Solvency Assessment and Management:

Steering Committee
Position Paper 451 (v 3)
Currency Risk

EXECUTIVE SUMMARY

This document discusses the structure and calibration of the currency risk sub-module of the
Market Risk Module The paper includes discussion of the Solvency II developments,
consideration of the approaches within other jurisdictions, highlights issues, considers
alternatives and recommends an approach going forward in SAM - incorporating feedback and
analysis of SA QIS 2 and feedback from SA QIS 3.
The task group recommends that an approach based on Solvency II be adopted for the
currency risk sub-module, for comparability across companies and consistency with the way
other sub-modules are treated. In line with feedback from SA QIS 1 to 3 as well as new
developments in Solvency II, it is proposed that the Solvency 2 approach be retained with the
following amendments /additions / clarifications:
 Where there is a direct causal relationship between currency risk and policyholder
behaviour, allowance for policyholder behaviour should be made within the currency risk
sub-module. Non-causal interrelationships should be reflected in the lapse sub-module
and the correlation matrices to be used. This is discussed more fully in Discussion
Document 48.
 Currency risk can be assumed to arise from events that apply to the industry as a whole
as opposed to company-specific events.
 Impairments should be made to the risk mitigating effect of risk mitigating contracts, as
specified in [Reference to Impairment of risk mitigating contracts within the Market risk
module/CDR module unless changed to explicit allowance in CDR module]
 The SA QIS 2 and 3 approach of grouping together all foreign currencies against the
Rand should be retained. This recognises the fact that the Rand may be less stable
against foreign currencies compared to the stability of the Euro against other foreign
currencies.
 Different size of stress, also dependent on the direction of the stress, as follows: a stress
of 50% in the case that a currency’s value increases relative to the Rand, and 30% in the
case that its value decreases against the Rand.
 The treatment of dual-listed shares is clarified and it is proposed that a different
approach be followed than in Solvency II. The treatment proposed is consistent with the
treatment of dual-listed shares in Position Paper 47 (equity Risk).

1
Position Paper 45 (v 3) was approved as a FINAL Position Paper by the SAM Steering Committee on 27 March
2015.
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

The main difference of this proposed approach with the SA QIS 2 specification (except for
the 6 points above) is that it is proposed to use the same % stresses to currencies pegged to
the Rand and other currencies. This is to reflect the very likely event in a 1-in-200 year
scenario that the currencies may be de-coupled. This is inconsistent with the way that
Solvency 2 treats exchange rate stresses between the Euro and currencies pegged to it
and/or between different currencies that are both pegged to the Euro.

1. INTRODUCTION AND PURPOSE

This document sets out the recommendations of the market risk working group with respect
to the capital to be held in respect of currency risk.

2. INTERNATIONAL STANDARDS: IAIS ICPs

IAIS is the international standards setting body for insurance supervisors. The FSB as a
member of the IAIS aims to adhere to these standards. The standards are principles based
and establish broad standards which are difficult to translate to the detailed capital
requirements of an individual risk module. However, the following are relevant within the
broad framework of the capital requirements, market risk and currency risk in particular
(reference: “Insurance Core Principles, Standards, Guidance and Assessment
Methodology – Consultation Draft February 2011”):

ICP 17 Capital Adequacy

The supervisor establishes capital adequacy requirements for solvency purposes so
that insurers can absorb significant unforeseen losses and to provide for degrees of
supervisory intervention.

Some sub-points in this standard that should be considered includes:

17.1 The supervisor requires that a total balance sheet approach is used in the assessment
of solvency to recognise the interdependence between assets, liabilities, regulatory capital
requirements and capital resources and to require that risks are appropriately recognised.

17.2 The supervisor establishes regulatory capital requirements at a sufficient level so that,
in adversity, an insurer’s obligations to policyholders will continue to be met as they fall due
and requires that insurers maintain capital resources to meet the regulatory capital
requirements.

17.6 The solvency requirements are open and transparent as to the regulatory capital
requirements that apply. It is explicit about the objectives of the regulatory capital
requirements and the bases on which they are determined. In determining regulatory capital
requirements, the supervisor allows a set of standardised and, if appropriate, other approved
more tailored approaches such as the use of (partial or full) internal models.

Page 2 of 20
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

17.7 The solvency requirements address all relevant and material categories of risk and are
explicit as to where risks are addressed, whether solely in technical provisions, solely in
regulatory capital requirements or if addressed in both, as to the extent to which the risks are
addressed in each. The requirements are also explicit as to how risks and their aggregation
are reflected in regulatory capital requirements.
Types of risks to be addressed
17.7.1 The solvency requirements should address all relevant and material
categories of risk - including as a minimum underwriting risk, credit risk, market risk,
operational risk and liquidity risk. This should include any significant risk
concentrations, for example, to economic risk factors, market sectors or individual
counterparties, taking into account both direct and indirect exposures and the
potential for exposures in related areas to become more correlated under stressed
circumstances.

17.8 The supervisor sets out appropriate target criteria for the calculation of regulatory
capital requirements, which underlie the calibration of a standardised approach…

3. EU DIRECTIVE ON SOLVENCY II: PRINCIPLES(LEVEL 1)

Relevant extracts from the Solvency II level 1 principles are provided below. As is the case
with the IAIS core principles, these requirements are in nature of a higher level than required
for the establishment of detailed principles in the currency risk sub-module of the market risk
module within the capital requirements. However, it provides the broad framework within
which these requirements are to be looked at and calculated.

Article 100
General provisions

Member States shall require that insurance and reinsurance undertakings hold eligible own
funds covering the Solvency Capital Requirement.
The Solvency Capital Requirement shall be calculated, either in accordance with the
standard formula in Subsection 2 or using an internal model, as set out in Subsection 3.

Article 101
Calculation of the Solvency Capital Requirement
1. The Solvency Capital Requirement shall be calculated in accordance with
paragraphs 2 to 5.
2. The Solvency Capital Requirement shall be calculated on the presumption that the
undertaking will pursue its business as a going concern.
3. The Solvency Capital Requirement shall be calibrated so as to ensure that all
quantifiable risks to which an insurance or reinsurance undertaking is exposed are
taken into account. It shall cover existing business, as well as the new business
expected to be written over the following 12 months. With respect to existing
business, it shall cover only unexpected losses.

Page 3 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

It shall correspond to the Value-at-Risk of the basic own funds of an insurance or

reinsurance undertaking subject to a confidence level of 99,5 % over a one-year
period.
4. The Solvency Capital Requirement shall cover at least the following risks:

(a) non-life underwriting risk;

(b) life underwriting risk;
(c) health underwriting risk;
(d) market risk;
(e) credit risk;
(f) operational risk.

Operational risk as referred to in point (f) of the first subparagraph shall include legal
risks, and exclude risks arising from strategic decisions, as well as reputation risks.
5. When calculating the Solvency Capital Requirement, insurance and reinsurance
undertakings shall take account of the effect of risk-mitigation techniques, provided
that credit risk and other risks arising from the use of such techniques are properly
reflected in the Solvency Capital Requirement.

Article 104
Design of the Basic Solvency Capital Requirement
1. The Basic Solvency Capital Requirement shall comprise individual risk modules, which
are aggregated in accordance with point (1) of Annex IV.
It shall consist of at least the following risk modules:
(a) non-life underwriting risk;
(b) life underwriting risk;
(c) health underwriting risk;
(d) market risk;
(e) counterparty default risk.
2. For the purposes of points (a), (b) and (c) of paragraph 1, insurance or reinsurance
operations shall be allocated to the underwriting risk module that best reflects the technical
nature of the underlying risks.
3. The correlation coefficients for the aggregation of the risk modules referred to in
paragraph 1, as well as the calibration of the capital requirements for each risk module, shall
result in an overall Solvency Capital Requirement which complies with the principles set out
in Article 101.
4. Each of the risk modules referred to in paragraph 1 shall be calibrated using a Value-at-
Risk measure, with a 99,5 % confidence level, over a one-year period.
Where appropriate, diversification effects shall be taken into account in the design of each
risk module.

Page 4 of 20
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

5.The same design and specifications for the risk modules shall be used for all insurance
and reinsurance undertakings, both with respect to the Basic Solvency Capital Requirement
and to any simplified calculations as laid down in Article 109.
6.With regard to risks arising from catastrophes, geographical specifications may, where
appropriate, be used for the calculation of the life, non-life and health underwriting risk
modules.
7. Subject to approval by the supervisory authorities, insurance and reinsurance
undertakings may, within the design of the standard formula, replace a subset of its
parameters by parameters specific to the undertaking concerned when calculating the life,
non-life and health underwriting risk modules.
Such parameters shall be calibrated on the basis of the internal data of the undertaking
concerned, or of data which is directly relevant for the operations of that undertaking using
standardised methods.
When granting supervisory approval, supervisory authorities shall verify the completeness,
accuracy and appropriateness of the data used.

Article 105
Calculation of the Basic Solvency Capital Requirement
1.The Basic Solvency Capital Requirement shall be calculated in accordance with
paragraphs 2 to 6.

5.The market risk module shall reflect the risk arising from the level or volatility of market
prices of financial instruments which have an impact upon the value of the assets and
liabilities of the undertaking. It shall properly reflect the structural mismatch between assets
and liabilities, in particular with respect to the duration thereof.
It shall be calculated, in accordance with point (4) of Annex IV, as a combination of the
capital requirements for at least the following sub-modules:

(e) the sensitivity of the values of assets, liabilities and financial instruments to
changes in the level or in the volatility of currency exchange rates (currency risk);

Article 109
Simplifications in the standard formula

Insurance and reinsurance undertakings may use a simplified calculation for a specific sub-
module or risk module where the nature, scale and complexity of the risks they face justifies
it and where it would be disproportionate to require all insurance and reinsurance
undertakings to apply the standardised calculation.
Simplified calculations shall be calibrated in accordance with Article 101(3).

4. MAPPING ANY PRINCIPLE (LEVEL 1) DIFFERENCES BETWEEN IAIS ICP & EU

DIRECTIVE

Page 5 of 20
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

No differences.

5. STANDARDS AND GUIDANCE (LEVELS 2 & 3)

5.1 IAIS standards and guidance papers

This was covered in section 2.

5.2 CEIOPS CPs (consultation papers)

The level 2 advice (formerly CP47, CP 77 and CP 70) covers the advice on market
risk excluding equity risk, simplifications and the calibration of the market risk module
respectively.

A delta-NAV approach is used, which ensures among other things, that the impact of
risk mitigation techniques adhering to certain standards (outside the scope of this
document) will be included in the capital requirement. The risk margin is excluded in
the valuation of the technical provisions for this purpose in the standard approach, to
avoid the circularity of calculating a risk margin which depends on the size of the
capital requirement (But where this can have a significant additional impact, it should
be treated using a capital add-on or internal model).

Furthermore, the revaluation of technical provisions should take account of any

relevant adverse change in the option take-up behaviour of policyholders in the
scenario tested as is required by former CP 49, and using the same definition of
option take-up behaviour as in that document. This is to ensure that the complex
interdependency between, for example, termination- and market risk are captured
accurately.

In particular, for currency risk:

4.34 Currency risk arises from changes in the level or volatility of currency exchange
rates.
4.35 Undertakings may be exposed to currency risk arising from various sources,
including their investment portfolios, as well as assets, liabilities and investments in
related undertakings. The design of the currency risk sub-module is intended to take
into account currency risk for an undertaking arising from all possible sources.

An explanation is provided as to why each currency should be treated separately in

the module, and not all currencies together as follows:
4.39 The QIS4 approach considered the effect of two scenarios (a rise and a fallin
exchange rates) on the net value of assets minus liabilities. The scenarios implicitly
assumed that all currencies experience the same rise or fall in reference to a local

Page 6 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

currency, whilst ignoring the interdependencies between currencies other than the
local currency.
4.40 As an example, consider the case of an insurer with regulatory accounts
denominated in EUR who has US$ denominated liabilities of value EUR 1million and
assets in £ sterling of value EUR 1 million at the 2007 yearend. In addition, suppose
that all other balance sheet items are denominated in EUR. In this case, the
scenarios fx upward and fx downward do not lead to a change in basic own funds,
because pound and dollar both are assumed to rise or fall in relation to euro.
According to the scenarios, the insurer seems to be perfectly hedged against
currency risk.
During the year 2008, the value of the US dollar liabilities have risen to EUR 1.05
million and the value of the pound sterling assets have fallen to EUR 0.77 million.
Consequently, there is a loss of basic own funds of EUR0.28 million, more than a
quarter of the initial exposure.
4.41 The example shows that significant currency risks may not be detected if the
QIS4 approach is applied. The two currency scenarios imply that the exchange rate
between pound and dollar is fixed. This is not a realistic assumption. Moreover, the
current approach incentivises a currency risk mismanagement as illustrated in the
above example: If no appropriate assets in US dollar are available to cover the dollar
liabilities, then the undertaking can reduce its capital requirement by covering the
liabilities with another foreign currency. However, if the dollar liabilities are covered
with euro assets then the resulting capital charge will be 20% of the liabilities.

The advice goes on to explain that the risk for each foreign currency should be
assessed by an upward and downward shock against the local currency, where the
local currency is the currency in which own funds are denominated, and relevant
currencies are those which can affect the amount of own funds should the exchange
rate with the local currency change.

No offset is allowed for a favourable move in a scenario (i.e. each currency in each
scenario will contribute a positive amount to the capital requirement or 0). The total
capital requirement per currency is the maximum of the requirements in the up- or
down- scenarios, and the total capital requirement is the sum of this over all relevant
currencies.

4.47 For the example presented above, the modification would require the analysis of
a dollar shock and a pound shock. The dollar shock would be an increase of the
dollar value compared to the euro by 20%. The pound shock would be a loss in value
of the pound of 20%. The resulting capital charge would be Mktfx = 0.2 million + 0.2
million = 0.4 million.
4.48 In situations where two foreign currencies are matched as in the example, the
proposed approach leads to the assumption that - compared to the local currency -
one exchange rate moves up and the other one down.
Consequently, the exchange rate between the two foreign currencies moves more
strongly than the assumed 20%. This could be considered to be a drawback of the
proposed approach. The effect could be avoided by allowing for diversification
between the shocks on different currencies. For example the results of the different
currency shocks could be aggregated with a correlation matrix.

Page 7 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

However, there are three arguments against such an amendment:

Firstly, it would be difficult to quantify the diversification between two exchange rates,
even if the simple approach is taken that the correlation factor for each pair of foreign
currencies is the same.
Secondly, the amendment would increase the complexity of the calculation. And
thirdly, situations as illustrated in the example can usually be avoided in practice.

No simplifications are proposed as the calculation is deemed relatively simple, and

for small undertakings it is unlikely that large currency positions across many
currencies would exist. For SAM, it is felt that the correlation between foreign
currencies relative to the Rand is much stronger than is the case for the Euro
(i.e. it is more likely that all foreign currencies would move in the same
direction relative to the Rand), and the direction of exposure for the majority of
insurers to one or more foreign currencies are likely to be the same.

The calibration of the module yielded the following (Summary):

4.88 Based on the assumptions contained in the explanatory text, CEIOPS has
calibrated the sub-module according to 99.5% VaR and a one year time horizon.
4.89 The currency risk sub-module will be calculated as set out in CP47. The
calibration of the required currency stresses is as follows, with the same magnitude
of stress applying in both upward and downward directions:
Danish Krone against any of EUR, Lithuanian litas or Estonian kroonat ±2.25%
Estonian Kroon against EUR or Lithuanian litas at ±0%
Latvian lats against any of EUR, Lithuanian litas or Estonian kroon at ±1%
Lithuanian litas against EUR or Estonian kroon at ±0%
Latvian lats against Danish Krone at 3.5%
All other currency pairs at ±25%.
4.90 The stress tests for currencies that are pegged to the euro revert to the standard
test of 25% when a country member of ERM II accepts euro as its currency or drops
out of the ERM II.

The full European calibration text from CEIOPS-SEC-40-10, the QIS 5 calibration
document, is given below:

3.1.3 Currency risk

3.129 CEIOPS-DOC-40/09 proposed a scenario-based approach for calculating the capital charge for
currency risk.

3.130 As set out in that paper, the capital charge arising from this sub-module will be Mktfx and will be
calculated based on two pre-defined scenarios: for each currency C, one scenario will consider a rise
Up
in the value of the foreign currency against the local currency and will deliver Mktfx,C ; the other
scenario will consider a fall in the value of the foreign currency against the local currency and will

Page 8 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

Down
deliver Mktfx,C . All of the undertaking's individual currency positions and its investment policy (e.g.
hedging arrangements,gearing etc.) should be taken into account. For each currency, the contribution
to the capital charge Mktfx,Cwill then be determined as the maximum of the results
Up Down
Mktfx,C andMktfx,C . The total capital charge Mktfx will be the sum over all currencies of Mktfx,C.

3.131 We note at this point that currency effects appear only in this sub-module. That is, the
calibration of the other market risk sub-modules has been carried out in such a way that currency
effects are stripped out.

3.132 The QIS3 technical specification document derived a 20% stress factor for currency risk, in
preference to the 25% stress factor proposed in QIS2.Furthermore, for QIS3 the implied stress factor
was derived assuming a diversified currency portfolio (i.e., 35% in USD, 24% in GBP, 13% in
Argentine Peso, 8% in JPY, 7% in SEK, 7% in CHF and 6% in AUD), which approximates the
currency positions held by Dutch financial institutions. In this exercise, currency exposure to emerging
markets was approximated by the Argentine Peso.

3.133 In our analysis, we show that the risk at the 99.5th percentile is exacerbated above the 20%
level proposed for QIS3 in portfolios whose composition is solely in currencies that suffered much
stronger moves. Furthermore, we use a currency portfolio diversified across 6 economies as a proxy
to currency exposures of emerging markets.

3.134 We use daily data to study the distribution of holding period rate of returns derived from EUR
and GBP currency pairs. Our data sample, sourced from Bloomberg, covers a daily period from
January 1971 to June 2009, a total of circa 10,000 observations across 14 currency pairs against
GBP. In addition, our sample consists of 14 currency pairs expressed against the EUR. For most
pairs, this sample covers a daily period spanning a period of 10 years starting in 1999 to 2009. We
compute annual holding period returns for the Japanese Yen (JPY), the Brazilian Real (BRL), the
Lithuanian Litas (LTL), the Indian Rupee (INR), the Chinese Yuan (CNY), the US, Hong Kong (HKD),
the Australian (AUD) and the New Zealand (NZD) Dollars, the Norwegian (NOK), Swedish (SEK) and
Danish (DKK) Krone, the Swiss Franc (CHF) and the British Pound (GBP).

3.135 Our proxy to emerging market economies is mainly a proxy to Pacific Basin economies. This is
a currency basket expressed against the EUR, and is equally distributed across CNY, INR, HKD,
AUD, BRL and ARS. We prefer to extend the definition of the emerging markets to include developed
economies, whilst including the dominant Latin American countries, Brazil and Argentina excluding
Mexico. The presence of the Australian and Hong Kong economy to our mix balances out the level of
the stress as we believe that insurance groups are more exposed to these economies across the
Pacific basin region. Below, we refer to this currency mix as EM.

3.136 We estimate the full probability density and especially the lower percentiles using non-
parametric methods as described in Silverman (1986). The figures below illustrate the standardised
probability density functions of a representative sample of six currency pairs against EUR, which are
implied from the annual holding period returns of the corresponding currency pairs.

3.137 QIS3 and 4 define a symmetric stress factor on the assumption that the percentage changes in
currency rates are normally distributed. A visual inspection of different standardised distributions,
which are plotted against the normal distribution shows that the data does not adhere to the laws of
normal distribution. Most distributions are skewed and exhibit excess kurtosis.

Page 9 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

3.138 The following two charts illustrate the 99.5th percentiles, left and right tail, of the annual holding
period returns of currency pairs against the EUR and GBP currencies, or for example Eastern
European currencies, were used instead of Pacific Basin economies respectively. The symmetric
band proposed by QIS3 is highlighted with a bold black line.

Page 10 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

3.139 The above results indicate that the year-on-year movements in currencies area symmetric at
the tail of the distribution and are likely to fall out of the symmetric 20% band. According to our results,
this is equally likely also for currency pairs against the British pound. Most breaches of the proposed
band occurred over 2008 to 2009 across both sets of currency pairs.

3.140 The following table illustrates the worst year-on-year percentage currency change estimated
within the period covered by our sample (1971 – 2009). In almost all cases, the currency pairs have
breached the proposed stress factor of 20%. Exceptions are the Danish krone and the Lithuanian
litas.

3.141 Given our analysis, we would not expect the symmetric stress factor of ± 20%to be a strict
representative of a 1 in 200 stress even for a well-diversified currency mix. In this particular case, if
we were to combine the above tabulated shocks with the specific currency mix proposed in QIS3
technical specification paper, the currency stress test is closer to 29%.

3.142 The level of the revised stress test crucially depends upon the choice of the optimal currency
weights, while the choice of the Argentine peso as a proxy to emerging markets introduces a degree
of bias as well as conservatism. We have carried out sensitivity testing on our result by varying both
sets of assumptions. The table below presents 16 sets of alternative choices of portfolio weights,
whilst we use a well diversified proxy of emerging markets termed EM.

3.143 The results demonstrate the sensitivity of our revised shock to the initial assumptions. Portfolio
1 represents the currency exposures of Dutch financial institutions, as proposed by QIS3 and
discussed above. Portfolio 2 tests the sensitivity of the revised shock to the Argentine peso, and uses
the alternative EM portfolio as a proxy to currency exposures across different markets. Portfolio 2
produces a 25% shock compared to the 29% shock produced by Portfolio 1.

Page 11 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

3.144 Ideally, we would prefer to have an average weight representing the average currency
exposures of European insurers to Pacific Basin economies. The lack of aggregate data encourages
further testing. We further carried out sensitivity analysis of our results to different weights of EM,
testing the sensitivity of the revised shock to concentration in EM of 10%, 20%, 25%,33% and 50% of
the total portfolio. The table below presents the results of this analysis.

Portfolio Weights

3.145 The sensitivity of the revised shock to alternative portfolio allocations to currencies in the
emerging market (EM) basket is analysed in portfolios 2, 4,6, 9, 10, 13 and 16. These results
demonstrate that the revised shock could vary from a maximum of 34% to a minimum of 20%. A
revised shock of 34%reflects a portfolio composition which is principally dominated by US dollar and
emerging market currency exposures. On the other hand a resulting minimum of 20% reflects a small
currency exposure of 10% to emerging market currencies, whilst maintaining all other allocations also
equal to 10%.

3.146 We have also investigated the sensitivity of the results to USD, JPY, CHF and GBP
concentrations as well as permutations of the portfolio in the absence of the emerging market basket.
In these cases, the results of the revised shock vary within 18% to 25%.

3.147 On the basis of the above sensitivity stress tests, we propose a revised stress factor of 25%.

3.148 We could further expand our emerging market basket to include other currency pairs and re-
test our proposed stress factor. Currency pairs that experience higher volatility than our proposed
basket may contribute positively and further increase the stress factor, whilst currencies with more
constrained volatility would not dramatically change our results.

3.149 In particular, we have investigated the inclusion of the Russian rouble and the Hungarian forint
in the currency basket, as proxy for eastern European currencies. However, there was no substantial
change in the overall results on the introduction of these two currencies.

Page 12 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

3.150 Exceptions to the above analysis are the member states of the European Exchange Rate
Mechanism (ERM II). The mechanism currently includes the Danish krone, the Estonian kroon, the
Lithuanian litas, and the Latvian lats:
 The Danish krone entered the ERM II in 1999, when the euro was created, and the
Denmark’s National bank keeps the exchange rate within a narrow range of ± 2.25% against
the central rate of EUR 1 = DKK 7.460
 The Lithuanian litas joined the ERM II on 28 June 2004.
 Latvia has a currency board arrangement, whose anchor switched from the IMF's SDR to the
euro on 1 January 2005.
 The Estonian kroon had been pegged to the German mark since its reintroduction on 20 June
1992, and is pegged to the euro since 1 January 1999. Estonia joined the ERM II on 28 June
2004.
For the latter 3 currencies, on the basis of ERM II the exchange rate is fixed within a broader nominal
band of ± 15%.

3.151 Moreover, for the three Baltic currencies the responsible national banks strictly control the
exchange rate to the euro:
 According to a commitment of the Bank of Lithuania the Lithuanian litas is pegged to the euro
with a fixed exchange rate of 3.4528 since 2 February 2002.
 As of 1 January 2005 the Latvian lats are pegged to the euro (at the rate 1 EUR = 0.702804
LVL. The Bank of Latvia performs interventions when the exchange rate of the lats exceeds
the normal fluctuation margins of ±1%.
 According to a commitment of the Bank of Estonia the Estonian kroon is pegged to the euro
with a fixed exchange rate of 15.6466 since 1 January 1999.

3.152 Based on these central bank commitments the currency stress for the Lithuanian litas and the
Estonian kroon against the euro can be neglected. The stress for the Latvian lats can be reduced to
1%.

3.153 The analysis set out above leads to the following proposal for calibration of the currency stress
scenario:

Page 13 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

3.154 The proposed currency tests for currencies that are pegged to the euro revert to the standard
test of 25% when a country member of ERM II accepts euro as its currency or drops out of the ERM
II.

QIS 2 Report

Nothing significant was highlighted in the report and currency risk constitutes a relatively minor part of
market risk overall.

QIS 3 feedback

There was a comment that the application of the full stress to Lesotho (and/or other currencies where
exchange rates are pegged to the Rand) is a bit harsh. This is dealt with elsewhere in this discussion
document and the task group remains of the opinion that it is very likely in a 1in200 year event that
currencies will be de-coupled and hence that the SA QIS 3 approach should be retained.

5.3 Other relevant jurisdictions (e.g. OSFI, APRA)

The approach of the Canadian regulator (OSFI) is as follows:

Up to date only quantitative impact studies are available on market risk, and little is
known as to the overall structure. The following seem to be what they are currently
thinking of, which seems to be quite similar to the Solvency II approach (although the
local currency is here specified as the Canadian dollar, which might be a draw-back
for those companies where the own funds are not denominated in that currency).
Currency market risk is the risk of economic loss due to changes in the amount and
timing of cash flows arising from changes in currency rates of exchange.
The solvency buffer for currency risk will be calculated by finding the net mismatch of
cash flow in each currency and measuring the effect of an immediate change of 20%
up and 20% down in that currency’s value against the Canadian dollar.

Australian approach (APRA)

The Australian approach is closer to the current treatment within Life Insurance CAR
in South Africa than Solvency II, and follows the specific structure of the Australian
regulation.

The Prudential Standards establish a two tier capital requirement on the statutory
funds of the life company with each tier considering the capital requirements in a
different set of circumstances. The first tier is intended to ensure the solvency of the
company. The second tier is intended to secure the financial soundness of the
company as a going-concern. It is expected, in most circumstances, that this second
tier will provide an additional buffer of capital above this minimum requirement.
However it will not always transpire that an additional buffer is necessary.

Page 14 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

This standard looks at the first tier capital requirement.

The Resilience Reserve

Mismatching of asset and liability exposures necessitates the provision of a reserve
for adverse movements in the asset values to the extent they will not be matched by
a corresponding movement in the liabilities.
When determining the impact of the various risks and adverse conditions on the
financial position on the fund, it is required to assess their impact consistently on all
assets and liabilities affected. This includes both beneficial and adverse combined
effects.

5.2.5 The Resilience Reserve as determined under the prescribed rules of Section 11
is based on the impact of market changes on the position of a statutory fund with a
simple asset and liability profile.
Where the fund is materially exposed to changes in investment market conditions
that are not captured by the application of the prescribed rules, a corresponding
additional provision must be made by the Actuary. The additional reserve must reflect
the purpose and principles of the Standard. It must provide a level of reserving that is
consistent with that applying under this Standard in respect of the changes in
investment market conditions explicitly considered under this Standard. For this
purpose, the Actuary may regard the prescribed requirements set out within this
Standard, when applied to the asset and liability profile of a typical life office, as
designed to provide a level of reserves which broadly meets the following
requirements:
a) Able to cover adverse changes in investment market conditions that would be
expected to arise once every 20 years;
b) Allowing a general time frame of 12 months in which the circumstances arise and
the actions under (c) and (d) below
follow;
c) The reserve required at the end of the period in (b) is able to be determined
assuming that a matched asset and liability profile is achieved and that the Solvency
Requirement of this Standard is otherwise satisfied at, or after, that time. This
includes making allowance for discretions in line with paragraph 3.2; and
d) Allowance for management corrective action to achieve a matched asset and
liability profile during the period in (b) is considered to be limited to highly reliable
actions only, with conservative response time allowances.

In this tier it seems that the “shock” for currency risk is a 10% decrease in all assets
denominated in a foreign currency.

This standard looks at the second tier capital requirement.

However, the prudent regulation of the life insurance industry requires that the level
of security offered to policy owners exceed that of a standard which secures
solvency. The Capital Adequacy Standard requires that each statutory fund has

Page 15 of 20
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

available sufficient additional assets to provide confidence in the longer term financial
strength of the fund. A fund that is capital adequate would have the ability to write
new business, in an unfettered manner, with the expectation of remaining solvent into
the future.

It seems that under this tier, there must be an additional 15% adverse movement in
the value of the assets denominated in foreign currencies.

5.4 Mapping of differences between above approaches (Level 2 and 3)

All approaches aim at relatively the same outcome, but the Solvency II approach seems to
be the most flexible and explicit.

6. ASSESSMENT OF AVAILABLE APPROACHES GIVEN THE SOUTH AFRICAN

CONTEXT

6.1 Discussion of inherent advantages and disadvantages of each approach

The Solvency II approach would be beneficial for promoting International Competition

and does not favour any particular currency as opposed to the Canadian approach.

Looking at South African data: (monthly exchange rates of the South African
rand/foreign currency)

The following graph visually demonstrates that there is a strong positive correlation
between the movements in different foreign currencies relative to the Rand. This is
daily data obtained from
http://www.oanda.com/currency/historical-rates/

Rand Exchange Rate 1990-2013

20

18

16

14

12
Rand

10

0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year

POUND US DOLLAR EURO AUS DOLLAR CHINESE YUAN

Page 16 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

The daily data were converted to year-on-year annual returns for each currency, and
the empirical 0.5th and 99.5th percentiles were derived, expressed as a percentage
and given in the table below. The period of investigation started in 1990 for the US
Dollar and from the points of data availability for the other currencies, up to the end of
March/start of April 2013.

POUND US DOLLAR EURO AUS DOLLAR CHINESE

percentile YUAN
0.005th -28.55 -32.07 -22.87 -21.40 -32.20
. 0.995th 45.41 53.33 49.51 50.64 64.04

Based on this, it is probably reasonable to use a stress of 50% in the case that a
currency’s value increases relative to the Rand, and 30% in the case that its value
decreases against the Rand. As can be expected, note that the range of answers
doesn’t differ as much from each other as that in the European calibration and seem
to be more extreme. Going forward, it is proposed that this analysis be extended to
cover some of the main African countries.

6.2 Impact of the approaches on EU 3rd country equivalence

Demonstrating equivalence would be easier using an approach similar to the

Solvency II approach. It is likely that much reliance can be placed on the principles
followed by CEIOPS in the calibration process.

6.3 Comparison of the approaches with the prevailing legislative framework

Within the current calculation of the OCAR (Ordinary CAR), one element is resilience
CAR which is calculated by stressing interest rates, as well as shocking the value of
a number of assets, e.g. equity, property etc. In this stress scenario, a minimum
shock of 20% is applied to all foreign assets including cash. However, in the case of
equity the shock on the value of the asset already exceeds the 20% and no further
allowance is prescribed for currency risk.

Market Risk is therefore modelled by this scenario. Furthermore, the capital

requirement for PGN110 embedded investment derivatives is also allowed for in this
scenario by recalculating the reserve under this scenario and subtracting the reserve
already held. This is then added to the resilience CAR component.

The current CAR aims at a 95% level of confidence, which is much below the 99.5%
used in Solvency II. The Solvency II approach more explicitly allows for this risk, risk
mitigation techniques and differing exposures possible.

Management action is easier to apply in the current CAR scenarios and can be
applied consistently for the whole scenario and allowing for interdependencies (the

Page 17 of 20
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

modular structure within Solvency II is a bit of a drawback compared to CAR when it

comes to allowing for management actions).

6.4 Conclusions on preferred approach

The task group recommends that an approach based on Solvency II be adopted for the
currency risk sub-module, for comparability across companies and consistency with the way
other sub-modules are treated. In line with feedback from SA QIS 1 to 3 as well as new
developments in Solvency II, it is proposed that the Solvency 2 approach be retained with the
following amendments /additions / clarifications:
 Where there is a direct causal relationship between currency risk and policyholder
benefits, allowance for policyholder behaviour should be made within the currency risk
sub-module. Non-causal interrelationships should be reflected in the lapse sub-module
and the correlation matrices to be used. This is discussed more fully in Discussion
Document 48.
 Currency risk can be assumed to arise from events that apply to the industry as a whole
as opposed to company-specific events.
 Impairments should be made to the risk mitigating effect of risk mitigating contracts, as
specified in [Reference to Impairment of risk mitigating contracts within the Market risk
module/CDR module unless changed to explicit allowance in CDR module]
 The SA QIS 2 and 3 approach of grouping together all foreign currencies against the
Rand should be retained. This recognises the fact that the Rand may be less stable
against foreign currencies compared to the stability of the Euro against other foreign
currencies.
 Different size of stress, also dependent on the direction of the stress as follows: a stress
of 50% in the case that a currency’s value increases relative to the Rand, and 30% in the
case that its value decreases against the Rand.
 The treatment of dual-listed shares is clarified and it is proposed that a different
approach be followed than in Solvency II. The treatment proposed is consistent with the
treatment of dual-listed shares in Position Paper 47 (equity Risk).

The main difference of this proposed approach with the SA QIS 2 specification (except for
the 6 points above) is that it is proposed to use the same % stresses to currencies pegged to
the Rand and other currencies. This is to reflect the very likely event in a 1-in-200 year
scenario that the currencies may be de-coupled. This is inconsistent with the way that
Solvency 2 treats exchange rate stresses between the Euro and currencies pegged to it
and/or between different currencies that are both pegged to the Euro.

Page 18 of 20
 Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

7. RECOMMENDATION

Currency risk

Currency risk (Mktfx)

Description

Currency risk arises from changes in the level or volatility of currency exchange
rates.

Insurers may be exposed to currency risk arising from various sources, including
their investment portfolios, as well as assets, liabilities and investments in related
undertakings. The design of the currency risk sub-module is intended to take into
account currency risk for an insurer arising from all possible sources.

The local currency is the currency in which the insurer prepares its financial
statements. All other currencies are referred to as foreign currencies. A foreign
currency is relevant for the scenario calculations if the amount of basic own funds
depends on the exchange rate between the foreign currency and the local currency.

Note that for each relevant foreign currency C, the currency position should include
any investment in foreign instruments where the currency risk is not hedged. This is
because the stresses for interest rate, equity, spread/credit default and property risks
have not been designed to incorporate currency risk.

Dual-listed shares which are listed on the JSE can be assumed to be denominated in
Rand and not sensitive to exchange rate movements. If, however, dual listed shares
have been purchased on an offshore exchange, the currency shocks apply. That is,
the currency of the listing determines whether the currency shock applies.

Non-listed equity and property should be assumed to be sensitive to the currency of

its location.

Impairments should be made to the risk mitigating effect of risk mitigating contracts

Dynamic policyholder behaviour should be allowed for in the calculation of Mktfx

where a causal relationship exists between the changes in the currency rates and
the behaviour under consideration.

100% of currency risk should be assumed to arise from an industry-wide event.

Input

The following input information is required:

BOF= Basic Own Funds

Page 19 of 20
Solvency Assessment and Management: Steering Committee
Position Paper 45 (v 3) Currency Risk

Output

The module delivers the following output:

Mktfx= Capital requirement for currency risk

Calculation

The capital requirement for currency risk is determined as the result of two pre-
defined scenarios separately:

Mktfxup= max(∆BOF| fxupward shock;0)

MktfxDown= max(∆BOF| fxdownward shock;0)

The scenario fxupward shock is an instantaneous rise in the value of 50% of all foreign
currencies against the local currency (Rand depreciates). The scenario fxdownward shock
is an instantaneous fall of 30% in the value of all the currencies against the local
currency (Rand appreciates).

For example, the “stressed” Dollar-Rand exchange rate in the upward stress
scenario is determined as

where Eup is the Dollar-Rand exchange rate after the shock and E is the current
Dollar-Rand exchange rate. At 31 December 2011 the exchange rate was
$0.1225/R1 and therefore in the scenario fx upward shock this becomes ($0.1225
/1.5)/R1.

The fxdownward shock is then Edown = E/(1-0.3)

All of the (re)insurer’s individual currency positions and its investment policy (e.g.
hedging arrangements, gearing etc.) should be taken into account. The result of the
scenarios should be determined under the condition that the value of future
discretionary benefits can change and that the undertaking is able to vary its
assumptions on future bonus rates in response to the shock being tested. The
resulting capital requirements are MktfxUp and MktfxDown.

The capital requirement Mktfx should be determined as the maximum of the values
MktfxUp and MktfxDown. This is subject to a floor of 0.
.

Page 20 of 20