International GAAP® 2019: Generally Accepted Accounting Practice under International Financial Reporting Standards
Page 752
change in credit quality and different input parameters for calculating ECLs for a
European government bond, which result in very different outcomes and volatility of the
IFRS 9 ECL allowance. It should also be stressed that the default rates provided by
external rating agencies are historical information. Entities need to understand the sources
of these historical default rates and update the data for current and forward-looking
information (see 5.9.3 above) when measuring ECLs or assessing credit deterioration.
Example 47.12: Use of credit ratings and/or CDS spreads to determine whether
there have been significant increases in credit risk and to
estimate expected credit losses
Introduction
A significant challenge in applying the IFRS 9 impairment requirements to quoted bonds is that the credit
ratings assigned by agencies such as Standard & Poor’s (S&P), and the historical experience of losses by
rating grade, can differ significantly with the view of the market, as reflected in, for instance, credit default
swap (CDS) spreads and bond spreads.
To illustrate the challenges of applying IFRS 9 to debt securities, we have examined how the ECL could be
determined for a real bond issued by a European government on 16 September 2008 and due to mature
in 2024. For three dates, we applied the IFRS 9 calculations to this bond, which is assumed to have been
acquired at inception. In January 2009, the Standard & Poor’s credit rating of the government was AA+, as
at origination, but by January 2012, its rating was downgraded to A. The bond was further downgraded to
BBB– in March 2014 before recovery to BBB in May 2014.
Three approaches
Shown below are three approaches:
• Approach 1: Use of S&P credit ratings both to determine whether the bond has significantly increased
in credit risk and to estimate ECLs.
• Approach 2: Use of S&P credit ratings to determine whether the bond has significantly increased in
credit risk and CDS spreads to estimate ECLs.
• Approach 3: Use of CDS spreads both to determine whether the bond has significantly increased in credit
risk and to estimate ECLs.
Based on the historical corporate PDs from each assessed S&P credit rating (approach 1) and based on the
CDS spreads (approaches 2 and 3), the loan loss percentages were calculated below. For the calculations, an
often used LGD of 60% was applied. (Because the LGD represents a percentage of the present value of the
gross carrying amount, this example does not illustrate the effect of the time value of money).
The percentage loss allowances were, as follows:
Credit
Historical 12-month
Lifetime Loss allowance (%)
ratings 12-month
PD based
PD based Approach 1 Approach 2 Approach 3
PD based
on CDS
on CDS
on ratings
spread
spread
1 January 2009
AA+
0.02%
0.44%
12.81%
–
–
–
31 January 2009
AA+
0.02%
1.84%
30.48%
0.01
1.10
18.29
31 January 2012
A
0.06%
4.96%
51,48%
0.04
2.98
30.89
31 March 2014
BBB–
0.31%
0.57%
23.01%
0.18
0.34
13.81
Approach 1
According to the credit ratings, the bond was investment grade throughout this period. Hence, using the low
risk simplification, the loss allowance would have been based on 12-month ECLs. Using the corporate
historical default rates implied by the credit ratings and an assumption of 60% LGD to calculate the ECLs,
the 12-month allowance would have increased from 0.01% on 31 January 2009 to 0.04% three years later,
Financial instruments: Impairment 3803
increasing to 0.18% by 31 March 2014. It should be stressed that the historical default rates implied by credit
ratings are historical rates for corporate debt and so they would not, without adjustment, satisfy the
requirements of the standard. IFRS 9 requires the calculation of ECLs, based on current conditions and
forecasts of future conditions, to be based on reasonable and supportable information. This is likely to include
market indicators such as CDS and bond spreads, as illustrated by Approach 2.
Approach 2
In contrast to Approach 1, using credit default swap spreads to calculate the ECLs and the same assumption of
60% LGD to calculate the ECLs, the 12-month allowance would have increased from 1.1% on 31 January 2009
to 2.98% three years later, declining to 0.34% by 31 March 2014. The default rates implied by the CDSs are
significantly higher than would have been expected given the ratings of these bonds. The loss allowances are,
correspondingly, very much higher and very volatile. It might be argued that CDS spreads are too responsive to
short term market sentiment to calculate long term ECLs, but it may appear difficult to find other reasonable and
supportable information to adjust these rates so as to dampen the effects of market volatility.
Approach 3
Credit ratings are often viewed by the market as lagging indicators. For these bonds, the ratings are difficult
to reconcile with the default probabilities as assessed by the markets. It might be argued that it is not sufficient
to focus only on credit ratings when assessing whether assets are low risk since, according to CDS spreads,
the bond was not low risk at any time in the period covered in this example, as it showed a significant increase
in 1 year PD after inception (based on CDS spreads). The 1 year PDs increased from 0.44% on issue to 1.84%
by 31 January 2009. Assessing the bond as requiring a lifetime ECL at all three dates, based on CDS spreads,
would have given much higher loss allowances of 18.29%, 30.89% and 13.81%.
The counter-view might be that CDS spreads are too volatile to provide a sound basis for determining
significant deterioration. Perhaps the best way to make the assessment of whether a bond has increased
significantly in credit risk is to use more than one source of data and to take account of the qualitative
indicators as described in the standard.
Conclusion
The calculated ECL figures differ significantly depending on the approach taken as to how to determine a
significant change in credit quality and the parameters used for the calculation. Those based on CDS spreads
are both large and very volatile, reflecting the investor uncertainty during the period, when the possibility of
default depended more on the political will of the European Union to maintain the integrity of the Eurozone
than the economic forecasts for the particular country. As a result, the disparity between the effect of the use
of credit grades and CDSs is probably more marked than for most other security investments. Nevertheless,
the same challenges will be found with other securities, albeit on a smaller scale.
6.4.2 Delinquency
As already described at 6.2.2 above, the standard allows use of past due information to
assess whether credit risk has increased significantly, if reasonable and
supportable
forward looking information (either at an individual or a collective level) is not available
without undue cost or effort. This is subject to the rebuttable presumption that there
has been a significant increase in credit risk if contractual payments are more than
30 days past due. [IFRS 9.5.5.11]. Similar to the low credit risk simplification (see 6.4.1
above), the Basel Committee guidance (see 7.1 below) considers that sophisticated banks
should not use days past due information as a primary indicator, because it is a lagging
indicator, but only as a backstop measure alongside other, earlier indicators.
Most sophisticated banks intend to follow this regulatory guidance. In addition, it is a
useful measure of the effectiveness of more forward looking primary criteria to monitor
the frequency that assets reach 30 days past due without having already been transferred
to stage 2.
3804 Chapter 47
6.4.3
12-month risk as an approximation for change in lifetime risk
In determining whether there has been a significant increase in credit risk, an entity
must assess the change in the risk of default occurring over the expected life of the
financial instrument. Despite this, the standard states that: ‘...changes in the risk of a
default occurring over the next 12 months may be a reasonable approximation...unless
circumstances indicate that a lifetime assessment is necessary’. [IFRS 9.B5.5.13].
The IASB observed in its Basis for Conclusions that changes in the risk of a default
occurring within the next 12 months generally should be a reasonable approximation of
changes in the risk of a default occurring over the remaining life of a financial instrument
and thus would not be inconsistent with the requirements. Also, some entities use a 12-
month PD measure for prudential regulatory requirements and these entities can
continue to use their existing systems and methodologies as a starting point for
determining significant increases in credit risk, thus reducing the costs of
implementation. [IFRS 9.BC5.178].
However, for some financial instruments, or in some circumstances, the use of changes
in the risk of default occurring over the next 12 months may not be appropriate to
determine whether lifetime ECLs should be recognised. For a financial instrument with
a maturity longer than 12 months, the standard gives the following examples:
[IFRS 9.B5.5.14]
• the financial instrument only has significant payment obligations beyond the next
12 months;
• changes in relevant macroeconomic or other credit-related factors occur that are
not adequately reflected in the risk of a default occurring in the next 12 months; or
• changes in credit-related factors only have an impact on the credit risk of the
financial instrument (or have a more pronounced effect) beyond 12 months.
At its meeting on 16 September 2015, the ITG members discussed the use of changes in 12-
month risk of default as a surrogate for changes in lifetime risk and commented as follows:
• An entity would be expected to complete a robust analysis up front in order to
support the conclusion that changes in the 12-month risk of a default occurring was
a reasonable approximation for the assessment of changes in the lifetime risk of
default occurring.
• The level of initial analysis required would depend on the specific type of financial
instrument being considered. Consequently in some cases, a qualitative analysis
would suffice whereas in less clear-cut cases, a quantitative analysis may be
necessary. Also, it may be appropriate to segregate portfolios (e.g. by maturity) in
order to facilitate the analysis for groups of similar financial instruments.
• An entity would need to be satisfied on an ongoing basis that the use of changes in
the 12-month risk of a default occurring continued to be a reasonable
approximation for changes in the lifetime risk of a default occurring.
At the meeting, the ITG members also discussed:
• The appropriate type of review that should be undertaken on an ongoing basis.
While a quantitative review would not necessarily be required, it would depend on
Financial instruments: Impairment 3805
the specific facts and circumstances. One way of approaching an ongoing review
would be as follows:
(a) identify the key factors that would affect the appropriateness of using changes
in the 12-month risk of a default occurring as an approximation of changes in
the lifetime risk of default occurring;
(b) monitor these factors on an ongoing basis as part of a qualitative review of
circumstances; and
(c) consider whether any changes in those factors indicated that changes in the
12-month risk of a default occurring was no longer an appropriate proxy for
changes in a lifetime risk of default occurring.
• If it were determined that changes in the 12-month risk of a default occurring were
no longer a reasonable approximation for the assessment of changes in the lifetime
risk of a default occurring, an entity would be required to determine an appropriate
approach to capture changes in the lifetime risk of a default occurring.
• It is important to emphasise that the guidance which permits an entity to use
changes in the 12-month risk of a default as an approximation for the lifetime risk
of default, is only relevant for the assessment of significant increases in credit risk
and does not relate to the measurement of ECLs. When an entity is required to
measure lifetime ECLs, that measurement must always reflect the lifetime risk of a
default occurring.
• IFRS 9 does not prescribe how an entity should determine whether the use of
changes in the 12-month risk of a default was an appropriate proxy for assessing
changes in the lifetime risk of a default. However, it was noted that entities are
required to disclose how they make the assessment of significant increases in credit
risk, in accordance with IFRS 7.
Most sophisticated banks are using the lifetime risk of default (or an annualised equivalent)
rather than the 12-month risk of default or the Basel risk of default for assessing whether
there has been a significant increase in credit risk. Movements in a 12-month risk of
default are, for most products and conditions, strongly correlated with movements in the
lifetime risk. However, these banks appreciate that 12-month PDs may need to be
adjusted or calibrated to reflect the longer term macroeconomic outlook. Also, there are
products such as interest-only mortgages and those with an introductory period in which
no repayments are required, where additional procedures may need to be implemented
in order to ensure that they are transferred to stage 2 appropriately.
6.4.4
Assessment at the counterparty level
As indicated by Example 7 in the Implementation Guidance of IFRS 9, assessment of
significant deterioration in credit risk can be made at the level of the counterparty rather
than the individual financial instrument. Such assessment at the counterparty level is
only allowed if the outcome would not be different to the outcome if the financial
instruments had been individually assessed. [IFRS 9.BC5.168]. In certain circumstances,
&nbs
p; assessment at the counterparty level would not be consistent with the impairment
requirements. Both these situations are illustrated in the example below, based on
Example 7 in the Implementation Guidance for the standard. [IFRS 9 IG Example 7 IE43-IE47].
3806 Chapter 47
Example 47.13: Counterparty assessment of credit risk
Scenario 1
In 2012 Bank A granted a loan of $10,000 with a contractual term of 15 years to Company Q when the
company had an internal credit risk rating of 4 on a scale of 1 (lowest credit risk) to 10 (highest credit risk).
The risk of a default occurring increases exponentially as the credit risk rating deteriorates so, for example,
the difference between credit risk rating grades 1 and 2 is smaller than the difference between credit risk
rating grades 2 and 3. In 2016, when Company Q had an internal credit risk rating of 6, Bank A issued another
loan to Company Q for $5,000 with a contractual term of 10 years. In 2019 Company Q fails to retain its
contract with a major customer and correspondingly experiences a large decline in its revenue. Bank A
considers that as a result of losing the contract, Company Q will have a significantly reduced ability to meet
its loan obligations and changes its internal credit risk rating to 8.
Bank A assesses credit risk on a counterparty level for credit risk management purposes and determines that
the increase in Company Q’s credit risk is significant. Although Bank A did not perform an individual
assessment of changes in the credit risk on each loan since its initial recognition, assessing the credit risk on
a counterparty level and recognising lifetime ECLs on all loans granted to Company Q, meets the objective
of the impairment requirements as stated in paragraph 5.5.4 of IFRS 9. This is because, even since the most
recent loan was originated, its credit risk has increased significantly. The counterparty assessment would
therefore achieve the same result as assessing the change in credit risk for each loan individually.
Scenario 2
Bank A granted a loan of $150,000 with a contractual term of 20 years to Company X in 2012 when the
company had an internal credit risk rating of 4. During 2016 economic conditions deteriorate and demand
for Company X’s products has declined significantly. As a result of the reduced cash flows from lower
sales, Company X could not make full payment of its loan instalment to Bank A. Bank A re-assesses