South African Insurance Outlook 2021 - Navigating the insurance - Deloitte
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
South African Insurance Outlook 2021 Navigating the insurance landscape 1
S A INSUR A NCE OUTLOOK | PER SPEC TI V E S ON THE INSUR A NCE L A NDSC A PE
Introduction
Contents Overview of the 2020 financial
and EV results
COVID-19 – Once bitten, twice shy.
A pandemic’s impact on stress
testing frameworks
Optimising the life insurance
regulatory balance sheet
Artificial intelligence and
machine learning
Using technology to combat
insurance fraud
Proactively managing conduct risk
Understanding a more complete
IFRS 17 picture
IFRS 17: Controls and financial
reporting under the new standard
VAT: Closing the compliance gap
2
S A INSUR A NCE OUTLOOK | INTRODUC TION
Introduction
Introduction Overview of the 2020 financial
and EV results
COVID-19 – Once bitten, twice shy.
Our South African Insurance Outlook 2021 publication The past year has shown that capital coverage of the We hope you enjoy reading our publication and look
A pandemic’s impact on stress
reflects on the past year, and shares some of our thoughts insurance industry has not been affected as much as might forward to hearing any thoughts or comments that you testing frameworks
on trends that will shape the industry in years to come. have been feared at the start of the pandemic. However, may have on any of the articles.
Inside is a collection of articles that were penned by it has highlighted the importance of a robust capital
Deloitte professionals who provide services to the South management and capital optimisation strategy. Optimising the life insurance
African insurance industry. Our team unpacks some options available to embed
Authors regulatory balance sheet
capital optimisation into your business operations. Andrew Warren
Our focus, this time, is on business, capital, governance Director, Actuarial and
Artificial intelligence and
and financial reporting priorities in the local insurance The pandemic has prompted change in a sector that Insurance Solutions, Insurance
machine learning
industry. was already dealing with systemic challenges. The Sector Advisory Leader,
Deloitte Africa
silver lining, though, was the industry’s response that
anwarren@deloitte.co.za
Not surprisingly, many of the articles comment on the led to unexpected improvements in some areas such +27 (011) 202 7423 Using technology to combat
impact of COVID-19 on the industry. The impact of the as customer satisfaction and communication. In this insurance fraud
pandemic and the lockdown response was a key driver publication we discuss how artificial intelligence (AI)
of the 2020 financial results of insurers, and we have continues to infiltrate every corner of the world, and how Gerdus Dixon Proactively managing conduct risk
unpacked the themes that have emerged in the results insurers are implementing machine learning methods that Partner, Audit and Assurance,
released recently by the listed insurers. The article explores underpin AI. Insurance Sector Audit Leader,
both the IFRS and embedded value consequences of the Deloitte Africa Understanding a more complete
increased (and decreased) claims rates and the impact of It would be remiss in 2021 for a publication like ours not gdixon@deloitte.co.za IFRS 17 picture
+27 (021) 427 5574
sales volumes and policy retention. to comment on IFRS 17. This financial reporting standard
will become mandatory for all insurers with financial years
While it is only natural to want to put the pandemic in the commencing on or after 1 January 2023. We comment on IFRS 17: Controls and financial
rear view mirror, as an industry we should take on board how insurers can manage their businesses using insights reporting under the new standard
learnings from the pandemic. The Deloitte team have gained from IFRS 17, rather than merely seeing the financial
brought insights to the question of how frequently we can reporting standard as a matter of compliance. And we
VAT: Closing the compliance gap
expect an event like COVID-19 to occur when compared know that the standard will bring changes to the financial
to the calibration of selected modules within the SAM control environment at insurers, and our team highlighted
regulatory regime. where management and audit committees should expect
change.
3
S A INSUR A NCE OUTLOOK | OV ERV IE W OF THE 2 02 0 FIN A NCI A L A ND EMBEDDED VA LUE R E SULT S OF THE L A RGE S T FI V E LI S TED INSUR A NCE GROUP S IN SOUTH A FR IC A
Introduction
Overview of the 2020 financial and embedded Overview of the 2020 financial
and EV results
value results of the largest five listed insurance COVID-19 – Once bitten, twice shy.
A pandemic’s impact on stress
groups in South Africa
testing frameworks
Optimising the life insurance
regulatory balance sheet
The completion of the 31 December 2020 financial reporting cycle by the listed
insurance groups in South Africa offers an opportunity for reflection. Their
Artificial intelligence and
results, achieved against the backdrop of a global pandemic, are scattered with
machine learning
references to muted new business volumes, increased claims and short-term
COVID-19 related provisions for poorer expected persistency and mortality.
Using technology to combat
While these features were an unavoidable reality for the 12-month reporting insurance fraud
period ended 31 December 2020, the same results also point to an industry
that delivered for its policyholders and the broader economy in uncertain times.
Proactively managing conduct risk
The results show the impact of premium holidays and reductions, significant
business interruption (BI) claims and interim relief payments to policyholders.
These customer relief measures, coupled with the wider assistance offered by Understanding a more complete
the industry in the form of relief funds, had a positive impact in South Africa, as IFRS 17 picture
well as the other countries where the groups have a foothold.
In this article we comment on themes evident in the International Financial IFRS 17: Controls and financial
Reporting Standards (IFRS) results, regulatory capital position and embedded reporting under the new standard
value (EV) results of the largest five insurance groups that collectively represent
more than 80% of the local industry’s premiums and assets. We analysed the
VAT: Closing the compliance gap
results in aggregate to form an industry view, rather than comment on the
results of the individual insurance groups.
4
S A INSUR A NCE OUTLOOK | OV ERV IE W OF THE 2 02 0 FIN A NCI A L A ND EMBEDDED VA LUE R E SULT S OF THE L A RGE S T FI V E LI S TED INSUR A NCE GROUP S IN SOUTH A FR IC A
Introduction
IFRS results and regulatory capital position Consolidated results of the five large listed insurance groups in South Africa
Three of the five insurance groups referenced in this as at and for the 12 months ended 31 December 2020 Overview of the 2020 financial
article have 31 December year ends, and two of the and EV results
groups have 30 June year ends. For the two groups
that have 30 June year ends we used their 2020 interim Momentum COVID-19 – Once bitten, twice shy.
results and historic announcements to calculate Metropolitan A pandemic’s impact on stress
pro forma results for a 12-month period ended 31 testing frameworks
December 2020. The adjacent table summarises the
IFRS results for the five insurance groups, on the
basis described. Where the article refers to “total” or
Optimising the life insurance
regulatory balance sheet
“aggregated” it is the sum of the five insurance groups.
Despite the local equity markets drop in value in March
Artificial intelligence and
2020 the markets recovered during the remainder of machine learning
the year to end relatively unchanged compared to the
start of 2020 (using SWIX as a reference). That recovery
allowed insurance groups, on an aggregated basis, to Using technology to combat
report a respectable 3.8% increase in assets. Insurance insurance fraud
groups are also impacted by the value of assets
throughout the year though. Old Mutual points out in
Proactively managing conduct risk
their results commentary that the average market levels
during 2020 were 6.7% lower than the prior year, which
negatively impacted asset-based fees for insurance Understanding a more complete
groups that manage and administer customer assets. IFRS 17 picture
The aggregated equity for the insurance groups
decreased by R15.6 billion, or 6.0%. The lower equity is IFRS 17: Controls and financial
mostly a function of the aggregated loss after tax of reporting under the new standard
R4.7 billion reported by the insurance groups as well
as ordinary dividends paid of R12.7 billion (2019: R16.2 VAT: Closing the compliance gap
billion). The dividend declarations during 2020 and
into 2021 were a mixed bag ranging from some groups
withholding dividends to maintaining past dividend
policies, but with a downward adjustment for 5
specific uncertainties.
S A INSUR A NCE OUTLOOK | OV ERV IE W OF THE 2 02 0 FIN A NCI A L A ND EMBEDDED VA LUE R E SULT S OF THE L A RGE S T FI V E LI S TED INSUR A NCE GROUP S IN SOUTH A FR IC A
Introduction
Despite the reduction in equity the insurance groups still contracted by 7% in 2020, and more severely in some For example, Santam as the largest short-term insurer in
reported relatively healthy regulatory Solvency Cover other African countries. the country established a claims provision of R3 billion,
Overview of the 2020 financial
Ratios (SCRs). Refer to the graph below illustrating some of net of reinsurance for its BI exposure. The industry’s and EV results
On an aggregated basis the insurance groups’ profit
the insurance group’s SCR ratios: accounting for BI claims at 31 December 2020 followed
before tax of R48.3 billion reported in 2019 reduced to
the rulings in both South Africa and the United Kingdom
R4.3 billion in 2020 (-91%). The financial results reflect the COVID-19 – Once bitten, twice shy.
SCR Cover Ratios courts that addressed the uncertainty around the
impact of: A pandemic’s impact on stress
application of BI clauses. These proceedings confirmed
260%
• Lower new business volumes as adviser productivity, in the testing frameworks
that cover should be provided for BI losses caused by the
211%
210%
199% absence of face-to-face sales, was significantly impacted government enforced national lockdown, provided there
181% 189% 185% 191%
180%
during the initial lock down period was an instance of COVID-19 within the defined radius of Optimising the life insurance
162%
160% the policyholder’s business. Insurers have accepted the regulatory balance sheet
• COVID-19 customer support mechanisms, such as premium
Cover ratio
holidays, and other direct financial support decisions of the courts, although further legal processes
110%
afoot to confirm the length of the indemnity period that
• Increased death claims – the Association for Savings and Artificial intelligence and
applies.
60% Investment South Africa (ASISA) noted that the South machine learning
African life assurance industry recorded 116 774 more A key feature of many of the results announcements is
10%
death claims in 2020 than it did in 2019, an increase the short-term provisions created for the anticipated
Momentum
Liberty Old Mutual
Metropolitan
Sanlam
of 37% impacts of worsening mortality, morbidity and persistency Using technology to combat
-40%
experience related to COVID-19. The provisions for those
insurance fraud
SCR Cover Ratios as at 31 Dec 2019 SCR Cover Ratios as at 31 Dec 2020 • Poorer persistency for life insurance policies – while most
insurers commented that the actual lapse experience in groups that disclosed them explicitly total more than
the current year tracked favourably against expectations R10 billion at 31 December 2020. Some groups noted that Proactively managing conduct risk
It is not always easy to make direct comparisons between
as policyholders opted to hold onto their insurance their provisions needed to be bolstered in the second
the IFRS results of the individual insurance groups as their
policies in uncertain times, the assumptions for short- part of 2020 as the actual experience was more severe
accounting policies for the recognition of negative reserves Understanding a more complete
term future terminations have been bolstered than their initial modelling, or additional information had
and revenue are often not consistent, and the level of IFRS 17 picture
become available. For example, Momentum Metropolitan
prudence applied in setting policyholder liabilities varies. • Significantly lower investment returns on shareholder Holdings note in their interim results announcement:
For example, in the current year Sanlam reported that it assets coupled with a reduction in asset-based fees on “Mortality data from the South African Medical Research
had previously created a pandemic reserve of customer assets (following on from lower assets under IFRS 17: Controls and financial
Council, indicates that recorded Covid-19 deaths are
R760 million that was now to be utilised. By contrast, management during the year) reporting under the new standard
understating the full mortality impact of the pandemic. In line
most other insurers did not previously hold any such
• Improved non-life personal lines underwriting results – lower with the SAMRC data, our mortality claims experience to date
reserve. Suffice to say that even if pandemic reserves were
personal lines claims volumes, particularly for motor has been more severe than our initial modelling. We therefore VAT: Closing the compliance gap
commonplace these reserves would unlikely have been
vehicles, in the first half of the year during the extended increased our COVID-19 provision…”.
sufficient to absorb all that 2020 brought to bear.
national lockdown enforced by government The short-term provisions generally allow for increased
The current year’s results were achieved against the mortality claims, higher terminations, reduced return-to-
• Relief payments to clients in the hospitality and leisure
backdrop of muted economic activity in Africa, even before 6
sector coupled with significant liabilities created to settle BI work experience on disability income claims in payment,
the pandemic hit. South Africa’s Gross Domestic Product as well as BI claims.
claims.
S A INSUR A NCE OUTLOOK | OV ERV IE W OF THE 2 02 0 FIN A NCI A L A ND EMBEDDED VA LUE R E SULT S OF THE L A RGE S T FI V E LI S TED INSUR A NCE GROUP S IN SOUTH A FR IC A
Introduction
Embedded Value results The slowdown in the economy and the pressure on household incomes further added to
The impact of these short-term COVID-19 assumptions and provisions is also evident in lower new business volumes, although arguably the increased need for life and disability Overview of the 2020 financial
the disclosed embedded value results, as can be seen in the graph below under the insurance during a pandemic may have had an offsetting impact. Several insurers indicated and EV results
‘Operating assumption and model changes’. The graph presents the aggregated position for marked increases in the sale of conventional annuity business.
the insurance groups. The negative impact was observed in both the adjusted net worth
VNB margins also decreased, with most insurers indicating higher per policy expenses COVID-19 – Once bitten, twice shy.
(as described in the paragraphs above) and value of in-force business (VIF).
A pandemic’s impact on stress
being the key driver. Insurers with fixed distributions costs (e.g. workforce agents)
testing frameworks
The aggregated EV decreased from R274.9 billion to R258.4 billion, or 6.0%. In addition to were harder hit due to their inability to reduce these costs as sales volumes reduced.
the impact of the short-term COVID-19 assumptions, economic/investment variances had Some insurance groups observed a shift towards more affordable products, as well as
a significant impact on reducing the aggregate VIF and therefore EV. policyholders reducing their cover or benefits, generally resulting in lower margins for new Optimising the life insurance
and exciting policies. regulatory balance sheet
Aggregated change in EV for the 12 months ending 31 Dec 2020
R’millions While there is largely consistency in how insurance groups went about setting up their
short-term COVID-19 provisions/reserves, it is unclear to what extent it pulls through to Artificial intelligence and
320 000 machine learning
the VNB assumptions. For example, there are limited disclosures on the extent to which
insurance groups allowed for changes in future mortality and persistency experiences,
300 000
related to COVID-19, in the VNB numbers. Using technology to combat
insurance fraud
Value of New Business and VNB Margins
280 000
274 876 for the 12 months ended 31 December 2020 and 2019
Proactively managing conduct risk
3 000 3.50%
258 449
260 000
2 500 2.98% 3.00% Understanding a more complete
2.58% 2.58%
IFRS 17 picture
2.50%
2 000
240 000
Rands (million)
Expected Operating Operational Economic Investment Extraordinary Foreign Other Capital 2.00%
EV Start VNB Dividends Other EV End
earnings assumption experience assumption experience expenses / currency
and model changes development movements
flows and
transfers 1 500 IFRS 17: Controls and financial
changes costs 1.50% reporting under the new standard
1 000
1.08%
1.00%
The Value of New Business (VNB) made a smaller contribution to the aggregate VIF 0.90%
0.84%
500
0.69% VAT: Closing the compliance gap
compared to previous years, with most insurance groups suffering a decrease in VNB, 0.50%
some significantly so. 0.10%
0 0.00%
Liberty Old Mutual Momentum Sanlam Discovery*
New business volumes deteriorated as traditional face-to-face distribution channels Metropolitan
took strain in generating sales, with mixed success in transitioning to digital channels.
VNB 12 months to 31 Dec 2019 VNB 12 months to 31 Dec 2020 New Business Margin Dec 2020 New Business Margin Dec 2019 7
*Discovery group VNB margin not disclosed.
S A INSUR A NCE OUTLOOK | OV ERV IE W OF THE 2 02 0 FIN A NCI A L A ND EMBEDDED VA LUE R E SULT S OF THE L A RGE S T FI V E LI S TED INSUR A NCE GROUP S IN SOUTH A FR IC A
Introduction
In summary
The record books may show 2020 as a year in which reported Overview of the 2020 financial
financial results were well below expectations, it is by no means and EV results
the full story. It was a year where the industry again showed its
resilience, while at the same time positively impacting the lives
COVID-19 – Once bitten, twice shy.
of its customers at a time of great financial need. Perhaps less
A pandemic’s impact on stress
obvious, it was also a year where the industry made meaningful testing frameworks
strides in changing its operating model through introducing
digital capabilities that will transform the industry for many years
to come. While it might be somewhat difficult for the authors
Optimising the life insurance
regulatory balance sheet
of this article to say, being an accountant and an actuary, but
perhaps 2020 is a year in which we need to look beyond just the
numbers to see the full picture.
Artificial intelligence and
machine learning
Authors
Using technology to combat
Gerdus Dixon insurance fraud
Partner, Audit and Assurance,
Insurance Sector Audit Leader,
Deloitte Africa Proactively managing conduct risk
gdixon@deloitte.co.za
+27 (021) 427 5574
Understanding a more complete
IFRS 17 picture
Carike Nel
Director, Actuarial and IFRS 17: Controls and financial
Insurance Solutions, reporting under the new standard
Deloitte Africa
canel@deloitte.co.za
+27 (021) 427 5358 VAT: Closing the compliance gap
8
S A INSUR A NCE OUTLOOK | COV ID -19 – ONCE BIT TEN, T W ICE SH Y
Introduction
COVID-19 – Once bitten, twice shy Overview of the 2020 financial
and EV results
A pandemic’s impact on stress testing frameworks COVID-19 – Once bitten, twice shy.
A pandemic’s impact on stress
Introduction were imposed, social distancing was the new norm and from its pre-crash high in December 2019 to its lowest testing frameworks
As we reflect on 2020 that seems to have gone by in a both fist bumps and elbow shakes were gaining traction. point in March 2020, which was a significant fall of 34%.
flash, we have seen the remarkable ability of the human At the time many felt like Henny Penny and were thinking But even this movement translates to only a 1-in-100 year Optimising the life insurance
race to adapt and innovate, with the insurance market that the sky was falling. But was it really a 1-in-200 year event according to the Standardised Formula regulatory balance sheet
being no different. Remote working was set up with near event when compared to the underlying calibrations of the calibrations.
perfect transition, that saw insurers adapting to change Standardised Formula?
and the use of technology on a level never seen before in With equities comprising only 14% of non-life insurers’ Artificial intelligence and
the insurance sector. However, the negative impact of the In answering this question, we look at the key risks that are investment portfolios on average1, the impact of the
machine learning
COVID-19 pandemic left its mark – from the tragic loss of captured by the Standardised Formula SCR, namely market falling equity market was not as significant as might
life to the significant economic and societal impacts and, of risk, life underwriting risk, non-life underwriting risk and have been expected. In contrast, life insurers were
Using technology to combat
course, the specific effects on the insurance sector. operational risk, and assess how the emergence of those more exposed with equities representing on average insurance fraud
risks under COVID-19 impacted on insurers’ solvency. 43% of their overall investment portfolio2. Despite this,
With the promulgation of the Solvency Assessment and however, most of these equity investments relate to
Management (SAM) regulatory framework on 1 July 2018, Market Risk investments made on behalf of policyholders through Proactively managing conduct risk
insurers have increasingly asked the question of what a In the market risk module we saw that equity risk, interest with-profits policies and linked business. With this
“1-in-200 year” event would actually look like. This paper rate risk and currency risk were the risks within the risk passed on to the policyholders, most SCR ratios of
Understanding a more complete
aims to explore how frequently we can expect an event like Standardised Formula that were most significantly affected life insurers were largely unaffected. Insurers who offer
IFRS 17 picture
COVID-19 to occur when compared to the calibration of by COVID-19. downside protection on their equity-backed policies
selected modules within the SAM Standardised Formula. saw a significant increase in their investment guarantee
It further goes on to recognise potential areas where • Equity Risk reserves, with the fall in the markets also resulting in an IFRS 17: Controls and financial
COVID-19 has highlighted shortfalls in the Solvency Capital According to the calibrations of the Standardised increase in the volatility of equity markets, with some of reporting under the new standard
Requirement (SCR) for consideration in insurers’ economic Formula, which consider annual movements of an this offset by the hedging strategies that were
capital modelling and broader Own Risk and Solvency insurer’s overall equity exposure, the MSCI World Index employed.
Assessment (ORSA). saw a 1-in-9 year event for the 12 months ending March VAT: Closing the compliance gap
2020 (13% fall from March 2019 to March 2020) while
The journey thus far the JSE All Share Index experienced a 1-in-10 year event Reference:
If we look back to March 2020 we can remember that over the same period (21% fall). However, we could also 1. Prudential Authority – Non-life industry experience 2018
2. Prudential Authority – An overview of the experience of life insurers in 9
equity markets were in freefall, worldwide lockdowns consider the intra-year drop in the JSE All Share Index South Africa for 2018
S A INSUR A NCE OUTLOOK | COV ID -19 – ONCE BIT TEN, T W ICE SH Y
Introduction
Observation Observation Observation
Overview of the 2020 financial
It is imperative that insurers are well prepared for Insurers need to understand the level of Insurers with long-dated liabilities need to assess and EV results
these extreme market movements, with a clearly diversification assumed in the calibrations of the their exposure to non-parallel movements in the yield
defined approach for investment decisions under Standardised Formula to identify areas where curve, e.g. tilts and changes in shape, as these are not
such conditions, allowing careful and objective economic capital requirements might need to considered within the Standardised Formula. This is COVID-19 – Once bitten, twice shy.
consideration when markets are in free fall, reducing deviate from the Standardised Formula. While especially important where asset-liability matching
A pandemic’s impact on stress
testing frameworks
the risk of knee jerk-reactions. This should include a the Standardised Formula does not allow for is not based on matching cash flows, but rather
focus on the hedging of investment guarantees, and diversification between different currencies, the past based on matching duration or overall movements in
stress testing the effectiveness of those hedges under year has made it clear that the volatility of the Rand is liabilities. This shortfall in the Standardised Formula Optimising the life insurance
extreme market movement scenarios. not the same for all foreign currencies, e.g. the USD/ was noted during SAM’s development and insurers’ regulatory balance sheet
ZAR exchange rate tends to be more volatile than risk management functions could benefit from
other exchange rates. revisiting the relevant position papers and discussion
Artificial intelligence and
• Currency Risk documents to understand the shortcomings of the
machine learning
Similar to equity risk, the Standardised Formula Standardised Formula, not only for interest rate risk,
calibrations (considering annual movements) suggest • Interest Rate Risk but also for other risk modules.
that the GBP/ZAR movement for the twelve months Interest rate risk was the most severely affected market Using technology to combat
to April 2020 equates to a 1-in-10 year event (19% risk module, with nominal yields reducing by up to 40% insurance fraud
depreciation). For the USD/ZAR, we saw a 1-in-17 year at short durations (equivalent to a 1-in-100 year event)
Observation
event over this same period (12% depreciation). and increasing by up to 80% at longer durations (which is
Many insurers have defaulted to using the PA’s risk Proactively managing conduct risk
However, as with equity risk, we could also consider the much more severe than a 1-in-200 year event).
free curve for other calculation bases, e.g. IFRS and
intra-year movement from the most recent strongest Subsequent to this volatility, the PA had updated the
Embedded Value reporting, but, after the volatility
position of the ZAR against the USD, in December 2019, constituent bonds used to derive the risk-free curve. Understanding a more complete
experienced during 2020, insurers were urgently
to the weakest position in April 2020, over which a 27% This update had very little impact at short durations, but IFRS 17 picture
considering alternative curves. With a variety of risk-
depreciation was experienced. Even this only translates significantly reduced the impacts at longer durations.
free curves available, it’s important for insurers to
to a roughly 1-in-30 year event when compared against Had the new bond constituents been used throughout
have a sufficiently deep understanding of any yield
the Standardised Formula calibrations. 2020 the impacts at long durations would be IFRS 17: Controls and financial
curve that is used, for example an understanding of
somewhere between a 1-in-20 and a 1-in-50 year event, reporting under the new standard
the curve construction methodology (interpolation
The average life and non-life insurer have limited depending on the duration. This less severe impact
and extrapolation), selection of bond constituents and
foreign exposures and hence the impact of the ZAR is also more in line with the observed movements in
whether historically the curve has displayed desirable VAT: Closing the compliance gap
deterioration had an insignificant effect on most government bond yields.
behaviour, especially during times of market stress.
insurers’ SCR ratios over the last year.
10S A INSUR A NCE OUTLOOK | COV ID -19 – ONCE BIT TEN, T W ICE SH Y
Introduction
Further to this there are also longer-term mortality recession caused between 2.2 and 3.0 million
Observation impacts from both lockdown and the associated economic job losses6 and 7, well in excess of the 1 million jobs lost Overview of the 2020 financial
Insurers can improve economic capital models recession, which have been estimated to be multiples of during the 2008 Global Financial Crisis8. Retrenchment and EV results
by recalibrating many of the market risk modules, the direct excess deaths, with this impact being spread risk might be negligible at an industry level9, but there are
using more recent and larger data sets than those over the next 10 years5. These longer-term impacts are a number of insurers with significant exposure thereto,
underlying the Standardised Formula. A typical COVID-19 – Once bitten, twice shy.
expected to be concentrated in lower income families leading to such insurers recognising large retrenchment
example is interest rate risk, where there is significant
A pandemic’s impact on stress
where poverty induced deaths are likely to occur. However, losses. This is the second round of such losses in less testing frameworks
experience available beyond the data set that was higher income families are also expected to be subject to than 15 years, illustrating that retrenchment experience
used to calibrate the Standardised Formula. Re- increased risk from at least a few factors, including delayed is highly volatile, but also that it behaves more like a
calibrating using more recent data could also better cancer diagnoses, emotional impacts from lockdown and short-lived catastrophe and less like the long-term
Optimising the life insurance
reflect changes in market behaviour, like the impact of regulatory balance sheet
potential long-term COVID-19 symptoms. To the extent upward stress included in the Standardised Formula.
technology and automated trading on equity markets. that these deaths occur in the most impoverished of This is especially relevant for business with shorter
communities the impact on the insurance industry would contract boundaries, where the impact of such a short-
Artificial intelligence and
be limited, but the loss of human life remains equally tragic. lived catastrophe might not be captured sufficiently by machine learning
Life Underwriting Risk When including the impact of these longer-term deaths the the Standardised Formula.
While pandemics have always been a classical stress test severity of the COVID-19 pandemic becomes undoubtably
for life insurers, COVID-19’s far reaching complexity could more severe than the 1-in-200 year event envisioned by the Using technology to combat
not have been captured in the simplicity of a theoretical Observation
Standardised Formula. insurance fraud
stress test. Insurers with any material retrenchment risk need
• Morbidity Risk to take great care in ensuring their economic capital
• Mortality Catastrophe Risk Life insurers were also subject to other claim variances, and ORSA stresses make an appropriate allowance Proactively managing conduct risk
COVID-19’s mortality impact has taken much longer including increases in temporary disability claims due to for the true nature of retrenchment risk. In light
to materialise than the three months assumed in the severe COVID-19 conditions, where waiting periods could of its volatility and potentially large and relatively
Standardised Formula. With vaccines now available frequent losses, risk appetite policies also need to be Understanding a more complete
be very short, as well as additional hospital cash claims
there is renewed hope that we can start estimating IFRS 17 picture
due to COVID-19 submissions, although this is expected reassessed to ensure there are appropriate risk limits
the pandemic’s ultimate impact. In this regard our to be more than offset by a reduction in submissions in place for retrenchment risk.
analysis is based on the estimated impact of a third arising from elective procedures. There could also be IFRS 17: Controls and financial
wave, without any fourth wave impact and after netting lockdown related impacts on morbidity claims, e.g. Reference: reporting under the new standard
off other lockdown related impacts like limited deaths from temporary changes in lifestyle activities and alcohol
3. Extrapolated from SAMRC Excess Deaths data
4. Swiss Re – Pandemic influenza: A 21st century model for mortality shocks
due to influenza and lower accidental deaths. The availability. 5. Business Tech – ‘Real and dire possibilities’ facing South Africa after lockdown:
excess deaths within the South African population is Dawie Roodt VAT: Closing the compliance gap
6. Statistics South Africa – Quarterly Labour Force Survey,
then estimated to reach anywhere between 180 000 and • Retrenchment Risk Quarter 2: 2020
250 000 by the end of 20213. According to the The lockdown induced recession is the worst economic 7. NIDS-CRAM – Overview and Findings, NIDS-CRAM Synthesis
Report Wave 1
Standardised Formula calibrations this could be expected contraction our country has faced, at least since 1960 8. Business Tech – South Africa lost 1 million jobs because of the
when economic growth data became available. This 2008 recession – here’s why this one could be even worse
to happen once every 250 to 370 years4. 11
9. Prudential Authority – An overview of the experience of life insurers in
South Africa for 2018S A INSUR A NCE OUTLOOK | COV ID -19 – ONCE BIT TEN, T W ICE SH Y
Introduction
• Lapse Risk and New Business Volumes As such, the below are short-term observations that were
The life insurance industry’s lapse experience is always observed across the non-life insurance market for the Observation
Overview of the 2020 financial
fascinating to observe and 2020 was no different. average insurer: One would expect that with new work-from-home and EV results
Despite severe and unprecedented economic hardship protocols, increased stress environments and
• Significant business interruption claims paid and
there were no massive increases in industry level lapses10, stretched resource capacity following the pandemic,
reserved for COVID-19 – Once bitten, twice shy.
definitely nothing that suggests we had a mass lapse operational risk would increase. However, the
• Reduced loss ratios of the motor line of business owing Standardised Formula doesn’t accurately capture this
A pandemic’s impact on stress
event on our hands. In fact, some insurers experienced testing frameworks
to the lockdown effect. Insurers that use the Standardised Formula
an improvement in lapse rates. This emphasises that any
experience item, like lapses, that depends on • Cash backs paid to policyholders to share in this as a proxy for economic capital as part of their ORSA
policyholder behaviour is notoriously difficult to predict improved motor experience process need to critically assess the appropriateness Optimising the life insurance
of the operational risk modules in light of the current regulatory balance sheet
under extreme conditions, as it might behave • Reduced cover from comprehensive to third party, fire
counterintuitively. and theft environment.
• Increased claims on accident and health, travel and Artificial intelligence and
New business volumes showed large reductions10, as property contents lines of business machine learning
disposable income came under pressure and advisor
• Increased expenses following work-from-home protocols
networks were restrained from travelling, placing at least
adopted
some upward pressure on per policy expenses. Using technology to combat
insurance fraud
Observation
Observation With the observation that there was very limited
ORSA scenarios representing extreme conditions Proactively managing conduct risk
impact on non-life insurers’ Standardised Formula
need to consider the possibility of policyholders SCRs, non-life insurers that use the Standardised
behaving in unexpected and counterintuitive ways, Formula as a proxy for economic capital as part Understanding a more complete
as this is not only plausible but could also notably of their ORSA process need to critically assess the IFRS 17 picture
change the outcomes of such scenarios. appropriateness of the non-life underwriting risk
modules in light of the current environment, both
from a claims and expenses perspective. IFRS 17: Controls and financial
Non-Life Underwriting Risk reporting under the new standard
With roughly 80% of non-life premiums being attributed to
the motor, property and liability lines of business11, we have Operational Risk
seen that, on the surface, there appears to be a limited As the Standardised Formula allowance for operational risk VAT: Closing the compliance gap
impact of the pandemic on non-life underwriting risk for is largely a premium and reserve exposure-based
the average non-life insurer, as the risk modules do not calculation, we have seen that, on average, the operational
specifically cater for the direct impacts of the COVID-19 risk allowance for insurers decreased relative to
Reference:
pandemic. expectations, in line with lower than expected business 10. Prudential Authority – Summary of QRT data 12
volumes. 11. Prudential Authority – Non-life industry experience 2018S A INSUR A NCE OUTLOOK | COV ID -19 – ONCE BIT TEN, T W ICE SH Y
Introduction
Looking Forward
With the pandemic not yet over, we have listed below some
Last year may have felt like more than a 1-in-200 event
across the risks the industry faced. But it is safe to say
Authors
Overview of the 2020 financial
items insurers should consider when assessing their top that, bar the remaining uncertainty surrounding business Lafras Eksteen and EV results
and emerging risks within the ORSA process. These items interruptions claims, the industry’s capital position was Senior Manager, Actuarial and
Insurance Solutions,
should also be considered as part of the post-stress profit more than adequate to absorb the severe impact, showing
Deloitte Africa COVID-19 – Once bitten, twice shy.
assessment for their Loss Absorbing Capacity of Deferred the resilience of the balance sheets under the new capital leksteen@deloitte.co.za A pandemic’s impact on stress
Taxes calculation. regime. What our analysis has confirmed, however, is that +27 (011) 209 8109 testing frameworks
not all of the risks and interrelationships of the risks can be
• Increased lapses and lower new business volumes owing catered for in a one-size-fits all standardised formula. It is
to suppressed economic growth and retrenchments important that insurers feed the insights and data gained
Optimising the life insurance
regulatory balance sheet
• Increased risk of defaults and widening of credit spreads during the pandemic into other elements of their risk Ricardo Govender
management framework, in particular their ORSAs. In this Senior Manager, Actuarial and
as the economy remains fragile
way they will be better informed about the effectiveness Insurance Solutions, Deloitte
Africa Artificial intelligence and
• Impact on the property market and property investments of various elements of their risk management strategies as
rgovender@deloitte.co.za machine learning
of a permanent shift towards remote working and they adapt and thrive in the increasingly uncertain world. +27 (011) 304 5953
e-commerce As Albert Einstein so eloquently put it: “In the middle of
• Fiscal and monetary policy impacts on the economic difficulty lies opportunity”. Using technology to combat
insurance fraud
environment and wider financial markets
• Longer-term mortality impacts which are still highly
uncertain Proactively managing conduct risk
• Impact on trade credit and credit life business over the
next few years following the economic impact of the Understanding a more complete
pandemic IFRS 17 picture
• Potential delays in transformation and other large-scale
programmes, including IFRS 17 implementation IFRS 17: Controls and financial
reporting under the new standard
• Potential long-term implications on staff skillsets
following prolonged remote working, school and
university disruption and the related implications for VAT: Closing the compliance gap
operational risk and scarce skills
• Changes in cyber and security risk related to prolonged
remote working and e-commerce.
13S A INSUR A NCE OUTLOOK | THE S A M DUS T H A S SE T TLED, TIME TO OP TIMI SE THE LIFE INSUR A NCE BA L A NCE SHEE T
Introduction
The SAM dust has settled, time to optimise the Overview of the 2020 financial
and EV results
life insurance balance sheet COVID-19 – Once bitten, twice shy.
A pandemic’s impact on stress
testing frameworks
Easy wins to improve life insurers’ regulatory capital positions
Optimising the life insurance
regulatory balance sheet
Introduction Priorities The clearly defined boundaries of the risk-based Standardised
We are about two years into reporting under Formula SCR provide a good starting point to optimising
the new Solvency Assessment and Management insurers’ capital consumption. Since capital optimisation is not Artificial intelligence and
(SAM) framework, and with the dust of Regulatory a once-off exercise, we also introduce a framework to embed
solvency
machine learning
implementation having settled many life insurers ratio capital optimisation across an organisation that considers the
are finding a steady rhythm of submitting the Economic Economic stakeholders and trade-offs mentioned here.
new regulatory returns. They have a more de-risking solvency Using technology to combat
ratio
insurance fraud
Performance metrics
hands-on understanding of the processes Optical Capital Optimisation, or Optimal Capital
Stakeholders
needed to produce, in particular, the new capital Consumption
measures, and have a better view of how these With the implementation of Solvency II leading South Africa’s Proactively managing conduct risk
Optimal
measures describe the risks of their businesses. solution implementation of SAM by a couple of years, we can leverage
Tax Earnings
We are seeing investments made to improve the efficiency volatility
various learnings from Europe. A particular aspect in this regard
reporting processes needed, but there is also an has been the approach to optimising regulatory capital without Understanding a more complete
opportunity to use the deeper understanding to any economic substance behind the optimisation. IFRS 17 picture
improve the performance of the business with
respect to capital consumption. IFRS and Liquidity Similar to certain tax shelters, this has been perceived as
EV earnings and cash IFRS 17: Controls and financial
generation “gaming” the system. Our view, however, is that a deep
reporting under the new standard
Capital optimisation requires trade-offs between understanding of capital optimisation is essential to understand
the different aspects summarised in the adjacent underlying risk drivers, which enables better risk management
diagram. The optimal solution considers all these Capital optimisation considerations and should thus be considered as part of an insurer’s Own Risk VAT: Closing the compliance gap
aspects across the different stakeholders, taking and Solvency Assessment. Better solvency ratios, based on a
into account their performance metrics and the sound understanding of risks assumed, also enables insurers
In this article we focus on the various options available to
business priorities. to offer more affordable products while still providing the
optimise regulatory capital ratios under SAM, also referred to as
appropriate risk-adjusted return to shareholders. 14
solvency ratios or Solvency Capital Requirement (SCR) covers.S A INSUR A NCE OUTLOOK | THE S A M DUS T H A S SE T TLED, TIME TO OP TIMI SE THE LIFE INSUR A NCE BA L A NCE SHEE T
Introduction
Which Levers Should Insurers Focus on?
The SAM standardised formula SCR is a complex Cost/Effort Modelling Optionality Reinsurance and Risk Transfer Balance Sheet Management and
Overview of the 2020 financial
Required Capital Structuring
calculation with many components underlying the and EV results
Market Risk, Life Underwriting Risk and Operational Risk Easy wins • Remove conservatism • Reduction of insurance liabilities • Composition of assets considering
calculations. Furthermore, diversification of components and/or SCR through traditional concentration and default risk
• Allowance for existing COVID-19 – Once bitten, twice shy.
has a dramatic impact of the resulting solvency ratio. management action framework reinsurance • SAM specific ALM strategy (positive A pandemic’s impact on stress
There are thus many levers insurers can pull to influence • Reinsurer credit rating, parent and negative liabilities) testing frameworks
• Interpretation of contract
their solvency ratio, within which we acknowledge two guarantee, reinsurance collateral
boundary • Minimise regulatory deductions
broad types of capital optimisation. The first type does from Own Funds in FSI 2.3 Section
• Illiquidity premium applied to • Concentration risk – use multiple Optimising the life insurance
not affect an insurer’s risk exposures per se, but rather reinsurers 8*
yield curve regulatory balance sheet
results in risk capital being modelled more accurately,
• Consider counterparty default
and we term this “Modelling Optionality”. The second
relaxations in FSI 2.2 Att. 3(b) *
type changes an insurers’ actual risk exposures, either
Artificial intelligence and
through risk transfer or risk reduction, and we have Moderate • Iterative risk margin (IRM) • Catastrophe risk reinsurance • Letters of guarantee machine learning
split this type into two categories, being those solutions effort • External rating model for • Mortality swap reinsurance • Use of Tier 2/3 Basic Own Funds or
relating to “Reinsurance and Risk Transfer” and those required company/counterparty CQS Ancillary Own Funds, as opposed
relating to “Balance Sheet Management and Capital • Mass lapse reinsurance Using technology to combat
mapping to just Tier 1 Basic Own Funds
Structuring”. insurance fraud
• “VIF” reinsurance solutions –
• Improved tax modelling, • Updates to management action
reduce cashflows uncertainty framework
Each of these categories are further grouped into: particularly maximising LACDT
Proactively managing conduct risk
• easy wins Honourable • Swap curve • Consider capital efficiency of • Company structures, subordinated
• those requiring moderate effort mentions reinsurance agreements debt, contingent loans
• Internal model
Understanding a more complete
• those requiring considerable investment, which we • Alternative risk transfer • Product design and contract
IFRS 17 picture
refer to as honourable mentions. agreements, e.g. insurance linked wording, e.g. new product offering
bonds • Capital efficient mergers/
These options are widely documented, so in this article acquisitions IFRS 17: Controls and financial
we do not describe them in detail, but rather assess the
reporting under the new standard
impacts of the various options on capital optimisation.
*Any use of “FSI” refers to the Financial Soundness Standards for Insurers, as published by the Prudential Authority
With a small- to medium-sized life insurer in mind, we determined the potential effect of selected solutions on the solvency
VAT: Closing the compliance gap
ratio, relative to the cost/effort and expertise required to implement those solutions. This was done through analysis
and judgement, as well as incorporating learnings from the European Solvency II regime. These results must be carefully
considered, as they depend on both an insurer’s specific business, as well as the skills, expertise and operational capabilities
15
available within the company.S A INSUR A NCE OUTLOOK | THE S A M DUS T H A S SE T TLED, TIME TO OP TIMI SE THE LIFE INSUR A NCE BA L A NCE SHEE T
Introduction
Our resultant findings for selected capital optimisation solutions are summarised in this diagram.
Overview of the 2020 financial
The potential impact on solvency ratio relative to cost/effort and internal expertise required and EV results
COVID-19 – Once bitten, twice shy.
More cost/effort required A pandemic’s impact on stress
testing frameworks
Honourable mentions
Bubble size represents the potential impact of each
5
option on the solvency ratio. Internal model
Optimising the life insurance
regulatory balance sheet
Tier 2 or 3 own funds
More internal expertise required
Product design and Artificial intelligence and
Internal Expertise Required
4
contract wording
machine learning
Easy wins
Using technology to combat
3
LACDT
insurance fraud
Management action VIF solutions Iterative risk margin
(contract boundary)
Proactively managing conduct risk
2
Mortality swap
Counterparty default
(CQS conservation) Illiquidity premium
Mass lapse reinsurance
Understanding a more complete
Asset liability matching IFRS 17 picture
1 Asset diversification
(concentration risk) Traditional reinsurance
Moderate effort IFRS 17: Controls and financial
Remove conservatism reporting under the new standard
(modelling)
0
0 1 2 3 4 5 VAT: Closing the compliance gap
Cost / Effort Required
As cost/effort and internal expertise increase we move towards the top right-hand corner of the graph. The size of each bubble gives an indication of the possible
improvement in solvency ratio relative to other initiatives. For example, implementing an internal regulatory capital model requires significant cost/effort and internal
16
expertise, but the potentially significant improvement in the solvency ratio might be worth the effort.S A INSUR A NCE OUTLOOK | THE S A M DUS T H A S SE T TLED, TIME TO OP TIMI SE THE LIFE INSUR A NCE BA L A NCE SHEE T
Introduction
Easy Wins insurers, is to fully understand the interplay between risk margin (IRM). As the name suggests, the IRM
Easy wins are summarised in the bottom left-hand assets held and components of the Market Risk module. calculates the risk margin and SCR iteratively. While this Overview of the 2020 financial
corner of the graph, as these are initiatives that can be Small tweaks, for example spreading cash assets across entails upfront effort, it has been proven to be cost- and EV results
implemented with relatively limited internal expertise multiple major banks, reduces concentration risk and can effective in the long term and can introduce significant
and minimal cost/effort. Many of the easy wins relate to significantly decrease the Market Risk capital requirement. solvency ratio improvements, particularly where there
COVID-19 – Once bitten, twice shy.
modelling initiatives. This can be understood with the are large negative reserves. While the IRM modelling
A pandemic’s impact on stress
context of where insurers’ capital thinking was grounded, Moderate Effort sophistication is not directly comparable to the actual risk testing frameworks
the Financial Soundness Valuation (FSV) framework. In the In the middle of the graph there are several classic risk transfer achieved through mass lapse reinsurance, both
FSV world conservatism in modelling was not only required management tools, for example, asset liability matching tools achieve similar outcomes in reducing the lapse risk
but was also common practice. Under SAM the liabilities that has long been used by insurers. This ranges from component of the SCR. Insurers could thus consider these
Optimising the life insurance
regulatory balance sheet
should be measured on a best estimate basis, however simple durational matching that can be done with less tools as alternatives to one another by comparing upfront
both implicit and sometimes explicit conservatism remains cost and expertise, all the way to complicated hedging cost/effort of the IRM approach, including regulatory
within some actuarial models and assumptions. Actuaries strategies. These provide protection against a wide variety application cost/effort, against the long-term cost of mass Artificial intelligence and
tend to include conservatism to allow for the uncertainty of movements in various financial variables. lapse reinsurance premiums. machine learning
in assumptions, model risk and data. Actuaries should do
more to ensure their numbers reflect a best estimate view. Reinsurance is a similarly well-established risk Honourable Mentions
Conservatism is particularly included in the valuation of transfer tool, starting with the transfer of mortality One of the biggest bubbles on the graph relates to the use Using technology to combat
new contracts. and morbidity risk through traditional reinsurance. of Tier 2 and tier 3 own funds, particularly the use of insurance fraud
Similarly, mortality swaps are an effective way to reduce Ancillary Own Funds in the form of subordinated debt and
Furthermore, some insurers choose more conservative longevity risk on annuity books and at the same time parental letters of guarantee. A parental guarantee can
Proactively managing conduct risk
Credit Quality Steps than can be justified. Similarly, insurers reduce cash flow volatility to better enable asset liability significantly improve the solvency ratio without requiring
might not be shortening contract boundaries for loss- matching. Reinsurance can also provide financing, like a capital injection. When applying a parental guarantee
making contracts, i.e. not allowing for the fact that such VIF solutions which entail transferring large portions of in the SCR calculations an allowance for default risk is Understanding a more complete
contracts can be assumed to be repriced at the expected premium to a reinsurer, thereby reducing Own Funds, required, however this allowance is generally small relative IFRS 17 picture
repricing date, hence reducing capitalisation of long-term but also significantly reducing most life underwriting risk to the maximum allowable increase in Own Funds, being
future losses. components, including lapse risk. The net effect of such a say 50% of SCR for Tier 2 Own Funds.
solution could well be an improved solvency ratio. IFRS 17: Controls and financial
A key modelling requirement under SAM is the loss One of the youngest additions to the reinsurers’ toolkit is Related to this are regulatory deductions from Own Funds, reporting under the new standard
absorbing capacity of deferred taxes which, with mass lapse reinsurance, which transfers a part of the including investments in an insurer’s own shares, in its
a moderate level of effort, could reduce the SCR by up loss arising from a mass lapse event. This is particularly holding company, cash and deposits at a bank within the VAT: Closing the compliance gap
to 28%. SAM also provides insurers with the option to beneficial for risk business with long contract boundaries same financial conglomerate, participation in financial and
increase the discount rate by an illiquidity premium, where the mass lapse SCR is sizeable. credit institutions and net deferred tax assets. Minimising
which can significantly reduce reserves for annuity these deductions will improve the solvency ratio.
business, albeit with a marginal increase in SCR. In South Africa insurers also have access to a
17
Another easy win, particularly at small/medium sized fundamentally different tool, application of an iterativeS A INSUR A NCE OUTLOOK | THE S A M DUS T H A S SE T TLED, TIME TO OP TIMI SE THE LIFE INSUR A NCE BA L A NCE SHEE T
Introduction
In the top right-hand corner of the graph is use of an
internal model for regulatory capital. An internal SAM Sponsorship Constraints Overview of the 2020 financial
Executive Sponsor/Business Owners Resources/Budget/Timescales
capital model requires significant cost/effort and internal and EV results
expertise to implement, but could lead to significant
improvement in the solvency ratio. These models are not 1 Agree objectives COVID-19 – Once bitten, twice shy.
very common in South Africa, especially the life insurance
A pandemic’s impact on stress
space. Historically it has been very difficult to get approval testing frameworks
2 Maintain ‘long-list’
for use of an internal model from the Prudential Authority.
Finally, the impact of new products and features should 3 Prioritised ‘short-list’ Optimising the life insurance
regulatory balance sheet
be fully understood before launch, including the impact
on diversification benefits of SCR components. In fact, by 4 Execution
launching products that target certain SCR components an
Artificial intelligence and
insurer can sell more policies without materially impacting Process machine learning
Control
its capital requirements, and thus improve Return on Equity. Communication and Stakeholder
Governance
Management
Ideas are cheap, execution is everything Using technology to combat
With so many capital optimisation tools available it is insurance fraud
A key aspect of such a framework is investigating various optimisation
far too easy for insurers to shoot from the hip, resulting
options and documenting these options succinctly in a log or an “ideas
in capital optimisation becoming a series of ad hoc and
hopper”. An ideas hopper would summarise key features of an optimisation Proactively managing conduct risk
sporadic decisions. This could result in sub-optimal
option, its impact on key metrics, as well as barriers to implementation
outcomes and/or unintended consequences for other
if any. The inclusion of barriers allows insurers to easily identify when
aspects of the business, which are costly to reverse after Understanding a more complete
previously unviable options become viable.
implementation. Truly effective capital optimisation, on the IFRS 17 picture
other hand, entails embedding it throughout the business
by establishing a capital optimisation framework which has Maintain the ‘ideas hopper’
buy-in from senior management, clearly defined objectives, IFRS 17: Controls and financial
appropriate controls, and well debated priorities and Capital Earnings Liquidity Volatility Timescale Resource
Execution reporting under the new standard
risk
processes that allow for efficient execution.
Option 1 +ve -ve n/a +ve 3 months low low
Option 2
VAT: Closing the compliance gap
Embedding capital optimisation should reach into the heart
of an organisation by touching its culture and making capital ...
optimisation a key factor in every business decision. Also, ...
capital optimisation should not just focus on maintaining a 18
certain solvency ratio, but also on stability of such ratios.You can also read
