IFRS 9: an analysis of Q2impacts and expectationsahead
15 September 2020
Agenda
IFRS 9: an analysis of Q2 impacts and expectations ahead2
IFRS 9: EY analysis of Q2 impacts and expectations ahead
Tara Kengla,IFRS 9 Solution Leader
Ernst & Young LLP
Introduction and Q&A
Laure Guégan,European FinancialServices IFRS Leader
Ernst & Young et Associés
Impacts on loan provisionsand related drivers andchallenges ahead
Catriona Early,Regulatory ReportingSpecialist
Ernst & Young LLP
Moratoria, guaranteedloans and the impact ofmacroeconomic scenarios
Marek Rozkrut,Head of Economic Analysisfor Europe
EY Polska
An outlook on themacroeconomicenvironment
An overview of theimpacts on loanprovisions and the keydrivers
Laure Guégan
3 IFRS 9: an analysis of Q2 impacts and expectations ahead
IFRS 9: an analysis of Q2 impacts and expectations ahead4
Analysis of the half-year 2020 IFRS 9 impacts
► Benchmark analysis based on the half-year earnings communication► Half-year reports International Accounting Standards (IAS) 34 financial statements,
earnings presentations and Pillar 3 regulatory disclosures► Areas of focus:
► Half-year 2020 expected credit loss (ECL) expense and COVID-19 pandemic effects► Cost of Risk (CoR)/loss rate ratios (ECL expense/gross loans)► Exposures and impairment allowances (movements and stage allocation)► Coverage ratios (ECL allowance/gross loans)► Macroeconomic assumptions (scenarios, weighting) and outlooks
1.183bAverage total assets
534bAvg. gross loans tocustomers
11.1bAverage allowance
14.2%Average CET1 ratio
medianaverage
outlier
25%
25%
25%
25%
Note: all reporting currencies have been converted to Euros for purposes of comparison
IFRS 9: an analysis of Q2 impacts and expectations ahead5
ECL in half-year (HY) 2020 already doubles full-year (FY) 2019 ECL expense
ECL expense in HY 2020compared to FY 2019 (€b)
ECL expense FY 2019ECL expense Q1 2020
ECL expense Q2 2020
2.7bAverage HY 2020ECL expense
IFRS 9: an analysis of Q2 impacts and expectations ahead6
The Cost of Risk (CoR) in HY 2020 has reached 100bps on average
2019 CoR
Bubble size reflects the proportionof S1 and S2 in the ECL expense
x4 x2
x1
NB. Re-calculated ratios whichmay differ from disclosed CoR
CoR HY 2020 (annualized)compared to FY 2019
ECL expense in HY 2020compared to FY 2019 (€b)
ECL expense FY 2019ECL expense Q1 2020
ECL expense Q2 2020
Average 1.0%
0.4%Average
Q2 2020 compared to Q1 2020► Comparing Q2 2020 to Q1 2020 shows some adjustments and
a slight reduction in the dispersion of banks► The five banks with the highest CoR in Q1 have decreased
(by 60 bp on average)► 10 banks have increased their CoR in Q2 (50bp on average)► A few significant shifts upwards driven by macroeconomic
scenarios (three banks increased by 90bp on average)
Between Q1 and Q2 2020, the dispersion between banks has slightly reduced
7
HY 2020 compared to FY 2019► The annualized CoR/loss rates for HY 2020 are still
scattered across a wide range from 30 bps to 205 bps► Same country trends as in Q1
► Highest levels of CoR: Spain and UK► Highest increase compared to 2019: UK (x 4.6)► Germany and France: x2.5 on average
► Various drivers :► Products► Geographical footprint► Macroeconomic outlook, scenarios and weights► Government support measures► Modeled versus non-modeled impacts► A few banks have significant single-name losses
► Two consistent trends across banks driving the sharpincrease in ECL expense :► Stage 1 (S1) and Stage 2 (S2) increase► Significant overlays
Q2 2020CoR
(annualized)
Q1 2020 CoR(annualized)
55%Q2 proportion of HY2020 ECL expense
IFRS 9: an analysis of Q2 impacts and expectations ahead
IFRS 9: an analysis of Q2 impacts and expectations ahead8
The HY20 ECL expense was driven by Stage 1 (S1) and Stage 2 (S2),which represent almost half of the impact
► The sharp increase in S1 and S2 ECL was driven byrevised macroeconomic scenarios and overlays
S1 and S2 portion of ECL expenseHY 2020 compared to FY 2019 (€b)
ECL expense FY 2019: S3ECL expense HY 2020: S1 / S2ECL expense HY2020: S3
ECL expense FY 2019: S1 / S2
3%S1 and S2 inFY 2019 ECL expense
44%S1 and S2in HY 2020 ECL expense
IFRS 9: an analysis of Q2 impacts and expectations ahead9
The HY20 ECL expense was driven by Stage 1 (S1) and Stage 2 (S2),which represent almost half of the impact
► The sharp increase in S1 and S2 ECL was driven byrevised macroeconomic scenarios and overlays
► Increased coverage ratio for performing exposures:
S1 and S2 portion of ECL expenseHY 2020 compared to FY 2019 (€b)
ECL expense FY 2019: S3ECL expense HY 2020: S1 / S2ECL expense HY2020: S3
ECL expense FY 2019: S1 / S2
HY 2020
Year-end2019
S1 / S2 ECL allowance / gross loans to customers
Average0.45%
Average0.64%
An increased use of overlays to supplement the models,involving a considerable level of expert judgment
10
► On average, overlays represent 20% of the HY 2020 ECLexpense, ranging from 5% to 65%► Often result in a net increase of the ECL balance, but
with very significant downward effects due to modeladjustments
► Some banks have included detailed disclosuresincluding the amount by main category, the modelsegments impacted and the framework used
► Comparisons are still difficult as overlays vary innature and disclosed amounts include some offsettingeffects
► Some significant overlays booked in Q1 have been“absorbed” by the revision of macroeconomic scenarios► The need for an overlay due to changed
macroeconomic conditions in Q1 was deemed nolonger necessary under the revised scenarios
► In some cases, the significant overlay recorded in thefirst quarter remained but was allocated across thedifferent markets / portfolios
► Banks describe three main types of overlays
1. Model adjustments (“in-model” and “post-model”adjustments)► Generally more significant for wholesale and low-risk counterparties
models► A number of IFRS 9 ECL models are now operating outside the
boundaries to which they were calibrated► Some models extrapolate the recent unseen variations in
macroeconomic loss drivers into unrealistically high default rates► Adjustments are made to more appropriate levels based on historical
benchmarks (peak default rates observed in previous crises andother stress scenario analysis)
► Wider-ranging model changes will take time to develop and needmore real data
2. Effect of government support measures► The mitigating effect of these measures is expected to result in both
a delay and a reduction in the peak level of default and loss rates► Several banks note that macroeconomic variables have been
averaged to ensure an appropriate probability of default (PD) output
3. Sectors► Adjustments have been applied to implement an appropriate
differentiation in the severity of projected default rate conditions fordifferent industry sectors
IFRS 9: an analysis of Q2 impacts and expectations ahead
0
25
50
75
100
125
150
175
200
225
UK
B1
UK
B2
UK
B3
UK
B4
UK
B5
Span
ish
B1
Span
ish
B2
Ital
ian
B1
Ital
ian
B2
Dut
ch B
1
Dut
ch B
2
Fren
ch B
1
Fren
ch B
2
Fren
ch B
3
Fren
ch B
4
Ger
man
B1
Ger
man
B2
Swis
s B1
CoR (1H2020)2020 Outlook (low)2020 Outlook (high)
IFRS 9: an analysis of Q2 impacts and expectations ahead11
Outlook suggests increased convergence by year-endbut with different country trends remaining
2020 outlook CoR compared to HY 2020 CoR (as disclosed) – in bps
► The average high end outlook is 105bps,and the low end is around 85bps
► The outlook is generally lower for thebanks with the highest CoR in HY 2020,suggesting they may have frontloadedmore the expected losses
► Two banks have an outlook below HY2020 and four banks have indicated thatthey expect less in the second half
► Three banks indicate a high-end outlookhigher than in first HY 2020 (suggestinga higher expense in the second half)
► For five banks, the outlook is very closeto the first HY 2020
► The outlook always includes a caveatthat it is only relevant if conditions don’tfurther deteriorate
► Some banks haven’t disclosed anyoutlook and stress the high level ofuncertainty in projecting CoR
CoR HY 2020 (annualized)2020 outlook (low)2020 outlook (high)
IFRS 9: an analysis of Q2 impacts and expectations ahead12
More banks are now applying the IFRS 9 transitional arrangementsin the calculation of their CET 1 ratio
Benefit of IFRS 9 transitional adjustment at Q2 2020 for CET1 ratio► On average, banks which apply the add-back experience a
benefit of 40 basis points (bps), with a maximum add-back of120 bps reported
► This is higher than the benefit at Q1, reflecting increasedECL allowance in Q2
► This is based on 13 banks that apply the transitionalarrangements (TA), all of which have internal rates-based(IRB) portfolios
► At Q2, there are an additional four banks that have adoptedthe TA. These banks cite the European Central Bank (ECB)
► The majority of banks apply the revised “quick fix” reliefpackage that was adopted into European law in June
0.1%
1.2%
0.8%
0.9%
0.1%
0.4%
0.7%
0.8%
0.1%
0.3%
0.2% 0.2%
0.1%
0.0%
0.2%
0.4%
0.6%
0.8%
1.0%
1.2%
UK
Ban
k 1
UK
Ban
k 2
UK
Ban
k 3
UK
Ban
k 4
UK
Ban
k 5
Span
ish
B 1
Ital
ian
B 1
Ital
ian
B 2
NLD
B 1
Fren
ch B
1
Fren
ch B
2
Fren
ch B
4
Ger
man
B 1
0.0
0.2
0.4
0.6
0.8
1.0
2018 2019 2020 2021 2022 2023 2024 2025
Revised proposal: add-back
Reviseddynamic add-
back
Static add-backand additional
static continues
IFRS 9: an analysis of Q2 impacts and expectations ahead13
A significant increase in Stage 2 exposures driven by forward-looking information
Stage 2% of [Stage 1 + Stage 2 exposures] ► On average, the proportion of Stage 2 almost doubles, butwith significant differences reflecting different product mixand how banks incorporate forward-looking information in theassessment of significant deterioration in credit risk (SICR)
► Legislative and non-legislative large scale moratoria did notautomatically trigger Stage 2 transfers
► Most often, banks have not changed their quantitative andqualitative criteria, but supplemented them with targetedsupplemental analysis for vulnerable sectors and borrowerswith weaker ratings before the crisis
2020 Stage 2% of [Stage 1 + Stage 2 exposures]2019 Stage 2% of [Stage 1 + Stage 2 exposures]
7.7%2019 Average Stage 2 %
11.7%Average 2020 Stage 2 %
14
Wholesale portfolios show the largest increase in Stage 2
Retail unsecured - Stage 2% of [S1 + S2 exposures] Retail secured - Stage 2% of [S1 + S2 exposures]
IFRS 9: an analysis of Q2 impacts and expectations ahead
Total retail - Stage 2% of [S1 + S2 exposures] Wholesale - Stage 2% of [S1 + S2 exposures]
15
However, increase in coverage ratio is morepronounced for retail unsecured portfolios
Retail unsecured - coverage ratio (ECL/gross loans) Retail secured - coverage ratio (ECL/gross loans)
IFRS 9: an analysis of Q2 impacts and expectations ahead
Total retail - coverage ratio (ECL/gross loans) Wholesale - coverage ratio (ECL/gross loans)
IFRS 9: an analysis of Q2 impacts and expectations ahead16
Stage 2 exposures in arrears have decreased,reflecting the effect of moratoria and other support measures
% Stage 2 exposures in arrears(from Pillar 3 disclosures)
30 days < Past due ≤ 90 days / Stage 2 loans at HY 202030 days < Past due ≤ 90 days / Stage 2 loans at FY 2019
► All banks stress that moratoriaand other support measures have“turned-off” some usual SICRtriggers
► Legislative and non-legislativelarge scale moratoria did notautomatically triggered Stage 2transfers
► The significant differences inStage 2 increase at HY 2020reveals differences in how banksincorporate forward-lookinginformation in SICR assessment
► Further transfers in Stage 2 aretherefore expected in the nextquarters when the moratoriacome to an end
IFRS 9: an analysis of Q2 impacts and expectations ahead17
Stage 3 ratio remained relatively stable,also reflecting the effect of moratoria and other support measures
Stage 3 % of total loans to customers ► On average, the proportion of Stage 3increased slightly from 2.84% to 3%
► The dispersion is still significant (from 1% to7%) but has decreased over the past years
► Legislative and non-legislative large scalemoratoria have not automatically beenconsidered as forbearance measures anddid not trigger a Stage 3 transfer
► Banks have to monitor carefully thetransition from moratoria to “back tonormal” to avoid massive shifts in Stage 3
2020 Stage 3 % of total loans to customers2019 Stage 3 % of total loans to customers
3%Average 2020 Stage 3 %
Moratoria, guaranteedloans and the impact ofmacroeconomic scenarios
Catriona Early
18 IFRS 9: an analysis of Q2 impacts and expectations ahead
26
60
17
26
6
91
33
8
18
7 3 4
17
4
-
5
4
8
1
12
7
-
21
25
33 21
40
4
6
6
13
8
27
6
8 10
1315
3
6%
14%
6%
12%
5%
13%12%
7%
10%
3%
7%6%
7%
1%
4%
-20%
-15%
-10%
-5%
0%
5%
10%
15%
-
20
40
60
80
100
120
UK Bank 1 UK Bank 2 UK Bank 3 UK Bank 4 UK Bank 5 Spanish B 1Spanish B 2 Italian B 1 Italian B 2 NLD B 1 NLD B 2 French B 1 French B 2 German B 1German B 2
% of
boo
k
Curr
ency
uni
t, b
Households SMENon-financial corporations (non-SME) Other% of total loans and advances subject to moratorium
Banks have started to disclose quantitative informationon moratoria and guaranteed loans - moratoria
IFRS 9: an analysis of Q2 impacts and expectations ahead19
Legislative and non-legislative moratoria
► Some significant differences in the allocationof moratoria between households, small andmedium sized entities (SME) and othercorporate companies
► On average, moratoria represents 7% of totalloans but with significant differences (from1% to 4% in Germany to 13% in Spain)
► The highest proportion of moratoria tend tobe more concentrated on households withlower average exposures (15-20k€ onaverage in Spain)
► Generally, the expected maturity is sixmonths, revealing a high level of uncertaintyin the short-term
► A number of banks stress that a number ofobligors have already resumed normalpayments revealing the efficiency of thesetemporary measures for at least a portion ofthe obligors
Loans and advances subject to legislative and non-legislative moratoria(based on Pillar 3 disclosures)
2.51.1 1.2 0.6
7.5 7.9
5.4
20.5
7.73.1
6.0
9.9
0.7 0.9 1.5
10.6
0.5 0.4
4.4
6.1
7.0
2.5
3.1
4.9
0.5 0.5
0.8
0
5
10
15
20
25
UK Bank 1 UK Bank 2 UK Bank 3 UK Bank 4 UK Bank 5 Spanish B 1 Spanish B 2 Italian B 1 Italian B 2 NLD B 1 NLD B 2 French B 1 French B 2 German B 1 German B 2 CH B 1
Curr
ency
uni
t, b
n
Households SME Non-financial corporations (non-SME) Other
Banks have started to disclose quantitative informationon moratoria and guaranteed loans – guaranteed loans
IFRS 9: an analysis of Q2 impacts and expectations ahead20
Guaranteed loans
► The amount of guaranteed loansreveal different implementationframeworks across countries: differenttargets (SME vs. other corporates),different structure (fully or partiallyguaranteed) and speed of execution
► In some countries, the guaranteedloans are passed-through to a state-owned bank and not reported on thebalance sheet
► In some schemes, no payment isrequired in the first year, resulting inthe same effect as moratoria (usualsigns of deterioration temporarilyturned-off)
Newly originated loans and advances provided under newly applicable public guarantee schemes (b€)(based on Pillar 3 disclosures)
0
2
4
6
8
10
12
14Agriculture and food
Automobiles and components
Aviation
Commodity traders
Construction, real estatedevelopment and
infrastructure
Health
Hospitality, leisure andtourismMetals and mining
Oil and gas
Retail (non-food retail)
Sports, Education and Media
Textiles
Transportation and logistics(excl. airlines)
Number of banks citing specific vulnerable exposures
No. of banks citing exposure, Q1 No. of banks citing exposure, Q2
Corporate exposures: vulnerable sector analysis
► Most banks (80%) provide significant informationon vulnerable sectors in the earnings presentation
► A number of banks stress that their sectorvulnerability is magnified by the COVID-19pandemic
► The content and format of information variessignificantly:► Sector coverage► Exposure values for vulnerable, high risk or
noted sectors► Interaction with geography► Credit quality and credit risk mitigation► Outlook (sensitivity analysis)
► Significant diversity in IFRS 9 ECL approaches:► Stage 2 transfers (individual or portfolio
approach)► Stressed ECL► Overlays
► Close interaction with macroeconomic outlookand expected impact of support measures
21 IFRS 9: an analysis of Q2 impacts and expectations ahead
Source: European Commission (EC), Summer 2020 Economic Forecast, July 2020; European Central Bank, Eurosystem Staff Macroeconomic Projections, September 2020; EuropeanCentral Bank, Eurosystem Staff Macroeconomic Projections, June 2020; Bank of England (BoE), Monetary Policy Report, August 2020; Bank of England, Monetary Policy Report, May2020
22
Annual GDP growth in the Eurozone (2019=100) Annual GDP growth in the UK (2019=100)
85
90
95
100
105
2019 2020 20222021
German B1
French B3French B1
Italian B1
ECB mild (Sep 2020)ECB baseline (Sep 2020)ECB severe (Sep 2020)ECB baseline (Jun 2020)
85
90
95
100
105
20222019 2020 2021
UK B1
UK B3
BoE (Aug 2020)
UK B4
BoE (May 2020)EC (Jul 2020)
ECB mild
ECB severe
Observations on forecasts disclosed by banksGDP forecasts compared to regulators revised forecasts
UK B2
French B4
0%
10%
20%
30%
40%
50%
60%
70%
80%
1 2 3 4 5
Wei
ghti
ng
Scenarios
Observations on forecasts disclosed by banksBanks have often adjusted their approach to multi-economic scenarios
IFRS 9: an analysis of Q2 impacts and expectations ahead23
► Banks have often adjusted their approach to multi-economic scenarios, some have removed some scenarios (now considered unrealistic)and others have increased the number of scenarios to reflect the increased level of uncertainty around the baseline
► The weight allocated to the baseline is on average 54% (compared to 51% at year-end 2019) but there are some significant differences :► Two banks increased it from 30% or 42.5% to 70%► Three banks kept it stable at 30%, 50% or 60%► Three banks decreased it, including the two banks with the highest weights in 2019 (from 80% and 74% to 60% and 65%)
Scenarios weightings year-end 2019 Scenarios weightings half-year 2020
0%
10%
20%
30%
40%
50%
60%
70%
80%
1 2 3 4 5
Wei
ghti
ng
ScenariosUpper case 1 Upper case 2 Base case Lower case 1 Lower case 2Upper case 1 Upper case 2 Base case Lower case 1 Lower case 2
The incremental reported ECL due to multiple economic scenariosremains a relatively small proportion of baseline ECL
IFRS 9: an analysis of Q2 impacts and expectations ahead24
9% 8% 8%
16%
27%
4%
105%
56%60%
53%
89%
15%
0%
20%
40%
60%
80%
100%
120%
UK Bank 1 UK Bank 2 UK Bank 3 UK Bank 5 CH B 1 NLD B 2
Sensitivity to macroeconomic scenarios% increase in ECL allowance
Reported versus Central (% Central)
Most downside versus Reported (% Reported)
► Banks stress there was an increasein ECL sensitivity observed in allscenarios
► A few banks disclose the ECLallowance calculated under severalmacroeconomic scenario weightedat 100%
► The reported ECL is theprobability-weighted outcome ofthese different modeled ECL
► On average, the reported ECLincreases the baseline ECLbalance (based on 100%weighting) by 12%
► On average, the most downsidescenario (based on 100%weighting) would increase thereported ECL balance (probability-weighted) by 63%
central (% central)
reported (% reported)
An outlook on themacroeconomicenvironment
Marek Rozkrut
25 IFRS 9: an analysis of Q2 impacts and expectations ahead
Many economies contracted already in Q1, with further deterioration in Q2 2020
26
Source: Eurostat; local statistical offices. * Data not seasonally adjusted
4.4
2.4
0.8 0.7 0.3
-0.1 -0.2 -0.4 -0.5-1.2 -1.5 -1.7 -2.2 -2.3 -2.4 -2.8 -3.2
-3.7 -3.7 -4.1
-5.6 -5.7
Latv
ia
Lith
uani
a
Fran
ce
USA
Irel
and
Ger
man
y
Net
herl
ands
Cypr
us
Slov
enia
Luxe
mbo
urg
Esto
nia
Mal
ta
Gre
ece
Finl
and
UK
Euro
zone
Port
ugal
Belg
ium
Aus
tria
Slov
akia
*
Spai
n
Ital
y
a) GDP growth in Q1 2020 (% YOY) in the Eurozone, UK and US – seasonally adjusted data
b) GDP growth in Q2 2020 (% YOY) in the Eurozone, UK and US – seasonally adjusted data
-3.7 -4.0-6.3 -6.5
-8.6 -9.1 -9.2-11.3 -11.9 -12.1 -12.9 -12.9
-14.4 -14.7 -15.2 -15.2 -16.3-17.7 -18.9
-21.7 -22.1
Irel
and
Lith
uani
a
Ger
man
y
Net
herl
ands
Latv
ia
Finl
and
Esto
nia
USA
Slov
enia
Cypr
us
Slov
akia
*
Aus
tria
Belg
ium
Euro
zone
Gre
ece
Mal
ta
Port
ugal
Ital
y
Fran
ce UK
Spai
n
IFRS 9: an analysis of Q2 impacts and expectations ahead
However, GDP contraction in Q2 2020 was less severe than forecasted – with theexception of Spain and UK
27
Source: Eurostat (September 8, 2020); European Commission, Summer 2020 Economic Forecast, July 2020
8.07.0
5.7
3.93.4 3.3 3.0 2.9
2.4 2.3 1.91.5
0.4 0.3 0.2
-1.5
-3.2
Belg
ium
Irel
and
Lith
uani
a
Esto
nia
Slov
akia
Latv
ia
Finl
and
Net
herl
ands
Italy
Aus
tria
Spai
n
Fran
ce
Euro
zone
Slov
enia
Port
ugal
Ger
man
y
UK
Difference for GDP growth in Q2 2020 between the initial Q2 2020 realization and the last forecast of the European Commissionin percentage point (PP)
IFRS 9: an analysis of Q2 impacts and expectations ahead
But no Eurozone member state is expectedto escape a deep recession in 2020
28Source: European Commission, Summer 2020 Economic Forecast, July 2020; European Central Bank, Eurosystem Staff Macroeconomic Projections, September 2020; IMF, WorldEconomic Outlook, June 2020 (for the US). * ECB projections (September 2020)
-0.1
-0.9
-1.0
-1.1
-1.3
-1.3
-1.9
-2.0
-2.3
-2.5
-2.7
-2.8
-2.9
-3.1
-3.4
-3.5
-3.7
-3.8
-3.9
-4.3
-4.4
-4.6
-5.8
Italy
Luxe
mbo
urg
Ger
man
y
Mal
ta
Latv
iaLi
thua
nia
Aus
tria
Slov
enia
Esto
nia
Slov
akia
Net
herl
ands
Irel
and
Cypr
usBe
lgiu
m
Fran
ce
Euro
zone
Gre
ece
Finl
and
USA U
KPo
rtug
alSp
ain
Euro
zone
*
b) Cumulative GDP growth in 2020-2021 (%)a) GDP growth in 2020 (% year on year (YOY))
-6.0
-6.2
-6.3
-6.3
-6.8
-7.0
-7.0
-7.1
-7.1
-7.7
-7.7
-8.0
-8.0
-8.5
-8.7
-8.8
-9.0
-9.0
-9.7
-9.8
-10.
6-1
0.9
-11.
2
Slov
akia
Mal
taLu
xem
bour
gFi
nlan
d
Aus
tria
Latv
ia
Ger
man
yN
ethe
rlan
ds
Gre
ece
Slov
enia
Lith
uani
aCy
prus
Spai
n
Esto
nia
USA
Euro
zone
*Ir
elan
dEu
rozo
neBe
lgiu
m UK
Port
ugal
Fran
ce
Italy
COVID-19 takes its economic toll: Eurozone experiencing a deep recession in 2020
Source: Eurostat (historical data); European Commission, Summer 2020 Economic Forecast, July 2020; Organisation for Economic & Cooperative Development (OECD), EconomicOutlook, June 2020; European Central Bank (ECB), Eurosystem Staff Macroeconomic Projections, September 2020; World Bank (WB), Global Economic Prospects, June 2020 ;International Monetary Fund (IMF), World Economic Outlook, June 2020. * Seasonally adjusted data
-9
-6
2019 202220212020
-3
0
6
3
9
Eurostat
ECBEC
OECDWBIMF
-15
-10
-5
0
5
10
15
20
Q1 2019 Q1 2021Q1 2020 Q1 2022
ECEurostat
OECD*ECB*
a) Quarterly GDP growth in the Eurozone (% YOY) b) Annual GDP growth in the Eurozone (% YOY)
29
d) Cumulative GDP growth in the Eurozone in 2020-2021 (%)c) GDP growth in the Eurozone (Q4 2019 = 100)
80859095
100
Q4 2019 Q2 2022Q2 2020 Q4 2022Q4 2020 Q4 2021Q2 2021
EC
ECB*OECD*
-3.1 -3.1 -3.2 -3.4-5.0
ECB(Sep 2020)
OECD(Jun 2020)
EC(Jul 2020)
IMF(Jun 2020)
WB(Jun 2020)
The impact of economic lockdown on GDP growth depends, among other things, on theeconomic structure, the number of disease cases and stringency of social distancing thataffect community mobility
30 Source: Eurostat; Google COVID-19 Community Mobility Reports; Oxford COVID-19 Government Response Tracker; https://www.worldometers.info/coronavirus/
4 90 1 52 7 103
-20
6 8-25
-10
-15
-5
0
HR
COVID-19 cases per one thousand inhabitants
BG
MTSI
NL
FI
AT
LT
RO
GDP
gro
wth
in 2
Q20
20(%
yoy
)
BE
CYCZ
DKPL
FR
DEGRHU
IT
EE
PT
SK
ES
SE
GB
LV
-45-15-10 -35
-20
-70
-5
0 -50-40 -65-60-5 -30 -55-25
-25
-15
-20
-10
0
Average mobility in retail and recreation in Q2 2020 (change vs. pre-pandemic period)
MT
HU
DEDKLV
GDP
gro
wth
in Q
2 20
20(%
YO
Y)
ATBE
BG
HRGR
CZ
EE LTFI
RO
FRIT
PL
NL
GB
SKSI
ES
SE
PT
5228 32 6836 4844
-10
-15
40-25
6056
-20
64
-5
0
FRIT GR
Government policy stringency index in Q2 2020 (quarter on quarter change)
BESIHU
ESGDP
gro
wth
in 2
Q20
20(%
yoy
)
AT
BG
HR
CY
ROCZ
PT
DKSE
EEFI
GB
DE
IELV
LT
NL
PL
SK
-15
8% 10%4%2%
-5
6%
-10
-25
-20
0
PT
PL
Share of sectors (in total VA):1) Manufacture of motor vehicles, trailers, semi-trailers and of other transport equipment,
2) Accommodation and food service activities, 3) Arts, entertainment and recreation
EE
HU
DK
FR
GDP
gro
wth
in Q
2 20
20(%
YO
Y)
AT
BE
BG
SIHR
LV
IT
CY
CZ
FI
GR
IE
NL
LT
DESK
ES
SE
GB
RO
y = -1.3678x - 3.4737R² = 0.3519
y = -1.0007x - 9.7556R² = 0.2667
y = -0.3605x + 5.1799R² = 0.1585
y = 0.2818x - 1.9948R² = 0.6668
a b
dc
The impact of economic lockdown on GDP growth depends, among other things, on theeconomic structure, the number of disease cases and stringency of social distancing thataffect community mobility
31 Source: Eurostat; Google COVID-19 Community Mobility Reports; Oxford COVID-19 Government Response Tracker; https://www.worldometers.info/coronavirus/
4 90 1 52 7 103
-20
6 8-25
-10
-15
-5
0
HR
COVID-19 cases per one thousand inhabitants
BG
MTSI
NL
FI
AT
LT
RO
GDP
gro
wth
in Q
2 20
20(%
YO
Y)
BE
CYCZ
DKPL
FR
DEGRHU
IT
EE
PT
SK
ES
SE
GB
LV
-45-15-10 -35
-20
-70
-5
0 -50-40 -65-60-5 -30 -55-25
-25
-15
-20
-10
0
MT
HU
DEDKLV
GDP
gro
wth
in Q
2 20
20(%
YO
Y)
ATBE
BG
HRGR
CZ
EE LTFI
RO
FRIT
PL
NL
GB
SKSI
ES
SE
PT
-15
8% 10%4%2%
-5
6%
-10
-25
-20
0
PT
PL
Share of sectors (in total VA):1) Manufacture of motor vehicles, trailers, semi-trailers and of other transport equipment,
2) Accommodation and food service activities, 3) Arts, entertainment and recreation
EE
HU
DK
FR
GDP
gro
wth
in Q
2 20
20(%
YO
Y)
AT
BE
BG
SIHR
LV
IT
CY
CZ
FI
GR
IE
NL
LT
DESK
ES
SE
GB
RO
y = -1.3678x - 3.4737R² = 0.3519
y = -1.0007x - 9.7556R² = 0.2667
y = 0.2818x - 1.9948R² = 0.6668
► Changes in mobility are more related to the number ofCOVID-19 cases rather than stringency of governmentpolicy
► While another nationwide lockdown is unlikely (not partof the baseline scenario) and survival rates of ICUpatients with COVID-19 have improved, a recentincrease in the number of new COVID-19 cases and alikely second wave are important risk factors
► Vaccine or an effective treatment would be game-changers
Average mobility in retail and recreation Q2 2020 (change vs. pre-pandemic period)
a b
c
The role of sectors most affected by the lockdown varies among Europeancountries
32
Source: Eurostat. *or latest available
0
1
2
3
4
5
6
Mal
ta
Spai
n
Net
herl
ands
Italy
Belg
ium
Ger
man
y
Cypr
us
Slov
akia
Finl
and
Euro
zone
Slov
enia
Lith
uani
a
Aus
tria UK
Fran
cePo
rtug
al
Esto
nia
Latv
iaG
reec
eIr
elan
d
Luxe
mbo
urg
2018
0
1
2
3
4
5
6
7
8
9
Cypr
us
Italy
Latv
ia
Spai
n
Port
ugal
Gre
ece
Aus
tria
Euro
zone UK
Fran
ce
Luxe
mbo
urg
Slov
enia
Esto
nia
Net
herl
ands
Irel
and
Finl
and
Ger
man
y
Lith
uani
a
Belg
ium
Slov
akia
Accommodation and food service activitiesArts, entertainment and recreation
a) Accommodation and food service activities; arts, entertainment andrecreation (% of total value added in 2018*)
b) Manufacture of motor vehicles, trailers, semi-trailers and of othertransport equipment (% of total value added in 2018*)
IFRS 9: an analysis of Q2 impacts and expectations ahead
… which partly explains differences in the performance of Eurozone economies
Contribution of value added generated by groups of sectors to the change in total value added in Q2 2020 (YOY, PP)
33
FranceEurozone Germany Netherlands-22.9
Italy
-14.7
UKSpain
-9.3
-18.9
-11.4
-22.0
-18.0
AgricultureIndustry
Information and communication
Business services
Finance
ConstructionTrade, transport, accommodation and restaurants
Real estate
Public administrationArts, entertainment and recreation
IFRS 9: an analysis of Q2 impacts and expectations ahead
Source: Eurostat
Unprecedented rescue packages adopted by governments significantly reduced the impact of theCOVID-19 pandemic on the labour market, which supported consumption
34
%
a) Annual change in employment and hours worked in Q2 2020
-2
0
2
4
6
8
10
12
Ger
man
yBe
lgiu
mJa
pan
Mal
taN
ethe
rlan
dsA
ustr
iaIta
lyRo
man
iaFr
ance UK
Finl
and
Euro
are
aEU
27Cy
prus
Swed
enPo
land
Hun
gary
Port
ugal
Slov
enia
Czec
hia
USA
Den
mar
kCr
oatia
Bulg
aria
Slov
akia
Gre
ece
Irel
and
Spai
nLi
thua
nia
Esto
nia
Latv
ia
2008 2020* **
b) Change in unemployment rate during the global financial crisis andduring the COVID-19 pandemic*
PP
Source: Eurostat; *maximum level of unemployment rate in 2010 compared to September 2008 level; **latest available data on unemployment rate (May-July 2020) compared to January 2020 level
-35
-30
-25
-20
-15
-10
-5
0
5
10
Gre
ece
Spai
nPo
rtug
alIta
ly UK
Fran
ceSl
ovak
iaEu
ro a
rea
Hun
gary
Cypr
us EUSl
oven
iaEs
toni
aIc
elan
dLi
thua
nia
Mal
taBu
lgar
iaCz
echi
aIr
elan
dLa
tvia
Ger
man
yN
ethe
rlan
dsLu
xem
bour
gD
enm
ark
Aus
tria
Swed
enFi
nlan
dN
orw
ayRo
man
iaPo
land
Croa
tiaemployment hours worked
► Rescue packages, adopted to protect companies and jobs, will result in record high government debts► Government policy will continue to play a critical role: extending vs. withdrawing stimulus measures
IFRS 9: an analysis of Q2 impacts and expectations ahead
Retail trade rebounded quickly (partly an effect of pent-up demand), but industrial production andconstruction remain significantly below the pre-pandemic trend
35 IFRS 9: an analysis of Q2 impacts and expectations ahead
Source: Eurostat
65
70
75
80
85
90
95
100
105
110
20202018 2019
-7%
a) Industrial production in Eurozoneby July 2020 (Feb 2020=100)
2018 2019 2020
-2%
b) Retail trade turnover in Eurozoneby July 2020 (Feb 2020=100)
65
90
70
75
80
95
85
100
105
110
2018 2019 2020
-7%
c) Construction in Eurozoneby June 2020 (Feb 2020=100)
The impact of the COVID-19 pandemic has varied among Eurozone economies
Source: Eurostat
Feb 20 Apr 20 Jun 20
a) Industrial production in selected countriesby July 2020 (Feb 2020=100)
36
Feb 20 Apr 20 Jun 20
b) Retail trade turnover in selected countriesby July 2020 (Feb 2020=100)
100
30
40
50
60
70
80
90
110
Apr 20Feb 20 Jun 20
c) Construction in selected countriesby July 2020 (Feb 2020=100)
Eurozone
FranceGermany
ItalySpainNetherlandsUK
EurozoneGermany
Netherlands
ItalyFrance
Spain
UK
EurozoneGermany
Netherlands
ItalyFrance
UK
Spain
IFRS 9: an analysis of Q2 impacts and expectations ahead
30
40
50
60
70
80
90
100
110
… and also among sectors and retail trade categories: from an economy-wide tosector recession
a) Volume of production in manufacturing inselected sectors, Euro area (Feb 2020=100)
37
b) Retail trade turnover in selected categoriesof stores, Euro area (Feb 2020=100)
c) Turnover in selected services categories,Euro area, (Feb 2020=100)
IFRS 9: an analysis of Q2 impacts and expectations ahead
0
20
40
60
80
100
120
Mar-20 May-20 Jul-20
Pharmaceuticals
Food and beverages
Motor vehiculesFurniture
Wearing apparel
Consumer electronics Food, beverages, tobacco
Textiles, clothing, footwareMedical, cosmetic, toilet articles
Household equipment, cultural and recreation goods
Computers, telecommunications equipment
0
20
40
60
80
100
120
Publishing
Land transport
Accommodation
Programming and broadcasting
Food and beverage services
Travel agencies, tour operators
Source: Eurostat
Feb-20 Apr-20 Jun-20 Mar-20 May-20 Jul-20Feb-20 Apr-20 Jun-20 Mar-20 May-20 Jul-20Feb-20 Apr-20 Jun-20
Many Eurozone countries are relatively strongly exposed to disruptions ininternational value chains
38
Source: EY Global Value Chain Model calculations based on updated World Input-Output data (2018)
155
20
250
30
300
10 20
35
35
40
40 45
25
45
50
55
70
60
65
5
10
15
DEU
HUN
HRV
FINMEX
POL
GRC
Share of foreign value added incorporated in gross output (production) in 2018 (%)
LUX
KOR
Shar
e of
dom
esti
c va
lue
adde
d ab
sorb
ed a
broa
d in
201
8(%
)
MLT
IRL
NLD
AUT
CZE
SVK
BELDNK
LTU
UK
EST
TWN
FRA
BGR
ROU
LVA
SVN
SWE
PRT
CHE
CANTUR
NOR
ROWESP
CYP
IDNAUS
JPNCHN
BRA
RUS
USAIND
ITA
The pace of recovery in many Eurozone economies will be strongly affected by changes in external demand
Eurozoneother countries
USAUK
45-degree line
IFRS 9: an analysis of Q2 impacts and expectations ahead
39
Key considerations underpinning macroeconomic estimates for banks
IFRS 9: an analysis of Q2 impacts and expectations ahead
► Incorporation of actual data into base case and scenarios/simulations► Consideration of:
► Number of new COVID-19 cases and the risk of second wave► Medical developments (effective treatment / vaccine)► Government support programmes: extending / withdrawing stimulus measures► Consumer and corporate responses (community mobility, confidence indicators, high-frequency data)
► Time profile — return to trend, level of trend given structural shifts► Sector and country specific analysis, including the role of external demand and exposure to supply
chain disruptions► Scenario weights, confidence in base case, balance of risks► Comparison to regulator “scenarios”
Summary and challengesahead
Laure Guégan
40 IFRS 9: an analysis of Q2 impacts and expectations ahead
Challenges ahead
41 IFRS 9: an analysis of Q2 impacts and expectations ahead
► Adjust and stabilise the process for estimating ECL in a new paradigm involving an increased level ofuncertainty and judgment► Unprecedented health and economic crisis with unprecedented level of government support measures
► Many unknowns: length and magnitude of the pandemic and the government support measures as well as middle and long-termconsequences on the economy
► Risk models have not been developed to provide reliable estimates under these previously unseen circumstances► Usual model correlations and calibration require some adjustments (macroeconomic models, PD models, loss given default (LGD))► Model recalibration requires observation of the full impacts of the COVID-19 pandemic are observed and will require more time
► The fast-moving environment requires more agility in the IFRS 9 processes to incorporate new information► High quality governance is critical around key areas of judgment (scenarios, model adjustments, overlays)
► Manage the transition from existing, widely available COVID-19 specific payment deferrals to tailoredforbearance arrangements► Prudential Regulation Authority (PRA) statement on COVID-19: IFRS 9 and capital requirements – guidance as COVID-19 specific
payment deferrals come to an end (26 August 2020):“At the end of the existing, widely available COVID-19 specific payment deferrals, if the borrowers involved are not able to resumepayments in full immediately with all deferred sums either paid in full or capitalised, tailored forbearance arrangements (…) should beconsidered. Such tailored forbearance arrangements are likely to be as good an indicator of SICR, credit impairments or defaults asforbearance was prior to the pandemic.”
Challenges ahead - how to assess significant deterioration of risk
Individual risk analysis
Because large-scale moratoria and guaranteed loans tend to“turn off” usual SICR backstops (forbearance, days past due),banks need to:► Consider other indicators to determine whether the
borrower’s difficulties are temporary (e.g., forced leave)versus longer-term (e.g., they have been made redundant)
► Define appropriate monitoring for new guaranteed loans► Assess implementation of governmental initiatives and to
what extent they will limit the defaults► Banks are likely to have less information for retail than
wholesale customers and need to design a holistic approach :► Economic conditions► History of missed payments or adverse credit bureau scores► Current data (employment status, current account activity etc.)► Use of portfolios approaches and application of expert judgment
► For wholesale exposures, ratings tend to react more quicklybased on updated financial information. However, portfolioand sectorial approaches remain critical
► Pooling of loans should consider factors such as:► Products: mortgages vs. retail unsecured loans, etc.
► Type of relief measures granted, initially and subsequently► Whether the borrower is asking for a renewal of a payment
holiday or it is a first request
► Geography (e.g., country-wide support measures)► Industry (travel, hospitality, entertainment and services
industries)
► Behavioral information (historic and current)► Information collected when granting or renewing payment
holidays► Revised macroeconomic assumptions should be incorporated
in the assessment► Overlay approaches may be needed to capture the
uncertainty, limitations of historical data, sector-specific risksand possible government actions
Collective or sectorial approaches
42 IFRS 9: an analysis of Q2 impacts and expectations ahead
Challenges ahead - key areas of IFRS 9 ECL disclosures
43
► How the COVID-19 pandemic hasbeen incorporated in themacroeconomic scenarios
► How new scenarios compare to theprevious ones
► Weights and underlying rationale► Sensitivity analysis and outlook for
the full year
► How government relief measureshave been reflected
► Overlays (articulation with Stage 2transfers and risk parameters)
► If and how the usual IFRS 9 ECLestimate process and governance ofthe bank has been adjusted for thepurpose of the half-year accounts
► Main features of the schemesimplemented by the bank: state-guaranteed loans, public/privatemoratoria (with or without waiver ofinterests)
► Accounting analysis of the schemes :► Initial fair value (FV) of the loans► Effective interest rate calculation► Effect of guarantees (integral or not)► Modification accounting
► Related exposures for the bank► Specific risk monitoring approaches► Expected effect of unwinding of public
and private moratoria► Expected impact on arrears► Expected impact on Stage 2 transfers
► Stage 3 losses (including single names)► COVID-19 pandemic scenarios
► Movements in “days past due” andarticulation with moratoria
► Vulnerable sectors: concentrations,portfolio approaches, overlays
► Stage 2 transfers:► How much and why?► Triggers: delinquency? forbearance? other?► Exceptions applied to usual indicators
► Portfolio approaches► Segmentation► Stage 2 transfers/adjusted risk parameters
► Overlays► Why, how and how much?► Product/business allocation► Movements
Macroeconomic scenarios andassumptions:
COVID-19 pandemic loans andgovernment relief measures: ECL movements and outlook
IFRS 9: an analysis of Q2 impacts and expectations ahead
THANK YOU…..and our next EY webcast in the ‘Accounting and regulatory’ series
44
The next EY webcast in this series will be: “The next and beyond in credit risk and capital management and scenario planning”We will cover:► the impact of the current environment on the future of credit risk and capital management► the link to stress testing and overall scenario planning► the impact on data, modeling, process and control considerations, including auditor expectations for the upcoming year-end► looking beyond to how will integrate across their loan loss, capital, stress testing and climate risk management activities
Invites will be distributed closer to the time of the event.
IFRS 9: an analysis of Q2 impacts and expectations ahead
EY | Assurance | Tax | Strategy & Transactions | Consulting
About EY
EY is a global leader in assurance, tax, strategy, transaction and consultingservices. The insights and quality services we deliver help build trust andconfidence in the capital markets and in economies the world over. Wedevelop outstanding leaders who team to deliver on our promises to all ofour stakeholders. In so doing, we play a critical role in building a betterworking world for our people, for our clients and for our communities.
EY refers to the global organization, and may refer to one or more, of themember firms of Ernst & Young Global Limited, each of which is a separatelegal entity. Ernst & Young Global Limited, a UK company limited byguarantee, does not provide services to clients. Information about how EYcollects and uses personal data and a description of the rights individualshave under data protection legislation are available via ey.com/privacy. Formore information about our organization, please visit ey.com.
© 2020 EYGM Limited.All Rights Reserved.
EYG no. 006383-20Gbl
ED None
This material has been prepared for general informational purposes only and is not intended to berelied upon as accounting, tax, legal or other professional advice. Please refer to your advisors forspecific advice.
ey.com/fscovid-19