4. 9 is about to
off globally
Lot of focus from investors, key
stakeholders and regulators on interim
financial statement before banks issue their
first annual financial statements applying
IFRS9
• Issuing of a separate transition document
to help users better understand the IFRS
9 impacts at, and beyond, adoption
• Given the lack of prescriptive
requirements, judgement will be required
by banks in designing disclosures for
reporting (different extent of disclosure
depending on bank size and nature of the
business)
• Early involvement of auditors to obtaining
views that will avoid potential issues later
in the annual reporting cycle.
5. The
New
normal
provision
for
undrawn
amounts
transfer
criteria
from
Stage 1 to
Stage 2
develop a
forward-
looking
assessment
• Monetary impact can be very
heterogeneous depending on portfolio risk
and type of product
• High impact on revolving product
• Lower impact on instalment loans and
mortgages (but strongly dependent on
product maturity)
• Greater ongoing volatility vs. IAS 39,
especially as forecast macroeconomic
conditions vary
From IAS39 to IFRS9
6. 6
modelling challenges
• Data infrastructure in place
• Analytical skills spread across the
teams or in the RM division
• The challenge is taking profit from
risk parameters that banks already
estimate and use (typically, for
credit risk management purposes),
aiming for maximum consistency
and minimum duplication of effort
For the lenders who developed
IRB framework …
• Retrieve of historical information and
build of Analytical Datamart
• Lack of predictive analytics skills
• Balancing of costs and benefits must
therefore be carefully assessed.
Start simple and planning a roadmap
for the evolution of the methodology
and IT infrastructure over the next
2-3 years
… For the others
7. operational
challenges
1 2 3 4
1. Reliability
IFRS 9 calculation
engine must be able to
process many
calculations against
large numbers of
accounts efficiently and
reliably.
2. Governance
Strong governance and
control elements to
manage changes to
data, models and
process flow. Changes
should be clearly
auditable and updates
to production models
should be tightly
controlled by managing
user permissions
3. Reporting
Robust reporting
capabilities to enable
mandatory
disclosures, internal
management
information, and ad hoc
analysis
4. Flexibility
Quick implementation
of new generation of
models and the
possible inclusion of
new data into the
process
8. Review your analytics toolkit:
…a few hints
TransferCriteria
• Statistically derive all migration factors (endogenous and
exogenous)
• Differentiate of trigger event by risk segment, counterparty
and product
• Perform some correlation analysis of each trigger possible
variable with the effective increase in the credit risk
• Determine of a threshold based on a historical analysis of
the triggers and the risk profile
• Involve business people for an integrated assessment that
is not the expression of only quantitative elements
9. Review your analytics toolkit:
…a few hints
ExpectedLoss
• Measure goodness of fit for alternative techniques to
project lifetime default estimations
• Different methodologies can be adopted for different
products
• Combine Markov Chain (which over-estimates after year 3)
and Vintage analysis (which overestimates in the first 3
years)
• Test alternative macroeconomic Indicators and adopt a
conservative approach in terms of variables
• Control for impairment system volatility to avoid impacts
on the value adjustment - and ultimately on income
statement
10. What
does
this
mean
for
Banks?
Banks will report a significant
increase of provisioning
figures (in a range of 30% -
60%, depending on products
maturity and forward looking
scenarios)
Phase1
Banks will seek for alternative
solution to limit impairment
increase and ensure full
compliance. Full review of risk
parameters, staging criteria and
forward looking approach
Phase2
Parallel run of alternative
solutions and comparison of
impact on provisioning.
Benchmarking and getting
ready for validation process
Phase3
2018 Q1
2018 Q2
2018 Q3
11. 1
Where do the savings
on impairment increase
come from?
Lifetime projection
of Risk parameters
Methodology
choice to estimate
lifetime PD
2
Staging
Optimization
Reduce to the
minimum the
‘scale’ effect due
to the migration
between Stage 1
and Stage 2
3
Reliable Forward
Looking modelling
Accuracy and
stability of
economic
forecasts is a key
factor to mitigate
impairment spikes
13. 13
Budget
Planning
Credit
Process
RM&Capital
Allocation
3-5 Years 1 Year 1 Day 1 Month
Industrial
Plan
Credit
Strategy
Capital
Planning
Budget
Goals
Policy
Rules
Capital
Adequacy
& Risk
Policies
Risk
Based
Pricing
Risk
Monitor
Lending
Process
Early
Warning
&
Collection
Targeting
Budget
Monitor
IFRS9 impact on planning and credit strategies to define cost of risk
and actions
14. 14
• Produce the risk
parameters for the new
impairment model
• Guarantee the
suitability for
accounting purposes
and the consistency
with other risk
measurement,
validation and capital
planning tools that
already exist within the
company
CRO
• Will be required to
review the areas where
the concept of lifetime
loss can be expected
to have material effects
(such as the bank’s
credit policies,
origination processes,
loan monitoring
procedures).
CLO
• Check that the tools
developed for the
implementation of the
new impairment model
are in line with the
bank’s Risk Appetite
Framework and the
scenarios already used
for strategic planning
• Update the ICAAP
process, and the bank’s
estimates of its overall
internal capital (both
present and future)
BoD
15. 15
• Identify the bank’s target portfolios, taking into
consideration also the risk of transition, over
multiple years, from Stage 1 to Stage 2
• Different segments (in terms of products, clients
and maturities) will have to be simulated over
multi-year horizons in order to gain a deeper
understanding of their potential, in terms of future
vulnerability to impairment risk
• Review their product catalog, especially for long-
term loans, which will be particularly penalized, in
terms of potential impairments, by the transition
to the lifetime approach embedded in IFRS9
Impact on Credit/Commercial
Strategies
16. 16
• Review pricing strategy to make it consistent with
the new cost of risk paradigm
• Loan application process and the approval of new
loans will have to take into account the new
lifetime PDs. to benefit from multi-year risk
estimates to achieve a better assessment of
creditworthiness in the origination phase.
Impact on Lending Strategies
17. 17
• Banks will have to identify those Early Warning indicators
that, while being consistent with the pre-existing risk
management criteria (reduce Default), may lead to a
transition across IFRS9 stages (reduce migration to Stage
2)
• Introduce leading indicators for the staging criteria
• Verify how present EW system predict Stage 1- Stage 2
migration
• Change le logic of priorities settings
• Introducing dedicated teams/treatments to manage high
transition risk cases
Impact on Credit Monitoring
Strategies (1/2)
19. 2018 will be crucial to
measure impacts of the new
principle and tune models
1
Methodology is not neutral
on the final bottom line
impact
2
Several internal processes
will be affected by the new
impairment calculation
3
20. Account level behavioural data
External Data
• Delinquency
status
• Balance
• Credit limit
• Loan term
• Score
• Exposure
• LGD (if not
available a
benchmark
value will be
provided);
• EW flags;
• Macro
economic
values
Bank ‘i’ data
Macroeconomic KPIs under
alterative scenarios
External Data
• Adverse
• Neutral
• Positive
Designer
• Designer of staging criteria;
• Macro economic scenario
setting & Simulation
• Lifetime risk parameters;
• 1y risk parameters;
• Forward Looking Parameters
• Staging Criteria
• Expected Loss
Risk KPIs developed on
Bureau data:
CRIF IFRS9 engine
IFRS9 analytical
data mart
Monthly batch updated
CRIF IFRS 9 ENGINE IS CURRENTLY
WORKING (in Italy) ON 160 BANKS
AND FINANCIAL ISTITUTIONS
• Reporting
• Expected Loss per line;
• Risk parameters and
stage per line
Containment of operational risk
No effort to provide IFRS9 EL
required to the client
No investment on engine
purchase
Sharing of maintenance
costs
CRIF answer: IFRS9 engine
21. CRIF answer: IFRS9 engine / embedded reporting
Main output variables
IFRS9
Contract
ECL by segment
Bucketing/ Staging criteria
1
Risk parameters (PD,LGD EAD)
Exposure
Type of contract
Expected Credit Loss
Ageing
Examples
PD next 12 m by segment
Cost of credit Expected by IFRS 9
3
2
3
Client
Segment
Credit risk status
Benchmarking4
Breakdown
1
1
2
Composition of Stages
PD, LGD, EAD by stage
3
ECL By stages
2
22. Meet The Team
To know more,
please contact
our domain
experts
Marco
Luca
AtriLuca
Principal Consultant
Based in Dubai (UAE)
l.calconi@crif.com
Marco
Head of Regulatory Risk
Based in Bologna (Italy)
m.macellari@crif.com
Atri
Analytics Director
Based in Pune (India)
a.basu@crif.com
23. IFRS9, it is not
just for your
CFO
Luca Calconi
CRIF Principal Consultant
01-February-2018