[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

Comparing bank and life insurer risks

How can I find out more?
This article is one of a series promoted by the
actuarial profession’s Action Group for Banking
(AGB) and explores situations where the actuarial
and banking worlds overlap to the benefit
of both sectors. The articles serve the following
broad purposes:
? Push: to demonstrate where actuarial skills
and thinking have been applied in a banking or
financial context to add real commercial value.
? Pull: to highlight tools and skills available in
banking and finance that may be of use to traditional
life and pensions actuaries.
Please contact either the author or Mark Symons
(mark.symons@actuaries.org.uk) for more information
on the article and the AGB’s activities. The growth in bancassurance over the
past two decades has increased the
interaction between actuaries, other life
insurance professionals, and bankers, a
trend that has further increased with the convergence
in regulatory approaches. While there
has been a fruitful exchange of ideas between
both, practitioners often find that this exchange
is hampered by the lack of understanding of the
different risks faced by the other. This article
seeks to address this gap by considering the similarities
and differences in the risks faced by a
typical UK bank (ABC Banking Corporation)
and life insurer (XYZ Life Insurance).
The core business of both is in essence very
similar: taking in funds from clients (depositors
and policyholders respectively), investing these
(by lending in the case of ABC), and making a
spread (interest margin or annual management
charge) on the return earned and that credited
back to clients. However, there are a number of
important differences.
First, ABC and XYZ serve different purposes. The
primary focus of XYZ is to act as a vehicle for longterm
saving. By contrast, ABC’s primary focus is
on lending, allowing individuals and companies
to crystallise future earnings, thereby providing
much needed liquidity to the economy.
Second, ABC maintains an extensive branch
network across the UK for distributing and servicing
its products. By contrast, XYZ has no highstreet
presence, instead distributing its products
through IFAs and other intermediaries.
Finally, while ABC’s accounts generally seek to
spread income over the lifetime of the business,
XYZ’s accounts include supplementary embedded
value reporting. This discounts all future
cashflows arising from its in-force portfolio,
effectively recognising expected future margins
upfront. The resulting ‘value in force’ (VIF) is
treated as an asset on the embedded value balance
sheet and changes in this are brought
though the embedded value P&L.
Market risk
Most of ABC’s cashflows are matched as the cost
of its funds and the rate it charges on loans are
variable, ie a rise in the former can be passed on
to borrowers. It has some loans which are fixedrate
which can be redeemed at par, thereby
exposing ABC to interest rate risk. However, rates
are generally fixed for less than five years which,
together with redemption penalties and hedging
strategies, helps to mitigate this risk.
ABC has a treasury function which is involved
in hedging interest-rate risk for the various business
units of ABC but which also accepts market
risk in respect of swaps and other derivative
transactions. While its portfolio is extensive, it has systems and controls capable of daily valuation
of positions and assessment of value at risk
(VaR) against set limits. This is overlaid with a
hedging programme to manage exposures and
the function could close out its positions within
a fortnight if it so desired.
By contrast, XYZ has considerable unhedged
market risk exposure. It has a large with-profits
portfolio which contains valuable guarantees, yet
these contracts are invested in volatile assets such
as equities and property to meet policyholders’
expectations. Worse, these expectations extend
to bonus additions to guarantees when investment
conditions are favourable. In mitigation,
XYZ can vary the asset mix backing these contracts
within limits of policyholder expectations,
selling equities when markets are falling in a similar
manner to that used in portfolio insurance.
These guarantees can be viewed as exotic,
path-dependent put options. Valuing these is
complicated by the fact that guarantee increments
and asset mix are set at an aggregate level
across all contracts. Therefore, while ABC’s treasury
function can value each derivative contract
on its own, XYZ must value all contracts with
guarantees simultaneously, ie its valuation
systems must process each contract in parallel
rather than sequentially.
Moreover, valuation must be by Monte-Carlo
sampling as the nature of the guarantees means
these are not tractable by formulaic means. Consequently
the time taken to run models, value
the guarantees, and assess VaR can be measured
in days and weeks rather than hours. Hedging
out such a complex position could take up to a
year or more, particularly as the Financial Services
Authority (FSA) would need to be satisfied
that any hedged investment strategy would not
prejudice policyholder expectations.
Additionally, for unit-linked business the
annual management charge XYZ levies will be
affected by market returns. Falling markets affect
not only the charge income arising in an
accounting year, but may also impair the VIF of
this business on the embedded value balance
sheet, exacerbating losses.
XYZ could hedge some of this VIF risk using
futures and forwards. However, the scope for
doing this is limited by the fact that the management
charges relate to a diverse and everchanging
mix of assets. Furthermore, there are
regulatory constraints to the use of derivatives
in life funds, as well as the residual liquidity risk
arising from having to meet margin calls when
any gain is in an illiquid VIF asset.
In one area of market risk ABC and XYZ are
very similar. Both have final salary schemes
which, while closed, continue to expose their
balance sheets to falls in scheme assets and the impact of bond yields on the value of liabilities.
Credit risk
ABC’s core business involves making loans to
retail and corporate clients. Such loans cannot
be readily traded and could potentially expose
ABC to the risk of counterparty failure for terms
of up to 25 years or more. ABC has extensive
systems and controls to mitigate this risk, and
regards its ability to assess such credit risk as a
core competence.
At the heart of its systems and controls are
detailed lending policies covering topics as
diverse as the environmental risks associated
with collateral to lending to hedge funds. This
is supplemented by detailed exposure limits for
different types of risk, including limits on term
(or ‘tenor’) and sector limits as well as credit scoring
models which are increasing in importance
with the Basel II regime.
ABC has the option to increase the spread on
its loans to reflect credit losses. However, the
scope to do this is limited by competition. Also,
an increasing number of its loans are tracker
mortgages where, while rates are variable, they
explicitly track the Bank of England base rate
and so the spread can be said to be fixed. Similarly,
spreads cannot be varied on hedged fixedrate
mortgages during the fixed-rate period.
XYZ invests in corporate bonds on behalf of
its investors, and uses these to back guaranteed
liabilities it faces. Such bonds are listed on securities
exchanges and exposure can be readily
traded. XYZ limits itself to investment grade
bonds and while it has criteria related to the
credit rating of these issues, neither its limit
structure nor its credit-risk policy is as comprehensive
as ABC’s. Given the transient nature of
exposures, the emphasis is more on managing
credit events such as downgrades, takeovers, etc
than the long-term risk of default.
XYZ is also exposed to credit risk in respect of
cash deposits; to the tenants of its property holdings
and in respect of over-the-counter derivatives
it enters into from time to time for hedging
or efficient portfolio management purposes. It
also has considerable exposure to reinsurance
counterparties. As for bonds, the limits on exposure
and policies governing these credit risks are
not as detailed as ABC’s.
The credit risk of ABC and XYZ overlap in two
areas: firstly ABC has a programme of securitising
its loans in which XYZ may sometimes
invest. Secondly, both are increasingly involved
in credit derivatives. XYZ is interested in the
enhanced yields offered by collateralised debt
obligations, while ABC is interested in the
income for accepting risk from the credit default
swaps which underpin these, as well as the ability
to hedge credit risk through these swaps.
Finally it is worth noting that recent market
concerns over credit risk, which have increased
the spread of LIBOR over base rate, and lead to
many corporate loan issues being withdrawn,
has also adversely affected life offices’ corporate
bond portfolios, highlighting the link between
corporate banking and life insurance credit-risk
Insurance risk
XYZ will obviously be affected by insurance risk
such as increased mortality rates leading to
losses on assurances and the longevity risk of
increased life expectancy on annuities.
ABC may be thought to be unaffected by insurance
risk, but death and disability will affect the
ability of individual borrowers to repay loans,
though this is mitigated if the borrower takes out
payment protection or other insurance.
More significantly, ABC is involved in equity
release mortgages, lending over the lifetime of
the borrower with no repayments and interest
rolling up during that time. Like other such
lenders it guarantees that on death or entry to
long-term care the accumulated loan will not be
greater than the value of the underlying property.
This ‘no negative equity’ guarantee is effectively
a put option on the value of the property
and as well as the market risk (relating to houseprice
inflation) involved, there is a longevity risk
as the exercise price of the option and its value
will increase the longer the borrower lives.
ABC also has a distribution deal with a general
insurance company to sell personal lines insurance
to its customers which includes a profit
share arrangement. This exposes ABC’s P&L
result to variations in claims on these products
which is a form of insurance risk.
Finally, both ABC and XYZ are exposed to the
longevity risk associated with their final salary
schemes. Recently, the Pensions Regulator questioned
the strength of longevity assumptions
used by final salary schemes, and the management
of both institutions are considering
strengthening assumptions in light of this, with
negative implications for the balance sheet value
of the schemes under IAS19.
Persistency risk
Most of XYZ’s contracts contain the option to
terminate the contract early and withdraw funds.
As noted, XYZ distributes through intermediaries
who are usually remunerated by an initial commission
payment. There is a risk that this and
other costs incurred on setting up a policy may
not be recouped on early withdrawal, although
commission may be partially clawed back.
In the past, XYZ had either discretion as to the
terms it offered to those withdrawing from its
with-profits policies; or the contract design
allowed for withdrawal penalties; or charges
were ‘front end’ loaded, ie with higher charges
levied in early years to recoup initial costs. This
allowed initial costs to be recouped, and in some
cases, early withdrawals could be a source of
profit. However, regulatory and competitive
pressures have limited the scope for discretion
and deterred surrender penalties for new contracts.
This has increased the risk that initial
costs are not recovered on early withdrawals.
There is also the loss of future profit margins on
withdrawal. This is significant as such margins
have been incorporated in the VIF asset on its
embedded value balance sheet. While the VIF calculation
will allow for early withdrawals, higher
than expected withdrawals will impair this VIF
asset, particularly if these trigger a wider revision
of withdrawal assumptions. This has happened of
late, as pensions simplification has led to higher
withdrawals on pension contracts. The resulting
charges to accounts have attracted considerable
negative comment on XYZ from analysts.
By contrast, ABC’s withdrawal levels draw less
comment, despite having higher levels on mortgages
than XYZ usually suffers on its products.
Withdrawals affect both the assets and liabilities
of ABC. On the liability side, withdrawals of
funds need to be replaced by attracting fresh
funds. This can be significant if retail funds have
to be replaced by more expensive wholesale
funds from other financial institutions, reducing
On the asset side, withdrawals take the form
of early loan redemptions. When a loan is
redeemed early, an income stream relating to
excess of loan interest over the cost of funds
ceases. To maintain profitability another income
stream needs to be found, ie ABC needs to make
another loan to replace any going off its books.
To do so, it will incur further costs, often having
to offer temporary discounts to its rates to attract
new custom. It then has the same issue with
recouping costs.
So early loan redemptions will impair the economic
value of ABC’s in-force loan book. However,
unlike XYZ with its embedded value
reporting, such value in force is not recognised
on ABC’s balance sheet. Therefore withdrawals
will not directly affect ABC’s balance sheet, nor
usually its share price, in the same way as life
companies, though the indirect impacts on the
cost of funds and interest rate spreads will affect
stockmarket valuations.
One aspect of persistency risk that does affect
ABC’s balance sheet directly is early redemptions
on fixed-rate loans, as these are redeemed at par
but the value of the loan may be greater than
par when interest rates are low. This prepayment
risk is closely monitored, but fixed-rate terms are
generally short and redemption penalties may
partially recoup some of the loss.
Note that the UK with its variable-rate
mortgages is different from the US and the
Continent, where mortgages are mostly fixed.
In the US in particular, prepayment experience
analysis is very sophisticated partly owing to the
involvement of investment banks in trading
mortgage securities such as collateralised
mortgage obligations.