International GAAP® 2019: Generally Accepted Accounting Practice under International Financial Reporting Standards
Page 746
considered ‘integral’ to the guaranteed loan.
Financial instruments: Impairment 3773
5.8.2
Cash flows from the sale of a defaulted loan
At its meeting in December 2015, the ITG also discussed whether cash flows that are
expected to be recovered from the sale on default of a loan could be included in the
measurement of ECLs. ITG members noted that:
• such cash flows should be included in the measurement of ECLs if:
(a) selling the loan is one of the recovery methods that the entity expects to
pursue in a default scenario;
(b) the entity is neither legally nor practically prevented from realising the loan
using that recovery method; and
(c) the entity has reasonable and supportable information upon which to base its
expectations and assumptions.
• in order to support an entity’s expectation that loan sales would be used as a
recovery method in a default scenario, an entity’s past practice would be an
important consideration. However future expectations, which may differ from past
practice, would also need to be considered. With respect to the amount of
recovery proceeds to be included in the measurement of ECLs, an entity should
consider relevant market related information relating to loan sale prices;
• in these circumstances, the inclusion of recovery sale proceeds in the
measurement of ECLs would be appropriate for financial instruments in all stages;
stage 1, 2 and 3 (see 3.1 above). This is because when measuring ECLs, IFRS 9
requires an entity to consider possible default scenarios for financial instruments
in all three stages;
• given that it is necessary to consider multiple scenarios in measuring ECLs (see 5.6
above), it is possible that an entity will need to calculate the cash flows based on
both a scenario in which the loan will be sold and one in which it will not. Expected
sale proceeds would only be relevant when considering the possibility that a credit
loss occurs (i.e. in a default scenario) and would not be relevant when considering
the possibility that no credit loss occurs (i.e. in a performing scenario). For example
if, in the case of a particular loan, an entity concluded that there was a 10 per cent
probability of default occurring, it would only be when considering the outcome
of this default scenario that expected sale proceeds would be considered. If, in that
default scenario, the entity expected to recover 30 per cent of the contractual cash
flows of the loan through sale proceeds but only 25 per cent through continuing to
hold, then the LGD would be 70 per cent rather than 75 per cent. In addition, the
expected sale proceeds should be net of selling costs.
5.8.3
Treatment of collection costs paid to an external debt collection agency
Questions have arisen on how the collection costs paid to an external debt collection
agency affect the measurement of ECLs. As an example, a bank engages the services of
an external debt collection agency to recover accounts that are 90 days past due on its
behalf and, in exchange, pays the agency a fee based on the amount recovered.
One view could be that in measuring the ECLs, all cash flows related to the recovery of
the asset should be considered in estimating the expected cash shortfalls. Therefore, the
recoveries and any incremental and directly attributable payments made to the external
3774 Chapter 47
agency should both be considered, irrespective of whether the collections costs are
deducted from the amount recovered or paid to the external debt collection agency
separately. This view is based on the fact that IFRS 9 specifies that the estimate of cash
flows from the realisation of collateral should include the costs of obtaining the
collateral. This can be read to suggest that all cash flows relating to the various ways of
recovering the asset should be considered in measuring the expected shortfall. An
analogy can also be made with the ITG discussion on sales of assets where selling costs
are to be included (see 5.8.2 above). This also reflects that if the cash flows expected to
be received are greater than those that would be realised if the agency were not used,
excluding the collection cost paid to the agency will understate the ECLs. Further, for
assets that are purchased credit-impaired, these costs may already be implicit in the fair
value at which they are acquired and so, unless these costs are included, the EIR would
be inflated.
Another view could be that only cash flows related to the contractual terms of the
financial instrument should be considered, apart from the costs of obtaining and selling
collateral. This would exclude costs of an external debt collection agency. This view is
based on the definition of a credit loss as ‘the difference between all contractual cash
flows that are due to an entity in accordance with the contracts and all the cash flows
that the entity expects to receive (i.e. net of all cash shortfalls). The cash flows that are
considered shall include cash flows from the sale of collateral held or other credit
enhancements that are integral to the contractual terms.’ Since the collection costs paid
to the agency are not part of the contractual terms of the financial instrument, nor a cost
for realising collateral, then they should not be considered in the calculation of ECLs.
The reason for treating costs of realising collateral differently is that the proceeds from
selling collateral is itself not a contractual cash flow of the instrument and the
commercial value of the collateral would include provision for such costs. The analogy
with sales of assets is considered irrelevant for the same reasons as they do not relate to
the recovery of contractual cash flows.
5.9
Reasonable and supportable information
IFRS 9 requires an entity to consider reasonable and supportable information that is
available without undue cost or effort at the reporting date about past events, current
conditions and forecasts of future economic conditions and that is relevant to the
estimate of ECLs, including the effect of expected prepayments. [IFRS 9.5.5.17(c), B5.5.51].
5.9.1
Undue cost or effort
The term undue cost or effort is not defined in the standard, although it is clear from
the guidance that information available for financial reporting purposes is considered to
be available without undue cost or effort. [IFRS 9.B5.5.49].
Beyond that, although the standard explains that entities are not required to undertake
an exhaustive search for information, it does include, as examples of relevant
information, data from risk management systems, as described in 5.9.2 below.
What is available without undue cost or effort would be an area subject to
management judgement in assessing the costs and associated benefits. This is
consistent with the guidance in International Financial Reporting Standard for Small
Financial instruments: Impairment 3775
and Medium-sized Entities (IFRS for SMEs) in relation to the application of undue
cost or effort. Paragraph 2.14B of the IFRS for SMEs states that considering whether
obtaining or determining the information necessary to comply with a requirement
would involve undue co
st or effort depends on the entity’s specific circumstances and
on management’s judgement of the costs and benefits from applying that requirement.
This judgement requires consideration of how the economic decisions of those that
are expected to use the financial statements could be affected by not having that
information. Applying a requirement would involve undue cost or effort by an SME if
the incremental cost (for example, valuers’ fees) or additional effort (for example,
endeavours by employees) substantially exceed the benefits that those that are
expected to use the SME’s financial statements would receive from having the
information. Paragraph 232 of the Basis for Conclusions to the IFRS for SMEs further
observes that:
• the undue cost or effort exemption is not intended to be a low hurdle. In particular,
the IASB observed that it would expect that if an entity already had, or could easily
and inexpensively acquire, the information necessary to comply with a
requirement, any related undue cost or effort exemption would not be applicable.
This is because, in that case, the benefits to the users of the financial statements of
having the information would be expected to exceed any further cost or effort by
the entity; and
• that an entity must make a new assessment of whether a requirement will involve
undue cost or effort at each reporting date.
If the reporting entity is a bank, there would presumably be a higher hurdle to determine
what credit risk information would require undue cost or effort, compared to a reporter
that is not a bank, given that the benefit to users of its financial statements would be also
expected to be higher. It is also an issue on which the Basel Committee has issued
guidance (see 7.1 below).
5.9.2
Sources of information
The standard states that the information used should include factors that are specific to
the borrower, general economic conditions and an assessment of both the current as
well as the forecast direction of conditions at the reporting date. Entities may use
various sources of data, both internal (entity-specific) data and external data that
includes internal historical credit loss experience, internal ratings, credit loss experience
of other entities for comparable financial instruments, and external ratings, reports and
statistics. Entities that have no, or insufficient, sources of entity-specific data may use
peer group experience for the comparable financial instrument (or groups of financial
instruments). [IFRS 9.B5.5.51].
Although the ECLs reflect an entity’s own expectations of credit losses, an entity should
also consider observable market information about the credit risk of particular financial
instruments. [IFRS 9.B5.5.54]. Therefore, although entities with in-house economic teams will
inevitably want to use their internal economic forecasts, while loss estimation models will
be built based on historical data, they should not ignore external market data.
3776 Chapter 47
5.9.3
Information about past events, current conditions and forecasts of
future economic conditions
One of the significant changes from the IAS 39 impairment requirements is that entities
are not only required to use historical information (e.g. their credit loss experience) that
is adjusted to reflect the effects of current conditions, but they are also required to
consider how forecasts of future conditions would affect their historical data. A
discussion of how this process needs to consider the existence of non-linearity in how
expected losses will change with varying economic conditions and the need to assess
multiple economic scenarios is included at 5.6 above. This section explores some of the
other challenges in forecasting future conditions and the consequent ECLs.
The degree of judgement that is required to estimate ECLs depends on the availability
of detailed information. An entity is not required to incorporate detailed forecasts of
future conditions over the entire expected life of a financial instrument. The standard
notes that as the forecast horizon increases, the availability of detailed information
decreases and the degree of judgement required to estimate ECLs increases. Therefore,
an entity is not required to perform a detailed estimate for periods that are far in the
future and may extrapolate projections from available, more detailed information.
[IFRS 9.B5.5.50]. Most banks apply a 3 to 5 year period over which macro-economic
variables are considered to be capable of being forecast.
Beyond the horizon to which economic conditions can be reliably forecast, the application
guidance suggests that entities may often be able to assume that economic conditions
revert to their long-term average. There are at least two methods of how this might be
done: either by reverting to the average immediately beyond the forecast horizon, or by
adjusting the forecast data to the long-term average over a few years. The latter would,
perhaps, more effectively make use of all reasonable and supportable information.
Historical information should be used as a starting point from which adjustments are
made to estimate ECLs on the basis of reasonable and supportable information that
incorporates both current and forward-looking information: [IFRS 9.B5.5.52]
• in most cases, adjustments would be needed to incorporate these effects that were not
present in the past or to remove these effects that are not relevant for the future; and
• in some cases, unadjusted historical information may be the best estimate,
depending on the nature of the historical information and when it was calculated,
compared to circumstances at the reporting date and the characteristics of the
financial instrument being considered. But it should not be assumed to be
appropriate in all circumstances. [IFRS 9.BC5.281].
Additionally, when considering whether historical credit losses should be adjusted, an
entity needs to consider various items, including:
• whether the historical data captures ECLs that are through-the-cycle (i.e. estimates
based on historical credit loss events and experience over the entire economic
cycle) or point-in-time (i.e. estimates based on information, circumstances and
events at the reporting date); and
• the period of time over which its historical data has been captured and the
corresponding economic conditions represented in that history. The historical data
period may reflect unusually benign or harsh conditions unless it is long enough.
Financial instruments: Impairment 3777
Meanwhile, products, customers and lending behaviours all change over time. When
using historical credit loss experience, it is important that information about
historical credit losses is applied to groups that are defined in a manner that is
consistent with the groups for which the historical credit losses were observed.
The estimates of changes in ECLs should be directionally consistent with changes in
related observable data from period to period (i.e. consistent with trends observed on
payment status and macroeconomic data such as changes in unemployment rates,
property prices, and commodity prices). Also, in order to reduce the differences
between an entity’s estimates and actual credit loss experience, the estimates of ECLs
should be back-tested and re-calibrated, i.e. an entity should regularly review its inputs,
assumptions, methodology and estimation techniques used as well as its groupings of
sub-portfolios with shared credit risk characteristics (see 6.5 below).
Back testing will be considerably more challenging for forecasts over several years than
may be the case for just the 12-month risk of default, because detailed information may
not be available over the forecast horizon and the degree of judgement increases as the
forecast horizon increases. [IFRS 9.B5.5.52, B5.5.53]. Also, economic forecasts are usually
wrong, as reality is much more complex than can ever be effectively modelled.
Therefore it is probably not a useful exercise to back test macroeconomic assumptions
against what actually transpires, but it is useful to back test whether, for a given
macroeconomic scenario, credit losses increased or decreased as expected.
In estimating ECLs, entities must consider how to bridge the gap between historical loss
experience and current expectations. Estimating future economic conditions is only the
first step of the exercise. Having decided what will happen to macroeconomic factors
such as interest rates, house prices, unemployment and GDP growth, entities then need
to decide how they translate into ECLs. This will need to reflect how such changes in
factors affected defaults in the past. However, it is possible that the forecast
combination of factors may have never been seen historically together.
We observe that banks are also trying to align IFRS 9 to their existing risk management
practices. Many banks are making use of their regulatory capital calculation and stress
testing frameworks for their IFRS 9 calculations. This manifests itself in many of the
individual decisions that banks have made in implementing IFRS 9 (e.g. definitions of
default and alignment to stress testing). It is likely that regulators and standard-setters
will concur with this approach. Basel PDs are used as a starting point and there is a need
for a different calibration for IFRS 9, in order to transform a Basel PD into an unbiased
point in time metric and include forward looking expectations. Stress testing resources,
previously working almost exclusively with capital issues, also play a role in calculating