IOPS Toolkit for Risk-based
Supervision
Module 2: Quantitative Assessment of Risk
Overview
•
•
•
•
•
Quantitative assessments can play an important part RBS - poor
results from quantitative tests imply higher levels of residual risk is
then factor into its overall risk analysis or risk score
Quantitative tools can provide a bridge between compliance and
risk-based approach to supervision
Quantitative tests can be undertaken by the supervisory authority
itself or by pension funds
Quantitative risk assessessment should focus on the risks relevant
for the type of fund (i.e. funding and solvency issues for DB funds,
investment returns and volatility and reliability of retirement
income for DC funds)
Not all risks can be measured in a quantitative fashion – the overall
risk assessment will always be a combination of quantitative and
qualitative factors
Models
Quantitative assessment tools make use of models – 3 main
problems:
•
•
•
what is modelled and how: including omitting factors which turn out to be
vital, the use of inappropriate data, only using recent data in the
development of models and their parameters, with models failing to take
account of extreme events and modellers all tending to use the same
accepted view
understanding the power and limitations of models: including their use
outside their sphere of applicability, hidden assumptions and simply
overestimating the power of models which are by their very nature
simplified representations of the real world
operational risks around the use of models: such as poor documentation,
lack of testing, and the misuse of data
Models
Recommendations include that models should:
•
sufficiently represent those aspects of the real world that are
relevant to the decision makers for which the information will be
used
•
include explanations of how the inputs are derived and what the
outputs are intended to represent
•
be fit for purpose in both theory and practice
•
include explanations of their significant limitations
Quantitative Regulations
•
Quantitative regulations are the starting point for quantitative risk analysis
•
These can be straight forward limits (such as minimum funding rules for DB
funds or investment restrictions for DC funds)
•
Alternatively, these quantitative regulations can be risk-based themselves
(e.g. factor based solvency rules, VaR calcuations)
•
Under a risk-based approach, supervisors need to consider not only whether
quantitative regulatory requirements (such as solvency stress and scenario
tests and value at risk (VaR) measures) are being met, but whether risks are
being identified and managed in such a way that the requirements will
continue to be met in the future
Quantitative Regulations
•
Risk-based supervision can incorporate quantitative regulations in 3 ways:
 The most simple fashion is to combine a ‘compliance’ and a ‘risk-based’
approach – compliance with quantitative restrictions is checked, and if
not met a lower score would be factored into the overall risk
assessment of the fund
 Alternatively, these straight forward quantitative requirements could be
made more ‘risk-based’ but testing whether compliance would still hold
in adverse circumstances (i.e. by stress testing) -the results of these
stress-tests would then be incorporated in the overall risk score
 Where the quantitative regulations are already risk-based compliance
with these risk-based regulations would be fed into the overall risk
score
Quantitative Regulations: DB Funds
•
•
•
•
Valuation requirements : these can be tested and incorporated into an
overall risk analysis by looking at the valuation assumptions, undertaking
sensitivity testing of changes in valuation assumptions, and stress testing of
risks such as high inflation
Minimum funding requirements: these can be incorporated into an overall
risk assessment either by checking for compliance (and scoring the fund
accordingly) or by stress testing the funding position to see if the minimum
requirements would be met under adverse circumstances (with the results
of such tests fed into the risk score)
Factor-based solvency margins: solvency margins can be risk-based by
requiring higher amounts of capital to be held against risky assets (such as
equities), thereby providing a buffer in case such assets decline in value.
Again either straight forward or stress-tested compliance with these margins
can be incorporated into an overall risk score
Stress-related solvency margins : require each entity to calculate the
additional amount of assets it would need to be able to meet its obligations
under a prescribed stress scenario or scenarios. The results are then fed into
the overall risk assessment
Risk-based Solvency Requirements
Country
Netherlands
Measurement
of
(Technical Provisions)
Treatment
of
Longevity Risk
Group
specific
mortality
table
adjusted for predicted
longevity
improvements, plus
buffer to address
uncertainty
in
predicted values.
Liabilities
Discount
Factors
Market
yield curve
measured
by
Euro
swap curve.
Minimum Solvency
Requirements
Solvency Buffers
5% of Technical
Provisions (from EU
IORP Directive).
Maximum probability of underfunding
within 1 year measured with stress
test: 2.5%.
Measured once per
year using current
market values.
Solvency buffers determined by risk
factors specific to each asset class.
Example of risk factors include yearly
decline in: equity 25-35% (depends on
type); currency 20%; real estate 15%.
Maximum period for
correction
of
deviations: 3 years.
Denmark
Fund-specific
mortality
table
approved by actuary
and supervisor.
Market
yield curve
measured
by
Euro
swap curve.
Traffic light stress
test
includes
assessment of the
impact of a 5%
improvement
in
longevity.
Mexico
No formal liabilities in DC plans.
Solvency
margin
defined by EU Life
Directive: 4% of
Technical Provisions
plus 0.3% of risk
bearing investments.
Measured every 6
months using current
market values.
Maximum period for correction of
deviations: 15 years.
Traffic light system is a stress test
rather than part of the formal solvency
rule, but results are taken into
consideration in the supervisory
assessment. Test defines 3 zones:
green, yellow, and red. Final outcome
depends on whether entity remains
solvent after test.
Example (year variations): listed
equity: red 12%, yellow 30%; interest
rate (medium duration): red +/- 0.85%,
yellow +/- 1.2%
Period of correction
from
minimum
required standards: 1
year.
No formal solvency requirements, but VaR limit designed to
limit downside risk for DC members. Historic VaR calculated
with rolling 550 day sample at 5% significant with different
limits imposed on the 2 portfolios. Price vector provided by 2
independent vendors.
Higher risk portfolio: 1% maximum daily loss.
Standard risk portfolio: 0.6% maximum daily loss.
Quantitative Regulations: DC Funds
•
•
•
•
Investment limits: compliance with these limits forms part of the overall risk
score
Minimum return requirements: solvency requirements backing these
guarantees would be measured and assessed in the same way as DB fund
promises
Value at risk limits: these assess the volatility of investment returns . Results
of such stress testing would be incorporated in the overall risk score. Where
such limits are themselves regulatory requirements, compliance would be
part of the overall risk assessment
Alternative risk measures: attempt to measure risk measurements against
long-term income requirements (such as replacement ratios), with
regulators devising optimal portfolios for achieving this target. The
performance of the actual portfolio of a pension fund could then be
assessed vs. this benchmark portfolio. Supervisors could then work this
analysis into their overall risk assessment via a ‘traffic light’ system - for
example a green light would indicate a pension fund with a portfolio
structure aligned with the benchmark and a good risk management system.
Use of Quantitative Restrictions for Defined Contribution Pension Plans
OECD
Australia
Denmark
Hungary
Mexico
Poland
Slovak republic
Sweden
Switzerland
Non-OECD
Chile
Colombia
Estonia
Israel
Russian
Federation
Quantitative investment
restrictions by asset
class
Minimum investment
return (absolute)
Quantitative risk
limits







































Techniques for Quantitative Risk Assessment
•
•
•
•
•
•
•
•
comparison of valuation assumptions – compare assumptions with peers to
identify if inappropriate
analysis of surplus –compare actual experience to assumptions to see if
ones using are accurate / appropriate
roll-forward calculations –financial position projected under certain
scenarios to assess exposure to adverse circumstances
duration analysis – project cash flows of assets and liabilities of fund and
compare interest rate sensitivity and timing mismatches
sensitivity testing – test sensitivity of valuation results to different
assumptions by recalculating results using different assumptions
deterministic stress testing – calculate the financial position of a pension
entity at current or future date to 1 or more adverse scenarios
stochastic stress testing –adverse scenarios computer generated not
predefined and distribution of results examined (i.e. likelihood that scenarios
adverse enough to create financial difficulty will occur)
value at risk (VaR) calculations – type of stochastic stress test measuring
adverse market movement with a specified probability
Techniques for Quantitative Risk Assessment
Technique
Comparison of assumptions
Analysis of surplus
Roll-forward calculations
Duration analysis
Sensitivity testing
Deterministic stress testing
Stochastic stress testing
Value at risk
DB Pension Plan
X
X
X
X
X
X
X
X
Insurer or Pension Company
X
X
X
X
X
X
X
X
DC Pension Plan
X
X
X
X
X
Integrating Quantitative Tools
DB
•
Solvency + funding ratios are the key quantitative tests for DB funds
– above 100% resulting in low risk score, below implying a higher
level
•
Roll forward + stress tests would tests these further – if funding
remains over 100% after stress testing a low risk score would result
•
ALM testing (matching assets and cash flows to liabilities) could also
be taken as a sign of robust risk management at a pension fund
Integrating Quantitative Tools
DC
•
VaR – measuring investment volatility can be used for DC funds,
though is controversial
•
ALM type measurements to see if DC funds can meet income
replacement targets are still under development
•
For DC funds offering guarantees, similar standard and solvency
stress tests to DB funds can be used
Quantitative Measurement of Nonfinancial Risk
Ratings varying between 0-1 could be given for risks such as:
•
•
DB plans having a number of complicating features, such as early
retirement benefits, indexation etc.
DC plans offering a large range of investment options, rather than
having just a few investment funds or “life-cycle”; having one fund
for all, but not allocating investment earnings on a market basis, but
“declaring” the rate on a non-transparent smoothing approach and
building up “reserves” ; high levels of outsourcing.