Allowance for credit losses

JPMorgan Chase’s allowance for loan losses covers the consumer, including credit card, portfolio segments (primarily scored); and wholesale (risk-rated) portfolio, and represents management’s estimate of probable credit losses inherent in the Firm’s loan portfolio. The allowance for loan losses includes an asset-specific component, a formula-based component and a component related to PCI loans, as described below. Management also estimates an allowance for wholesale and consumer lending-related commitments using methodologies similar to those used to estimate the allowance on the underlying loans. During 2014, the Firm did not make any significant changes to the methodologies or policies used to determine its allowance for credit losses; such policies are described in the following paragraphs.

The asset-specific component of the allowance relates to loans considered to be impaired, which includes loans that have been modified in TDRs as well as risk-rated loans that have been placed on nonaccrual status. To determine the asset-specific component of the allowance, larger loans are evaluated individually, while smaller loans are evaluated as pools using historical loss experience for the respective class of assets. Scored loans (i.e., consumer loans) are pooled by product type, while risk-rated loans (primarily wholesale loans) are segmented by risk rating.

The Firm generally measures the asset-specific allowance as the difference between the recorded investment in the loan and the present value of the cash flows expected to be collected, discounted at the loan’s original effective interest rate. Subsequent changes in impairment are reported as an adjustment to the provision for loan losses. In certain cases, the asset-specific allowance is determined using an observable market price, and the allowance is measured as the difference between the recorded investment in the loan and the loan’s fair value. Impaired collateral-dependent loans are charged down to the fair value of collateral less costs to sell and therefore may not be subject to an asset-specific reserve as are other impaired loans. See Note 14 for more information about charge-offs and collateral-dependent loans.

The asset-specific component of the allowance for impaired loans that have been modified in TDRs incorporates the effects of foregone interest, if any, in the present value calculation and also incorporates the effect of the modification on the loan’s expected cash flows, which considers the potential for redefault. For residential real estate loans modified in TDRs, the Firm develops product-specific probability of default estimates, which are applied at a loan level to compute expected losses. In developing these probabilities of default, the Firm considers the relationship between the credit quality characteristics of the underlying loans and certain assumptions about home prices and unemployment, based upon industry-wide data. The Firm also considers its own historical loss experience to date based on actual redefaulted modified loans. For credit card loans modified in TDRs, expected losses incorporate projected redefaults based on the Firm’s historical experience by type of modification program. For wholesale loans modified in TDRs, expected losses incorporate redefaults based on management’s expectation of the borrower’s ability to repay under the modified terms.

The formula-based component is based on a statistical calculation to provide for incurred credit losses in performing risk-rated loans and all consumer loans, except for any loans restructured in TDRs and PCI loans. See Note 14 for more information on PCI loans.

For scored loans, the statistical calculation is performed on pools of loans with similar risk characteristics (e.g., product type) and generally computed by applying loss factors to outstanding principal balances over an estimated loss emergence period. The loss emergence period represents the time period between the date at which the loss is estimated to have been incurred and the ultimate realization of that loss (through a charge-off). Estimated loss emergence periods may vary by product and may change over time; management applies judgment in estimating loss emergence periods, using available credit information and trends.

Loss factors are statistically derived and sensitive to changes in delinquency status, credit scores, collateral values and other risk factors. The Firm uses a number of different forecasting models to estimate both the PD and the loss severity, including delinquency roll rate models and credit loss severity models. In developing PD and loss severity assumptions, the Firm also considers known and anticipated changes in the economic environment, including changes in home prices, unemployment rates and other risk indicators.

A nationally recognized home price index measure is used to estimate both the PD and the loss severity on residential real estate loans at the metropolitan statistical areas (“MSA”) level. Loss severity estimates are regularly validated by comparison to actual losses recognized on defaulted loans, market-specific real estate appraisals and property sales activity. The economic impact of potential modifications of residential real estate loans is not included in the statistical calculation because of the uncertainty regarding the type and results of such modifications.

For risk-rated loans, the statistical calculation is the product of an estimated PD and an estimated LGD. These factors are differentiated by risk rating and expected maturity. In assessing the risk rating of a particular loan, among the factors considered are 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. These factors are based on an evaluation of historical and current information, and involve subjective assessment and interpretation. Emphasizing one factor over another or considering additional factors could impact the risk rating assigned by the Firm to that loan. PD estimates are based on observable external through-the-cycle data, using credit-rating agency default statistics. LGD estimates are based on the Firm’s history of actual credit losses over more than one credit cycle. Estimates of PD and LGD are subject to periodic refinement based on changes to underlying external and Firm-specific historical data.

Management applies judgment within an established framework to adjust the results of applying the statistical calculation described above. The determination of the appropriate adjustment is based on management’s view of loss events that have occurred but that are not yet reflected in the loss factors and that relate to current macroeconomic and political conditions, the quality of underwriting standards and other relevant internal and external factors affecting the credit quality of the portfolio. For the scored loan portfolios, adjustments to the statistical calculation are made in part by analyzing the historical loss experience for each major product segment. Factors related to unemployment, home prices, borrower behavior and lien position, the estimated effects of the mortgage foreclosure-related settlement with federal and state officials and uncertainties regarding the ultimate success of loan modifications are incorporated into the calculation, as appropriate. For junior lien products, management considers the delinquency and/or modification status of any senior liens in determining the adjustment. In addition, for the risk-rated portfolios, any adjustments made to the statistical calculation take into consideration model imprecision, deteriorating conditions within an industry, product or portfolio type, geographic location, credit concentration, and current economic events that have occurred but that are not yet reflected in the factors used to derive the statistical calculation.

Management establishes an asset-specific allowance for lending-related commitments that are considered impaired and computes a formula-based allowance for performing consumer and wholesale lending-related commitments. These are computed using a methodology similar to that used for the wholesale loan portfolio, modified for expected maturities and probabilities of drawdown.

Determining the appropriateness of the allowance is complex and requires judgment by management about the effect of matters that are inherently uncertain. Subsequent evaluations of the loan portfolio, in light of the factors then prevailing, may result in significant changes in the allowances for loan losses and lending-related commitments in future periods. At least quarterly, the allowance for credit losses is reviewed by the Chief Risk Officer, the Chief Financial Officer and the Controller of the Firm and discussed with the Risk Policy and Audit Committees of the Board of Directors of the Firm. As of December 31, 2014, JPMorgan Chase deemed the allowance for credit losses to be appropriate (i.e., sufficient to absorb probable credit losses inherent in the portfolio).

Allowance for credit losses and loans and lending-related commitments by impairment methodology

The table below summarizes information about the allowance for loan losses, loans by impairment methodology, the allowance for lending-related commitments and lending-related commitments by impairment methodology.



	2014
Year ended December 31, (in millions)	Consumer, excluding credit card			Credit card				Wholesale			Total
Allowance for loan losses
Beginning balance at January 1,	$	8,456		$	3,795			$	4,013		$	16,264
Gross charge-offs	2,132			3,831				151			6,114
Gross recoveries	(814		)	(402		)		(139		)	(1,355		)
Net charge-offs/(recoveries)	1,318			3,429				12			4,759
Write-offs of PCI loans^(a)	533			—				—			533
Provision for loan losses	414			3,079				(269		)	3,224
Other	31			(6		)		(36		)	(11		)
Ending balance at December 31,	$	7,050		$	3,439			$	3,696		$	14,185

Allowance for loan losses by impairment methodology
Asset-specific^(b)	$	539		$	500		^(c)	$	87		$	1,126
Formula-based	3,186			2,939				3,609			9,734
PCI	3,325			—				—			3,325
Total allowance for loan losses	$	7,050		$	3,439			$	3,696		$	14,185

Loans by impairment methodology
Asset-specific	$	12,020		$	2,029			$	637		$	14,686
Formula-based	236,263			125,998				323,861			686,122
PCI	46,696			—				4			46,700
Total retained loans	$	294,979		$	128,027			$	324,502		$	747,508

Impaired collateral-dependent loans
Net charge-offs	$	133		$	—			$	21		$	154
Loans measured at fair value of collateral less cost to sell	3,025			—				326			3,351

Allowance for lending-related commitments
Beginning balance at January 1,	$	8		$	—			$	697		$	705
Provision for lending-related commitments	5			—				(90		)	(85		)
Other	—			—				2			2
Ending balance at December 31,	$	13		$	—			$	609		$	622

Allowance for lending-related commitments by impairment methodology
Asset-specific	$	—		$	—			$	60		$	60
Formula-based	13			—				549			562
Total allowance for lending-related commitments	$	13		$	—			$	609		$	622

Lending-related commitments by impairment methodology
Asset-specific	$	—		$	—			$	103		$	103
Formula-based	58,153			525,963				471,953			1,056,069
Total lending-related commitments	$	58,153		$	525,963			$	472,056		$	1,056,172

(a)

Write-offs of PCI loans are recorded against the allowance for loan losses when actual losses for a pool exceed estimated losses that were recorded as purchase accounting adjustments at the time of acquisition. A write-off of a PCI loan is recognized when the underlying loan is removed from a pool (e.g., upon liquidation). During the fourth quarter of 2014, the Firm recorded a $291 million adjustment to reduce the PCI allowance and the recorded investment in the Firm’s PCI loan portfolio, primarily reflecting the cumulative effect of interest forgiveness modifications. This adjustment had no impact to the Firm’s Consolidated statements of income.


(b)	Includes risk-rated loans that have been placed on nonaccrual status and loans that have been modified in a TDR.


(c)	The asset-specific credit card allowance for loan losses is related to loans that have been modified in a TDR; such allowance is calculated based on the loans’ original contractual interest rates and does not consider any incremental penalty rates.



(table continued from previous page)
2013													2012
Consumer, excluding credit card			Credit card				Wholesale			Total			Consumer, excluding credit card			Credit card				Wholesale			Total

$	12,292		$	5,501			$	4,143		$	21,936		$	16,294		$	6,999			$	4,316		$	27,609
2,754			4,472				241			7,467			4,805			5,755				346			10,906
(847		)	(593		)		(225		)	(1,665		)	(508		)	(811		)		(524		)	(1,843		)
1,907			3,879				16			5,802			4,297			4,944				(178		)	9,063
53			—				—			53			—			—				—			—
(1,872		)	2,179				(119		)	188			302			3,444				(359		)	3,387
(4		)	(6		)		5			(5		)	(7		)	2				8			3
$	8,456		$	3,795			$	4,013		$	16,264		$	12,292		$	5,501			$	4,143		$	21,936


$	601		$	971		^(c)	$	181		$	1,753		$	729		$	1,681		^(c)	$	319		$	2,729
3,697			2,824				3,832			10,353			5,852			3,820				3,824			13,496
4,158			—				—			4,158			5,711			—				—			5,711
$	8,456		$	3,795			$	4,013		$	16,264		$	12,292		$	5,501			$	4,143		$	21,936


$	13,785		$	3,115			$	845		$	17,745		$	13,938		$	4,762			$	1,475		$	20,175
221,609			124,350				307,412			653,371			218,945			123,231				304,728			646,904
53,055			—				6			53,061			59,737			—				19			59,756
$	288,449		$	127,465			$	308,263		$	724,177		$	292,620		$	127,993			$	306,222		$	726,835


$	235		$	—			$	37		$	272		$	973		$	—			$	77		$	1,050
3,105			—				362			3,467			3,272			—				445			3,717


$	7		$	—			$	661		$	668		$	7		$	—			$	666		$	673
1			—				36			37			—			—				(2		)	(2		)
—			—				—			—			—			—				(3		)	(3		)
$	8		$	—			$	697		$	705		$	7		$	—			$	661		$	668


$	—		$	—			$	60		$	60		$	—		$	—			$	97		$	97
8			—				637			645			7			—				564			571
$	8		$	—			$	697		$	705		$	7		$	—			$	661		$	668


$	—		$	—			$	206		$	206		$	—		$	—			$	355		$	355
56,057			529,383				446,026			1,031,466			60,156			533,018				434,459			1,027,633
$	56,057		$	529,383			$	446,232		$	1,031,672		$	60,156		$	533,018			$	434,814		$	1,027,988