•

Wealth Management.

Wealth management loans includes loans extended to private bank clients, including wealth management and other clients. These loans are used to finance investments in both financial and nonfinancial assets, bridge cash flow timing gaps or provide liquidity for other needs. Substantially all of such loans are secured by securities, residential real estate, commercial real estate or other assets.

•

Commercial Real Estate.

Commercial real estate loans include originated loans (other than those extended to private bank clients) that are directly or indirectly secured by hotels, retail stores, multifamily housing complexes and commercial and industrial properties. Commercial real estate loans also includes loans extended to clients who warehouse assets that are directly or indirectly backed by commercial real estate. In addition, commercial real estate includes loans purchased by the firm.

•

Residential Real Estate.

Residential real estate loans primarily includes loans extended by the firm to clients (other than those extended to private bank clients) who warehouse assets that are directly or indirectly secured by residential real estate and loans purchased by the firm.

•

Installment.

Installment loans are unsecured and are originated by the firm.

•

Credit Cards.

Credit card loans are loans made pursuant to revolving lines of credit issued to consumers by the firm.

•

Other.

Other loans primarily includes loans extended to clients who warehouse assets that are directly or indirectly secured by consumer loans, including auto loans and private student loans. Other loans also includes unsecured consumer and credit card loans purchased by the firm.

Loans accounted for at amortized cost included PCI loans with a carrying value of $1.62 billion (outstanding principal balance of $3.23 billion and accretable yield of $220 million) as of December 2019, which were secured by commercial and residential real estate. In January 2020, the firm elected the fair value option for these PCI loans in accordance with ASU

No. 2016-13.

These loans were primarily transferred to trading assets. See Note 3 for further information about adoption of this ASU.

Credit Quality

Risk Assessment.

The firm’s risk assessment process includes evaluating the credit quality of its loans. For corporate loans and a majority of wealth management, real estate and other loans, the firm performs credit reviews which include initial and ongoing analyses of its borrowers, resulting in an internal credit rating. A credit review is an independent analysis of the capacity and willingness of a borrower to meet its financial obligations and is performed on an annual basis or more frequently if circumstances change that indicate that a review may be necessary. The determination of internal credit ratings also incorporates assumptions with respect to the nature of and outlook for the borrower’s industry and the economic environment.

The table below presents gross loans by an internally determined public rating agency equivalent or other credit metrics and the concentration of secured and unsecured loans.



$ in millions	Investment- Grade	Non-Investment- Grade	Other/ Unrated	Total
As of December 2020
Accounting Method
Amortized cost	$ 33,532	$58,250	$11,783	$103,565
Fair value	2,084	5,925	5,616	13,625
Held for sale	224	2,152	423	2,799
Total	$35,840	$66,327	$17,822	$119,989
Loan Type
Corporate	$ 9,478	$38,704	$ 477	$ 48,659
Wealth management	22,098	5,331	5,594	33,023
Real estate:
Commercial	1,792	17,480	1,018	20,290
Residential	636	3,852	1,262	5,750
Consumer:
Installment	–	–	3,823	3,823
Credit cards	–	–	4,270	4,270
Other	1,836	960	1,378	4,174
Total	$35,840	$66,327	$17,822	$119,989
Secured	83%	90%	46%	82%
Unsecured	17%	10%	54%	18%
Total	100%	100%	100%	100%
As of December 2019
Accounting Method
Amortized cost	$30,266	$51,222	$ 9,148	$ 90,636
Fair value	2,844	5,174	6,368	14,386
Held for sale	323	4,368	632	5,323
Total	$33,433	$60,764	$16,148	$110,345
Loan Type
Corporate	$10,507	$35,509	$ 291	$ 46,307
Wealth management	20,001	3,576	4,363	27,940
Real estate:
Commercial	306	15,997	1,440	17,743
Residential	244	4,600	2,114	6,958
Consumer:
Installment	–	–	4,747	4,747
Credit cards	–	–	1,858	1,858
Other	2,375	1,082	1,335	4,792
Total	$33,433	$60,764	$16,148	$110,345
Secured	83%	91%	54%	83%
Unsecured	17%	9%	46%	17%
Total	100%	100%	100%	100%

In the table above, other/unrated loans include installment and credit card loans of $8.09 billion as of December 2020 and $6.61 billion as of December 2019 for which an important credit-quality indicator is the Fair Isaac Corporation (FICO) credit score (which measures a borrower’s creditworthiness by considering factors such as payment and credit history). FICO credit scores are periodically refreshed by the firm to assess the updated creditworthiness of the borrower. See “Vintage” below for information about installment and credit card loans by FICO credit scores. The vast majority of the remaining loans of $9.73 billion as of December 2020 and $9.54 billion as of December 2019 included in the other/unrated category are secured. These loans primarily consist of wealth management loans backed by residential real estate and securities, and purchased real estate-backed loans. The firm’s risk assessment process for such loans includes reviewing certain key metrics, such as

loan-to-value

ratio, delinquency status, collateral values, expected cash flows and other risk factors.

The firm also assigns a regulatory risk rating to its loans based on the definitions provided by the U.S. federal bank regulatory agencies. Total loans included 85% of loans as of December 2020 and 91% of loans as of December 2019 that were rated

pass/non-criticized.

Vintage.

The table below presents gross loans accounted for at amortized cost (excluding installment and credit card loans) by an internally determined public rating agency equivalent or other credit metrics and origination year for term loans.



	As of December 2020

$ in millions		Investment- Grade	Non-Investment- Grade	Other/ Unrated	Total
2020		$ 1,978	$ 7,545	$ 140	$ 9,663
2019		889	6,106	–	6,995
2018		2,076	3,555	–	5,631
2017		851	3,083	–	3,934
2016		268	1,262	–	1,530
2015 or earlier		351	2,073	–	2,424
Revolving		2,662	11,891	48	14,601
Corporate		9,075	35,515	188	44,778
2020		497	313	–	810
2019		723	403	–	1,126
2018		298	87	–	385
2017		377	30	–	407
2016		22	20	–	42
2015 or earlier		531	264	–	795
Revolving		18,077	2,085	1,424	21,586
Wealth management		20,525	3,202	1,424	25,151
2020		848	3,071	55	3,974
2019		76	1,965	–	2,041
2018		137	2,164	25	2,326
2017		26	1,734	12	1,772
2016		–	165	9	174
2015 or earlier		–	775	526	1,301
Revolving		461	5,047	–	5,508
Commercial real estate		1,548	14,921	627	17,096
2020		402	976	115	1,493
2019		–	90	271	361
2018		–	123	249	372
2017		9	83	152	244
2016		–	1	–	1
2015 or earlier		–	–	70	70
Revolving		225	2,470	–	2,695
Residential real estate		636	3,743	857	5,236
2020		242	84	466	792
2019		–	67	29	96
2018		–	46	–	46
2017		–	8	–	8
Revolving		1,506	664	99	2,269
Other		1,748	869	594	3,211
Total		$33,532	$58,250	$3,690	$95,472
Percentage of total		35%	61%	4%	100%

In the table above, revolving loans which converted to term loans were not material as of December 2020.

The table below presents gross installment loans by refreshed FICO credit scores and origination year and gross credit card loans by refreshed FICO credit scores.



	As of December 2020

$ in millions		Greater than or equal to 660	Less than 660	Total
2020		$1,321	$ 38	$1,359
2019		1,225	132	1,357
2018		792	150	942
2017		128	30	158
2016		6	1	7
Installment		3,472	351	3,823
Credit cards		3,398	872	4,270
Total consumer		$6,870	$1,223	$8,093
Percentage of total:
Installment		91%	9%	100%
Credit cards		80%	20%	100%
Total consumer		85%	15%	100%

In the table above, credit card loans consist of revolving lines of credit.

Credit Concentrations.

The table below presents the concentration of gross loans by region.



$ in millions	Carrying Value	Americas	EMEA	Asia	Total
As of December 2020
Corporate	$ 48,659	60%	31%	9%	100%
Wealth management	33,023	88%	10%	2%	100%
Commercial real estate	20,290	71%	19%	10%	100%
Residential real estate	5,750	88%	9%	3%	100%
Consumer:
Installment	3,823	100%	–	–	100%
Credit cards	4,270	100%	–	–	100%
Other	4,174	81%	17%	2%	100%
Total	$119,989	75%	19%	6%	100%
As of December 2019
Corporate	$ 46,307	60%	31%	9%	100%
Wealth management	27,940	88%	9%	3%	100%
Commercial real estate	17,743	69%	21%	10%	100%
Residential real estate	6,958	90%	9%	1%	100%
Consumer:
Installment	4,747	100%	–	–	100%
Credit cards	1,858	100%	–	–	100%
Other	4,792	87%	12%	1%	100%
Total	$110,345	73%	21%	6%	100%

In the table above:

•

EMEA represents Europe, Middle East and Africa.

•

The top five industry concentrations for corporate loans as of December 2020 were 17% for diversified industrials (17% as of December 2019), 17% for technology, media & telecommunications (17% as of December 2019), 13% for funds (9% as of December 2019), 12% for natural resources & utilities (12% as of December 2019), and 10% for financial institutions (10% as of December 2019).

Nonaccrual and Past Due Loans.

Loans accounted for at amortized cost (other than credit card loans) are placed on nonaccrual status when it is probable that the firm will not collect all principal and interest due under the contractual terms, regardless of the delinquency status or if a loan is past due for 90 days or more, unless the loan is both well collateralized and in the process of collection. At that time, all accrued but uncollected interest is reversed against interest income and interest subsequently collected is recognized on a cash basis to the extent the loan balance is deemed collectible. Otherwise, all cash received is used to reduce the outstanding loan balance. A loan is considered past due when a principal or interest payment has not been made according to its contractual terms. Credit card loans are not placed on nonaccrual status and accrue interest until the loan is paid in full or is

charged-off.

In certain circumstances, the firm may modify the original terms of a loan agreement by granting a concession to a borrower experiencing financial difficulty, typically in the form of a modification of loan covenants, but may also include forbearance of interest or principal, payment extensions or interest rate reductions. These modifications, to the extent significant, are considered troubled debt restructurings (TDRs). Loan modifications that extend payment terms for a period of less than 90 days are generally considered insignificant and therefore not reported as TDRs.

In response to the global outbreak of the coronavirus

(COVID-19)

pandemic, the firm adopted the relief issued under the Coronavirus Aid, Relief, and Economic Security (CARES) Act, as amended, and certain interpretive guidance issued by the U.S. banking agencies that provides for certain modified loans that would otherwise meet the definition of a TDR to not be classified as such. As of December 2020, the firm had $184 million of loans accounted for at amortized cost that were not classified as TDRs as a result of this relief and interpretive guidance.

The table below presents information about past due loans.



$ in millions	30-89 days	90 days or more	Total
As of December 2020
Corporate	$ –	$294	$294
Wealth management	58	34	92
Commercial real estate	49	183	232
Residential real estate	4	23	27
Consumer:
Installment	42	16	58
Credit cards	46	31	77
Other	20	4	24
Total	$219	$585	$804
Total divided by gross loans at amortized cost			0.8%
As of December 2019
Corporate	$197	$ 42	$239
Wealth management	13	15	28
Commercial real estate	54	123	177
Residential real estate	19	18	37
Consumer:
Installment	71	29	100
Credit cards	35	4	39
Other	6	1	7
Total	$395	$232	$627
Total divided by gross loans at amortized cost			0.7%

The table below presents information about nonaccrual loans.



	As of December

$ in millions		2020	2019
Corporate		$2,651	$1,122
Wealth management		61	52
Commercial real estate		649	175
Residential real estate		25	143
Installment		44	38
Other		122	–
Total		$3,552	$1,530
Total divided by gross loans at amortized cost		3.4%	1.7%

In the table above:

•

Nonaccrual loans included $533 million as of December 2020 and $429 million as of December 2019 of loans that were 30 days or more past due.

•

Loans that were 90 days or more past due and still accruing were not material as of both December 2020 and December 2019.

•

Nonaccrual loans included $315 million as of December 2020 and $251 million as of December 2019 of corporate and commercial real estate loans that were modified in a troubled debt restructuring. The firm’s lending commitments related to these loans were not material as of both December 2020 and December 2019. Installment loans that were modified in a troubled debt restructuring were not material as of both December 2020 and December 2019.

Allowance for Credit Losses

The firm’s allowance for credit losses consists of the allowance for losses on loans and lending commitments accounted for at amortized cost. Loans and lending commitments accounted for at fair value or accounted for at the lower of cost or fair value are not subject to an allowance for credit losses.

The firm adopted ASU

No. 2016-13

in January 2020, which replaced the incurred credit loss model for recognizing credit losses with the CECL model. As a result, the firm’s allowance for credit losses effective January 2020 reflects management’s estimate of credit losses over the remaining expected life of such loans and also considers forecasts of future economic conditions. Prior to January 2020, the allowance for credit losses reflected probable incurred credit losses. See Note 3 for further information about the adoption of CECL.

To determine the allowance for credit losses, the firm classifies its loans and lending commitments accounted for at amortized cost into wholesale and consumer portfolios. Prior to January 2020, the firm also had PCI loans which were classified as a separate portfolio. These portfolios represent the level at which the firm has developed and documented its methodology to determine the allowance for credit losses. The allowance for credit losses is measured on a collective basis for loans that exhibit similar risk characteristics using a modeled approach and asset-specific basis for loans that do not share similar risk characteristics.

Under CECL, the allowance for credit losses takes into account the weighted average of a range of forecasts of future economic conditions over the expected life of the loan and lending commitments. The expected life of each loan or lending commitment is determined based on the contractual term adjusted for extension options or demand features. The forecasts include baseline, favorable and adverse economic scenarios over a three-year period. For loans with expected lives beyond three years, the model reverts to historical loss information based on a

non-linear

modeled approach. The forecasted economic scenarios consider a number of risk factors relevant to the wholesale and consumer portfolios described below. The firm applies judgment in weighing individual scenarios each quarter based on a variety of factors, including the firm’s internally derived economic outlook, market consensus, recent macroeconomic conditions and industry trends.

The allowance for credit losses also includes qualitative components which allow management to reflect the uncertain nature of economic forecasting, capture uncertainty regarding model inputs, and account for model imprecision and concentration risk.

Management’s estimate of credit losses entails judgment about loan collectability at the reporting dates, and there are uncertainties inherent in those judgments. The allowance for credit losses is subject to a governance process that involves review and approval by senior management within the firm’s independent risk oversight and control functions. Personnel within the firm’s independent risk oversight and control functions are responsible for forecasting the economic variables that underlie the economic scenarios that are used in the modeling of expected credit losses. While management uses the best information available to determine this estimate, future adjustments to the allowance may be necessary based on, among other things, changes in the economic environment or variances between actual results and the original assumptions used.

The table below presents gross loans and lending commitments accounted for at amortized cost by portfolio.



	As of December

	2020			2019

$ in millions		Loans	Lending Commitments		Loans	Lending Commitments
Wholesale
Corporate		$ 44,778	$127,756		$41,129	$127,226
Wealth management		25,151	2,314		20,116	2,198
Commercial real estate		17,096	4,154		12,803	3,207
Residential real estate		5,236	1,804		4,965	759
Other		3,211	4,841		3,396	3,029
Consumer
Installment		3,823	4		4,747	12
Credit cards		4,270	21,640		1,858	13,669
PCI		–	–		1,622	–
Total		$103,565	$162,513		$90,636	$150,100

In the table above, wholesale loans included $3.51 billion as of December 2020 and $1.49 billion as of December 2019 of nonaccrual loans for which the allowance for credit losses was measured on an asset-specific basis. The allowance for credit losses on these loans was $649 million as of December 2020 and $207 million as of December 2019. These loans included $584 million as of December 2020 and $754 million as of December 2019 of loans which did not require a reserve as the loan was deemed to be recoverable.

See Note 18 for further information about lending commitments.

The following is a description of the methodology used to calculate the allowance for credit losses:

Wholesale.

The allowance for credit losses for wholesale loans and lending commitments that exhibit similar risk characteristics is measured using a modeled approach. These models determine the probability of default and loss given default based on various risk factors, including internal credit ratings, industry default and loss data, expected life, macroeconomic indicators, the borrower’s capacity to meet its financial obligations, the borrower’s country of risk and industry, loan seniority and collateral type. For lending commitments, the methodology also considers probability of drawdowns or funding. In addition, for loans backed by real estate, risk factors include the

loan-to-value

ratio, debt service ratio and home price index. The most significant inputs to the forecast model for wholesale loans and lending commitments include unemployment rates, GDP, credit spreads, commercial and industrial delinquency rates, short- and long-term interest rates, and oil prices.

The allowance for loan losses for wholesale loans that do not share similar risk characteristics, such as nonaccrual loans or loans in a troubled debt restructuring, is calculated using the present value of expected future cash flows discounted at the loan’s original effective rate, the observable market price of the loan or the fair value of the collateral.

Wholesale loans are

charged-off

against the allowance for loan losses when deemed to be uncollectible.

Consumer.

The allowance for credit losses for consumer loans that exhibit similar risk characteristics is calculated using a modeled approach which classifies consumer loans into pools based on borrower-related and exposure-related characteristics that differentiate a pool’s risk characteristics from other pools. The factors considered in determining a pool are generally consistent with the risk characteristics used for internal credit risk measurement and management and include key metrics, such as FICO credit scores, delinquency status, loan vintage and macroeconomic indicators. The most significant inputs to the forecast model for consumer loans include unemployment rates and delinquency rates. The expected life of revolving credit card loans is determined by modeling expected future draws and the timing and amount of repayments allocated to the funded balance. The firm does not recognize an allowance for credit losses on credit card lending commitments as they are cancellable by the firm.

The allowance for credit losses for consumer loans that do not share similar risk characteristics, such as loans in a troubled debt restructuring, is calculated using the present value of expected future cash flows discounted at the loan’s original effective rate.

Installment loans are

charged-off

when they are 120 days past due. Credit card loans are

charged-off

when they are 180 days past due.

Allowance for Credit Losses Rollforward

The table below presents information about the allowance for credit losses.



$ in millions	Wholesale		Consumer		PCI		Total
Year Ended December 2020
Allowance for loan losses
Beginning balance, reported	$ 879		$ 393		$ 169		$1,441
Impact of CECL adoption	452		444		(169	)	727
Beginning balance, adjusted	1,331		837		–		2,168
Net charge-offs	(615	)	(292	)	–		(907	)
Provision	2,108		745		–		2,853
Other	(240	)	–		–		(240	)
Ending balance	$2,584		$1,290		$ –		$3,874
Allowance ratio	2.7%		15.9%		–		3.7%
Net charge-off ratio	0.6%		4.2%		–		0.9%
Allowance for losses on lending commitments
Beginning balance, reported	$ 361		$ –		$ –		$ 361
Impact of CECL adoption	(48	)	–		–		(48	)
Beginning balance, adjusted	313		–		–		313
Provision	244		–		–		244
Ending balance	$ 557		$ –		$ –		$ 557
Year Ended December 2019
Allowance for loan losses
Beginning balance	$ 658		$ 292		$ 116		$1,066
Net charge-offs	(121	)	(317	)	(52	)	(490	)
Provision	469		418		103		990
Other	(127	)	–		2		(125	)
Ending balance	$ 879		$ 393		$ 169		$1,441
Allowance ratio	1.1%		6.0%		10.4%		1.6%
Net charge-off ratio	0.2%		6.2%		3.2%		0.6%
Allowance for losses on lending commitments
Beginning balance	$ 286		$ –		$ –		$ 286
Provision	75		–		–		75
Ending balance	$ 361		$ –		$ –		$ 361

In the table above:

•

Other represents the reduction to the allowance related to loans and lending commitments transferred to held for sale.

•

The allowance ratio is calculated by dividing the allowance for loan losses by gross loans accounted for at amortized cost.

•

The net

charge-off

ratio is calculated by dividing net charge-offs by average gross loans accounted for at amortized cost.

Allowance for Credit Losses Rollforward Commentary

Year Ended December 2020.

The allowance for credit losses increased by $2.63 billion during 2020.

The impact of CECL adoption for wholesale and consumer loans was driven by the fact that the allowance under CECL covers expected credit losses over the full expected life of the loan portfolios and also considers forecasts of expected future economic conditions. The impact of CECL adoption for PCI loans was as a result of the firm electing to apply the fair value option for such loans.

The provision for credit losses for wholesale and consumer loans reflected the continued impact of the

COVID-19

pandemic on economic conditions, which resulted in higher modeled expected losses and lower recoveries. The allowance for loan losses ratio for wholesale loans increased to 2.7% as of December 2020 compared with 1.1% as of December 2019, while the allowance for loan losses ratio for consumer loans increased to 15.9% as of December 2020 compared with 6.0% as of December 2019. The increase in the allowance for loan losses ratios reflected both the impact of adopting the CECL standard, as well as higher provision for credit losses.

When modeling expected credit losses, the firm employs a weighted, multivariate forecast, which includes baseline, adverse and favorable economic scenarios. As of December 2020, the forecasted economic scenarios were most heavily weighted towards the baseline and adverse scenarios. The forecast model incorporated adjustments to reflect the impact of

COVID-19-related

economic support programs provided by national governments.

The table below presents the forecasted range (across the baseline, adverse and favorable scenarios) of the U.S. unemployment and U.S. GDP growth rates used in the forecast model as of December 2020.



	U.S. Unemployment Rate	Growth/(Decline) in U.S. GDP
Forecast for the quarter ended:
June 2021	6.0% to 12.0%	(0.2)% to (8.1)%
December 2021	5.3% to 8.3%	2.7% to (4.1)%
June 2022	4.9% to 6.9%	4.4% to (1.8)%

In the table above:

•

U.S. unemployment rate represents the rate forecasted as of the respective

quarter-end.

•

Growth/(decline) in U.S. GDP represents the change in quarterly U.S. GDP relative to the U.S. GDP for the fourth quarter of 2019

(pre-pandemic

levels).

•

Recovery of quarterly U.S. GDP to its

pre-pandemic

levels in the three scenarios ranges from the quarters ending September 2021 to March 2023.

•

While the U.S. unemployment and U.S. GDP growth rates are significant inputs to the forecast model, the model contemplates a variety of other inputs across a range of scenarios to provide a forecast of future economic conditions. Given the complex nature of the forecasting process, no single economic variable can be viewed in isolation and independently of other inputs.

In addition, the provision for credit losses for wholesale loans was impacted by asset-specific provisions and ratings downgrades primarily related to borrowers in the diversified industrials, technology, media & communications and natural resources industries. Besides the weaker economic outlook related to the

COVID-19

pandemic, the provision for credit losses for consumer loans for 2020 was also impacted by the growth of the credit card portfolio.

Net charge-offs for 2020 for wholesale loans were substantially all related to corporate loans and net charge-offs for consumer loans were primarily related to installment loans.

Year Ended December 2019.

The allowance for credit losses increased by $450 million during the year ended December 2019.

The provision for credit losses for wholesale loans was substantially all related to corporate loans for the year ended December 2019. The provision for credit losses related to consumer loans was primarily related to installment loans for the year ended December 2019.

Net charge-offs for wholesale loans were primarily related to corporate loans for the year ended December 2019. Net charge-offs for consumer loans were substantially all related to installment loans for the year ended December 2019.

Fair Value of Loans by Level

The table below presents loans held for investment accounted for at fair value under the fair value option by level within the fair value hierarchy.



$ in millions	Level 1	Level 2	Level 3	Total
As of December 2020
Loan Type
Corporate	$ –	$ 1,822	$ 929	$ 2,751
Wealth management	–	7,809	63	7,872
Commercial real estate	–	857	1,104	1,961
Residential real estate	–	234	260	494
Other	–	225	322	547
Total	$ –	$10,947	$2,678	$13,625
As of December 2019
Loan Type
Corporate	$ –	$ 2,472	$ 752	$ 3,224
Wealth management	–	7,764	60	7,824
Commercial real estate	–	1,285	591	1,876
Residential real estate	–	571	221	792
Other	–	404	266	670
Total	$ –	$12,496	$1,890	$14,386

The gains as a result of changes in the fair value of loans held for investment for which the fair value option was elected were $151 million for 2020 and $355 million for 2019. These gains were included in other principal transactions.

See Note 4 for an overview of the firm’s fair value measurement policies and the valuation techniques and significant inputs used to determine the fair value of loans.

Significant Unobservable Inputs

The table below presents the amount of level 3 loans, and ranges and weighted averages of significant unobservable inputs used to value such loans.

As of December 2020

As of December 2019

$ in millions

Amount or

Range

Weighted

Average

Amount or

Range

Weighted

Average

Corporate

Level 3 assets

$929

$752

Yield

1.1% to 45.2%

12.4%

1.9% to 26.3%

9.5%

Recovery rate

15.0% to 58.0%

31.0%

13.5% to 78.0%

44.4%

Duration (years)

1.5 to 5.3

3.4

3.7 to 5.8

3.9

Commercial real estate

Level 3 assets

$1,104

$591

Yield

4.5% to 19.3%

11.0%

7.0% to 16.0%

9.3%

Recovery rate

3.0% to 99.8%

66.5%

5.9% to 85.2%

48.6%

Duration (years)

0.3 to 4.8

2.6

0.2 to 5.3

3.5

Residential real estate

Level 3 assets

$260

$221

Yield

2.0% to 14.0%

12.1%

1.1% to 14.0%

11.5%

Duration (years)

0.6 to 2.6

1.7

1.1 to 4.8

4.0

Wealth management and other

Level 3 assets

$385

$326

Yield

2.8% to 18.7%

8.0%

3.9% to 16.0%

9.9%

Duration (years)

0.9 to 5.5

4.1

1.6 to 6.7

3.7

In the table above:

•

Ranges represent the significant unobservable inputs that were used in the valuation of each type of loan.

•

Weighted averages are calculated by weighting each input by the relative fair value of the loan.

•

The ranges and weighted averages of these inputs are not representative of the appropriate inputs to use when calculating the fair value of any one loan. For example, the highest yield for residential real estate loans is appropriate for valuing a specific residential real estate loan but may not be appropriate for valuing any other residential real estate loan. Accordingly, the ranges of inputs do not represent uncertainty in, or possible ranges of, fair value measurements of level 3 loans.

•

Increases in yield or duration used in the valuation of level 3 loans would have resulted in a lower fair value measurement, while increases in recovery rate would have resulted in a higher fair value measurement as of both December 2020 and December 2019. Due to the distinctive nature of each level 3 loan, the interrelationship of inputs is not necessarily uniform within each product type.

•

Loans are valued using discounted cash flows.

Level 3 Rollforward

The table below presents a summary of the changes in fair value for level 3 loans.



	Year Ended December

$ in millions		2020		2019
Beginning balance		$1,890		$1,990
Net realized gains/(losses)		72		46
Net unrealized gains/(losses)		87		85
Purchases		670		249
Sales		(50	)	(14	)
Settlements		(727	)	(795	)
Transfers into level 3		836		444
Transfers out of level 3		(100	)	(115	)
Ending balance		$2,678		$1,890

In the table above:

•

Changes in fair value are presented for loans that are classified in level 3 as of the end of the period.

•

Net unrealized gains/(losses) relates to loans that were still held at

period-end.

•

Purchases includes originations and secondary purchases.

•

Transfers between levels of the fair value hierarchy are reported at the beginning of the reporting period in which they occur. If a loan was transferred to level 3 during a reporting period, its entire gain or loss for the period is classified in level 3.

The table below presents information, by loan type, for loans included in the summary table above.



	Year Ended December

$ in millions		2020		2019
Corporate
Beginning balance		$ 752		$ 659
Net realized gains/(losses)		22		5
Net unrealized gains/(losses)		(22	)	(27	)
Purchases		277		151
Sales		(38	)	–
Settlements		(125	)	(298	)
Transfers into level 3		163		290
Transfers out of level 3		(100	)	(28	)
Ending balance		$ 929		$ 752
Commercial real estate
Beginning balance		$ 591		$ 677
Net realized gains/(losses)		24		20
Net unrealized gains/(losses)		60		28
Purchases		334		11
Sales		(5	)	(9	)
Settlements		(366	)	(229	)
Transfers into level 3		466		94
Transfers out of level 3		–		(1	)
Ending balance		$1,104		$ 591
Residential real estate
Beginning balance		$ 221		$ 290
Net realized gains/(losses)		13		15
Net unrealized gains/(losses)		10		26
Purchases		48		58
Sales		(2	)	(5	)
Settlements		(78	)	(137	)
Transfers into level 3		48		60
Transfers out of level 3		–		(86	)
Ending balance		$ 260		$ 221
Wealth management and other
Beginning balance		$ 326		$ 364
Net realized gains/(losses)		13		6
Net unrealized gains/(losses)		39		58
Purchases		11		29
Sales		(5	)	–
Settlements		(158	)	(131	)
Transfers into level 3		159		–
Ending balance		$ 385		$ 326

Level 3 Rollforward Commentary

Year Ended December 2020.

The net realized and unrealized gains on level 3 loans of $159 million (reflecting $72 million of net realized gains and $87 million of net unrealized gains) for 2020 included gains of $135 million reported in other principal transactions and $24 million reported in interest income.

The drivers of the net unrealized gains on level 3 loans for 2020 were not material.

Transfers into level 3 loans during 2020 reflected transfers of certain loans backed by commercial real estate, corporate loans, and wealth management and other loans from level 2, principally due to reduced price transparency as a result of a lack of market evidence, including fewer market transactions in these instruments.

Transfers out of level 3 loans during 2020 reflected transfers of certain corporate loans to level 2, principally due to duration and yield inputs no longer being significant to the valuation of these loans and increased price transparency as a result of increased market evidence, including market transactions in these instruments.

Year Ended December 2019.

The net realized and unrealized gains on level 3 loans of $131 million (reflecting $46 million of net realized gains and $85 million of net unrealized gains) for 2019 included gains of $98 million reported in other principal transactions and $33 million reported in interest income.

The drivers of the net unrealized gains on level 3 loans for 2019 were not material.

Transfers into level 3 loans during 2019 primarily reflected transfers of certain corporate loans from level 2, principally due to reduced price transparency as a result of a lack of market evidence.

The drivers of transfers out of level 3 loans during 2019 were not material.

Estimated Fair Value

The table below presents the estimated fair value of loans that are not accounted for at fair value and in what level of the fair value hierarchy they would have been classified if they had been included in the firm’s fair value hierarchy.



	Carrying Value		Estimated Fair Value
$ in millions	Carrying Value			Level 2	Level 3	Total
As of December 2020
Amortized cost		$99,691		$52,793	$48,512	$101,305
Held for sale		$ 2,799		$ 1,541	$ 1,271	$ 2,812
As of December 2019
Amortized cost		$89,195		$52,091	$37,095	$ 89,186
Held for sale		$ 5,323		$ 4,157	$ 1,252	$ 5,409