XML 152 R34.htm IDEA: XBRL DOCUMENT

Expected credit loss measurement

12 Months Ended

Dec. 31, 2020

Disclosure Of Provision Matrix [Line Items]

Disclosure Of Financial Assets Explanatory

Additional information

Note 20 Expected credit loss measurement

a) Expected credit losses in the period

Total net credit loss expenses were USD 694 million in 2020, reflecting net credit loss expenses of USD 266 million related to stage 1 and 2 positions and USD 429 million net credit loss expenses related to credit-impaired (stage 3) positions.

Stage 1 and 2 net credit loss expenses of USD 266 million were primarily driven by a net expense of USD 200 million from updating the forward-looking scenarios and their associated weightings, factoring in updated macroeconomic assumptions to reflect the effects of the COVID-19 pandemic, with approximately half from the baseline scenario and half from the severe downside scenario. The main drivers included updated GDP and unemployment assumptions in Switzerland and the US, primarily impacting Large corporate clients and, to a lesser extent, Private clients with mortgages, Real estate financing and SME clients. These scenario updates impacted remeasurements for stage 1 and 2 positions without stage transfers and triggered exposure movements between stages, primarily from stage 1 to stage 2 as probabilities of default increased.

In addition to the scenario related effects, stage 1 and 2 expenses of USD 73 million arose from new transactions, net of releases from derecognized transactions, primarily from Large corporate clients and SME clients. A further USD 32 million stage 1 and 2 net release of expenses resulted from a number of model updates, primarily impacting Financial intermediaries, Real estate financing and SME clients. The remaining stage 1 and 2 expenses of USD 24 million mainly reflect the effects of post-model adjustments for selected exposures to Swiss SME clients, as well as remeasurements within the loan book, mainly in the Investment Bank.

The changes in the macroeconomic environment in the second half of 2020 generally included more optimistic forward-looking assumptions for both the baseline and severe downside scenarios compared with those applied in the first half of the year. Management applied a post-model expense adjustment of USD 117 million to offset the stage 1 and 2 releases that would have otherwise arisen, deeming them to be premature given the high degree of prevailing uncertainties and the wide range of reasonable possible outcomes.

›Refer to Note 20b for more information

Stage 3 net expenses of USD 429 million were recognized across a number of defaulted positions. In the Investment Bank, stage 3 net expenses of USD 217 million were recognized, of which USD 81 million related to an exposure to a client in the travel sector. In Personal & Corporate Banking, stage 3 net expenses of USD 128 million were recognized, of which USD 59 million related to a case of fraud at a commodity trade finance counterparty, which affected a number of lenders, including UBS. In Global Wealth Management, stage 3 net expenses of USD 40 million were recognized, primarily across a small number of collateralized and securities-based lending positions. In Group Functions, stage 3 expenses of USD 42 million were recognized from one energy-related exposure in the Non-core and Legacy Portfolio.

Credit loss (expense) / release
USD million	Global Wealth Management	Personal & Corporate Banking	AssetManagement	Investment Bank	Group Functions	Total
For the year ended 31.12.20
Stages 1 and 2	(48)	(129)	0	(88)	0	(266)
Stage 3	(40)	(128)	(2)	(217)	(42)	(429)
Total credit loss (expense) / release	(88)	(257)	(2)	(305)	(42)	(694)

For the year ended 31.12.19
Stages 1 and 2	3	23	0	(4)	0	22
Stage 3	(23)	(44)	0	(26)	(7)	(100)
Total credit loss (expense) / release	(20)	(21)	0	(30)	(7)	(78)

For the year ended 31.12.18
Stages 1 and 2	0	0	0	(9)	(1)	(9)
Stage 3	(15)	(56)	0	(29)	(8)	(109)
Total credit loss (expense) / release	(15)	(56)	0	(38)	(8)	(118)

b) Changes to ECL models, scenarios, scenario weights and key inputs

Refer to Note 1a for information about the principles governing ECL models, scenarios, scenario weights and key inputs applied.

During 2020, management carefully considered guidance issued by supervisory authorities concerning the interpretation of key elements of IFRS 9, Financial instruments, in the context of COVID-19.

Governance

Comprehensive cross-functional and cross-divisional governance processes are in place and used to discuss and approve scenario updates and weights, to assess whether significant increases in credit risk resulted in stage transfers, to review model outputs and to reach conclusions regarding post-model adjustments.

Model changes

During 2020, the probability of default (PD) and loss given default (LGD) models applied to Financial intermediaries, Large corporate clients, Real estate financing and SME clients were revised to reflect updates to PD and LGD risk drivers and macroeconomic dependencies.

The model updates resulted in a USD 32 million decrease in ECL allowances, primarily in Personal & Corporate Banking across Financial intermediaries, Real estate financing and SME clients.

Scenario and key input updates

During 2020, the four scenarios and related macroeconomic factors that were applied at the end of 2019 were reviewed in light of the economic and political conditions and prevailing uncertainties through a series of governance meetings, with input from UBS risk and finance experts across the regions and business divisions. Scenario assumptions are benchmarked against external data, e.g., from Bloomberg Consensus, Oxford Economics and the International Monetary Fund World Economic Outlook (IMF WEO). The hypothetical scenarios, in particular the upside and mild downside scenarios, were viewed less plausible. Given the considerable uncertainties associated with the economic conditions, an exceptional interim design of these scenarios was not deemed appropriate. Therefore, management concluded that the probability weights of the upside and the mild downside scenarios would be set to zero.

The baseline scenario, which is aligned to the economic and market assumptions used for UBS’s business planning purposes, and the severe downside scenario, which is the Group’s binding stress scenario, were updated throughout 2020 using the most recent available macroeconomic and market information.

The baseline scenario updates during the first half of 2020 assumed a deterioration of GDP in relevant markets, especially in the US and in Switzerland, increasing unemployment, including a sharp increase in the US to previously unseen levels, lower equity prices and higher market volatility. House prices were assumed to be largely flat in Switzerland over 2020 but to decrease in the US. Overall, only modest economic improvements were expected from the second half of 2020. The severe downside assumptions were considered to be consistent with assumptions for COVID-19-related disruption but to a significantly more adverse degree than what was considered under the baseline scenario, with a full year contraction expected to continue into 2021 and only a moderate recovery starting from the end of 2021.

Improvements in macroeconomic forward-looking assumptions started from the third quarter 2020, with the fourth quarter 2020 in particular including more optimistic assumptions for the baseline, with increased GDP growth forecasts and lower unemployment levels in the US and in Switzerland in particular, given improvements in economic activity as well as greater optimism regarding the availability and effective distribution of vaccines and continued government support. In addition, the assumptions for the severe downside scenario were made less pessimistic in the second half of 2020.

The table on the following page details the key assumptions for the baseline and severe downside scenarios applied as of 31 December 2020. The outlook of the one-year and three-year cumulative GDP growth rates in the baseline are significantly higher than those seen at the end of 2019, as the economy is expected to recover from the sharp contractions seen in mid-2020. However, GDP levels are expected to remain below 31 December 2019 levels until 2022 in the US and Switzerland, and until 2023 in the Eurozone. The GDP growth rates in the severe downside scenario are also higher, to reflect the recovery from the weaker starting levels. Under the baseline scenario, US unemployment is expected to decline to 5.5% by the end of the first year and to 4.5% by the end of the third year. Unemployment rates in the Eurozone and Switzerland are expected to rise modestly in the first year in the baseline scenario but to recover by the end of the third year. The severe downside scenario includes marked increases in unemployment.

Scenario weights and post-model adjustments

As a consequence of the exceptional circumstances and prevailing uncertainties during 2020 and as at 31 December 2020, the weight allocations shifted significantly since 2019, with the baseline scenario weighted at 70% and the severe downside scenario at 30% through the end of the third quarter of 2020, to best reflect management’s sentiment regarding the boundaries of economic outcomes. During the fourth quarter of 2020, changes in the macro-economic environment generally included more optimistic forward-looking assumptions as stated above. However, developments as at 31 December 2020, including an increase in infection and hospitalization rates, as well as strict lockdowns in many jurisdictions, led to a continued high level of uncertainty in relation to the effects of the pandemic and its impact on the global economy. These developments gave rise to questions around whether the assumptions will play out as forecasted. As a consequence, in the fourth quarter 2020, management decreased the weight placed on the baseline scenario from 70% to 60% and increased the weight placed on the severe downside scenario from 30% to 40%, and applied additionally a post-model adjustment of USD 117 million to offset the stage 1 and 2 ECL releases which would have otherwise arisen from the scenario update effects.

ECL scenario

Assigned weights in %

31.12.20

31.12.19

Upside

0.0

7.5

UBS baseline

60.0

42.5

Mild downside

0.0

35.0

Severe downside

40.0

15.0

Scenario assumptions	One year		Three years cumulative
31.12.20	Baseline	Severe downside	Baseline	Severe downside
Real GDP growth (% change)
United States	2.7	(5.9)	9.1	(3.8)
Eurozone	2.5	(8.7)	9.9	(10.3)
Switzerland	3.3	(6.6)	9.0	(5.7)
Consumer price index (% change)
United States	1.7	(1.2)	5.5	0.4
Eurozone	1.4	(1.3)	3.9	(1.7)
Switzerland	0.3	(1.8)	0.9	(1.6)
Unemployment rate (end-of-period level, %)¹
United States	5.5	12.1	4.5	9.9
Eurozone	9.5	14.1	8.0	16.4
Switzerland	3.8	6.1	3.2	6.8
Fixed income: 10-year government bonds (change in yields, basis points)
USD	22.0	(50.0)	46.0	(15.0)
EUR	4.0	(35.0)	21.0	(25.0)
CHF	13.0	(70.0)	31.0	(35.0)
Equity indices (% change)
S&P 500	(2.9)	(50.2)	(1.7)	(40.1)
EuroStoxx 50	3.8	(57.6)	13.5	(50.4)
SPI	(0.8)	(53.6)	5.8	(44.2)
Swiss real estate (% change)
Single-Family Homes	3.4	(17.0)	7.1	(30.0)
Other real estate (% change)
United States (S&P / Case-Shiller)	2.5	(15.3)	9.2	(28.7)
Eurozone (House Price Index)	1.1	(22.9)	7.2	(35.4)
1 2020 unemployment rate is presented as an end-of-period level. 2019 unemployment rate was presented as a change in levels. The 2020 change in level would have been: One year shock in the baseline scenario: United States: -3.5%, Eurozone: 0.4% and Switzerland: 0.4% and for the global crisis scenario: United States: 3.1%, Eurozone: 5.0% and Switzerland: 2.6%. Three year shock in the baseline scenario: United States: -4.5%, Eurozone: -1.2% and Switzerland: -0.2% and for the global crisis scenario: United States: 0.9%, Eurozone: 7.2% and Switzerland: 3.4%

Scenario assumptions	One year		Three years cumulative
31.12.19	Baseline	Severe downside	Baseline	Severe downside
Real GDP growth (% change)
United States	1.9	(6.4)	6.4	(4.3)
Eurozone	1.0	(9.1)	2.8	(10.8)
Switzerland	1.5	(7.0)	4.8	(6.2)
Consumer price index (% change)
United States	1.8	(1.2)	6.2	0.4
Eurozone	1.3	(1.3)	4.3	(1.7)
Switzerland	0.8	(1.8)	2.7	(1.6)
Unemployment rate (change, percentage points)
United States	(0.4)	5.7	(0.5)	5.6
Eurozone	(0.1)	5.6	(0.2)	7.9
Switzerland	0.1	2.6	0.3	3.6
Fixed income: 10-year government bonds (change in yields, basis points)
USD	0.2	(100.0)	10.1	(75.0)
EUR	8.4	(30.0)	28.2	(20.0)
CHF	9.5	(70.0)	30.0	(35.0)
Equity indices (% change)
S&P 500	3.5	(53.0)	9.5	(42.9)
EuroStoxx 50	0.5	(60.0)	4.4	(52.9)
SPI	1.4	(56.2)	5.3	(46.8)
Swiss real estate (% change)
Single-Family Homes	0.1	(15.2)	2.3	(27.0)
Other real estate (% change)
United States (S&P / Case-Shiller)	4.0	(13.3)	16.7	(23.4)
Eurozone (House Price Index)	1.2	(23.0)	2.2	(33.2)

c) Development of ECL allowances and provisions

The ECL allowances and provisions recognized in the period are impacted by a variety of factors, such as:

–origination of new instruments during the period;

–effect of passage of time as the ECLs on an instrument for the remaining lifetime decrease (all other factors remaining the same);

–discount unwind within ECLs as it is measured on a present value basis;

–derecognition of instruments in the period;

–change in individual asset quality of instruments;

–effect of updating forward-looking scenarios and the respective weights;

–movements from a maximum 12-month ECL to the recognition of lifetime ECLs (and vice versa) following transfers between stages 1 and 2;

–movements from stages 1 and 2 to stage 3 (credit-impaired status) when default has become certain and probability of default (PD) increases to 100% (or vice versa);

–changes in models or updates to model parameters; and

–foreign exchange translations for assets denominated in foreign currencies and other movements.

The following table explains the changes in the ECL allowances and provisions for on- and off-balance sheet financial instruments and other credit lines in scope of ECL requirements between the beginning and the end of the period due to the factors listed on the previous page.

Development of ECL allowances and provisions
USD million	Total	Stage 1	Stage 2	Stage 3
Balance as of 31 December 2019	(1,029)	(181)	(160)	(688)
Net movement from new and derecognized transactions¹	(28)	(90)	17	46
of which: Private clients with mortgages	(2)	(3)	2	0
of which: Real estate financing	(3)	(5)	2	0
of which: Large corporate clients	(32)	(29)	(4)	0
of which: SME clients	(16)	(14)	(3)	0
of which: Other	26	(39)	20	46
of which: Securities financing transactions REIT	32	(1)	15	17
of which: Loans to financial advisors	9	(1)	9	0
of which: Lombard loans	23	(6)	0	29
of which Financial intermediaries	(20)	(15)	(5)	0
Remeasurements with stage transfers²	(427)	45	(134)	(338)
of which: Private clients with mortgages	(19)	(2)	(17)	0
of which: Real estate financing	(6)	3	(9)	0
of which: Large corporate clients	(224)	34	(83)	(175)
of which: SME clients	(43)	(1)	(11)	(31)
of which: Other	(134)	11	(14)	(131)
of which: Securities financing transactions REIT	(36)	0	(18)	(19)
of which: Loans to financial advisors	(12)	7	(7)	(11)
of which: Lombard loans	(36)	0	0	(36)
of which Commodity Trade Finance	(59)	0	0	(59)
Remeasurements without stage transfers³	(271)	(88)	(47)	(136)
of which: Private clients with mortgages	(34)	(19)	(8)	(7)
of which: Real estate financing	(14)	(4)	(11)	1
of which: Large corporate clients	(149)	(53)	(17)	(79)
of which: SME clients	(13)	0	(7)	(6)
of which: Other	(60)	(11)	(4)	(44)
of which: Loans to financial advisors	(18)	(12)	(3)	(3)
of which: Lombard loans	(3)	6	0	(9)
of which: Credit cards	(12)	0	0	(12)
Model changes⁴	32	21	11	0
Total ECL allowance movements with profit or loss impact⁵	(694)	(112)	(154)	(429)
Write-offs, FX and other movements (without profit or loss impact)⁶	254	(14)	(19)	287
Balance as of 31 December 2020	(1,468)	(306)	(333)	(829)
1 Represents the increase and decrease in allowances and provisions resulting from financial instruments (including guarantees and facilities) that were newly originated, purchased or renewed and from the final derecognition of loans or facilities on their maturity date or earlier. 2 Represents the remeasurement between 12-month and lifetime ECL due to stage transfers. 3 Represents the change in allowances and provisions related to changes in model inputs or assumptions, including changes in forward-looking macroeconomic conditions, changes in the exposure profile, PD and LGD changes, and unwinding of the time value. 4 Represents the change in the allowances and provisions related to changes in models and methodologies. 5 Includes ECL movements from new and derecognized transactions, remeasurement changes, model and methodology changes. 6 Represents the decrease in allowances and provisions resulting from write-offs of the ECL allowance against the gross carrying amount when all or part of a financial asset is deemed uncollectible or forgiven and movements in foreign exchange rates.

In 2020, ECL allowances and provisions increased by USD 694 million from net credit loss expenses impacting profit or loss:

–a USD 28 million net increase from new and derecognized transactions that resulted from a USD 90 million stage 1 increase primarily in Large corporate clients and SME clients, offset by a USD 63 million net release from stage 2 and 3 transactions, driven by transactions that were terminated before their contractual maturity, mainly in Lombard lending and Securities financing transactions Real estate investment trusts (SFT-REITs);

–a USD 697 million net increase from book quality movements that resulted from a USD 427 million net increase from transactions moving from stages 1 and 2 into stages 2 and 3, respectively, of which approximately half related to Large corporate clients, with further substantial effects from Commodity trade finance, SME clients, SFT REITs and Lombard loans, and USD 271 million from remeasurements without stage transfers, approximately half relating to Large corporate clients, and another significant portion relating to real estate related lending, primarily due to the updates of macroeconomic factors;

–a USD 32 million net decrease that resulted from a number of model revisions, primarily impacting Financial intermediaries, Real estate financing and SME clients, from updates to the PD and LGD risk drivers and macroeconomic dependencies.

In addition to the movements impacting profit or loss, allowances decreased by USD 346 million as a result of a number of write offs. A further USD 75 million allowance increase resulted from foreign exchange movements, almost entirely due to the Swiss franc strengthening against the US dollar.

The following table explains the changes in the ECL allowances and provisions for on- and off-balance sheet financial instruments and other credit lines in scope of ECL requirements between the beginning and the end of the prior period due to the factors listed earlier in this note.


USD million	Total	Stage 1	Stage 2	Stage 3
Balance as of 31 December 2018	(1,054)	(176)	(183)	(695)
Net movement from new and derecognized transactions¹	(53)	(66)	10	3
of which: Private clients with mortgages	(1)	(4)	3	0
of which: Real estate financing	(3)	(5)	2	0
of which: Large corporate clients	(6)	(14)	8	0
of which: SME clients	(16)	(14)	(2)	0
Remeasurements with stage transfers²	(125)	14	(35)	(105)
of which: Private clients with mortgages	(5)	1	(5)	(1)
of which: Real estate financing	5	4	1	0
of which: Large corporate clients	(45)	4	(11)	(38)
of which: SME clients	(64)	2	(11)	(55)
Remeasurements without stage transfers³	73	31	41	1
of which: Private clients with mortgages	22	2	30	(9)
of which: Real estate financing	1	0	0	1
of which: Large corporate clients	(24)	(10)	0	(14)
of which: SME clients	35	9	10	17
Model changes⁴	26	17	9	0
Total ECL allowance movements with profit or loss impact⁵	(78)	(4)	25	(100)
Write-offs, FX and other movements (without profit or loss impact)⁶	105	(1)	(2)	108
Balance as of 31 December 2019	(1,029)	(181)	(160)	(688)
1 Represents the increase and decrease in allowances and provisions resulting from financial instruments (including guarantees and facilities) that were newly originated, purchased or renewed and from the final derecognition of loans or facilities on their maturity date or earlier. 2 Represents the remeasurement between 12-month and lifetime ECL due to stage transfers. 3 Represents the change in allowances and provisions related to changes in model inputs or assumptions, including changes in forward-looking macroeconomic conditions, changes in the exposure profile, PD and LGD changes, and unwinding of the time value. 4 Represents the change in the allowances and provisions related to changes in models and methodologies. 5 To align to the table format for the 2020 ECL allowance and provision movement, UBS has adjusted the 2019 table format. Includes ECL movements from new and derecognized transactions, remeasurement changes, model and methodology changes. 6 Represents the decrease in allowances and provisions resulting from write-offs of the ECL allowance against the gross carrying amount when all or part of a financial asset is deemed uncollectible or forgiven and movements in foreign exchange rates.

As explained in Note 1a, the assessment of an SICR considers a number of qualitative and quantitative factors to determine whether a stage transfer between stage 1 and stage 2 is required. The primary assessment considers changes in probability of default (PD) based on rating analyses and economic outlook. Additionally, UBS considers counterparties that have moved to a credit watch list and those with payments that are at least 30 days past due.

Stage 2 classification by trigger	ECL allowances / provisions as of 31 December 2020
USD million	Stage 2	of which: PD layer	of which: watch list	of which: ≥30 days past due
On-and off-balance sheet	(333)	(252)	(41)	(40)
of which: Private clients with mortgages	(93)	(83)	0	(11)
of which: Real estate financing	(53)	(45)	(2)	(6)
of which: Large corporate clients	(110)	(89)	(20)	0
of which: SME clients	(38)	(16)	(16)	(5)
of which: Financial intermediaries and hedge funds	(19)	(19)	0	0
of which: Loans to financial advisors	(5)	0	(1)	(4)
of which: Credit cards	(14)	0	0	(14)
of which: Other	(2)	0	(2)	0

d) Maximum exposure to credit risk

The tables below and on the following page provide the Group’s maximum exposure to credit risk for financial instruments subject to ECL requirements and the respective collateral and other credit enhancements mitigating credit risk for these classes of financial instruments.

The maximum exposure to credit risk includes the carrying amounts of financial instruments recognized on the balance sheet subject to credit risk and the notional amounts for off-balance sheet arrangements. Where information is available, collateral is presented at fair value. For other collateral, such as real estate, a reasonable alternative value is used. Credit enhancements, such as credit derivative contracts and guarantees, are included at their notional amounts. Both are capped at the maximum exposure to credit risk for which they serve as security. The “Risk management and control” section of this report describes management’s view of credit risk and the related exposures, which can differ in certain respects from the requirements of IFRS.

Maximum exposure to credit risk
	31.12.20
		Collateral¹				Credit enhancements¹			Exposure to credit risk after collateral and credit enhancements
USD billion	Maximum exposure to credit risk	Cash collateral received	Collateralized by securities	Secured by real estate	Other collateral²	Netting	Credit derivative contracts	Guarantees
Financial assets measured at amortized cost on the balance sheet
Cash and balances at central banks	158.2								158.2
Loans and advances to banks³	15.4		0.1						15.3
Receivables from securities financing transactions	74.2	0.0	67.1		7.0				0.0
Cash collateral receivables on derivative instruments^4,5	32.7					21.1			11.6
Loans and advances to customers⁶	379.5	25.8	118.2	194.6	21.7			4.4	14.8
Other financial assets measured at amortized cost	27.2	0.1	0.2		1.3				25.5
Total financial assets measured at amortized cost	687.3	26.0	185.7	194.6	30.1	21.1	0.0	4.4	225.5
Financial assets measured at fair value through other comprehensive income – debt	8.3								8.3
Total maximum exposure to credit risk reflected on the balance sheet in scope of ECL	695.6	26.0	185.7	194.6	30.1	21.1	0.0	4.4	233.7
Guarantees⁷	17.0	0.7	5.0	0.2	1.7			2.5	7.0
Loan commitments⁷	41.2	0.0	4.2	2.1	6.8		0.4	2.4	25.3
Forward starting transactions, reverse repurchaseand securities borrowing agreements	3.2		3.2						0.0
Committed unconditionally revocable credit lines	40.1	0.1	10.3	6.2	2.7			0.0	20.7
Total maximum exposure to credit risk not reflected on the balance sheet, in scope of ECL	101.6	0.8	22.7	8.5	11.2	0.0	0.4	4.9	53.0
	31.12.19
		Collateral¹				Credit enhancements¹			Exposure to credit risk after collateral and credit enhancements
USD billion	Maximum exposure to credit risk	Cash collateral received	Collateralized by securities	Secured by real estate	Other collateral²	Netting	Credit derivative contracts	Guarantees
Financial assets measured at amortized cost on the balance sheet
Cash and balances at central banks	107.1								107.1
Loans and advances to banks³	12.4		0.0						12.4
Receivables from securities financing transactions	84.2		77.6		5.8				0.8
Cash collateral receivables on derivative instruments^4,5	23.3					14.4			8.9
Loans and advances to customers⁶	326.8	18.4	101.4	174.7	17.1			1.1	14.0
Other financial assets measured at amortized cost	23.0	0.1	0.4	0.0	1.3				21.1
Total financial assets measured at amortized cost	576.8	18.6	179.4	174.7	24.3	14.4	0.0	1.1	164.4
Financial assets measured at fair value through other comprehensive income – debt	6.3								6.3
Total maximum exposure to credit risk reflected on the balance sheet in scope of ECL	583.2	18.6	179.4	174.7	24.3	14.4	0.0	1.1	170.7
Guarantees⁷	18.1	1.0	3.0	0.1	1.7			2.5	9.8
Loan commitments⁷	27.5	0.2	1.9	1.3	5.8		0.2	0.2	18.0
Forward starting transactions, reverse repurchaseand securities borrowing agreements	1.7		1.7						0.0
Committed unconditionally revocable credit lines	35.1	0.3	8.3	4.9	3.6			0.0	17.9
Total maximum exposure to credit risk not reflected on the balance sheet, in scope of ECL	82.3	1.5	14.9	6.3	11.0	0.0	0.2	2.8	45.7
1 Of which: USD 1,983 million for 31 December 2020 (31 December 2019: USD 1,720 million) relates to total credit-impaired financial assets measured at amortized cost and USD 154 million for 31 December 2020 (31 December 2019: USD 27 million) to total off-balance sheet financial instruments and other credit lines for credit-impaired positions. 2 Includes but is not limited to life insurance contracts, inventory, mortgage loans, gold and other commodities. 3 Loans and advances to banks include amounts held with third-party banks on behalf of clients. The credit risk associated with these balances may be borne by those clients. 4 Included within Cash collateral receivables on derivative instruments are margin balances due from exchanges or clearing houses. Some of these margin balances reflect amounts transferred on behalf of clients who retain the associated credit risk. 5 The amount shown in the “Netting” column represents the netting potential not recognized on the balance sheet. Refer to Note 22 for more information. 6 Collateral arrangements generally incorporate a range of collateral, including cash, securities, property and other collateral. 7 The amount shown in the “Guarantees” column includes sub-participations.

e) Financial assets subject to credit risk by rating category

The table below shows the credit quality and the maximum exposure to credit risk based on the Group’s internal credit rating system and year-end stage classification. Under IFRS 9, the credit risk rating reflects the Group’s assessment of the probability of default of individual counterparties, prior to substitutions. The amounts presented are gross of impairment allowances.

›Refer to the “Risk management and control” section of this report for more details regarding the Group’s internal grading system

Financial assets subject to credit risk by rating category
USD million	31.12.20
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired (defaulted)	Total gross carrying amount	ECL allowances	Net carrying amount (maximum exposure to credit risk)
Financial assets measured at amortized cost
Cash and balances at central banks	156,250	1,981	0	0	0	0	158,231	0	158,231
of which: stage 1	156,250	1,981	0	0	0	0	158,231	0	158,231
Loans and advances to banks	543	12,129	1,344	1,182	260	1	15,460	(16)	15,444
of which: stage 1	543	12,074	1,277	1,145	231	0	15,269	(9)	15,260
of which: stage 2	0	55	67	37	29	0	189	(5)	184
of which: stage 3	0	0	0	0	0	1	1	(1)	0
Receivables from securities financing transactions	22,998	16,009	15,367	17,995	1,842	0	74,212	(2)	74,210
of which: stage 1	22,998	16,009	15,367	17,995	1,842	0	74,212	(2)	74,210
Cash collateral receivables on derivative instruments	8,196	13,477	7,733	3,243	88	0	32,737	0	32,737
of which: stage 1	8,196	13,477	7,733	3,243	88	0	32,737	0	32,737
Loans and advances to customers	5,813	214,307	67,270	69,217	21,038	2,943	380,589	(1,060)	379,528
of which: stage 1	5,813	212,970	63,000	59,447	15,860	0	357,090	(142)	356,948
of which: stage 2	0	1,338	4,269	9,770	5,178	0	20,556	(215)	20,341
of which: stage 3	0	0	0	0	0	2,943	2,943	(703)	2,240
Other financial assets measured at amortized cost	15,404	4,018	280	6,585	481	560	27,327	(133)	27,194
of which: stage 1	15,404	4,015	269	6,334	389	0	26,410	(34)	26,377
of which: stage 2	0	3	11	251	91	0	357	(9)	348
of which: stage 3	0	0	0	0	0	560	560	(90)	469
Total financial assets measured at amortized cost	209,204	261,922	91,993	98,223	23,709	3,505	688,556	(1,211)	687,345
On-balance sheet financial instruments
Financial assets measured at FVOCI – debt instruments	3,212	5,014	0	32	0	0	8,258	0	8,258
Total on-balance sheet financial instruments	212,417	266,936	91,993	98,255	23,709	3,505	696,815	(1,211)	695,603

Off-balance sheet positions subject to expected credit loss by rating category
USD million	31.12.20
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired(defaulted)	Total off - balance sheet exposure(maximum exposure to credit risk)	ECL provisions
Off-balance sheet financial instruments
Guarantees	3,482	4,623	3,522	4,293	991	170	17,081	(63)
of which: stage 1	3,482	4,219	2,688	3,558	739	0	14,687	(14)
of which: stage 2	0	404	834	736	252	0	2,225	(15)
of which: stage 3	0	0	0	0	0	170	170	(34)
Irrevocable loan commitments	3,018	14,516	8,583	9,302	5,850	104	41,372	(142)
of which: stage 1	3,018	13,589	6,873	8,739	4,676	0	36,894	(74)
of which: stage 2	0	927	1,711	563	1,174	0	4,374	(68)
of which: stage 3	0	0	0	0	0	104	104	0
Forward starting reverse repurchase and securities borrowing agreements	82	150	0	3,015	0	0	3,247	0
Total off-balance sheet financial instruments	6,583	19,289	12,105	16,610	6,840	273	61,700	(205)
Other credit lines
Committed unconditionally revocable credit lines	574	13,505	5,958	8,488	11,501	108	40,134	(50)
of which: stage 1	574	12,940	4,517	6,609	10,593	0	35,233	(29)
of which: stage 2	0	565	1,441	1,879	908	0	4,792	(21)
of which: stage 3	0	0	0	0	0	108	108	0
Irrevocable committed prolongation of existing loans	14	1,349	931	632	357	0	3,282	(2)
of which: stage 1	14	1,349	930	630	355	0	3,277	(2)
of which: stage 2	0	1	1	2	1	0	5	0
of which: stage 3	0	0	0	0	0	0	0	0
Total other credit lines	588	14,854	6,889	9,119	11,858	109	43,416	(52)
1 Refer to the “Internal UBS rating scale and mapping of external ratings” table in the “Risk management and control” section of this report for more information on rating categories.

Financial assets subject to credit risk by rating category
USD million	31.12.19
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired (defaulted)	Total gross carrying amount	ECL allowances	Net carrying amount (maximum exposure to credit risk)
Financial assets measured at amortized cost
Cash and balances at central banks	105,195	1,873	0	0	0	0	107,068	0	107,068
of which: stage 1	105,195	1,873	0	0	0	0	107,068	0	107,068
Loans and advances to banks	309	9,832	1,326	687	298	1	12,454	(6)	12,447
of which: stage 1	309	9,832	1,326	677	228	0	12,371	(4)	12,367
of which: stage 2	0	0	0	10	71	0	81	(1)	80
of which: stage 3	0	0	0	0	0	1	1	(1)	0
Receivables from securities financing transactions	21,089	16,889	14,366	28,815	3,088	0	84,246	(2)	84,245
of which: stage 1	21,089	16,889	14,366	28,815	3,088	0	84,246	(2)	84,245
Cash collateral receivables on derivative instruments	4,899	10,553	5,033	2,765	39	0	23,289	0	23,289
of which: stage 1	4,899	10,553	5,033	2,765	39	0	23,289	0	23,289
Loans and advances to customers	1,744	174,982	59,240	70,528	18,748	2,308	327,550	(764)	326,786
of which: stage 1	1,744	174,328	56,957	62,435	14,117	0	309,581	(82)	309,499
of which: stage 2	0	655	2,283	8,093	4,631	0	15,661	(123)	15,538
of which: stage 3	0	0	0	0	0	2,308	2,308	(559)	1,749
Other financial assets measured at amortized cost	13,031	1,560	390	7,158	312	672	23,123	(143)	22,980
of which: stage 1	13,031	1,549	381	6,747	280	0	21,988	(35)	21,953
of which: stage 2	0	11	9	412	32	0	463	(13)	451
of which: stage 3	0	0	0	0	0	672	672	(95)	576
Total financial assets measured at amortized cost	146,267	215,690	80,354	109,952	22,485	2,981	577,730	(915)	576,815
On-balance sheet financial instruments
Financial assets measured at FVOCI – debt instruments	5,854	450	0	41	0	0	6,345	0	6,345
Total on-balance sheet financial instruments	152,120	216,139	80,354	109,994	22,485	2,981	584,075	(915)	583,159

Off-balance sheet positions subject to expected credit loss by rating category
USD million	31.12.19
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired(defaulted)	Total off - balance sheet exposure(maximum exposure to credit risk)	ECL provisions
Off-balance sheet financial instruments
Guarantees	857	4,932	6,060	5,450	761	82	18,142	(42)
of which: stage 1	857	4,931	6,048	5,218	704	0	17,757	(8)
of which: stage 2	0	1	12	233	57	0	304	(1)
of which: stage 3	0	0	0	0	0	82	82	(33)
Irrevocable loan commitments	2,548	10,068	4,862	5,859	4,160	50	27,547	(35)
of which: stage 1	2,548	10,068	4,862	5,722	3,878	0	27,078	(30)
of which: stage 2	0	0	0	137	282	0	419	(5)
of which: stage 3	0	0	0	0	0	50	50	0
Forward starting reverse repurchase and securities borrowing agreements	0	672	50	936	0	0	1,657	0
Total off-balance sheet financial instruments	3,405	15,672	10,972	12,245	4,922	132	47,347	(77)
Other credit lines
Committed unconditionally revocable credit lines	632	12,459	6,231	7,169	8,554	46	35,092	(34)
of which: stage 1	628	12,422	6,120	6,789	7,889	0	33,848	(17)
of which: stage 2	4	37	111	380	665	0	1,197	(17)
of which: stage 3	0	0	0	0	0	46	46	0
Irrevocable committed prolongation of existing loans	25	1,399	870	633	359	4	3,289	(3)
of which: stage 1	25	1,399	870	633	359	0	3,285	(3)
of which: stage 2	0	0	0	0	0	0	0	0
of which: stage 3	0	0	0	0	0	4	4	0
Total other credit lines	657	13,858	7,101	7,801	8,913	50	38,381	(37)
1 Refer to the “Internal UBS rating scale and mapping of external ratings” table in the “Risk management and control” section of this report for more information on rating categories.

f) Sensitivity information

As outlined in Note 1a, ECL estimates involve significant uncertainties at the time they are made.

ECL model

The models applied to determine point-in-time PDs and LGDs rely on market and statistical data, which has been found to correlate well with historically observed defaults in sufficiently homogeneous segments. The risk sensitivities for each of the IFRS 9 ECL reporting segments to such factors are summarized in Note 9.

Forward-looking scenarios

Depending on the scenario selection and related macro-economic assumptions for the risk factors, the components of the relevant weighted average ECL change. This is particularly relevant for interest rates, which can move in both directions under a given growth assumption (for example, low growth with high interest rates in a stagflation scenario, versus low growth and falling interest rates in a recession). Management generally look for scenario narratives that reflect the key risk drivers of a given credit portfolio.

As forecasting models are complex, due to the combination of multiple factors, simple what-if analyses involving a change of individual parameters do not necessarily provide realistic information on the exposure of segments to changes in the macroeconomy. Portfolio-specific analyses based on their key risk factors would also not be meaningful, as potential compensatory effects in other segments would be ignored. The table below indicates some sensitivities to ECLs if a key macroeconomic variable for the forecasting period is amended across all scenarios with all other factors remaining unchanged.

Potential effect on stage 1 and stage 2 positions from changing key parameters as at 31 December 2020


USD million	Baseline	Severe downside	Weighted average
Change in key parameters
Fixed income: 10-year government bonds (absolute change)
–0.5%	(1.36)	(1.84)	(1.93)
+0.5%	2.10	3.19	3.23
+1.00%	5.69	6.86	7.19
Unemployment rate (absolute change)
–1.00%	(7.40)	(63.01)	(27.83)
–0.5%	(3.78)	(33.54)	(15.67)
+0.5%	4.15	36.97	16.99
+1.00%	8.50	75.93	33.74
Real GDP growth (relative change)
-2.00%	3.72	16.14	9.10
-1.00%	1.86	9.84	5.09
+1.00%	(1.46)	(3.30)	(2.36)
+2.00%	(2.97)	(9.44)	(5.93)
House Price Index (relative change)
–5.00%	8.04	144.34	51.46
–2.50%	3.45	65.80	23.28
+2.50%	(2.79)	(56.60)	(19.09)
+5.00%	(5.16)	(105.61)	(35.29)
Equity (S&P500, EuroStoxx, SMI) (relative change)
–10.00%	3.94	9.66	6.78
–5.00%	1.91	4.29	3.34
+5.00%	(8.30)	(4.23)	(7.27)
+10.00%	(10.14)	(8.58)	(10.22)

Sensitivities can be more meaningfully assessed in the context of coherent scenarios with consistently developed macroeconomic factors. The table on the previous page outlines favorable and unfavorable effects, based on reasonably possible alternative changes to the economic conditions for stage 1 and stage 2 positions. The ECL impact is calculated for material portfolios and disclosed for each scenario.

The forecasting horizon is limited to three years, with a model-based mean reversion of PD and LGD assumed thereafter. Changes to these timelines may have an effect on ECLs: depending on the cycle, a longer or shorter forecasting horizon will lead to different annualized lifetime PD and average LGD estimations. This is currently not deemed to be material for UBS, as a large proportion of loans, including mortgages in Switzerland, have maturities that are within the forecasting horizon.

Scenario weights

ECL is sensitive to changing scenario weights, in particular if narratives and parameters are selected that are not close to the baseline scenario, highlighting the non-linearity of credit losses.

As shown in the table on the bottom of this page, the ECL for stage 1 and stage 2 positions would have been USD 442 million (31 December 2019: USD 234 million) instead of USD 639 million (31 December 2019: USD 341 million) if ECL had been determined solely on the baseline scenario. The weighted average ECL therefore amounts to 145% (31 December 2019: 149%) of the baseline value.

Stage allocation and SICR

The determination of what constitutes a significant increase in credit risk (SICR) is based on management judgment as explained in Note 1a. Changing the SICR trigger will have a direct effect on ECLs, as more or fewer positions would be subject to lifetime ECLs under any scenario.

The relevance of the SICR trigger on overall ECL is demonstrated in the table below with the indication that the ECL allowances and provisions for stage 1 and stage 2 positions would have been USD 1,336 million if all non-impaired positions across the portfolio had been measured for lifetime ECLs irrespective of their actual SICR status. This amount compares to actual stage 1 and 2 allowances and provisions of USD 639 million as of 31 December 2020.

Maturity profile

The maturity profile of the assets is an important driver for changes in ECL due to transfers to stage 2 and from stage 2 to stage 1. The current maturity profile of most lending books is relatively short; hence a movement to stage 2 may have a limited effect on ECLs. A significant portion of our lending to SMEs is documented under multi-purpose credit agreements, which allow for various forms of utilization but are unconditionally cancelable by UBS at any time. The relevant maturity for drawings under such agreements with a fixed maturity is the respective term, or a maximum of 12 months in stage 1. For unused credit lines and all drawings that have no fixed maturity (e.g., current accounts), UBS generally applies a 12-month maturity from the reporting date, given the credit review policies, which require either continuous monitoring of key indicators and behavioral patterns for smaller positions or an annual formal review for any other limit. The ECLs for these products is sensitive to shortening or extending the maturity assumption.

Potential effect on stage 1 and stage 2 positions from changing scenario weights or moving to an ECL lifetime calculation as at 31 December 2020

	Actual ECL allowances and provisions (as per Note 9)		Pro forma ECL allowances and provisions, assuming application of 100% weighting				Pro forma ECL allowances and provisions, assuming all positions being subject to lifetime ECL
Scenarios	Weighted average		Baseline		Severe downside		Weighted average
USD million, except where indicated	ECL	in % of baseline	ECL	in % of baseline	ECL	in % of baseline	ECL	in % of baseline
Segmentation
Private clients with mortgages	(131)	244	(54)	100	(302)	562	(385)	717
Real estate financing	(76)	138	(55)	100	(123)	224	(131)	237
Large corporate clients	(206)	149	(138)	100	(298)	216	(307)	222
SME clients	(74)	115	(64)	100	(93)	144	(129)	200
Other segments	(152)	116	(131)	100	(183)	140	(385)	294
Total	(639)	145	(442)	100	(999)	226	(1,336)	302

UBS AG

Disclosure Of Provision Matrix [Line Items]

Disclosure Of Financial Assets Explanatory

Additional information

Note 20 Expected credit loss measurement

a) Expected credit losses in the period

Total net credit loss expenses were USD 695 million in 2020, reflecting net credit loss expenses of USD 266 million related to stage 1 and 2 positions and USD 429 million net credit loss expenses related to credit-impaired (stage 3) positions.

›Refer to Note 20b for more information

Stage 3 net expenses of USD 429 million were recognized across a number of defaulted positions. In the Investment Bank, stage 3 net expenses of USD 217 million were recognized, of which USD 81 million related to an exposure to a client in the travel sector. In Personal & Corporate Banking, stage 3 net expenses of USD 128 million were recognized, of which USD 59 million related to a case of fraud at a commodity trade finance counterparty, which affected a number of lenders, including UBS AG. In Global Wealth Management, stage 3 net expenses of USD 40 million were recognized, primarily across a small number of collateralized and securities-based lending positions. In Group Functions, stage 3 expenses of USD 42 million were recognized from one energy-related exposure in the Non-core and Legacy Portfolio.

Credit loss (expense) / release
USD million	Global Wealth Management	Personal & Corporate Banking	AssetManagement	Investment Bank	Group Functions	Total
For the year ended 31.12.20
Stages 1 and 2	(48)	(129)	0	(88)	0	(266)
Stage 3	(40)	(128)	(2)	(217)	(42)	(429)
Total credit loss (expense) / release	(88)	(257)	(2)	(305)	(42)	(695)

For the year ended 31.12.19
Stages 1 and 2	3	23	0	(4)	0	22
Stage 3	(23)	(44)	0	(26)	(7)	(100)
Total credit loss (expense) / release	(20)	(21)	0	(30)	(7)	(78)

For the year ended 31.12.18
Stages 1 and 2	0	0	0	(9)	0	(9)
Stage 3	(15)	(56)	0	(29)	(8)	(109)
Total credit loss (expense) / release	(15)	(56)	0	(38)	(8)	(117)

b) Changes to ECL models, scenarios, scenario weights and key inputs

Refer to Note 1a for information about the principles governing ECL models, scenarios, scenario weights and key inputs applied.

During 2020, management carefully considered guidance issued by supervisory authorities concerning the interpretation of key elements of IFRS 9, Financial instruments, in the context of COVID-19.

Governance

Model changes

The model updates resulted in a USD 32 million decrease in ECL allowances, primarily in Personal & Corporate Banking across Financial intermediaries, Real estate financing and SME clients.

Scenario and key input updates

During 2020, the four scenarios and related macroeconomic factors that were applied at the end of 2019 were reviewed in light of the economic and political conditions and prevailing uncertainties through a series of governance meetings, with input from UBS AG risk and finance experts across the regions and business divisions. Scenario assumptions are benchmarked against external data, e.g., from Bloomberg Consensus, Oxford Economics and the International Monetary Fund World Economic Outlook (IMF WEO). The hypothetical scenarios, in particular the upside and mild downside scenarios, were viewed less plausible. Given the considerable uncertainties associated with the economic conditions, an exceptional interim design of these scenarios was not deemed appropriate. Therefore, management concluded that the probability weights of the upside and the mild downside scenarios would be set to zero.

The baseline scenario, which is aligned to the economic and market assumptions used for UBS AG’s business planning purposes, and the severe downside scenario, which is the Group’s binding stress scenario, were updated throughout 2020 using the most recent available macroeconomic and market information.

Scenario weights and post-model adjustments

ECL scenario

Assigned weights in %

31.12.20

31.12.19

Upside

0.0

7.5

UBS AG baseline

60.0

42.5

Mild downside

0.0

35.0

Severe downside

40.0

15.0

Scenario assumptions	One year		Three years cumulative
31.12.20	Baseline	Severe downside	Baseline	Severe downside
Real GDP growth (% change)
United States	2.7	(5.9)	9.1	(3.8)
Eurozone	2.5	(8.7)	9.9	(10.3)
Switzerland	3.3	(6.6)	9.0	(5.7)
Consumer price index (% change)
United States	1.7	(1.2)	5.5	0.4
Eurozone	1.4	(1.3)	3.9	(1.7)
Switzerland	0.3	(1.8)	0.9	(1.6)
Unemployment rate (end-of-period level, %)¹
United States	5.5	12.1	4.5	9.9
Eurozone	9.5	14.1	8.0	16.4
Switzerland	3.8	6.1	3.2	6.8
Fixed income: 10-year government bonds (change in yields, basis points)
USD	22.0	(50.0)	46.0	(15.0)
EUR	4.0	(35.0)	21.0	(25.0)
CHF	13.0	(70.0)	31.0	(35.0)
Equity indices (% change)
S&P 500	(2.9)	(50.2)	(1.7)	(40.1)
EuroStoxx 50	3.8	(57.6)	13.5	(50.4)
SPI	(0.8)	(53.6)	5.8	(44.2)
Swiss real estate (% change)
Single-Family Homes	3.4	(17.0)	7.1	(30.0)
Other real estate (% change)
United States (S&P / Case-Shiller)	2.5	(15.3)	9.2	(28.7)
Eurozone (House Price Index)	1.1	(22.9)	7.2	(35.4)
1 2020 unemployment rate is presented as an end-of-period level. 2019 unemployment rate was presented as a change in levels. The 2020 change in level would have been: One year shock in the baseline scenario: United States: -3.5%, Eurozone: 0.4% and Switzerland: 0.4% and for the global crisis scenario: United States: 3.1%, Eurozone: 5.0% and Switzerland: 2.6%. Three year shock in the baseline scenario: United States: -4.5%, Eurozone: -1.2% and Switzerland: -0.2% and for the global crisis scenario: United States: 0.9%, Eurozone: 7.2% and Switzerland: 3.4%

Scenario assumptions	One year		Three years cumulative
31.12.19	Baseline	Severe downside	Baseline	Severe downside
Real GDP growth (% change)
United States	1.9	(6.4)	6.4	(4.3)
Eurozone	1.0	(9.1)	2.8	(10.8)
Switzerland	1.5	(7.0)	4.8	(6.2)
Consumer price index (% change)
United States	1.8	(1.2)	6.2	0.4
Eurozone	1.3	(1.3)	4.3	(1.7)
Switzerland	0.8	(1.8)	2.7	(1.6)
Unemployment rate (change, percentage points)
United States	(0.4)	5.7	(0.5)	5.6
Eurozone	(0.1)	5.6	(0.2)	7.9
Switzerland	0.1	2.6	0.3	3.6
Fixed income: 10-year government bonds (change in yields, basis points)
USD	0.2	(100.0)	10.1	(75.0)
EUR	8.4	(30.0)	28.2	(20.0)
CHF	9.5	(70.0)	30.0	(35.0)
Equity indices (% change)
S&P 500	3.5	(53.0)	9.5	(42.9)
EuroStoxx 50	0.5	(60.0)	4.4	(52.9)
SPI	1.4	(56.2)	5.3	(46.8)
Swiss real estate (% change)
Single-Family Homes	0.1	(15.2)	2.3	(27.0)
Other real estate (% change)
United States (S&P / Case-Shiller)	4.0	(13.3)	16.7	(23.4)
Eurozone (House Price Index)	1.2	(23.0)	2.2	(33.2)

c) Development of ECL allowances and provisions

The ECL allowances and provisions recognized in the period are impacted by a variety of factors, such as:

–origination of new instruments during the period;

–effect of passage of time as the ECLs on an instrument for the remaining lifetime decrease (all other factors remaining the same);

–discount unwind within ECLs as it is measured on a present value basis;

–derecognition of instruments in the period;

–change in individual asset quality of instruments;

–effect of updating forward-looking scenarios and the respective weights;

–movements from a maximum 12-month ECL to the recognition of lifetime ECLs (and vice versa) following transfers between stages 1 and 2;

–movements from stages 1 and 2 to stage 3 (credit-impaired status) when default has become certain and probability of default (PD) increases to 100% (or vice versa);

–changes in models or updates to model parameters; and

–foreign exchange translations for assets denominated in foreign currencies and other movements.

Development of ECL allowances and provisions
USD million	Total	Stage 1	Stage 2	Stage 3
Balance as of 31 December 2019	(1,029)	(181)	(160)	(688)
Net movement from new and derecognized transactions¹	(28)	(90)	17	46
of which: Private clients with mortgages	(2)	(3)	2	0
of which: Real estate financing	(3)	(5)	2	0
of which: Large corporate clients	(32)	(29)	(4)	0
of which: SME clients	(16)	(14)	(3)	0
of which: Other	26	(39)	20	46
of which: Securities financing transactions REIT	32	(1)	15	17
of which: Loans to financial advisors	9	(1)	9	0
of which: Lombard loans	23	(6)	0	29
of which Financial intermediaries	(20)	(15)	(5)	0
Remeasurements with stage transfers²	(427)	45	(134)	(338)
of which: Private clients with mortgages	(19)	(2)	(17)	0
of which: Real estate financing	(6)	3	(9)	0
of which: Large corporate clients	(224)	34	(83)	(175)
of which: SME clients	(43)	(1)	(11)	(31)
of which: Other	(134)	11	(14)	(131)
of which: Securities financing transactions REIT	(36)	0	(18)	(19)
of which: Loans to financial advisors	(12)	7	(7)	(11)
of which: Lombard loans	(36)	0	0	(36)
of which Commodity Trade Finance	(59)	0	0	(59)
Remeasurements without stage transfers³	(271)	(88)	(47)	(136)
of which: Private clients with mortgages	(34)	(19)	(8)	(7)
of which: Real estate financing	(14)	(4)	(11)	1
of which: Large corporate clients	(149)	(53)	(17)	(79)
of which: SME clients	(13)	0	(7)	(6)
of which: Other	(60)	(11)	(4)	(44)
of which: Loans to financial advisors	(18)	(12)	(3)	(3)
of which: Lombard loans	(3)	6	0	(9)
of which: Credit cards	(12)	0	0	(12)
Model changes⁴	32	21	11	0
Total ECL allowance movements with profit or loss impact⁵	(694)	(112)	(154)	(429)
Write-offs, FX and other movements (without profit or loss impact)⁶	254	(14)	(19)	287
Balance as of 31 December 2020	(1,468)	(306)	(333)	(829)
1 Represents the increase and decrease in allowances and provisions resulting from financial instruments (including guarantees and facilities) that were newly originated, purchased or renewed and from the final derecognition of loans or facilities on their maturity date or earlier. 2 Represents the remeasurement between 12-month and lifetime ECL due to stage transfers. 3 Represents the change in allowances and provisions related to changes in model inputs or assumptions, including changes in forward-looking macroeconomic conditions, changes in the exposure profile, PD and LGD changes, and unwinding of the time value. 4 Represents the change in the allowances and provisions related to changes in models and methodologies. 5 Includes ECL movements from new and derecognized transactions, remeasurement changes, model and methodology changes. 6 Represents the decrease in allowances and provisions resulting from write-offs of the ECL allowance against the gross carrying amount when all or part of a financial asset is deemed uncollectible or forgiven and movements in foreign exchange rates.

In 2020, ECL allowances and provisions increased by USD 694 million from net credit loss expenses impacting profit or loss:

The following table explains the changes in the ECL allowances and provisions for on- and off-balance sheet financial instruments and other credit lines in scope of ECL requirements between the beginning and the end of the prior period due to the factors listed earlier in this note.


USD million	Total	Stage 1	Stage 2	Stage 3
Balance as of 31 December 2018	(1,054)	(176)	(183)	(695)
Net movement from new and derecognized transactions¹	(53)	(66)	10	3
of which: Private clients with mortgages	(1)	(4)	3	0
of which: Real estate financing	(3)	(5)	2	0
of which: Large corporate clients	(6)	(14)	8	0
of which: SME clients	(16)	(14)	(2)	0
Remeasurements with stage transfers²	(125)	14	(35)	(105)
of which: Private clients with mortgages	(5)	1	(5)	(1)
of which: Real estate financing	5	4	1	0
of which: Large corporate clients	(45)	4	(11)	(38)
of which: SME clients	(64)	2	(11)	(55)
Remeasurements without stage transfers³	73	31	41	1
of which: Private clients with mortgages	22	2	30	(9)
of which: Real estate financing	1	0	0	1
of which: Large corporate clients	(24)	(10)	0	(14)
of which: SME clients	35	9	10	17
Model changes⁴	26	17	9	0
Total ECL allowance movements with profit or loss impact⁵	(78)	(4)	25	(100)
Write-offs, FX and other movements (without profit or loss impact)⁶	105	(1)	(2)	108
Balance as of 31 December 2019	(1,029)	(181)	(160)	(688)
1 Represents the increase and decrease in allowances and provisions resulting from financial instruments (including guarantees and facilities) that were newly originated, purchased or renewed and from the final derecognition of loans or facilities on their maturity date or earlier. 2 Represents the remeasurement between 12-month and lifetime ECL due to stage transfers. 3 Represents the change in allowances and provisions related to changes in model inputs or assumptions, including changes in forward-looking macroeconomic conditions, changes in the exposure profile, PD and LGD changes, and unwinding of the time value. 4 Represents the change in the allowances and provisions related to changes in models and methodologies. 5 To align to the table format for the 2020 ECL allowance and provision movement, UBS has adjusted the 2019 table format. Includes ECL movements from new and derecognized transactions, remeasurement changes, model and methodology changes. 6 Represents the decrease in allowances and provisions resulting from write-offs of the ECL allowance against the gross carrying amount when all or part of a financial asset is deemed uncollectible or forgiven and movements in foreign exchange rates.

As explained in Note 1a, the assessment of an SICR considers a number of qualitative and quantitative factors to determine whether a stage transfer between stage 1 and stage 2 is required. The primary assessment considers changes in probability of default (PD) based on rating analyses and economic outlook. Additionally, UBS AG considers counterparties that have moved to a credit watch list and those with payments that are at least 30 days past due.

Stage 2 classification by trigger	ECL allowances / provisions as of 31 December 2020
USD million	Stage 2	of which: PD layer	of which: watch list	of which: ≥30 days past due
On-and off-balance sheet	(333)	(252)	(41)	(40)
of which: Private clients with mortgages	(93)	(83)	0	(11)
of which: Real estate financing	(53)	(45)	(2)	(6)
of which: Large corporate clients	(110)	(89)	(20)	0
of which: SME clients	(38)	(16)	(16)	(5)
of which: Financial intermediaries and hedge funds	(19)	(19)	0	0
of which: Loans to financial advisors	(5)	0	(1)	(4)
of which: Credit cards	(14)	0	0	(14)
of which: Other	(2)	0	(2)	0

d) Maximum exposure to credit risk

The tables below and on the following page provide UBS AG’s maximum exposure to credit risk for financial instruments subject to ECL requirements and the respective collateral and other credit enhancements mitigating credit risk for these classes of financial instruments.

Maximum exposure to credit risk
	31.12.20
		Collateral¹				Credit enhancements¹			Exposure to credit risk after collateral and credit enhancements
USD billion	Maximum exposure to credit risk	Cash collateral received	Collateralized by securities	Secured by real estate	Other collateral²	Netting	Credit derivative contracts	Guarantees
Financial assets measured at amortized cost on the balance sheet
Cash and balances at central banks	158.2								158.2
Loans and advances to banks³	15.3		0.1						15.2
Receivables from securities financing transactions	74.2	0.0	67.1		7.0				0.0
Cash collateral receivables on derivative instruments^{4, 5}	32.7					21.1			11.6
Loans and advances to customers⁶	381.0	27.0	118.2	194.6	21.7		0.0	4.4	15.1
Other financial assets measured at amortized cost	27.2	0.1	0.2	0.0	1.3				25.5
Total financial assets measured at amortized cost	688.7	27.2	185.7	194.6	30.1	21.1	0.0	4.4	225.6
Financial assets measured at fair value through other comprehensive income – debt	8.3								8.3
Total maximum exposure to credit risk reflected on the balance sheet in scope of ECL	697.0	27.2	185.7	194.6	30.1	21.1	0.0	4.4	233.9
Guarantees⁷	17.0	0.7	5.0	0.2	1.7			2.5	7.0
Loan commitments⁷	41.2	0.0	4.2	2.1	6.8		0.4	2.4	25.3
Forward starting transactions, reverse repurchaseand securities borrowing agreements	3.2		3.2						0.0
Committed unconditionally revocable credit lines	42.0	0.1	10.3	6.2	2.7			0.0	22.7
Total maximum exposure to credit risk not reflected on the balance sheet, in scope of ECL	103.5	0.8	22.7	8.5	11.2	0.0	0.4	4.9	54.9
	31.12.19
		Collateral¹				Credit enhancements¹			Exposure to credit risk after collateral and credit enhancements
USD billion	Maximum exposure to credit risk	Cash collateral received	Collateralized by securities	Secured by real estate	Other collateral²	Netting	Credit derivative contracts	Guarantees
Financial assets measured at amortized cost on the balance sheet
Cash and balances at central banks	107.1								107.1
Loans and advances to banks³	12.4		0.0						12.3
Receivables from securities financing transactions	84.2		77.6		5.8				0.8
Cash collateral receivables on derivative instruments^{4, 5}	23.3					14.4			8.9
Loans and advances to customers⁶	328.0	19.4	101.4	174.7	17.1			1.1	14.3
Other financial assets measured at amortized cost	23.0	0.1	0.4	0.0	1.3				21.2
Total financial assets measured at amortized cost	578.0	19.5	179.4	174.7	24.3	14.4	0.0	1.1	164.6
Financial assets measured at fair value through other comprehensive income – debt	6.3								6.3
Total maximum exposure to credit risk reflected on the balance sheet in scope of ECL	584.3	19.5	179.4	174.7	24.3	14.4	0.0	1.1	171.0
Guarantees⁷	18.1	1.0	3.0	0.1	1.7			2.5	9.8
Loan commitments⁷	27.5	0.2	1.9	1.3	5.8		0.2	0.2	18.0
Forward starting transactions, reverse repurchaseand securities borrowing agreements	1.7		1.7						0.0
Committed unconditionally revocable credit lines	36.9	0.3	8.3	4.9	3.6			0.0	19.8
Total maximum exposure to credit risk not reflected on the balance sheet, in scope of ECL	84.2	1.5	14.9	6.3	11.0	0.0	0.2	2.8	47.6
1 Of which: USD 1,983 million for 31 December 2020 (31 December 2019: USD 1,720 million) relates to total credit-impaired financial assets measured at amortized cost and USD 154 million for 31 December 2020 (31 December 2019: USD 27 million) to total off-balance sheet financial instruments and other credit lines for credit-impaired positions. 2 Includes but is not limited to life insurance contracts, inventory, mortgage loans, gold and other commodities. 3 Loans and advances to banks include amounts held with third-party banks on behalf of clients. The credit risk associated with these balances may be borne by those clients. 4 Included within Cash collateral receivables on derivative instruments are margin balances due from exchanges or clearing houses. Some of these margin balances reflect amounts transferred on behalf of clients who retain the associated credit risk. 5 The amount shown in the “Netting” column represents the netting potential not recognized on the balance sheet. Refer to Note 22 for more information. 6 Collateral arrangements generally incorporate a range of collateral, including cash, securities, property and other collateral. 7 The amount shown in the “Guarantees” column includes sub-participations.

e) Financial assets subject to credit risk by rating category

›Refer to the “Risk management and control” section of this report for more details regarding the Group’s internal grading system

Financial assets subject to credit risk by rating category
USD million	31.12.20
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired (defaulted)	Total gross carrying amount	ECL allowances	Net carrying amount (maximum exposure to credit risk)
Financial assets measured at amortized cost
Cash and balances at central banks	156,250	1,981	0	0	0	0	158,231	0	158,231
of which: stage 1	156,250	1,981	0	0	0	0	158,231	0	158,231
Loans and advances to banks	543	12,029	1,344	1,182	260	1	15,360	(16)	15,344
of which: stage 1	543	11,974	1,277	1,145	231	0	15,170	(9)	15,160
of which: stage 2	0	55	67	37	29	0	189	(5)	184
of which: stage 3	0	0	0	0	0	1	1	(1)	0
Receivables from securities financing transactions	22,998	16,009	15,367	17,995	1,842	0	74,212	(2)	74,210
of which: stage 1	22,998	16,009	15,367	17,995	1,842	0	74,212	(2)	74,210
Cash collateral receivables on derivative instruments	8,196	13,477	7,733	3,243	88	0	32,737	0	32,737
of which: stage 1	8,196	13,477	7,733	3,243	88	0	32,737	0	32,737
Loans and advances to customers	5,813	215,755	67,270	69,217	21,038	2,943	382,036	(1,060)	380,977
of which: stage 1	5,813	214,418	63,000	59,447	15,860	0	358,538	(142)	358,396
of which: stage 2	0	1,338	4,269	9,770	5,178	0	20,556	(215)	20,341
of which: stage 3	0	0	0	0	0	2,943	2,943	(703)	2,240
Other financial assets measured at amortized cost	15,404	4,043	280	6,585	481	560	27,352	(133)	27,219
of which: stage 1	15,404	4,040	269	6,334	389	0	26,435	(34)	26,401
of which: stage 2	0	3	11	251	91	0	357	(9)	348
of which: stage 3	0	0	0	0	0	560	560	(90)	469
Total financial assets measured at amortized cost	209,204	263,295	91,993	98,223	23,709	3,505	689,929	(1,211)	688,717
On-balance sheet financial instruments
Financial assets measured at FVOCI – debt instruments	3,212	5,014	0	32	0	0	8,258	0	8,258
Total on balance sheet financial instruments	212,417	268,309	91,993	98,255	23,709	3,505	698,187	(1,211)	696,976

Off-balance sheet positions subject to expected credit loss by rating category
USD million	31.12.20
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired(defaulted)	Total off - balance sheet exposure(maximum exposure to credit risk)	ECL provisions
Off-balance sheet financial instruments
Guarantees	3,482	4,623	3,522	4,293	991	170	17,081	(63)
of which: stage 1	3,482	4,219	2,688	3,558	739	0	14,687	(14)
of which: stage 2	0	404	834	736	252	0	2,225	(15)
of which: stage 3	0	0	0	0	0	170	170	(34)
Irrevocable loan commitments	3,018	14,516	8,583	9,302	5,850	104	41,372	(142)
of which: stage 1	3,018	13,589	6,873	8,739	4,676	0	36,894	(74)
of which: stage 2	0	927	1,711	563	1,174	0	4,374	(68)
of which: stage 3	0	0	0	0	0	104	104	0
Forward starting reverse repurchase and securities borrowing agreements	82	150	0	3,015	0	0	3,247	0
Total off balance sheet financial instruments	6,583	19,289	12,105	16,610	6,840	273	61,700	(205)
Other credit lines
Committed unconditionally revocable credit lines	574	15,448	5,958	8,488	11,501	108	42,077	(50)
of which: stage 1	574	14,883	4,517	6,609	10,593	0	37,176	(29)
of which: stage 2	0	565	1,441	1,879	908	0	4,792	(21)
of which: stage 3	0	0	0	0	0	108	108	0
Irrevocable committed prolongation of existing loans	14	1,349	931	632	357	0	3,282	(2)
of which: stage 1	14	1,349	930	630	355	0	3,277	(2)
of which: stage 2	0	1	1	2	1	0	5	0
of which: stage 3	0	0	0	0	0	0	0	0
Total other credit lines	588	16,797	6,889	9,119	11,858	109	45,359	(52)
1 Refer to the “Internal UBS rating scale and mapping of external ratings” table in the “Risk management and control” section of this report for more information about rating categories.

Financial assets subject to credit risk by rating category
USD million	31.12.19
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired (defaulted)	Total gross carrying amount	ECL allowances	Net carrying amount (maximum exposure to credit risk)
Financial assets measured at amortized cost
Cash and balances at central banks	105,195	1,873	0	0	0	0	107,068	0	107,068
of which: stage 1	105,195	1,873	0	0	0	0	107,068	0	107,068
Loans and advances to banks	309	9,764	1,326	687	298	1	12,386	(6)	12,379
of which: stage 1	309	9,764	1,326	677	228	0	12,303	(4)	12,298
of which: stage 2	0	0	0	10	71	0	81	(1)	80
of which: stage 3	0	0	0	0	0	1	1	(1)	0
Receivables from securities financing transactions	21,089	16,889	14,366	28,815	3,088	0	84,246	(2)	84,245
of which: stage 1	21,089	16,889	14,366	28,815	3,088	0	84,246	(2)	84,245
Cash collateral receivables on derivative instruments	4,899	10,553	5,033	2,765	39	0	23,289	0	23,289
of which: stage 1	4,899	10,553	5,033	2,765	39	0	23,289	0	23,289
Loans and advances to customers	1,744	176,189	59,240	70,528	18,748	2,308	328,756	(764)	327,992
of which: stage 1	1,744	175,534	56,957	62,435	14,117	0	310,787	(82)	310,705
of which: stage 2	0	655	2,283	8,093	4,631	0	15,661	(123)	15,538
of which: stage 3	0	0	0	0	0	2,308	2,308	(559)	1,749
Other financial assets measured at amortized cost	13,030	1,592	390	7,158	312	672	23,154	(143)	23,012
of which: stage 1	13,030	1,581	381	6,747	280	0	22,019	(35)	21,985
of which: stage 2	0	11	9	412	32	0	463	(13)	451
of which: stage 3	0	0	0	0	0	672	672	(95)	576
Total financial assets measured at amortized cost	146,267	216,860	80,354	109,952	22,485	2,981	578,899	(915)	577,985
On-balance sheet financial instruments
Financial assets measured at FVOCI – debt instruments	5,854	450	0	41	0	0	6,345	0	6,345
Total on balance sheet financial instruments	152,120	217,309	80,354	109,994	22,485	2,981	585,245	(915)	584,329

Off-balance sheet positions subject to expected credit loss by rating category
USD million	31.12.19
Rating category¹	0–1	2–3	4–5	6–8	9–13	Credit-impaired(defaulted)	Total off - balance sheet exposure(maximum exposure to credit risk)	ECL provisions
Off-balance sheet financial instruments
Guarantees	857	4,932	6,060	5,450	761	82	18,142	(42)
of which: stage 1	857	4,931	6,048	5,218	704	0	17,757	(8)
of which: stage 2	0	1	12	233	57	0	304	(1)
of which: stage 3	0	0	0	0	0	82	82	(33)
Irrevocable loan commitments	2,548	10,068	4,862	5,859	4,160	50	27,547	(35)
of which: stage 1	2,548	10,068	4,862	5,722	3,878	0	27,078	(30)
of which: stage 2	0	0	0	137	282	0	419	(5)
of which: stage 3	0	0	0	0	0	50	50	0
Forward starting reverse repurchase and securities borrowing agreements	0	672	50	936	0	0	1,657	0
Total off balance sheet financial instruments	3,405	15,672	10,972	12,245	4,922	132	47,347	(77)
Other credit lines
Committed unconditionally revocable credit lines	632	14,346	6,231	7,169	8,554	46	36,979	(34)
of which: stage 1	632	14,309	6,120	6,789	7,885	0	35,735	(17)
of which: stage 2	0	37	111	380	669	0	1,197	(17)
of which: stage 3	0	0	0	0	0	46	46	0
Irrevocable committed prolongation of existing loans	25	1,399	870	633	359	4	3,289	(3)
of which: stage 1	25	1,399	870	633	359	0	3,285	(3)
of which: stage 2	0	0	0	0	0	0	0	0
of which: stage 3	0	0	0	0	0	4	4	0
Total other credit lines	657	15,745	7,101	7,801	8,913	50	40,268	(37)
1 Refer to the “Internal UBS rating scale and mapping of external ratings” table in the “Risk management and control” section of this report for more information about rating categories.

f) Sensitivity information

As outlined in Note 1a, ECL estimates involve significant uncertainties at the time they are made.

ECL model

Forward-looking scenarios

Potential effect on stage 1 and stage 2 positions from changing key parameters as at 31 December 2020


USD million	Baseline	Severe downside	Weighted average
Change in key parameters
Fixed income: 10-year government bonds (absolute change)
–0.5%	(1.36)	(1.84)	(1.93)
+0.5%	2.10	3.19	3.23
+1.00%	5.69	6.86	7.19
Unemployment rate (absolute change)
–1.00%	(7.40)	(63.01)	(27.83)
–0.5%	(3.78)	(33.54)	(15.67)
+0.5%	4.15	36.97	16.99
+1.00%	8.50	75.93	33.74
Real GDP growth (relative change)
-2.00%	3.72	16.14	9.10
-1.00%	1.86	9.84	5.09
+1.00%	(1.46)	(3.30)	(2.36)
+2.00%	(2.97)	(9.44)	(5.93)
House Price Index (relative change)
–5.00%	8.04	144.34	51.46
–2.50%	3.45	65.80	23.28
+2.50%	(2.79)	(56.60)	(19.09)
+5.00%	(5.16)	(105.61)	(35.29)
Equity (S&P500, EuroStoxx, SMI) (relative change)
–10.00%	3.94	9.66	6.78
–5.00%	1.91	4.29	3.34
+5.00%	(8.30)	(4.23)	(7.27)
+10.00%	(10.14)	(8.58)	(10.22)

The forecasting horizon is limited to three years, with a model-based mean reversion of PD and LGD assumed thereafter. Changes to these timelines may have an effect on ECLs: depending on the cycle, a longer or shorter forecasting horizon will lead to different annualized lifetime PD and average LGD estimations. This is currently not deemed to be material for UBS AG, as a large proportion of loans, including mortgages in Switzerland, have maturities that are within the forecasting horizon.

Scenario weights

ECL is sensitive to changing scenario weights, in particular if narratives and parameters are selected that are not close to the baseline scenario, highlighting the non-linearity of credit losses.

Stage allocation and SICR

Maturity profile

The maturity profile of the assets is an important driver for changes in ECL due to transfers to stage 2 and from stage 2 to stage 1. The current maturity profile of most lending books is relatively short; hence a movement to stage 2 may have a limited effect on ECLs. A significant portion of our lending to SMEs is documented under multi-purpose credit agreements, which allow for various forms of utilization but are unconditionally cancelable by UBS AG at any time. The relevant maturity for drawings under such agreements with a fixed maturity is the respective term, or a maximum of 12 months in stage 1. For unused credit lines and all drawings that have no fixed maturity (e.g., current accounts), UBS AG generally applies a 12-month maturity from the reporting date, given the credit review policies, which require either continuous monitoring of key indicators and behavioral patterns for smaller positions or an annual formal review for any other limit. The ECLs for these products is sensitive to shortening or extending the maturity assumption.

Potential effect on stage 1 and stage 2 positions from changing scenario weights or moving to an ECL lifetime calculation as at 31 December 2020

	Actual ECL allowances and provisions (as per Note 9)		Pro forma ECL allowances and provisions, assuming application of 100% weighting				Pro forma ECL allowances and provisions, assuming all positions being subject to lifetime ECL
Scenarios	Weighted average		Baseline		Severe downside		Weighted average
USD million, except where indicated	ECL	in % of baseline	ECL	in % of baseline	ECL	in % of baseline	ECL	in % of baseline
Segmentation
Private clients with mortgages	(131)	244	(54)	100	(302)	562	(385)	717
Real estate financing	(76)	138	(55)	100	(123)	224	(131)	237
Large corporate clients	(206)	149	(138)	100	(298)	216	(307)	222
SME clients	(74)	115	(64)	100	(93)	144	(129)	200
Other segments	(152)	116	(131)	100	(183)	140	(385)	294
Total	(639)	145	(442)	100	(999)	226	(1,336)	302