v3.26.1
IFRS 9 Impairment
6 Months Ended
Jun. 30, 2026
IFRS 9 Impairment [Abstract]  
IFRS 9 Impairment IFRS 9 Impairment
Model overview
During the first half of 2026, Deutsche Bank continued to apply the same IFRS 9 impairment models and methodologies,
key assumptions and risk management activities as disclosed in the Annual Report 2025, except a change to the
methodology on how the refinancing risk overlay has been quantified. This change represents a change in estimate and
contributed to the overlay releases described in the IFRS 9 Expected Credit Losses section below. As outlined in the
Annual Report 2025, the Group leverages existing models used for the determination of capital demand under the Basel
Internal Ratings Based Approach and internal risk management practices to calculate the bank’s ECL.
The latest developments and key uncertainties in the first half of 2026 and their consideration in the ECL calculation,
based on the bank’s ongoing credit risk management activities and governance framework, are described in the section
‘Key risk themes’ in this report. Activities targeted at assessing the appropriateness of the ECL calculation include regular
emerging risk reviews as well as portfolio deep dives, day-to-day risk management on the level of individual borrowers,
and regular model validations. The Group also considers each reporting period if there are any potential model
imprecisions or uncertainties included in the model that require an overlay. Lastly, the Group presents its sensitivity
analysis regarding forward-looking information as a key assumption.
Forward-looking information
The tables below contain the macroeconomic variables (MEVs) included in the application of forward-looking
information feeding the IFRS 9 model as of June 30, 2026, and as of December 31, 2025. At each reporting date, the
consensus data include the latest macroeconomic developments.
Macroeconomic variables applied
as of June 20261,2
Year 1
(4 quarter avg)
Year 2
(4 quarter avg)
GDP - USA
1.95%
2.11%
GDP - Eurozone
0.44%
1.23%
GDP - Germany
0.73%
1.38%
GDP - Italy
0.54%
0.86%
GDP - Developing Asia
4.96%
4.54%
GDP - Emerging Markets
4.02%
4.10%
Unemployment - USA
4.36%
4.27%
Unemployment - Eurozone
6.30%
6.11%
Unemployment - Germany
3.84%
3.72%
Unemployment - Italy
5.38%
5.54%
Unemployment - Spain
9.96%
9.62%
Unemployment - Japan
2.61%
2.51%
Real Estate Prices - CRE Index USA
314.50
317.33
Real Estate Prices - CRE Index Eurozone
111.91
112.86
Real Estate Prices - House Price Index USA
327.76
336.43
Real Estate Prices - House Price Index Germany
155.11
156.01
Real Estate Prices - House Price Index Spain
2,405.97
2,491.88
Equity - S&P500
7,556
8,004
Equity - Eurostoxx50
6,178
6,408
Equity - DAX40
25,271
26,835
Equity - MSCI EAFE
1,481
1,551
Equity - MSCI Asia
2,526
2,641
Equity - Nikkei
63,543
65,949
Credit - High Yield Index
307.55
366.97
Credit - CDX High Yield
339.97
394.10
Credit - CDX IG
57.98
67.02
Credit - CDX Emerging Markets
171.13
204.21
Credit - ITX Europe 125
60.13
67.26
Commodity - WTI
78.11
69.67
Commodity - Gold
4,692.37
4,677.44
1MEVs as of June 18, 2026, which remained consistent as of June 30, 2026
2Year 1 equals second quarter of 2026 to first quarter of 2027, Year 2 equals second quarter of 2027 to first quarter of 2028
as of December 20251,2
Year 1
(4 quarter avg)
Year 2
(4 quarter avg)
GDP - USA
1.87%
1.97%
GDP - Eurozone
1.16%
1.44%
GDP - Germany
0.65%
1.54%
GDP - Italy
0.60%
0.91%
GDP - Developing Asia
4.45%
4.78%
GDP - Emerging Markets
3.85%
4.19%
Unemployment - USA
4.42%
4.29%
Unemployment - Eurozone
6.30%
6.18%
Unemployment - Germany
3.75%
3.66%
Unemployment - Italy
6.14%
6.22%
Unemployment - Spain
10.37%
10.05%
Unemployment - Japan
2.49%
2.45%
Real Estate Prices - CRE Index USA
300.74
301.87
Real Estate Prices - CRE Index Eurozone
110.44
111.75
Real Estate Prices - House Price Index USA
331.21
340.69
Real Estate Prices - House Price Index Germany
157.28
158.82
Real Estate Prices - House Price Index Spain
2,213.53
2,264.16
Equity - S&P500
6,942
7,366
Equity - Eurostoxx50
5,793
6,086
Equity - DAX40
24,453
25,886
Equity - MSCI EAFE
1,288
1,351
Equity - MSCI Asia
2,068
2,160
Equity - Nikkei
50,891
53,099
Credit - High Yield Index
308.27
348.99
Credit - CDX High Yield
333.97
370.05
Credit - CDX IG
54.64
62.78
Credit - CDX Emerging Markets
149.82
179.86
Credit - ITX Europe 125
56.42
62.27
Commodity - WTI
61.07
59.01
Commodity - Gold
3,976.94
4,189.01
1MEVs as of December 8, 2025, which remained consistent as of December 31, 2025
2Year 1 equals fourth quarter of 2025 to third quarter of 2026, Year 2 equals fourth quarter of 2026 to third quarter of 2027
Overlays applied to the IFRS 9 model output
The Group regularly reviews the IFRS 9 methodology and processes, key inputs into the ECL calculation and discusses
upcoming model changes, potential model imprecisions or other estimation uncertainties, for example in the
macroeconomic environment to determine if any overlays are required. Moreover, regular reviews for evolving or
emerging risks are performed, especially in the current macroeconomic and geopolitical environment. Measures applied
include client surveys and interviews, along with analysis of portfolios across businesses, regions and sectors. In addition,
the Group regularly reviews and validates key model inputs and assumptions (including those in feeder models) and
ensures where expert judgments are applied that these are in line with the Group’s risk management framework.
To ensure that Deutsche Bank’s ECL model accounted for the uncertainties in the macroeconomic environment
throughout the first half of 2026, the Group continued to review emerging risks, assessed potential baseline and
downside impacts and required actions to manage the bank’s credit strategy and risk appetite. As a result of these
reviews the broader macroeconomic overlay for the uncertainties and second order impacts related to the Middle East
conflict recognized as of March 31, 2026 has been partially retained given the overall development of MEVs.
As of June 30, 2026, management overlays amounted to € 158 million, compared to € 156 million at the end of 2025
(which resulted in an increase of Allowance for credit losses in both periods). Overlays as of June 30, 2026 included the
aforementioned overlay for uncertainties and second order impacts related to the Middle East conflict, an overlay for
observations from the bank’s portfolio reviews and credit risk assessments as well as expected impacts from model and
process refinements.
Overall assessment of ECLs
Results from the above reviews and development of key portfolio indicators are regularly discussed at the Credit Risk
Appetite and Management Forum, the Risk and Finance Credit Loss Provisioning Forum and Group Risk Committee.
Where necessary, actions and measures are taken to mitigate the risks. Client ratings are regularly reviewed to reflect the
latest macroeconomic developments and where potentially significant increases in credit risks are identified, clients are
moved to the watchlist (Stage 2), forbearance measures may be negotiated, and credit limits and collateralization are
reviewed. Overall, the Group believes that based on its day-to-day risk management activities and regular reviews of
emerging risks it has adequately provided for its ECL.
Model sensitivity
The Group has identified three key model assumptions included in the IFRS 9 model. These include forward looking
macroeconomic variables, the quantitative criteria for determining if a borrower has incurred a significant increase in
credit risk and transferred to Stage 2, and the LGD setting on homogenous portfolios in Stage 3. The bank also provides
sensitivity analysis on the potential impact if these key assumptions applied in the ECL model were to deviate from the
bank’s base case expectations. The sensitivity of the quantitative criteria for determining if a borrower has incurred a
significant increase in credit risk and transferred to Stage 2 and the sensitivity of LGD settings on portfolios in Stage 3
have not materially changed versus amounts disclosed in the Annual Report 2025.
Macroeconomic Variables
The sensitivity of the ECL model with respect to potential changes in projections for key MEVs is shown in the tables
below, which provides ECL impacts from downward and upward shifts applied separately to each group of MEVs as of
June 30, 2026, and December 31, 2025. The magnitude of the shifts is selected in the range of one standard deviation,
which is a statistical measure of the dispersion of the values of a random variable. Each of these groups consists of MEVs
from the same category:
GDP growth rates: includes U.S., Eurozone, Germany, Italy, Developing Asia, Emerging Markets
Unemployment rates: includes U.S., Eurozone, Germany, Italy, Japan, Spain
Equities: S&P500, Eurostoxx50, DAX40, Nikkei, MSCI Asia, MSCI EAFE
Credit spreads: ITX Europe 125, High Yield Index, CDX IG, CDX High Yield, CDX Emerging Markets
Real Estate: CRE Index USA, CRE Index Eurozone, House Price Index USA, House Price Index Germany, House Price
Index Italy (until 2024 only), House Price Index Spain
Commodities: WTI oil price, Gold price
Although interest rates and inflation are not included in the above set of MEVs as separate risk drivers, their overall
economic impact is reflected by other macroeconomic variables, such as GDP growth rates, unemployment, equities and
credit spreads, since higher rates and inflation typically filter through these forecasts and are thus reflected in the ECL
model and below sensitivity analysis in an implicit way.
In addition, the sensitivity analysis only includes the impact of the aggregated MEV group (i.e., potential correlations
between different MEV groups or the impact of management overlays is not taken into consideration). Interdependencies
and timing of MEV changes following the Middle East conflict can have further impacts on ECL. ECL quantification for
Stage 3 is not included where a model-based process is not followed.
As of June 30, 2026, the sensitivity impact, which does not consider overlays, is lower, compared to December 31, 2025,
mainly due to portfolio changes, lower ECL as well as improvements of base MEV projections which the analyses were
based on.
IFRS 9 – Sensitivities of Forward-Looking Information applied on Stage 1 and Stage 2 – Group Level
Jun 30, 2026
Upward sensitivity
Downward sensitivity
Upward shift
ECL impact
in € m.
Downward shift
ECL impact
in € m.
GDP growth rates
1pp
(71.9)
(1pp)
82.7
Unemployment rates
(0.5pp)
(48.8)
0.5pp
52.4
Real estate prices¹
5%
(31.6)
(5%)
37.2
Equities
10%
(16.7)
(10%)
23.5
Credit spreads
(40%)
(19.0)
40%
21.0
Commodities²
10%
(8.1)
(10%)
8.9
Dec 31, 2025
Upward sensitivity
Downward sensitivity
Upward shift
ECL impact
in € m.
Downward shift
ECL impact
in € m.
GDP growth rates
1pp
(76.3)
(1pp)
87.2
Unemployment rates
(0.5pp)
(49.4)
0.5pp
51.6
Real estate prices¹
5%
(35.3)
(5%)
40.1
Equities
10%
(17.1)
(10%)
23.7
Credit spreads
(40%)
(19.5)
40%
21.6
Commodities²
10%
(6.9)
(10%)
7.4
1 For a more severe stress test relating to the CRE portfolio that also takes into consideration existing and potential exposure in Stage 3, reference is made to the Key Risk
Theme section on Commercial Real Estate above
2Here the sign of the shift applies to oil prices changes. Gold price changes have the opposite sign. 1pp (percentage point), e.g., GDP shifts from 3% to 4%; 1% (percentage
change), e.g., Real estate price shifts from 100 to 101
IFRS 9 Expected Credit Losses
Provision for credit losses was € 460 million in the second quarter of 2026, or 38 basis points of average loans, down by
11% from € 519 million in the first quarter of 2026 and up from € 423 million in the prior year quarter. Provision for credit
losses included an impact of approximately € 120 million from the planned exit of certain non-performing exposures as
part of capital-accretive de-risking actions, partially offset by overlay releases.
Provision for non-performing (Stage 3) loans was € 562 million, up from € 300 million in the prior year quarter, primarily
reflecting additional provisions following the aforementioned targeted asset sales in the Investment Bank, the majority of
which are related to Commercial Real Estate, as well as Private Bank which benefited from a release on model updates in
the prior year period. Provision for performing loans (Stage 1 and 2) was € (102) million compared to € 123 million in the
prior year quarter, primarily due to releases of overlays.
Provisions for credit losses for the first six months of 2026 was € 979 million, up from € 894 million reported in the prior
year period mainly in the Investment Bank relating to Commercial Real Estate and Private Bank.