misleading results if it   Individual models are not the magic trick that solves all
   is assumed to be ‘normal’ and hence projected             problems when dealing with limited data, noise and a
   forward.                                                  large volatility in outcomes. However, they do offer a
                                                             much more transparent and robust analysis with the
2. How should insurers reflect the various clauses in the
                                                             biggest advantage that each individual (large) claim
   reinsurance programmes in such a modelling
                                                             can be discussed with the claims manager. This
   technique?
                                                             probably makes this technique much more defendable
3. How should insurers deal with issues such as layering     from a ‘use’ perspective. For each individual claim
   of reinsurance and clash covers?                          ultimate paths can be created and discussed with the
                                                             claims manager, reinsurance manager and legal
Individual models                                            department to assess the quality of the modelled
                                                             results. Experience, non-quantitative data (e.g. medical
In individual models, each claim that has a potential to
                                                             reports, court uncertainty) can be used to discuss the
create a reinsurance asset is projected to its ultimate
                                                             feasibility of modelled results.
position and gross and reinsurance recoverable cash
flows are calculated. Such techniques have the major
advantage that the actual reinsurance programme can          Which approach to take?
be applied and hence all uncertainty of the results is       The use of aggregate methods to calculate the value of
due to the uncertainty of the gross (individual) model.      reinsurance assets is defendable as a proxy method but
Many assumptions are actually hidden in aggregate            only when all above considerations have been evaluated
modelling (e.g. if one assumes that the past experience      and in case the reinsurance assets are minimal compared
can be projected into the future), while in individual       to gross estimates. In other cases it is worth analysing
modelling many assumptions need to be defined                 individual methods and comparing the results between
explicitly. For example:                                     aggregate and individual models. Certainly when the
                                                             impact on the Net Asset Value (and hence the solvency
                                                             ratio) is material, it is advised to apply individual
  selected and modelled to ultimate and how should           methods to have a more accurate result.
  these levels be changed to reflect claims inflation?
                                                             Case study
  (unreported or below the selected threshold) which         The case study was conducted on a European Motor
  have the potential to create a reinsurance asset when      book for a company which had an individual line of
  they are developed to ultimate?
                                                             but where the priority changed every year. The analysis
                                                             was done using a stochastic ‘nearest neighbour’
Applying such a modelling technique makes the results        method where each (potential) large loss was projected
much more transparent and allows for individual stress       to ultimate and then sent to the various reinsurance
testing on elements such as the impact of one                programmes to calculate the netting effect. The
individual claim and parameters.                             modelling was done using ReMetrica, Aon Benfield’s
                                                             dynamic financial analysis modelling tool. No discount
                                                             effect was introduced.




                                                                                                                    39
Solvency II Revealed




Figure 3: Accounting for Reinsurance Assets
                                                           119.4
                                                                             -12%
         100.0                                                                          105.0
                                           +31%
                        -9%
                                  91.2




                                                                                                D. Impact of interest sharing
                                                                                                   clause added

                                                                   C. Ultimate effect added. The scenario’s left and right
                                                                     were obtained by assuming fgu losses were 5% lower
                                                                     (left) or higher (right) then assumed in the best estimate.

                                          B. Accounting value of Reinsurance assets
                                             including effect of stabilisation clause

                      A. Accounting value of Reinsurance Assets
                         based on reported claims
Source: Aon Benfield




The first bar (A) reflects the current accounting situation           Two stress tests were conducted to calculate the effect
of the reinsurance assets. Accounting valuation was done            on the value of the reinsurance asset in case From
on paid recoverables + EOY estimate of the recoverable
based on the incurred value as at EOY.
                                                                    Finally the accounting value of reinsurance assets did
If the company has valued the reinsurance assets using              not include any effect of a legal interest sharing clause
the gross incurred value but kept account of the effect             which made the final best estimate drop to 105% (from
of the stabilisation clause and the payment pattern of              the original 100%).
large losses, the value of the reinsurance assets would
                                                                    The impact of clauses and individual claims data (and
drop by 9%. This is reflected in bar B. In reality
                                                                    their relative position towards settlement) causes a roller
however, inflation will have an impact on the ultimate
                                                                    coaster on the best estimate assumption. The example
cash flows and — in general — large losses tend to
                                                                    was chosen as each individual effect has a large impact
increase over time due to incomplete information (as
                                                                    and therefore demonstrates the importance of taking
shown in one of the above graphs).
                                                                    each into account. The fact that the final result was close
After conducting a (stochastic) ultimate calculation of             to the initial accounting value should not be generalised
all individual gross losses that have the potential to              since the ultimate is well below or above accounting
breach the reinsurance programme, the best estimate is              value in many analysed portfolios.
changed to 119.4 which is above the accounting value.




40
                                                                                                              Aon Benfield




Risk and Capital Modelling for
Solvency II: A Pillar of Strength

Many insurers have chosen to reap the benefits of using an internal model for Pillar 1,
albeit in consideration of the investment in time and resources. However, internal models
can also play a positive role under Pillar II as part of the Own Risk and Solvency
Assessment (ORSA) to demonstrate to regulators that risk is being effectively managed.



Introduction                                                 For firms that have chosen to build partial or full
Solvency II is putting the spotlight on capital modelling    internal capital models, the ORSA necessitates a longer
and creating both opportunities and challenges. There        term view of risk. Whereas Pillar I requires risks to be
is significant growth in the adoption of partial or full      considered on a one-year time horizon, insurers prefer
internal models as insurers seek a more representative       to model the total risk emerging in run-off when
capital assessment than the Standard Formula. But            making risk management decisions.
there is also an increase in complexity as companies         With this in mind, Aon Benfield updated ReMetrica to
using internal models need to ensure their models can        allow firms to generate both one year and longer-term
satisfy each pillar of Solvency II as well as existing and   views using a single model and thereby avoid
emerging accounting standards. In addition, they must        duplication of effort.
ensure that these models are actually used in making
key decisions about the business.                            In addition, Pillar II requires firms with internal models
                                                             to prove to the regulator that adequate model
So far in the Solvency II project, both regulators and       governance processes exist around the model and its
insurers have focused much of their attention on Pillar I    data. To help address this requirement, ReMetrica’s
to calculate solvency capital requirements. Efforts are      Enterprise Edition incorporates new components to
now well advanced, with firms having stepped through          help companies keep track of their models as they
several Quantitative Impact Studies and have chosen          move one from iteration to the next. Enterprise Edition
either the prescriptive Standard Formula or a bespoke        allows companies to control which users have access to
internal capital model using a Dynamic Financial             the which parts of a model, record who has changed
Analysis tool such as Aon Benfield’s ReMetrica.               what and when, and compare one iteration of a model
                                                             with another with reporting functions.
Onus on the ORSA
                                                             The ORSA has sometimes, mistakenly, been viewed as a
Attention is now shifting towards demonstrating
                                                             simple box-ticking exercise, supplementing the Standard
wider risk management and governance capabilities to
                                                             Formula with a handful of template documents and
the regulator to satisfy Pillar II and, in particular, the
                                                             forms. Although understandable given the lack of
completion of the Own Risk and Solvency Assessment,
                                                             prescriptive guidance from EIOPA, this approach misses
                                                             a principal aim of Pillar II – to prove to the regulator that
regulator (EIOPA) due imminently. EIOPA has
                                                             a risk management culture permeates the organisation,
previously defined the ORSA only in broad, non-
                                                             and influences both everyday decision-making and
prescriptive terms, defining it as ”the entirety of the
                                                             longer-term business strategy. A regulator could impose
processes and procedures employed to identify,
                                                             a capital loading if they do not see sufficient evidence of
assess, monitor, manage, and report the short and
                                                             a strong risk mitigation strategy.
long term risks a (re)insurance undertaking faces or
may face and to determine the own funds necessary
to ensure that the undertaking’s overall solvency
needs are met at all times.”



                                                                                                                        41
Solvency II Revealed




To this end, some firms using the Standard Formula for
calculating capital are also turning to tools such as
ReMetrica to model and analyse their key risks for Pillar II
under a range of business scenarios and time frames.
This particularly applies where risks, such as natural
catastrophes, are not adequately captured by the

Standard Formula assumptions or the risk interactions
are complex.

Modelling key risks in such cases allows firms to
demonstrate these are being analysed, measured and
monitored. In turn, this provides a tangible, quantitative
output to inform business decisions, thereby allowing a
firm to realise real business value from the investment in
meeting the requirements of Solvency II. It also provides
a less burdensome entry point into capital modelling,
with some firms likely to evolve these models into a
partial or full internal model.

Conclusion
Solvency II is increasing both the take-up of models and
the range of decisions influenced by models. Tools such
as ReMetrica are well-suited to the disciplined, analytical
approach to risk management required for Pillar II, with
a clear direction towards wider usage and acceptance of
modelling key risks.




42
                                                                                                                                              Aon Benfield




Contact Information
Gareth Haslip
Head of Risk & Capital Strategy EMEA
+44 20 7522 8137
gareth.haslip@aonbenfield.com

Marc Beckers
Head of Aon Benfield Analytics EMEA
+44 7931 472 999
marc.beckers@aonbenfield.com

John Moore
Head of International Analytics
Aon Benfield Analytics
+44 20 752 3973
john.moore@aonbenfield.com




                 Scan here to access Aon Benfield’s thought leadership publications including the Solvency II Revealed report.




About Aon Benfield
Aon Benfield, a division of Aon Corporation (NYSE: AON), is the world’s leading reinsurance intermediary and full-service capital
advisor. We empower our clients to better understand, manage and transfer risk through innovative solutions and personalized
access to all forms of global reinsurance capital across treaty, facultative and capital markets. As a trusted advocate, we deliver
local reach to the world’s markets, an unparalleled investment in innovative analytics, including catastrophe management,
actuarial and rating agency advisory. Through our professionals’ expertise and experience, we advise clients in making optimal
capital choices that will empower results and improve operational effectiveness for their business. With more than 80 offices in
50 countries, our worldwide client base has access to the broadest portfolio of integrated capital solutions and services. To learn
how Aon Benfield helps empower results, please visit aonbenfield.com.




This document is intended for general information purposes only and should not be construed as advice or opinions on any specific facts or circumstances.
The comments in this summary are based upon Aon Benfield’s preliminary analysis of publicly available information. The content of this document is
made available on an “as is” basis, without warranty of any kind. Aon Benfield disclaims any legal liability to any person or organization for loss or damage
caused by or resulting from any reliance placed on that content. Aon Benfield reserves all rights to the content of this document.


                                                                                                                                                         43

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:0
posted:3/31/2013
language:English
pages:44
dominic.cecilia dominic.cecilia http://
About