Sie sind auf Seite 1von 16

White Paper

IFRS 9 Impairment Modeling


Regulatory Perspectives & Modeling Approaches

Moodys Analytics Knowledge Services maknowledgeservices.com


October 2016
Contents
1. Background ............................................................................................................................................. 3

2. Expected Credit Loss Model under IFRS 9 ........................................................................................ 4

3. Scope ........................................................................................................................................................ 5

4. Model Implementation ......................................................................................................................... 6

4.1. Measuring ECL ........................................................................................................................................ 6

4.1.1. General Considerations ................................................................................................. 6

4.1.2. Information to Use ........................................................................................................... 7

4.1.3. Challenges .......................................................................................................................... 7

4.1.4. Overcoming the Challenges ......................................................................................... 8

4.2. Assessing Significant Increase in Risk .......................................................................................... 11

4.2.1. General Considerations ............................................................................................... 11

4.2.2. Information to Use ........................................................................................................ 11

4.2.3. Challenges ........................................................................................................................ 12

4.2.4. Overcoming the Challenges ...................................................................................... 13

5. Impact of ECL Model ........................................................................................................................... 14

5.1. Basel Framework vs. IFRS 9 .............................................................................................................. 14

5.2. IRB Banks vs. Non-IRB Banks .......................................................................................................... 14

5.3. Impact on Regulatory Capital ......................................................................................................... 14

6. System and Data Requirements ........................................................................................................ 15

6.1. System Requirements ........................................................................................................................ 15

6.2. Additional Data Requirements ....................................................................................................... 15

Author
Veer Bahadur Singh
vee.singh@moodys.com

2
1. Background
The International Accounting Standards Board (IASB) introduced International Financial Reporting Standards 9 (IFRS 9),
Financial Instruments in July 2014, replacing International Accounting Standard 39 (IAS 39) Financial Instruments:
Recognition and Measurement. In 2005, the IASB and the Financial Accounting Standards Board (FASB), which sets US
national standards, had begun working toward a long-term objective of improving and simplifying the reporting for
financial instruments, upon requests from users of financial statements and other interested parties. These interested
parties raised concerns that the requirements in the IAS 39 were difficult to apply and interpret. They requested the IASB
to develop a new Standard for the reporting of financial instruments that was
principle-based and less complex.
We believe the
implementation of the The final version of the standard combines the classification and measurement,
impairment, and hedge accounting phases of the IASBs project to replace
new ECL model will pose IAS 39. IAS 39 had many different classification categories and associated
certain practical challenges impairment models based on the incurred loss approach.

in terms of system/ On the basis of the asset classifications, IFRS 9 introduced a single forward-
looking expected credit loss (ECL) model, replacing the multiple impairment
information requirements models in IAS 39, which would be applied to all financial instruments that
and cost of implementation are subject to impairment accounting, thereby removing a major source of
complexity in IAS 39.

This ECL model will likely result in the timely recognition of loan losses, addressing the issue of delayed recognition of
credit losses in the existing accounting standard, which was identified during the recent financial crisis.

We believe the implementation of the ECL model will likely pose certain practical challenges in terms of system/information
requirements and implementation cost(s). In this paper, we cover the new impairment model in detail, focusing on
the practical challenges financial institutions would face and how existing available information and resources can be
efficiently utilized. A few solutions to address the challenges are discussed.

Timeline for Implementation

The effective date for the implementation of IFRS 9 (published in July 2014) for firms is January 1, 2018 or earlier.

3
2. Expected Credit Loss Model under IFRS 9
IFRS 9 presents a three-stage model for estimating expected losses on the basis of changes in credit quality since initial
recognition. Figure 1 depicts impairment recognition and interest revenue calculation in the three-stage approach.

Figure 1: Three-Stage Expected Credit Loss Model under IFRS 9

Increase in credit risk since initial recognition

---------------------------------------------

---------------------------------------------
Stage 1: Stage 2: Stage 3:

Performing (Initial recognition*) Under-Performing (Assets with Non-Performing (Credit-impaired


significant increase in credit risk since financial assets)
initial recognition)
Impairment Recognition

12-month expected credit losses Lifetime expected credit losses Lifetime expected credit losses

Interest Revenue
Effective Interest on gross Effective Interest on gross carrying Effective Interest on amortized
carrying amount amount cost i.e. gross carrying amount
adjusted for the loss allowance
*There is a separate guidance for purchased or originated credit-impaired financial assets (See IFRS 9, Paragraph 5.5.13)

As soon as a financial instrument is purchased or originated, 12-month expected credit losses are recognized in profit or
loss. Moreover, a loss allowance is established. This serves as a proxy for initial expectations of credit losses. For all other
financial instruments, e.g. already on books, the rules below are followed.

Stage 1: 12-month ECL


Financial instruments that fall in this category are those whose credit risk has not appreciated significantly since initial
recognition or those that had a low credit risk on the reporting date

For these assets, 12-month ECL is recognized, and interest revenue is calculated on the gross carrying amount of the
asset, that is, without adjustment for credit allowance

12-month ECL indicates the part of expected credit losses over the lifetime of an instrument that mimics expected
credit losses from payment defaults within 12 months following the reporting date. It does not represent cash shortfall
expected over the next 12 months; instead, it is the entire credit loss on an asset multiplied by the probability that the
loss will occur in the next 12 months.

Stage 2: Lifetime ECL


Financial instruments classified in this category are those whose credit risk increased significantly since initial
recognition (except they have low credit risk on the reporting date), but do not have objective evidence
of impairment

Lifetime ECL is recognized in these assets, but interest revenue is still calculated on the gross carrying amount of
the asset

Lifetime expected credit losses measure the expected present value of losses that may arise if a borrower defaults on his
or her obligation during the life of the financial instrument.

1. It is not applicable for purchased or originated credit-impaired financial assets (see IFRS 9, Paragraph 5.5.13)

2. See IFRS 9, Appendix A, Defined terms

4
Stage 3: Lifetime ECL
Financial assets that have clear evidence of impairment as on the reporting date (credit-impaired assets are classified
in this category)

Lifetime ECL of such assets is recognized, and interest revenue is computed on the net carrying amount (net of credit
allowance)

Why switch from 12-month ECL to Lifetime ECL?


In the new model, ECL would be recognized from the time financial instruments are purchased or originated. A trigger
event (such as default or unlikelihood of payment) is not required for an ECL to be recognized (as was the case with the
previous incurred loss model under IAS 39). The amount recognized as an ECL is considered a function of credit-quality
change since initial recognition to reflect the link between an ECL and the pricing of a financial instrument. Unless there
is a substantial increase in risk since initial recognition, estimating lifetime ECL ignores the link between pricing and
initial credit loss expectations. A true economic loss occurs when ECL exceeds initial expectations, i.e. when a lender has
stopped receiving compensation for the level of credit risk to which it is exposed.

Note:
The ECL model takes into account the relative assessment of credit risk while moving an instrument from 12-month
ECL to lifetime ECL. Therefore, two loans with the same credit risk on the reporting date may be allocated to different
stages, depending on the credit risk at the time of initial recognition. On a similar note, different loans from the same
counterparty could be categorized into different stages, depending on the credit risk each loan had at the time of
origination. For financial institutions to be able to address these intricacies, increased level of sophistication would be
required in data storage and reporting.

3. Scope
The IFRS 9 new impairment model is applied to:
Debt instruments (such as bank deposits, loans, debt securities, and trade receivables) measured at amortized cost
or FVTOCI

Loan commitments and financial guarantees not measured at fair value through profit or loss (FVTPL)

Lease receivables

Contract assets

5
4. Model Implementation
Period over which to Estimate ECL: While measuring
The two major components of the IFRS 9 ECL model are
ECL, the period to be considered is the maximum
measurement of expected credit losses and assessment
contractual horizon in which an entity is exposed to
of a significant increase in credit risk. In this section,
credit risk.
we discuss the requirements, information to be used,
challenges, and solutions to these challenges.
For loans with a fixed term and repayment structure
4.1. Measuring ECL (e.g., term loans), the period over which ECL should be
estimated is the period that remains before the term
4.1.1. General Considerations
ends. But for instruments that have a loan and undrawn
Credit losses are the present value of all cash shortfalls. commitments, financial institutions ability to demand
Expected credit losses are a probability-weighted estimate repayment and cancel undrawn commitment does not
of the losses in credit during the expected life of a financial limit their exposure to credit losses, e.g. for revolving
instrument. As expected, credit losses depend on the facilities such as credit cards, lenders continue to
timing and amount of payments. A credit loss arises even if extend credit unless they realize credit risk on the
an entity expects to be paid in full after the contractual due borrower has increased significantly. For such loans,
date. Therefore, while measuring expected credit losses, financial institutions should use historical information
financial institutions should consider the following. (from similar instruments with a significant increase in
credit risk) to measure the length of time it takes for
Probability-weighted Outcome: The ECL should not defaults to occur. Further, financial institutions should
be estimated from either a worst-case scenario or a consider credit risk management actions (such as
best-case scenario. Rather, the estimate should reflect reduction or removal of undrawn limits) once credit
the possibility of a credit loss even if the most likely risk on the financial instruments increases.
outcome based on current and historical information
does not indicate a credit loss.
Definition of Default: When defining defaults to
determine risk of a default, financial institutions should
This requirement calls for the evaluation of a range apply a default definition consistent with the definition
of possible outcomes. Moreover, financial institutions used for internal credit risk management purposes
have to compute 12-month ECL and lifetime ECL on for the relevant financial instrument. They should also
the basis of the current classification stage. For this consider qualitative indicators (e.g., covenants) when
purpose, forward-looking models for lifetime PD appropriate. However, it is assumed that a default does
incorporating macroeconomic conditions (e.g. HPI, not occur later than 90 days past due (DPD) unless
GDP, and FX) would be required for all portfolios. an entity has enough information to support a more
Here forward-looking information is not just limited delayed default definition.
to macroeconomic indicators, but also includes
information that is easily available and provides
Use reasonable and supportable information
a fair idea about the future condition of an obligor, available without undue cost or effort.
indicating the obligors likely creditworthiness. To
evaluate possible outcomes, probabilities need to be Figure 2: Summary of General Requirements in
embedded in the different scenarios. Measurement of ECL

12-month PD can be easily obtained by using the


lifetime PD model. Lifetime LGD models would also Current
be required. Conditions

Time Value of Money: An ECL should be discounted Historical Forward looking


to the reporting date by using the effective interest Information Information
rate determined at initial recognition. If a financial
instrument has a variable interest rate, the current
ECL
effective interest rate should be used to discount A Probability
expected credit losses. Weighted
Outcome + Time
Value of Money

3. See IFRS 9, paragraph B5.4.5

6
4.1.2. Information to Use 4.1.3. Challenges
Financial institutions should use reasonable and fact-based The major challenge financial institutions will likely face
information (such as historical events, current economic is in the calculation of lifetime and 12-month ECL for all
conditions, and forecasts of economic conditions) available portfolios. For this purpose, the following models need to
as on the reporting date, without incurring undue cost, be developed:
for computing ECL. They are not required to incorporate
forecasts of economic conditions over the entire expected Lifetime probability of default (PD) and exposure as
life of an instrument. Forecasts depend on the available default (EAD) models would be required to calculate
information. As the forecast horizon expands, reliability lifetime ECL for all portfolios (e.g., Commercial
and specificity of the information used to measure ECL will & Industrial [C&I], Commercial Real Estate [CRE],
decrease. Therefore, the degree of judgment required to project finance, banks, sovereigns, retail, etc.)
estimate ECL increases. Consequently, for periods far from
the reporting date, financial institutions may extrapolate The 12-month ECL model would require 12-month PD
available and detailed information. and EAD

LGD models would be the same for both 12-month and


The standard does not state how and what kind of lifetime ECL models. Forward-looking information may
extrapolation should be used when forecasts are be considered for LGD models that may change the
not available. Management would need to decide LGD at different points in time in the future
what extrapolation is to be used, e.g. it may choose
to use just the last available forecast for all the In general, all models should incorporate forward-
remaining periods or may use the average ECL. For looking information, including macroeconomic
some portfolios, e.g. retails loans, it may be simpler conditions (such as HPI, GDP, unemployment rate, and
to use the median or average life cycle, calculated industrial production growth)
from historical information, for the periods for which
economic forecasts are not available after adjusting
for current conditions. IFRS 9 does not specifically mention that PDs used
in assessing credit risk and calculating ECL should
IFRS 9 does not prescribe any particular data be point-in-time (PIT) or through-the-cycle (TTC).
requirements; therefore, financial institutions may use However, it clearly stresses all the information
various sources of data (internal or external). used should be forward looking and incorporate
current economic conditions. It also states historical
information should be leveraged, but adjusted to
It should be noted that although the model is forward-
reflect current conditions.
looking, historical information is perceived to be a
significant anchor or base for calculating ECL. However, Considering the above facts, it is clear risk measures,
adjustments should be made to historical data on the i.e. PD, LGD, and EAD, must reflect current conditions
basis of current observable data to reflect the effects of and need not be TTC or adjusted for a downturn, as in
current and forecast conditions. the case of Basel rules.

7
4.1.4. Overcoming the Challenges

12-month PD, LGD and EAD Models:


In this case, the existing Basel framework can be utilized with some adjustments.
Definitions of ECL and defaults in IFRS 9 are almost the same as those in Basel unless there are supportive proofs for
using more delayed or early defaults.

PDs used under a Basel framework are through-the-cycle (TTC) PDs, whereas under the IFRS 9 framework, we
require point-in-time (PIT) PDs. Therefore, TTC PDs would need to be converted to PIT PDs. According to Basel
regulations, PDs used for calculating capital should be stable throughout the economic cycle, whereas PIT PDs
reflect current economic conditions and are highly sensitive to changes in economic conditions. Figure 3 provides
an example of the difference between PIT PD and TTC PD. It can be observed that during a downturn, i.e. 2008-
2010, PIT PDs are higher than TTC PDs, whereas during a benign period, i.e. 2003-2007, PIT PDs are lower.

Figure 3: TTC PDs vs. PIT PDs


3.0%
PIT PD
TTC PD
2.5%

2.0%

1.5%

1.0%

0.5%

0.0%
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

LGD and EAD models from the Basel framework can be used directly, but post-adjustments, e.g. in the case of LGD,
downturn adjustments need to eliminated and forward-looking information needs to be incorporated
to the extent possible.

Our Solution for Conversion of TTC PD to PIT PD

1. Using Credit Quality Index (Z-Index)


IFRS 9 does not provide any information on how to convert TTC PD to PIT PD. We suggest leveraging the credit quality
index (Z-index) or one parameter representation of the credit transition matrix at any point in time to convert TTC PD to
PIT PD. In this methodology, the TTC transition matrix is represented by a normal distribution (see Figure 4, left), and by
using a Z-shift (right), the PIT matrix can be obtained.

Figure 4: Credit Quality Index (Z-Index)

0.45
0.4

----------------------------------------------

. 0.40
Bad Days Good Days
Ba 0.35 (Z- -2) (Z-2)
0.3

A
0.30
Probability

0.25
B
0.2

Firm 0.20
C remains
Baa 0.15
Default Area
0.1

Aa 0.10
Default Area Aaa
0.05
0.0

0.00
-4 -2 0 2 4 -5 -4 -3 -2 -1 0 1 2 3 4 5
Credit-change indicator
8
The Z-index is calculated such that once applied to each state in a periods transition matrix, it would minimize the
difference between the probabilities of historical PIT transitions and TTC transitions (see Figure 5).

Figure 5: Calculating Z-Index (Illustration)

Corresponding TTC matrix using inverse Shift resulted matrix by Z and then convert
TTC matrix probability function for a standard normal it again to probability function such that it
distribution would be approximately equal to PIT matrix

X1 X2 D X1 X2 D X1 X2 D

X1 91% 8% 1% X1 (inf,-1.35) (-1.35,-2.38) (-2.38,-inf) X1 PIT11 PIT12 PIT13

X2 7% 87% 6% X2 (inf,-1.56) (1.56,-1.63) (-1.63,-inf) X2 PIT21 PIT22 PIT23

Each Pij can be represented as P(i,j) =


(xij+1)-(Xij), is normal cumulative distribution. Z index

Using the above approach, Z-index for each year is calculated, and a regression model is used to estimate the
relationship between Z-index and macroeconomic variables. By using macroeconomic forecasts, Z-index is
estimated and applied to the TTC transition matrix to obtain the PIT transition matrix, which in turn gives PIT PDs for
each grade.

Figure 6: Estimate and Forecast of Z-Index (Illustration)

1.5

0.5

0
91
92
93
94
95
96
97
98
99
00
01
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20

-0.5

-1

-1.5

2. Modelling Default Rates


In this approach, we can model default rates for each grade/score bucket separately. However, it has a major
drawback this methodology does not take into account intergrade transitions. Therefore, we do not support this
methodology for IFRS 9 purposes.

9
Lifetime PD and EAD Models:
Financial institutions would need to develop lifetime PD and EAD models for all portfolios. For term loans, EAD models
would not be required, but for revolving facilities, lifetime EAD models need to be developed. To obtain lifetime PD,
the simplest approach is to obtain a PD term structure and then use it to obtain lifetime PD. For example, if we have
PDs P1 and P2 for year 1 and year 2, respectively, then the two-year cumulative PD can be calculated as:
Cum.PD=1-(1-P)*(1-P)
Alternatively, to obtain lifetime PD, one can obtain PD transition matrices at each point in time that are independent
of each other and then use these matrices to calculate the cumulative multi-year transition matrix. For example, if
T1 and T2 are PD transition matrices for year 1 and year 2, respectively, then the two-year transition matrix can be
calculated as:
2 Year Transition matrix=T*T, here * denotes matrix multiplication
where * denotes matrix multiplication.
We recommend using the transition matrix approach, as it also incorporates the effect of downgrades and upgrades from
time 1 to time 2. For example, if an obligor is in grade B as on the reporting date and PD for grade B for year 1 and year
2 is 2% and 3%, respectively, the cumulative two-year PD will be 4.94% (= 1 [1 - 0.02] * [1 - 0.03]). This PD does not take
into account the fact that from year 1 to year 2, the obligor may move to grade A, for which PD for year 2 is 1%. In this
case, the PD applicable for year 2 will change and the cumulative PD will be 2.98% (= 1 [1 - 0.02] * [1 - 0.01]).

Our Solution for Lifetime PD Model

1. Using Credit Quality Index


By using the methodology used to convert TTC PDs to PIT PDs, we can obtain PIT transition matrices at each point in
time in the future, using reasonable and fact-based forecasts. Depending on the expected life of financial instruments,
the lifetime transition matrix can then be calculated as:

=
=1
2. Model Downgrade, Upgrade, and Default Events
In this method, rating actions are modeled, e.g. B1 to B2 and B1 to B3 are grouped into one rating action B1
downgraded. Once the probabilities of rating actions are estimated, they are further distributed into more granular
rating transitions based on a static conditional empirical probability distribution. For example, the estimated probability
of B1 downgraded will be distributed into B1 to B2 and B1 to B3 using the historical average transition probability
distribution of B1 to B2 and B1 to B3. Figure 7 shows illustrative rating actions to be modeled.

Figure 7: Rating Actions

Starng Ending Rang


Rang B1 B2 B3 B4 C WR DEF
B1 1.B1_DN 1.B1_DN 1.B1_DN 2.B1_C 3.B1_WR 4.B1_DEF
B2 5.B2_UP 6.B2_DN 6.B2_DN 7.B2_C 8.B2_WR 9.B2_DEF
B3 10.B3_UP 10.B3_UP 11.B3_DN 12.B3_C 13.B3_WR 14.B3_DEF
B4 15.B4_UP 15.B4_UP 15.B4_UP 16.B4_C 17.B4_WR 18.B4_DEF
C 19.W_UP 19.W_UP 19.W_UP 19.W_UP 20.W_WR 21.W_DEF
WR
DEF
WR: withdrawn, DEF: default

10
A multinomial logistic regression is chosen for modeling. Such a regression has many advantages compared with
other methods. The estimates of standard errors and p values are straightforward. Moreover, a multinomial logistic
regression requires fewer assumptions than other models, as it does not assume normality, linearity, or homogeneity of
variance for the independent variables.

Figure 8 shows rating paths for an obligor in rating A. For multi-horizon forecasting, we would need to compute all
possible rating paths. The final transition matrix would be computed by using the average of the probabilities of all
rating paths.

Figure 8: Rating Paths


A_Up (10%)
Aa

A 100% A (81%) A

A_Dn (8%) Baa

D
Def (1%)

4.2. Assessing Significant Increase in Risk

4.2.1. General Considerations


Financial institutions should assess, at each reporting date, whether financial-instrument credit risk has surged
significantly since initial recognition. While making this assessment, it should be noted that a change in the default
risk, i.e. a change in PD, over the remaining expected life should be considered instead of a change in the expected
credit losses.

Financial institutions may assume the financial-instrument credit risk has not enhanced drastically since initial
recognition if the financial instrument has a low risk as on the reporting date.

If reasonable and fact-based information is available on the reporting date, financial institutions cannot solely rely
on past due information. But if forward-looking information is available at a cost or effort, past due information may
be used.

IFRS 9 puts a rebuttable presumption that credit risk on a financial asset would be considered to have increased
significantly since initial recognition if contractual payments are more than 30 DPD. This assumption will always hold
true unless financial institutions can prove, through reasonable and fact-based information, the risk has not increased
significantly at 30 DPD. Examples include non-payment due to oversight (and not because of financial constraints)
or historical evidence brought by financial institutions that shows no correlation between 30 DPD and a significant
increase in default risk.

Generally, there will be a substantial increase in credit risk before a financial asset becomes credit-impaired or an actual
default occurs. Therefore, financial institutions cannot align the timing of significant increase with when financial asset
is considered credit-impaired.

11
4.2.2. Information to Use
Financial institutions should consider reasonable and fact-based information available at no cost or effort. They are not
required to undertake an exhaustive search for information while determining whether credit risk has increased. Both
internal (institution-specific) and external information can be used. For example, for determining the risk for obligors in
a low-default portfolio, financial institutions may use internal grades consistent with the globally understood definition
of low credit risk. An actual or expected internal credit rating downgrade for a borrower can be considered a significant
increase in credit risk.

Internal credit ratings are considered more reliable when they are supported by default studies or mapped to external
ratings. Collateral information should not be used while assessing default risk. An assessment of a significant increase
in default risk may be made on a collective basis, e.g. on the basis of a group or subgroup of financial instruments with
similar characteristics. At times, it may be required to group individual instruments to identify a substantial increase
in credit risk on a timely basis, such as identifying instruments from a particular region that have been most adversely
affected by changing economic conditions even though this increase in risk might not have been reflected on individuals
credit risk.

4.2.3. Challenges
IFRS 9 recommends using lifetime PD over the remaining life of an instrument while assessing a significant increase in
credit risk since initial recognition. However, 12-month PD can be used if it is expected to provide similar results.

It should be noted that while using 12-month PD for an assessment, 12-month PD at initial recognition cannot be
directly compared with 12-month PD on the reporting date. The comparison should be made between 12-month PD on
the reporting date and 12-month PD on the reporting date calculated at the time of initial recognition. Similarly, while
comparing lifetime PD on the remaining expected life, PD for the same periods should be compared, e.g. if a financial
instrument has a maturity period of 5 years and 3 years are left (before it matures) as on the reporting date, then while
comparing lifetime PD, we would have to compare PD for this remaining 3 years as of the reporting date with that as of
initial recognition. Figure 9 provides an illustrative example of an assessment of a considerable increase in default risk.
Therefore, financial institutions will need to compute lifetime PD for financial instruments and will require systems to
store information at origination for most financial instruments.

Figure 9: Assessing significant increase in default risk

12-Month Point-in-Time PD Lifetime PD


1.8% 8.0%
PDs NOT comparable
1.6%
7.0%
1.4%
6.0%
1.2%
5.0%
1.0%
4.0%
0.8%
Comparable Points 3.0%
0.6%

0.4% 2.0%
As of Origination As of Reporting Date As of Origination As of Reporting Date
0.2% 1.0%

0.0% 0.0%
t=0 1 2 3 4 5 6 7 8 9 10 t=0 1 2 3 4 5 6 7 8 9 10

For some loans (e.g., bullet loans or CRE loans with a maturity of say 10 years and amortization term of say 30 years or
interest-only loans, which can be refinanced at maturity), it is quite clear that 12-month PD is not sufficient to assess a
significant increase in default risk. But for usual C&I term loans, it may appear that 12-month PD is a good alternative;
however, even in these cases, sometimes 12-month PD may not be a good method to ascertain risk. For example, in the
left chart in Figure 9, 12-month PDs appear to be almost the same, i.e. default risk has not increased significantly, but the
forecast as of the reporting date suggests a downturn after a year or so. Hence, when we compare lifetime PD, we observe
a significant increase in default risk since origination (as shown in the diagram on the right). This suggests 12-month
PD may not be sufficient in all conditions even though financial instrument characteristics suggest otherwise.

For the same counterparty, multiple loans may be treated differently, depending on the credit risk at the time of
origination of the loans. This, again, requires enhanced risk management systems.

12
4.2.4. Overcoming Challenges
Given so many difficulties in assessing a significant increase in credit risk, IFRS 9 provides a few guidelines and examples
for implementation, including the following:

A significant change in price indicators of credit risk for similar financial instruments with the same
terms and same counterparty as on (or close to) the reporting date from the date of origination can
be considered a significant increase in credit risk

If more stringent requirements such as covenants, increase collaterals or guarantees, or higher debt coverage
because of a change in credit risk would have been applied if instruments were to be newly originated, it can be
considered an increase in credit risk

Significant changes in market indicators of credit risk for a particular financial instrument or similar instruments with
the same expected life, e.g. CRE loans in a particular market. Market indicators of credit risk include:
Credit spread
Credit default swap prices for borrower
The length of time or the extent to which the fair value of a financial asset has been less than its amortized cost
Other market information related to a borrower, such as changes in the price of a borrowers debt or
equity instruments

A significant increase in credit risk of other financial instruments of the same borrower may also be considered

For a portfolio of instruments of similar characteristics, the maximum credit risk can be defined. If any instrument
shows credit risk higher than the threshold at the reporting date, the instrument can be categorized under stage 2

Assessment of significant increase in credit risk using lifetime PD or PIT PD calculated using lifetime PD models

In addition to the above guidelines, we believe financial institutions can use profitability or return concepts, such as
RAROC, and compare them with the hurdle rate, at the instrument level, to determine when profitability or return falls
below the threshold RAROC (or hurdle rate). This approach has the advantage of using one metric across different
portfolios and assumes a bank uses some profitability criteria to decide on deals and their pricing.

13
5. Impact of ECL Model
The scope of the IFRS 9 standard is not industry specific. For non-financial institutions, a catalyst (e.g., simplified approach
for trade and lease receivables) is likely to considerably reduce implementation efforts. For financial institutions, new
requirements are expected to involve revisions in not only accounting policies but also credit models to properly measure
the credit risk of financial instruments in a portfolio and calculate ECL.

5.1. Basel Framework vs. IFRS 9


Table 1: Basel EL Model Requirement vs. IFRS 9 ECL Model Requirements

Key Requirements Basel IFRS 9


Default definition 90+ DPD or less including others Should comply with internal
e.g. bankruptcy, charge definition, but should not be more
than 90 DPD, unless supportable
evidences available
12 months or lifetime (depending
Default window 12 month
upon the instrument's stage)
Five years for retail and 7 years for
Minimum data requirement corporate and others, 7 years for No such requirement
LGD
Generally hybrid of TTC and PIT,
PD Type PIT (Forward-looking)
calibrated as historical long run PD
LGD Type Downturn Adjustment to LGD No downturn adjustment
EAD 12 month EAD 12 month or lifetime EAD
Lifetime PD or 12-month
EL PD*LGD*EAD PD*Present Value of Cash
Shortfall
5.2. IRB Banks vs. Non-IRB Banks
Table 2: IRB Banks vs Non-IRB Banks

Key Requirements IRB Banks Non-IRB Banks


Leverage existing Basel models
Have to develop something new,
12-Months PD Model with adjustments e.g. adjustment for
both models and infrastructure
conservatism
Lifetime PD Models Need to develop Need to develop
LGD Models Leverage on existing models Need to develop
Leverage on existing models,
EAD Models develop lifetime EAD models Need to develop
especially for revolving facilities
Models forportfoliosnot
Need to develop Need to develop
covered under IRB
Will be less as can utilize existing
Modelling Challenges Challenges could be huge
setup

5.3. Impact on Regulatory Capital


The new IFRS 9 ECL models will significantly increase loss provisions, as financial institutions will have to assign ECL to
every instrument in a portfolio. The increased provisions will likely reduce Tier 1 capital ratio. Management will need to
take this into consideration during capital planning.

4. See IFRS 9, Paragraph 5.5.15

14
6. System and Data Requirements
6.1. System Requirements 6.2. Additional Data Requirements
IFRS 9 will be challenging in terms of system requirements, Loan-level data would be required including data for
particularly for financial institutions. In general, the corresponding counterparties, such as risk ratings,
implementation of the following requirements will be a demographic information, rating at origination,
challenge: default information, and payment information for
building models.
All the information required for credit risk assessment
would have to be stored (for most instruments) at Instrument-specific information, e.g. loan term, would
the time of origination. This may require a significant be required for credit risk assessment.
change in information-gathering systems.
Any computation of a significant increase in risk would
New models would need to be built for lifetime PD/EAD rely on the easy availability of pricing and market
and 12-month PD (if required). Systems would need to information.
be modified for the implementation of new models, as
they would be different from existing Basel models.

For the assessment of a significant increase in risk,


banks with highly diversified portfolios would require
a complex system. Moreover, for portfolios with
similar credit risk characteristics, the process may be
simpler, but for a diversified portfolio, it would be
relatively complex.

15
About Our Risk Management Services
Our Risk Management Services are uniquely positioned to help financial institutions by providing support, guidance, and thought
leadership on the rapidly evolving regulatory landscape through unrivaled, seamless and end-to-end solutions.

Our cross-functional teams of domain and technology experts help in the development and support of risk management systems,
while our consultants are involved in risk reporting, model development/validation, stress testing, data management, formulation of
trading strategies, pricing, and other regulatory tasks.

About Moodys Analytics Knowledge Services


We enable organizations to innovate and transform in an ever-changing world. Our global pool of domain experts enables businesses
to innovate, scale, and reduce costs. Providing customized solutions is in our DNA solutions that have a tangible impact on our
clients top-line and bottom-line, enabling them to achieve more with less.

Our clients include leading bulge-bracket financial institutions, asset managers, Fortune 100 corporations, mid-tier companies,
boutique investment banks, and funds. We support over 150 institutional clients through our team of 2,600+ employees and our
global delivery centers, located in close proximity to our clients.

To learn how we can help you, please email contact_maks@moodys.com

maknowledgeservices.com

2016 Moodys Corporation, Moodys Investors Service, Inc., Moodys Analytics, Inc. and/or their licensors and affiliates (collectively, MOODYS). All rights reserved.

SP23703/101215/IND-103