days or more, unless the loan is both well-secured and in the process of collection. A loan is determined to be past due when the minimum payment is not received from the borrower by the contractually specified due date or for certain loans (e.g., residential real estate loans), when a monthly payment is due and unpaid for 30 days or more. Finally, collateral-dependent loans are typically maintained on nonaccrual status.

On the date a loan is placed on nonaccrual status, all interest accrued but not collected is reversed against interest income. In addition, the amortization of deferred amounts is suspended. Interest income on nonaccrual loans may be recognized as cash interest payments are received (i.e., on a cash basis) if the recorded loan balance is deemed fully collectible; however, if there is doubt regarding the ultimate collectibility of the recorded loan balance, all interest cash receipts are applied to reduce the carrying value of the loan (the cost recovery method). For consumer loans, application of this policy typically results in the Firm recognizing interest income on nonaccrual consumer loans on a cash basis.

A loan may be returned to accrual status when repayment is reasonably assured and there has been demonstrated performance under the terms of the loan or, if applicable, the terms of the restructured loan.

As permitted by regulatory guidance, credit card loans are generally exempt from being placed on nonaccrual status; accordingly, interest and fees related to credit card loans continue to accrue until the loan is charged off or paid in full.

Allowance for loan losses

The allowance for loan losses represents the estimated expected credit losses in the held-for-investment loan portfolio at the balance sheet date and is recognized on the balance sheet as a contra asset, which brings the amortized cost to the net carrying value. Changes in the allowance for loan losses are recorded in the provision for credit losses on the Firm’s Consolidated statements of income. Refer to Note 13 for further information on the Firm’s accounting policies for the allowance for loan losses.

Charge-offs

Consumer loans are generally charged off or charged down to the lower of the amortized cost or the net realizable value of the underlying collateral (i.e., fair value less estimated costs to sell), with an offset to the allowance for loan losses, upon reaching specified stages of delinquency in accordance with standards established by the FFIEC. Residential real estate loans, unmodified credit card loans and scored business banking loans are generally charged off no later than 180 days past due. Scored auto and modified credit card loans are charged off no later than 120 days past due.

Certain consumer loans are charged off or charged down to their net realizable value earlier than the FFIEC charge-off standards in certain circumstances as follows:

•Loans modified in a TDR that are determined to be collateral-dependent.

•Loans to borrowers who have experienced an event that suggests a loss is either known or highly certain are subject to accelerated charge-off standards (e.g., residential real estate and auto loans are charged off or charged down within 60 days of receiving notification of a bankruptcy filing).

•Auto loans upon repossession of the automobile.

Other than in certain limited circumstances, the Firm typically does not recognize charge-offs on the government-guaranteed portion of loans.

Wholesale loans are charged off when it is highly certain that a loss has been realized. The determination of whether to recognize a charge-off includes many factors, including the prioritization of the Firm’s claim in bankruptcy, expectations of the workout/restructuring of the loan and valuation of the borrower’s equity or the loan collateral.

When a loan is charged down to the lower of its amortized cost or the estimated net realizable value of the underlying collateral, the determination of the fair value of the collateral depends on the type of collateral (e.g., securities, real estate). In cases where the collateral is in the form of liquid securities, the fair value is based on quoted market prices or broker quotes. For illiquid securities or other financial assets, the fair value of the collateral is generally estimated using a discounted cash flow model.

For residential real estate loans, collateral values are based upon external valuation sources. When it becomes likely that a borrower is either unable or unwilling to pay, the Firm utilizes a broker’s price opinion, appraisal and/or an automated valuation model of the home based on an exterior-only valuation (“exterior opinions”), which is then updated at least every twelve months, or more frequently depending on various market factors. As soon as practicable after the Firm receives the property in satisfaction of a debt (e.g., by taking legal title or physical possession), the Firm generally obtains an appraisal based on an inspection that includes the interior of the home (“interior appraisals”). Exterior opinions and interior appraisals are discounted based upon the Firm’s experience with actual liquidation values as compared with the estimated values provided by exterior opinions and interior appraisals, considering state-specific factors.

For commercial real estate loans, collateral values are generally based on appraisals from internal and external valuation sources. Collateral values are typically updated every six to twelve months, either by obtaining a new appraisal or by performing an internal analysis, in accordance with the Firm’s policies. The Firm also considers both borrower- and market-specific factors, which may

result in obtaining appraisal updates or broker price opinions at more frequent intervals.

Loans held-for-sale

Loans held-for-sale are measured at the lower of cost or fair value, with valuation changes recorded in noninterest revenue. For consumer loans, the valuation is performed on a portfolio basis. For wholesale loans, the valuation is performed on an individual loan basis.

Interest income on loans held-for-sale is accrued and recognized based on the contractual rate of interest.

Loan origination fees or costs and purchase price discounts or premiums are deferred in a contra loan account until the related loan is sold. The deferred fees or costs and discounts or premiums are an adjustment to the basis of the loan and therefore are included in the periodic determination of the lower of cost or fair value adjustments and/or the gain or loss recognized at the time of sale.

Because these loans are recognized at the lower of cost or fair value, the Firm’s allowance for loan losses and charge-off policies do not apply to these loans. However, loans held-for-sale are subject to the nonaccrual policies described above.

Loans at fair value

Loans for which the fair value option has been elected are measured at fair value, with changes in fair value recorded in noninterest revenue.

Interest income on these loans is accrued and recognized based on the contractual rate of interest. Changes in fair value are recognized in noninterest revenue. Loan origination fees are recognized upfront in noninterest revenue. Loan origination costs are recognized in the associated expense category as incurred.

Because these loans are recognized at fair value, the Firm’s allowance for loan losses and charge-off policies do not apply to these loans. However, loans at fair value are subject to the nonaccrual policies described above.

Refer to Note 3 for further information on the Firm’s elections of fair value accounting under the fair value option. Refer to Note 2 and Note 3 for further information on loans carried at fair value and classified as trading assets.

Loan classification changes

Loans in the held-for-investment portfolio that management decides to sell are transferred to the held-for-sale portfolio at the lower of cost or fair value on the date of transfer. Credit-related losses are charged against the allowance for loan losses; non-credit related losses such as those due to changes in interest rates or foreign currency exchange rates are recognized in noninterest revenue.

In the event that management decides to retain a loan in the held-for-sale portfolio, the loan is transferred to the held-for-investment portfolio at amortized cost on the date of transfer. These loans are subsequently assessed for impairment based on the Firm’s allowance methodology. Refer to Note 13 for a further discussion of the methodologies used in establishing the Firm’s allowance for loan losses.

Loan modifications

The Firm seeks to modify certain loans in conjunction with its loss mitigation activities. Through the modification, JPMorgan Chase grants one or more concessions to a borrower who is experiencing financial difficulty in order to minimize the Firm’s economic loss and avoid foreclosure or repossession of the collateral, and to ultimately maximize payments received by the Firm from the borrower. The concessions granted vary by program and by borrower-specific characteristics, and may include interest rate reductions, term extensions, payment delays, principal forgiveness, or the acceptance of equity or other assets in lieu of payments. Such modifications are accounted for and reported as TDRs. Loans with short-term and other insignificant modifications that are not considered concessions are not TDRs.

Loans, except for credit card loans, modified in a TDR are generally placed on nonaccrual status, although in many cases such loans were already on nonaccrual status prior to modification. These loans may be returned to performing status (the accrual of interest is resumed) if the following criteria are met: (i) the borrower has performed under the modified terms for a minimum of six months and/or six payments, and (ii) the Firm has an expectation that repayment of the modified loan is reasonably assured based on, for example, the borrower’s debt capacity and level of future earnings, collateral values, LTV ratios, and other current market considerations. In certain limited and well-defined circumstances in which the loan is current at the modification date, such loans are not placed on nonaccrual status at the time of modification.

Loans modified in TDRs are generally measured for impairment using the Firm’s established asset-specific allowance methodology, which considers the expected re-default rates for the modified loans. A loan modified in a TDR generally remains subject to the asset-specific component of the allowance throughout its remaining life, regardless of whether the loan is performing and has been returned to accrual status. Refer to Note 13 for further

discussion of the methodology used to estimate the Firm’s asset-specific allowance.

The Firm has granted various forms of assistance to customers and clients impacted by the COVID-19 pandemic, including payment deferrals and covenant modifications. The majority of the Firm’s COVID-19 related loan modifications have not been considered TDRs because:

•they represent short-term or other insignificant modifications, whether under the Firm’s regular loan modification assessments or as permitted by regulatory guidance, or

•the Firm has elected to apply the option to suspend the application of accounting guidance for TDRs as provided by the CARES Act and extended by the Consolidated Appropriations Act.

To the extent that certain modifications do not meet any of the above criteria, the Firm accounts for them as TDRs.

As permitted by regulatory guidance, the Firm does not place loans with deferrals granted due to COVID-19 on nonaccrual status where such loans are not otherwise reportable as nonaccrual. The Firm considers expected losses of principal and accrued interest associated with all COVID-19 related loan modifications in its allowance for credit losses.

Assistance provided in response to the COVID-19 pandemic could delay the recognition of delinquencies, nonaccrual status, and net charge-offs for those customers who would have otherwise moved into past due or nonaccrual status.

Foreclosed property

The Firm acquires property from borrowers through loan restructurings, workouts, and foreclosures. Property acquired may include real property (e.g., residential real estate, land, and buildings) and commercial and personal property (e.g., automobiles, aircraft, railcars, and ships).

The Firm recognizes foreclosed property upon receiving assets in satisfaction of a loan (e.g., by taking legal title or physical possession). For loans collateralized by real property, the Firm generally recognizes the asset received at foreclosure sale or upon the execution of a deed in lieu of foreclosure transaction with the borrower. Foreclosed assets are reported in other assets on the Consolidated balance sheets and initially recognized at fair value less estimated costs to sell. Each quarter the fair value of the acquired property is reviewed and adjusted, if necessary, to the lower of cost or fair value. Subsequent adjustments to fair value are charged/credited to noninterest revenue. Operating expense, such as real estate taxes and maintenance, are charged to other expense.

In response to the COVID-19 pandemic, the Firm has temporarily suspended certain foreclosure activities. This could delay recognition of foreclosed properties until the foreclosure moratoriums are lifted.

Loan portfolio

The Firm’s loan portfolio is divided into three portfolio segments, which are the same segments used by the Firm to determine the allowance for loan losses: Consumer, excluding credit card; Credit card; and Wholesale. Within each portfolio segment the Firm monitors and assesses the credit risk in the following classes of loans, based on the risk characteristics of each loan class.

In conjunction with the adoption of CECL, the Firm revised its loan classes. Prior-period amounts have been revised to conform with the current presentation:

•The consumer, excluding credit card portfolio segment’s residential mortgage and home equity loans and lending-related commitments have been combined into a residential real estate class.

•Upon adoption of CECL, the Firm elected to discontinue the pool-level accounting for PCI loans and to account for these loans on an individual loan basis. PCI loans are considered PCD loans under CECL and are subject to the Firm’s nonaccrual and charge-off policies. PCD loans are now reported in the consumer, excluding credit card portfolio segment’s residential real estate class.

•Risk-rated business banking and auto dealer loans and lending-related commitments held in CCB were reclassified from the consumer, excluding credit card portfolio segment, to the wholesale portfolio segment, to align with the methodology applied when determining the allowance. The remaining scored auto and business banking loans and lending-related commitments have been combined into an auto and other class.

•The wholesale portfolio segment’s classes, previously based on the borrower’s primary business activity, have been revised to align with the loan classifications as defined by the bank regulatory agencies, based on the loan’s collateral, purpose, and type of borrower.

Consumer, excluding

credit card

Credit card

Wholesale(c)

• Residential real estate(a)

• Auto and other(b)

• Credit card loans

• Secured by real estate

• Commercial and industrial

• Other(d)

(a)Includes scored mortgage and home equity loans held in CCB and AWM, and scored mortgage loans held in CIB and Corporate.

(b)Includes scored auto and business banking loans and overdrafts.

(c)Includes loans held in CIB, CB, AWM, Corporate as well as risk-rated business banking and auto dealer loans held in CCB for which the wholesale methodology is applied when determining the allowance for loan losses.

(d)Includes loans to financial institutions, states and political subdivisions, SPEs, nonprofits, personal investment companies and trusts, as well as loans to individuals and individual entities (predominantly Wealth Management clients within AWM). Refer to Note 14 for more information on SPEs.

The following tables summarize the Firm’s loan balances by portfolio segment.


December 31, 2020	Consumer, excluding credit card			Credit card			Wholesale			Total(b)(c)
(in millions)
Retained		$	302,127		$	143,432		$	514,947		$	960,506
Held-for-sale		1,305			784			5,784			7,873
At fair value(a)		15,147			—			29,327			44,474
Total		$	318,579		$	144,216		$	550,058		$	1,012,853

December 31, 2019	Consumer, excluding credit card				Credit card			Wholesale			Total(b)(c)
(in millions)
Retained		$	294,999		$	168,924		$	481,678		$	945,601
Held-for-sale		3,002			—			4,062			7,064
At fair value(a)		19,816			—			25,139			44,955
Total		$	317,817		$	168,924		$	510,879		$	997,620

(a)In the third quarter of 2020, the Firm reclassified certain fair value option elected lending-related positions from trading assets to loans. Prior-period amounts have been revised to conform with the current presentation.

(b)Excludes $2.9 billion of accrued interest receivables at both December 31, 2020 and 2019. The Firm wrote off accrued interest receivables of $121 million and $50 million for the years ended December 31, 2020 and 2019, respectively.

(c)Loans (other than those for which the fair value option has been elected) are presented net of unamortized discounts and premiums and net deferred loan fees or costs. These amounts were not material as of December 31, 2020 and 2019.

The following tables provide information about the carrying value of retained loans purchased, sold and reclassified to held-for-sale during the periods indicated. Loans that were reclassified to held-for-sale and sold in a subsequent period are excluded from the sales line of this table.


		2020
Year ended December 31, (in millions)	Consumer, excluding credit card				Credit card			Wholesale			Total
Purchases		$	3,474	(b)(c)		$	—		$	1,159		$	4,633
Sales		352				—			17,916			18,268
Retained loans reclassified to held-for-sale(a)		2,084				787			1,580			4,451


		2019
Year ended December 31, (in millions)	Consumer, excluding credit card				Credit card			Wholesale			Total
Purchases		$	1,282	(b)(c)		$	—		$	1,291		$	2,573
Sales		30,474				—			23,445			53,919
Retained loans reclassified to held-for-sale(a)		9,188				—			2,371			11,559


		2018
Year ended December 31, (in millions)	Consumer, excluding credit card				Credit card			Wholesale			Total
Purchases		$	2,543	(b)(c)		$	—		$	2,354		$	4,897
Sales		9,984				—			16,741			26,725
Retained loans reclassified to held-for-sale(a)		36				—			2,276			2,312

(a)Reclassifications of loans to held-for-sale are non-cash transactions.

(b)Predominantly includes purchases of residential real estate loans, including the Firm’s voluntary repurchases of certain delinquent loans from loan pools as permitted by Government National Mortgage Association (“Ginnie Mae”) guidelines for the years ended December 31, 2020, 2019 and 2018. The Firm typically elects to repurchase these delinquent loans as it continues to service them and/or manage the foreclosure process in accordance with applicable requirements of Ginnie Mae, FHA, RHS, and/or VA.

(c)Excludes purchases of retained loans sourced through the correspondent origination channel and underwritten in accordance with the Firm’s standards. Such purchases were $15.3 billion, $16.6 billion and $18.6 billion for the years ended December 31, 2020, 2019 and 2018, respectively.

Gains and losses on sales of loans

Net gains/(losses) on sales of loans and lending-related commitments (including adjustments to record loans and lending-related commitments held-for-sale at the lower of cost or fair value) recognized in noninterest revenue was $(43) million for the year ended December 31, 2020 of which $(36) million was related to loans. Net gains on sales of loans was $394 million for the year ended December 31, 2019. Gains and losses on sales of loans was not material for the year ended December 31, 2018. In addition, the sale of loans may also result in write downs, recoveries or changes in the allowance recognized in the provision for credit losses.

Consumer, excluding credit card loan portfolio

Consumer loans, excluding credit card loans, consist primarily of scored residential mortgages, home equity loans and lines of credit, auto and business banking loans, with a focus on serving the prime consumer credit market. The portfolio also includes home equity loans secured by junior liens, prime mortgage loans with an interest-only payment period, and certain payment-option loans that may result in negative amortization.

The following table provides information about retained consumer loans, excluding credit card, by class.


December 31, (in millions)	2020		2019
Residential real estate	$	225,302	$	243,317
Auto and other(a)	76,825		51,682
Total retained loans	$	302,127	$	294,999

(a)At December 31, 2020, included $19.2 billion of loans in Business Banking under the PPP.

Delinquency rates are the primary credit quality indicator for consumer loans. Loans that are more than 30 days past due provide an early warning of borrowers who may be experiencing financial difficulties and/or who may be unable or unwilling to repay the loan. As the loan continues to age, it becomes more clear whether the borrower is likely to be unable or unwilling to pay. In the case of residential real estate loans, late-stage delinquencies (greater than 150 days past due) are a strong indicator of loans that will ultimately result in a foreclosure or similar liquidation transaction. In addition to delinquency rates, other credit quality indicators for consumer loans vary based on the class of loan, as follows:

•For residential real estate loans, the current estimated LTV ratio, or the combined LTV ratio in the case of junior lien loans, is an indicator of the potential loss severity in the event of default. Additionally, LTV or combined LTV ratios can provide insight into a borrower’s continued willingness to pay, as the delinquency rate of high-LTV loans tends to be greater than that for loans where the borrower has equity in the collateral. The geographic distribution of the loan collateral also provides insight as to the credit quality of the portfolio, as factors such as the regional economy, home price changes and specific events such as natural disasters, will affect credit quality. The borrower’s current or “refreshed” FICO score is a secondary credit quality indicator for certain loans, as FICO scores are an indication of the borrower’s credit payment history. Thus, a loan to a borrower with a low FICO score (less than 660) is considered to be of higher risk than a loan to a borrower with a higher FICO score. Further, a loan to a borrower with a high LTV ratio and a low FICO score is at greater risk of default than a loan to a borrower that has both a high LTV ratio and a high FICO score.

•For scored auto and business banking loans, geographic distribution is an indicator of the credit performance of the portfolio. Similar to residential real estate loans, geographic distribution provides insights into the portfolio performance based on regional economic activity and events.

Residential real estate

The following table provides information on delinquency, which is the primary credit quality indicator for retained residential real estate loans.


(in millions, except ratios)	December 31, 2020																											December 31, 2019
	Term loans by origination year																		Revolving loans						Total			Total
	2020			2019			2018			2017			2016			Prior to 2016			Within the revolving period			Converted to term loans
Loan delinquency(a)(b)
Current	$	55,562		$	31,820		$	13,900		$	20,410		$	27,978		$	50,232		$	7,370		$	15,792		$	223,064		$	239,979
30–149 days past due	9			25			20			22			29			674			21			245			1,045			1,910
150 or more days past due	3			14			10			18			18			844			22			264			1,193			1,428
Total retained loans	$	55,574		$	31,859		$	13,930		$	20,450		$	28,025		$	51,750		$	7,413		$	16,301		$	225,302		$	243,317
% of 30+ days past due to total retained loans(c)	0.02		%	0.12		%	0.22		%	0.20		%	0.17		%	2.86		%	0.58		%	3.12		%	0.98		%	1.35		%

(a)Individual delinquency classifications include mortgage loans insured by U.S. government agencies as follows: current included $36 million and $17 million; 30–149 days past due included $16 million and $20 million; and 150 or more days past due included $24 million and $26 million at December 31, 2020 and 2019, respectively.

(b)At December 31, 2020, loans under payment deferral programs offered in response to the COVID-19 pandemic which are still within their deferral period and performing according to their modified terms are generally not considered delinquent.

(c)At December 31, 2020 and 2019, residential real estate loans excluded mortgage loans insured by U.S. government agencies of $40 million and $46 million, respectively, that are 30 or more days past due. These amounts have been excluded based upon the government guarantee.

Approximately 35% of the total revolving loans are senior lien loans; the remaining balance are junior lien loans. The lien position the Firm holds is considered in the Firm’s allowance for credit losses. Revolving loans that have been converted to term loans have higher delinquency rates than those that are still within the revolving period. That is primarily because the fully-amortizing payment that is generally required for those products is higher than the minimum payment options available for revolving loans within the revolving period.

Nonaccrual loans and other credit quality indicators

The following table provides information on nonaccrual and other credit quality indicators for retained residential real estate loans.


(in millions, except weighted-average data)	December 31, 2020			December 31, 2019
Nonaccrual loans(a)(b)(c)(d)(e)	$	5,313		$	2,780
90 or more days past due and government guaranteed(f)	33			38

Current estimated LTV ratios(g)(h)
Greater than 125% and refreshed FICO scores:
Equal to or greater than 660	$	10		$	31
Less than 660	18			38
101% to 125% and refreshed FICO scores:
Equal to or greater than 660	72			134
Less than 660	65			132
80% to 100% and refreshed FICO scores:
Equal to or greater than 660	2,365			5,953
Less than 660	435			764
Less than 80% and refreshed FICO scores:
Equal to or greater than 660	208,457			219,469
Less than 660	12,072			14,681
No FICO/LTV available	1,732			2,052
U.S. government-guaranteed	76			63
Total retained loans	$	225,302		$	243,317

Weighted average LTV ratio(g)(i)	54		%	55		%
Weighted average FICO(h)(i)	763			758

Geographic region(j)
California	$	73,444		$	82,147
New York	32,287			31,996
Florida	13,981			13,668
Texas	13,773			14,474
Illinois	13,130			15,587
Colorado	8,235			8,447
Washington	7,917			8,990
New Jersey	7,227			7,752
Massachusetts	5,784			6,210
Connecticut	5,024			4,954
All other(k)	44,500			49,092
Total retained loans	$	225,302		$	243,317

(a)Includes collateral-dependent residential real estate loans that are charged down to the lower of amortized cost or the fair value of the underlying collateral less costs to sell. The Firm reports, in accordance with regulatory guidance, residential real estate loans that have been discharged under Chapter 7 bankruptcy and not reaffirmed by the borrower (“Chapter 7 loans”) as collateral-dependent nonaccrual TDRs, regardless of their delinquency status. At December 31, 2020, approximately 7% of Chapter 7 residential real estate loans were 30 days or more past due, respectively.

(b)At December 31, 2020, nonaccrual loans included $1.6 billion of PCD loans. Prior to the adoption of CECL, nonaccrual loans excluded PCI loans as the Firm recognized interest income on each pool of PCI loans as each of the pools was performing.

(c)Generally, all consumer nonaccrual loans have an allowance. In accordance with regulatory guidance, certain nonaccrual loans that are considered collateral-dependent have been charged down to the lower of amortized cost or the fair value of their underlying collateral less costs to sell. If the value of the underlying collateral improves subsequent to the charge down, the related allowance may be negative.

(d)Interest income on nonaccrual loans recognized on a cash basis was $161 million and $166 million for the years ended December 31, 2020 and 2019, respectively.

(e)Generally excludes loans under payment deferral programs offered in response to the COVID-19 pandemic. Includes loans to customers that have exited COVID-19 payment deferral programs and are 90 or more days past due, predominantly all of which were also at least 150 days past due and therefore considered collateral-dependent. Collateral-dependent loans are charged down to the lower of amortized cost or fair value of the underlying collateral less costs to sell.

(f)These balances are excluded from nonaccrual loans as the loans are guaranteed by U.S government agencies. Typically the principal balance of the loans is insured and interest is guaranteed at a specified reimbursement rate subject to meeting agreed-upon servicing guidelines. At December 31, 2020 and 2019, these balances included $33 million and $34 million, respectively, of loans that are no longer accruing interest based on the agreed-upon servicing guidelines. For the remaining balance, interest is being accrued at the guaranteed reimbursement rate. There were no loans that were not guaranteed by U.S. government agencies that are 90 or more days past due and still accruing interest at December 31, 2020 and 2019.

(g)Represents the aggregate unpaid principal balance of loans divided by the estimated current property value. Current property values are estimated, at a minimum, quarterly, based on home valuation models using nationally recognized home price index valuation estimates incorporating actual data to the extent available and forecasted data where actual data is not available. Current estimated combined LTV for junior lien home equity loans considers all available lien positions, as well as unused lines, related to the property.

(h)Refreshed FICO scores represent each borrower’s most recent credit score, which is obtained by the Firm on at least a quarterly basis.

(i)Excludes loans with no FICO and/or LTV data available.

(j)The geographic regions presented in the table are ordered based on the magnitude of the corresponding loan balances at December 31, 2020.

(k)At December 31, 2020 and 2019, included mortgage loans insured by U.S. government agencies of $76 million and $63 million, respectively. These amounts have been excluded from the geographic regions presented based upon the government guarantee.

Loan modifications

Modifications of residential real estate loans, where the Firm grants concessions to borrowers who are experiencing financial difficulty are generally accounted for and reported as TDRs. Loans with short-term or other insignificant modifications that are not considered concessions are not TDRs nor are loans for which the Firm has elected to apply the option to suspend the application of accounting guidance for TDRs as provided by the CARES Act and extended by the Consolidated Appropriations Act. The carrying value of new TDRs was $819 million, $490 million and $736 million for the years ended December 31, 2020, 2019 and 2018, respectively. There were no additional commitments to lend to borrowers whose residential real estate loans have been modified in TDRs.

Nature and extent of modifications

The Firm’s proprietary modification programs as well as government programs, including U.S. GSE programs, generally provide various concessions to financially troubled borrowers including, but not limited to, interest rate reductions, term or payment extensions and delays of principal and/or interest payments that would otherwise have been required under the terms of the original agreement.

The following table provides information about how residential real estate loans were modified in TDRs under the Firm’s loss mitigation programs described above during the periods presented. This table excludes Chapter 7 loans where the sole concession granted is the discharge of debt, loans with short-term or other insignificant modifications that are not considered concessions, and loans for which the Firm has elected to apply the option to suspend the application of accounting guidance for TDRs as provided by the CARES Act and extended by the Consolidated Appropriations Act.



Year ended December 31,	2020		2019		2018
Number of loans approved for a trial modification	5,522		5,872		7,175
Number of loans permanently modified	6,850		4,918		7,853
Concession granted:(a)
Interest rate reduction	50	%	77	%	54	%
Term or payment extension	49		71		62
Principal and/or interest deferred	14		13		29
Principal forgiveness	2		5		7
Other(b)	66		63		51

(a)Represents concessions granted in permanent modifications as a percentage of the number of loans permanently modified. The sum of the percentages exceeds 100% because predominantly all of the modifications include more than one type of concession. Concessions offered on trial modifications are generally consistent with those granted on permanent modifications.

(b)Includes variable interest rate to fixed interest rate modifications and payment delays that meet the definition of a TDR for the years ended December 31, 2020, 2019 and 2018.

Financial effects of modifications and redefaults

The following table provides information about the financial effects of the various concessions granted in modifications of residential real estate loans under the loss mitigation programs described above and about redefaults of certain loans modified in TDRs for the periods presented. The following table presents only the financial effects of permanent modifications and do not include temporary concessions offered through trial modifications. This table also excludes Chapter 7 loans where the sole concession granted is the discharge of debt, loans with short-term or other insignificant modifications that are not considered concessions, and loans for which the Firm has elected to apply the option to suspend the application of accounting guidance for TDRs as provided by the CARES Act and extended by the Consolidated Appropriations Act.


Year ended December 31, (in millions, except weighted - average data)	2020			2019			2018
Weighted-average interest rate of loans with interest rate reductions – before TDR	5.09		%	5.68		%	5.50		%
Weighted-average interest rate of loans with interest rate reductions – after TDR	3.28			3.81			3.60
Weighted-average remaining contractual term (in years) of loans with term or payment extensions – before TDR	22			20			21
Weighted-average remaining contractual term (in years) of loans with term or payment extensions – after TDR	39			39			38
Charge-offs recognized upon permanent modification	$	5		$	1		$	2
Principal deferred	16			19			30
Principal forgiven	5			7			17
Balance of loans that redefaulted within one year of permanent modification(a)	$	199		$	166		$	161

(a)Represents loans permanently modified in TDRs that experienced a payment default in the periods presented, and for which the payment default occurred within one year of the modification. The dollar amounts presented represent the balance of such loans at the end of the reporting period in which such loans defaulted. For residential real estate loans modified in TDRs, payment default is deemed to occur when the loan becomes two contractual payments past due. In the event that a modified loan redefaults, it will generally be liquidated through foreclosure or another similar type of liquidation transaction. Redefaults of loans modified within the last twelve months may not be representative of ultimate redefault levels.

At December 31, 2020, the weighted-average estimated remaining lives of residential real estate loans permanently modified in TDRs were 6 years. The estimated remaining lives of these loans reflect estimated prepayments, both voluntary and involuntary (i.e., foreclosures and other forced liquidations).

Active and suspended foreclosure

At December 31, 2020 and 2019, the Firm had residential real estate loans, excluding those insured by U.S. government agencies, with a carrying value of $846 million and $1.2 billion, respectively, that were not included in REO, but were in the process of active or suspended foreclosure.

Auto and other

The following table provides information on delinquency, which is the primary credit quality indicator for retained auto and other consumer loans.


	December 31, 2020																												December 31, 2019
(in millions, except ratios)	Term Loans by origination year																			Revolving loans
	2020				2019			2018			2017			2016			Prior to 2016			Within the revolving period			Converted to term loans			Total			Total
Loan delinquency(a)
Current	$	46,169		(b)	$	12,829		$	7,367		$	4,521		$	2,058		$	742		$	2,517		$	158		$	76,361		$	51,005
30–119 days past due	97				107			77			53			42			23			30			17			446			667
120 or more days past due	—				—			—			1			—			1			8			8			18			10
Total retained loans	$	46,266			$	12,936		$	7,444		$	4,575		$	2,100		$	766		$	2,555		$	183		$	76,825		$	51,682
% of 30+ days past due to total retained loans	0.21		%		0.83		%	1.03		%	1.18		%	2.00		%	3.13		%	1.49		%	13.66		%	0.60		%	1.31		%

(a)At December 31, 2020, loans under payment deferral programs offered in response to the COVID-19 pandemic which are still within their deferral period and performing according to their modified terms are generally not considered delinquent.

(b)At December 31, 2020, included $19.2 billion of loans in Business Banking under the PPP. PPP loans are guaranteed by the SBA. Other than in certain limited circumstances, the Firm typically does not recognize charge-offs, classify as nonaccrual nor record an allowance for loan losses on these loans.

Nonaccrual and other credit quality indicators

The following table provides information on nonaccrual and other credit quality indicators for retained auto and other consumer loans.


(in millions, except ratios)	Total Auto and other
	December 31, 2020		December 31, 2019
Nonaccrual loans(a)(b)(c)	151		146

Geographic region(d)
California	$	12,302	$	7,795
New York	8,824		3,706
Texas	8,235		5,457
Florida	4,668		3,025
Illinois	3,768		2,443
New Jersey	2,646		1,798
Arizona	2,465		1,347
Ohio	2,163		1,490
Pennsylvania	1,924		1,721
Colorado	1,910		1,247
All other	27,920		21,653
Total retained loans	$	76,825	$	51,682

(a)There were no loans that were 90 or more days past due and still accruing interest at December 31, 2020 and 2019.

(b)All nonaccrual auto and other consumer loans generally have an allowance. Certain nonaccrual loans that are considered collateral-dependent have been charged down to the lower of amortized cost or the fair value of their underlying collateral less costs to sell. If the value of the underlying collateral improves subsequent to the charge down, the related allowance may be negative.

(c)Interest income on nonaccrual loans recognized on a cash basis was not material for the years ended December 31, 2020 and 2019.

(d)The geographic regions presented in this table are ordered based on the magnitude of the corresponding loan balances at December 31, 2020.

Loan modifications

Certain other consumer loan modifications are considered to be TDRs as they provide various concessions to borrowers who are experiencing financial difficulty. Loans with short-term or other insignificant modifications that are not considered concessions are not TDRs.

The impact of these modifications, as well as new TDRs, were not material to the Firm for the years ended December 31, 2020, 2019 and 2018. Additional commitments to lend to borrowers whose loans have been modified in TDRs as of December 31, 2020 and 2019 were not material.

Credit card loan portfolio

The credit card portfolio segment includes credit card loans originated and purchased by the Firm. Delinquency rates are the primary credit quality indicator for credit card loans as they provide an early warning that borrowers may be experiencing difficulties (30 days past due); information on those borrowers that have been delinquent for a longer period of time (90 days past due) is also considered. In addition to delinquency rates, the geographic distribution of the loans provides insight as to the credit quality of the portfolio based on the regional economy.

While the borrower’s credit score is another general indicator of credit quality, the Firm does not view credit scores as a primary indicator of credit quality because the borrower’s credit score tends to be a lagging indicator. The

distribution of such scores provides a general indicator of credit quality trends within the portfolio; however, the score does not capture all factors that would be predictive of future credit performance. Refreshed FICO score information, which is obtained at least quarterly, for a statistically significant random sample of the credit card portfolio is indicated in the following table. FICO is considered to be the industry benchmark for credit scores.

The Firm generally originates new card accounts to prime consumer borrowers. However, certain cardholders’ FICO scores may decrease over time, depending on the performance of the cardholder and changes in the credit score calculation.

The following table provides information on delinquency, which is the primary credit quality indicator for retained credit card loans.


(in millions, except ratios)	December 31, 2020									December 31, 2019
	Within the revolving period			Converted to term loans(b)			Total			Total
Loan delinquency(a)
Current and less than 30 days past due and still accruing	$	139,783		$	1,239		$	141,022		$	165,767
30–89 days past due and still accruing	997			94			1,091			1,550
90 or more days past due and still accruing	1,277			42			1,319			1,607
Total retained loans	$	142,057		$	1,375		$	143,432		$	168,924
Loan delinquency ratios
% of 30+ days past due to total retained loans	1.60		%	9.89		%	1.68		%	1.87		%
% of 90+ days past due to total retained loans	0.90			3.05			0.92			0.95

(b)Represents TDRs.

Other credit quality indicators

The following table provides information on other credit quality indicators for retained credit card loans.


(in millions, except ratios)	December 31, 2020			December 31, 2019
Geographic region(a)
California	$	20,921		$	25,783
Texas	14,544			16,728
New York	11,919			14,544
Florida	9,562			10,830
Illinois	8,006			9,579
New Jersey	5,927			7,165
Ohio	4,673			5,406
Pennsylvania	4,476			5,245
Colorado	4,092			4,763
Michigan	3,553			4,164
All other	55,759			64,717
Total retained loans	$	143,432		$	168,924
Percentage of portfolio based on carrying value with estimated refreshed FICO scores
Equal to or greater than 660	85.9		%	84.0		%
Less than 660	13.9			15.4
No FICO available	0.2			0.6

(a)The geographic regions presented in the table are ordered based on the magnitude of the corresponding loan balances at December 31, 2020.

Loan modifications

The Firm may offer one of a number of loan modification programs granting concessions to credit card borrowers who are experiencing financial difficulty. The Firm grants concessions for most of the credit card loans under long-term programs. These modifications involve placing the customer on a fixed payment plan, generally for 60 months, and typically include reducing the interest rate on the credit card. Substantially all modifications under the Firm’s long-term programs are considered to be TDRs. Loans with short-term or other insignificant modifications that are not considered concessions are not TDRs.

If the cardholder does not comply with the modified payment terms, then the credit card loan continues to age and will ultimately be charged-off in accordance with the Firm’s standard charge-off policy. In most cases, the Firm does not reinstate the borrower’s line of credit.

Financial effects of modifications and redefaults

The following table provides information about the financial effects of the concessions granted on credit card loans modified in TDRs and redefaults for the periods presented. For all periods disclosed, new enrollments were less than 1% of total retained credit card loans.


Year ended December 31, (in millions, except weighted-average data)	2020			2019			2018
Balance of new TDRs(a)	$	818		$	961		$	866
Weighted-average interest rate of loans – before TDR	18.04		%	19.07		%	17.98		%
Weighted-average interest rate of loans – after TDR	4.64			4.70			5.16
Balance of loans that redefaulted within one year of modification(b)	$	110		$	148		$	116

(a)Represents the outstanding balance prior to modification.

(b)Represents loans modified in TDRs that experienced a payment default in the periods presented, and for which the payment default occurred within one year of the modification. The amounts presented represent the balance of such loans as of the end of the quarter in which they defaulted.

For credit card loans modified in TDRs, payment default is deemed to have occurred when the borrower misses two consecutive contractual payments. Defaulted modified credit card loans remain in the modification program and continue to be charged off in accordance with the Firm’s standard charge-off policy.

Wholesale loan portfolio

Wholesale loans include loans made to a variety of clients, ranging from large corporate and institutional clients to high-net-worth individuals.

The primary credit quality indicator for wholesale loans is the internal risk rating assigned to each loan. Risk ratings are used to identify the credit quality of loans and differentiate risk within the portfolio. Risk ratings on loans consider the PD and the LGD. The PD is the likelihood that a loan will default. The LGD is the estimated loss on the loan that would be realized upon the default of the borrower and takes into consideration collateral and structural support for each credit facility.

Management considers several factors to determine an appropriate internal risk rating, including the obligor’s debt capacity and financial flexibility, the level of the obligor’s earnings, the amount and sources for repayment, the level and nature of contingencies, management strength, and the industry and geography in which the obligor operates. The Firm’s internal risk ratings generally align with the qualitative characteristics (e.g., borrower capacity to meet financial commitments and vulnerability to changes in the economic environment) defined by S&P and Moody’s, however the quantitative characteristics (e.g., PD and LGD) may differ as they reflect internal historical experiences and assumptions. The Firm generally considers internal ratings with qualitative characteristics equivalent to BBB-/Baa3 or higher as investment grade, and these ratings have a lower PD and/or lower LGD than non-investment grade ratings.

Noninvestment-grade ratings are further classified as noncriticized and criticized, and the criticized portion is further subdivided into performing and nonaccrual loans, representing management’s assessment of the collectibility of principal and interest. Criticized loans have a higher PD than noncriticized loans. The Firm’s definition of criticized aligns with the U.S. banking regulatory definition of criticized exposures, which consist of special mention, substandard and doubtful categories.

Risk ratings are reviewed on a regular and ongoing basis by Credit Risk Management and are adjusted as necessary for updated information affecting the obligor’s ability to fulfill its obligations.

As noted above, the risk rating of a loan considers the industry in which the obligor conducts its operations. As part of the overall credit risk management framework, the Firm focuses on the management and diversification of its industry and client exposures, with particular attention paid to industries with actual or potential credit concern. Refer to Note 4 for further detail on industry concentrations.

The following tables provide information on internal risk rating, which is the primary credit quality indicator for retained wholesale loans.


December 31, (in millions, except ratios)	Secured by real estate						Commercial and industrial							Other(b)						Total retained loans
	2020			2019			2020				2019			2020			2019			2020				2019
Loans by risk ratings
Investment-grade	$	90,147		$	96,611		$	71,917		(a)	$	80,489		$	217,209		$	186,344		$	379,273		(a)	$	363,444
Noninvestment- grade:
Noncriticized	26,129			22,493			57,870				60,437			33,053			27,591			117,052				110,521
Criticized performing	3,234			1,131			10,991				4,399			1,079			1,126			15,304				6,656
Criticized nonaccrual	483			183			1,931				844			904			30			3,318				1,057
Total noninvestment- grade	29,846			23,807			70,792				65,680			35,036			28,747			135,674				118,234
Total retained loans	$	119,993		$	120,418		$	142,709			$	146,169		$	252,245		$	215,091		$	514,947			$	481,678
% of investment-grade to total retained loans	75.13		%	80.23		%	50.39		%		55.07		%	86.11		%	86.63		%	73.65		%		75.45		%
% of total criticized to total retained loans	3.10			1.09			9.05				3.59			0.79			0.54			3.62				1.60
% of criticized nonaccrual to total retained loans	0.40			0.15			1.35				0.58			0.36			0.01			0.64				0.22


	Secured by real estate
(in millions)	December 31, 2020																		December 31, 2019
	Term loans by origination year												Revolving loans
	2020		2019		2018		2017		2016		Prior to 2016		Within the revolving period		Converted to term loans		Total		Total
Loans by risk ratings
Investment-grade	$	16,560	$	19,575	$	12,192	$	11,017	$	13,439	$	16,266	$	1,098	$	—	$	90,147	$	96,611
Noninvestment-grade	3,327		4,339		4,205		2,916		2,575		11,994		489		1		29,846		23,807
Total retained loans	$	19,887	$	23,914	$	16,397	$	13,933	$	16,014	$	28,260	$	1,587	$	1	$	119,993	$	120,418


	Commercial and industrial
(in millions)	December 31, 2020																			December 31, 2019
	Term loans by origination year													Revolving loans
	2020			2019		2018		2017		2016		Prior to 2016		Within the revolving period		Converted to term loans		Total		Total
Loans by risk ratings
Investment-grade	$	21,211	(a)	$	7,304	$	2,934	$	1,748	$	1,032	$	1,263	$	36,424	$	1	$	71,917	$	80,489
Noninvestment-grade	15,060			8,636		5,131		2,104		497		2,439		36,852		73		70,792		65,680
Total retained loans	$	36,271		$	15,940	$	8,065	$	3,852	$	1,529	$	3,702	$	73,276	$	74	$	142,709	$	146,169


	Other(b)
(in millions)	December 31, 2020																		December 31, 2019
	Term loans by origination year												Revolving loans
	2020		2019		2018		2017		2016		Prior to 2016		Within the revolving period		Converted to term loans		Total		Total
Loans by risk ratings
Investment-grade	$	31,389	$	10,169	$	6,994	$	6,206	$	3,553	$	12,595	$	145,524	$	779	$	217,209	$	186,344
Noninvestment-grade	5,009		2,220		1,641		550		146		636		24,710		124		35,036		28,747
Total retained loans	$	36,398	$	12,389	$	8,635	$	6,756	$	3,699	$	13,231	$	170,234	$	903	$	252,245	$	215,091

(a)At December 31, 2020, included $8.0 billion of loans under the PPP, of which $7.4 billion is included in commercial and industrial. PPP loans are guaranteed by the SBA and considered investment-grade. Other than in certain limited circumstances, the Firm typically does not recognize charge-offs, classify as nonaccrual nor record an allowance for loan losses on these loans.

(b)Includes loans to financial institutions, states and political subdivisions, SPEs, nonprofits, personal investment companies and trusts, as well as loans to individuals and individual entities (predominantly Wealth Management clients within AWM). Refer to Note 14 for more information on SPEs.

The following table presents additional information on retained loans secured by real estate within the Wholesale portfolio, which consists of loans secured wholly or substantially by a lien or liens on real property at origination. Multifamily lending includes financing for acquisition, leasing and construction of apartment buildings. Other commercial lending largely includes financing for acquisition, leasing and construction, largely for office, retail and industrial real estate. Included in secured by real estate loans is $6.4 billion and $6.3 billion as of December 31, 2020 and 2019, respectively, of construction and development loans made to finance land development and on-site construction of commercial, industrial, residential, or farm buildings.


December 31, (in millions, except ratios)	Multifamily						Other Commercial						Total retained loans secured by real estate
	2020			2019			2020			2019			2020			2019
Retained loans secured by real estate	$	73,078		$	73,840		$	46,915		$	46,578		$	119,993		$	120,418
Criticized	1,144			340			2,573			974			3,717			1,314
% of total criticized to total retained loans secured by real estate	1.57		%	0.46		%	5.48		%	2.09		%	3.10		%	1.09		%
Criticized nonaccrual	$	56		$	28		$	427		$	155		$	483		$	183
% of criticized nonaccrual loans to total retained loans secured by real estate	0.08		%	0.04		%	0.91		%	0.33		%	0.40		%	0.15		%

The following table provides additional information about retained wholesale loans, including geographic distribution, delinquency and net charge-offs.


	Secured by real estate						Commercial and industrial						Other						Total retained loans
December 31, (in millions)	2020			2019			2020			2019			2020			2019			2020			2019
Loans by geographic distribution(a)
Total U.S.	$	116,990		$	117,836		$	109,273		$	111,954		$	180,583		$	150,512		$	406,846		$	380,302
Total non-U.S.	3,003			2,582			33,436			34,215			71,662			64,579			108,101			101,376
Total retained loans	$	119,993		$	120,418		$	142,709		$	146,169		$	252,245		$	215,091		$	514,947		$	481,678
Loan delinquency(b)
Current and less than 30 days past due and still accruing	$	118,894		$	120,119		$	140,100		$	144,839		$	249,713		$	214,641		$	508,707		$	479,599
30–89 days past due and still accruing	601			115			658			449			1,606			415			2,865			979
90 or more days past due and still accruing(c)	15			1			20			37			22			5			57			43
Criticized nonaccrual	483			183			1,931			844			904			30			3,318			1,057
Total retained loans	$	119,993		$	120,418		$	142,709		$	146,169		$	252,245		$	215,091		$	514,947		$	481,678
Net charge-offs/(recoveries)	$	10		$	44		$	737		$	335		$	52		$	36		$	799		$	415
% of net charge-offs/(recoveries) to end-of-period retained loans	0.01		%	0.04		%	0.52		%	0.23		%	0.02		%	0.02		%	0.16		%	0.09		%

(a)The U.S. and non-U.S. distribution is determined based predominantly on the domicile of the borrower.

(b)The credit quality of wholesale loans is assessed primarily through ongoing review and monitoring of an obligor’s ability to meet contractual obligations rather than relying on the past due status, which is generally a lagging indicator of credit quality. Generally excludes loans under payment deferral programs offered in response to the COVID-19 pandemic.

(c)Represents loans that are considered well-collateralized and therefore still accruing interest.

Nonaccrual loans

The following table provides information on retained wholesale nonaccrual loans.


December 31, (in millions)	Secured by real estate				Commercial and industrial				Other				Total retained loans
	2020		2019		2020		2019		2020		2019		2020		2019
Nonaccrual loans(a)
With an allowance	$	351	$	169	$	1,667	$	688	$	800	$	28	$	2,818	$	885
Without an allowance(b)	132		14		264		156		104		2		500		172
Total nonaccrual loans(c)	$	483	$	183	$	1,931	$	844	$	904	$	30	$	3,318	$	1,057

(a)Loans that were modified in response to the COVID-19 pandemic continue to be risk-rated in accordance with the Firm’s overall credit risk management framework. As of December 31, 2020, predominantly all of these loans were considered performing.

(b)When the discounted cash flows, collateral value or market price equals or exceeds the amortized cost of the loan, the loan does not require an allowance. This typically occurs when the loans have been partially charged off and/or there have been interest payments received and applied to the loan balance.

(c)Interest income on nonaccrual loans recognized on a cash basis were not material for the years ended December 31, 2020 and 2019.

Loan modifications

Certain loan modifications are considered to be TDRs as they provide various concessions to borrowers who are experiencing financial difficulty. Loans with short-term or other insignificant modifications that are not considered concessions are not TDRs nor are loans for which the Firm has elected to apply the option to suspend the application of accounting guidance for TDRs as provided by the CARES Act and extended by the Consolidated Appropriations Act. The carrying value of TDRs was $954 million and $501 million as of December 31, 2020 and 2019, respectively. The carrying value of new TDRs was $734 million, $407 million and $718 million for the years ended December 31, 2020, 2019 and 2018, respectively. The impact of these modifications, as well as new TDRs, were not material to the Firm for the years ended December 31, 2020, 2019 and 2018.