hsba201502236k4.htm

Table 24: Credit risk exposure - RWA density by region

	RWA density
	Europe	Asia	MENA	North America	Latin America	Total
	%	%	%	%	%	%

IRB advanced approach	34	33	40	49	39	36
Retail:
- secured by mortgages on immovable property SME1	24	-	-	-	-	21
- secured by mortgages on immovable property non-SME	6	10	-	96	-	25
- qualifying revolving retail	20	26	-	31	-	23
- other SME1	45	-	-	50	-	45
- other non-SME	17	22	-	80	-	26
Total retail	12	14	-	90	-	25
Central governments and central banks	16	14	46	10	35	17
Institutions	38	25	28	26	67	30
Corporates2	45	59	-	54	-	52
Securitisation positions3	115	46	-	12	-	106
Non-credit obligation assets	51	14	40	77	41	26

IRB foundation approach	67	-	60	-	-	65
Central governments and central banks	-	-	-	-	-	-
Institutions	-	-	-	-	-	-
Corporates	67	-	60	-	-	65

	Europe	Asia	MENA	North America	Latin America	Total
	%	%	%	%	%	%

Standardised approach	27	67	79	108	96	60
Central governments and central banks	3	5	10	174	226	10
Institutions	76	37	43	-	-	37
Corporates	98	90	95	99	102	94
Retail	72	75	75	72	71	74
Secured by mortgages on immovable property	36	35	41	36	37	36
Exposures in default	126	128	118	143	134	129
Regional governments or local authorities	-	-	-	-	72	57
Equity	192	236	126	100	172	204
Other4	65	72	89	64	160	74


At 31 December 2014	33	43	64	54	77	43

IRB advanced approach	31	30	43	54	32	35
Retail:
- secured on real estate property	6	8	-	128	-	34
- qualifying revolving retail	21	24	-	34	-	23
- SMEs1	49	3	-	63	-	48
- other retail	21	23	-	50	-	23

Total retail	14	12	-	119	-	32
Central governments and central banks	14	13	49	10	30	16
Institutions	36	18	23	14	48	22
Corporates2	55	56	-	52	-	55
Securitisation positions3	47	40	-	15	-	44

IRB foundation approach	59	-	55	-	-	58
Corporates	59	-	55	-	-	58

Standardised approach	19	57	79	87	88	49
Central governments and central banks	-	1	1	10	-	-
Institutions	3	38	53	-	-	34
Corporates	84	89	96	89	99	91
Retail	79	75	75	78	74	76
Secured on real estate property	41	49	56	92	60	56
Past due items	122	100	124	124	141	131
Regional governments or local authorities	-	-	100	-	92	93
Equity	124	100	100	100	100	105
Other items6	61	32	69	85	64	48


At 31 December 2013	28	39	66	57	72	40

For footnotes, see page 36.

Key points

· The CRD IV requirement to report exposure gross of any cash collateral has caused a reduction in RWA density primarily within corporates in Europe as a result of the increase in exposure value.

· Retail IRB density has improved due to the continued run-off of the US CML retail mortgage portfolio resulting in an improved residual portfolio. The IRB RWA densities remain high in North America compared to other regions due to the continuing challenging conditions in the US mortgage market.

· Standardised institutions RWA density in Europe has worsened due to immaterial movements on a small portfolio (exposure is US$0.2bn).

Table 25: Credit risk exposure - by industry sector

	Exposure value
	Personal	Manu- facturing	Inter- national trade and services	Property and other business activities	Government and public admin- istration	Other commercial	Financial	Non- customer assets	Total
	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn

IRB advanced approach	404.2	140.4	149.2	181.1	113.1	88.4	464.9	52.5	1,593.8
Retail:
- secured by mortgages on immovable property SME1	0.5	-	0.2	2.4	-	-	-	-	3.1
- secured by mortgages on immovable property non-SME	288.7	-	-	0.1	-	-	0.1	-	288.9
- qualifying revolving retail	66.2	-	-	-	-	-	-	-	66.2
- other SME1	-	0.9	2.5	7.3	0.8	2.1	0.3	-	13.9
- other non-SME	47.1	-	-	-	0.2	-	-	-	47.3
Total retail	402.5	0.9	2.7	9.8	1.0	2.1	0.4	-	419.4
Central governments and central banks	-	-	0.1	-	94.7	-	232.6	-	327.4
Institutions	-	-	-	-	0.7	-	129.7	-	130.4
Corporates2	1.7	139.5	146.4	171.3	16.7	86.3	63.9	-	625.8
Securitisation positions3	-	-	-	-	-	-	38.3	-	38.3
Non-credit obligation assets	-	-	-	-	-	-	-	52.5	52.5

IRB foundation approach	0.2	8.9	6.0	1.5	0.5	4.9	3.8	-	25.8
Central governments and central banks	-	-	-	-	-	-	0.1	-	0.1
Institutions	-	-	-	-	-	-	0.1	-	0.1
Corporates	0.2	8.9	6.0	1.5	0.5	4.9	3.6	-	25.6

Standardised approach	88.0	63.0	52.0	46.2	89.0	44.0	187.7	20.6	590.5
Central governments or central banks	-	-	-	-	62.4	-	119.3	7.6	189.3
Institutions	-	-	-	-	-	-	30.1	-	30.1
Corporates	5.4	61.6	49.4	42.3	22.2	41.9	17.3	-	240.1
Retail	43.9	0.7	1.5	1.0	0.2	0.4	0.2	-	47.9
Secured by mortgages on immovable property	36.8	0.1	0.1	1.5	-	0.1	-	-	38.6
Exposures in default	1.9	0.6	0.8	0.6	0.1	0.6	0.1	-	4.7
Regional governments or local authorities	-	-	-	-	0.8	-	0.3	-	1.1
Equity	-	-	-	0.4	-	-	3.8	9.0	13.2
Other4	-	-	0.2	0.4	3.3	1.0	16.6	4.0	25.5


At 31 December 2014	492.4	212.3	207.2	228.8	202.6	137.3	656.4	73.1	2,210.1

	Exposure value
	Personal	Manu- facturing	Inter- national trade and services	Property and other business activities	Government and public admin- istration	Other commercial	Financial	Non- customer assets	Total
	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn

IRB advanced approach	426.7	118.9	113.8	151.7	107.2	73.8	476.7	-	1,468.8
Retail:
- secured on real estate property	310.7	-	-	-	-	-	-	-	310.7
- qualifying revolving retail	66.9	-	-	-	-	-	-	-	66.9
- SMEs1	-	0.9	1.7	14.2	0.4	0.9	0.5	-	18.6
- other retail	46.7	-	-	-	0.1	-	-	-	46.8

Total retail	424.3	0.9	1.7	14.2	0.5	0.9	0.5	-	443.0
Central governments and central banks	-	-	-	-	90.4	0.2	251.1	-	341.7
Institutions	-	-	-	-	0.2	-	129.8	-	130.0
Corporates2	2.4	118.0	112.1	137.5	16.1	72.7	49.9	-	508.7
Securitisation positions3	-	-	-	-	-	-	45.4	-	45.4

IRB foundation approach	-	8.6	5.9	1.1	0.4	4.2	3.4	-	23.6
Corporates	-	8.6	5.9	1.1	0.4	4.2	3.4	-	23.6

Standardised approach	89.4	58.9	50.7	44.0	81.0	46.2	238.8	58.7	667.7
Central governments and central banks	-	-	-	-	56.9	-	163.1	-	220.0
Institutions	-	-	-	-	-	-	35.2	-	35.2
Corporates	3.2	57.5	47.4	35.1	21.1	44.1	13.4	-	221.8
Retail	42.5	1.0	1.9	1.2	0.2	0.6	0.3	-	47.7
Secured on real estate property	41.3	0.1	1.1	7.0	-	0.9	-	-	50.4
Past due items	2.4	0.3	0.3	0.4	0.1	0.6	-	-	4.1
Regional governments or local authorities	-	-	-	-	0.8	-	-	-	0.8
Equity	-	-	-	-	-	-	3.3	-	3.3
Other items6	-	-	-	0.3	1.9	-	23.5	58.7	84.4


At 31 December 2013	516.1	186.4	170.4	196.8	188.6	124.2	718.9	58.7	2,160.1

For footnotes, see page 36.

Key points

· The CRD IV requirement to report exposure gross of any cash collateral has resulted in an increase of exposure value across a number of industries.

· The decrease in exposure to financial is primarily driven by reduced central bank exposures under both the standardised approach and the IRB advanced approach due primarily to lower deposits with central banks.

· Higher corporate lending, including term and trade-related lending, has led to an increase of exposure across industry sectors, primarily manufacturing, property and other business services and other commercial within corporate under the IRB advanced approach.

· The decrease in personal sector is primarily due to the continued run-off of the US CML retail mortgage portfolio.

Table 26: Credit risk exposure - by residual maturity

	Exposure value
	Less than 1 year	Between 1 and 5 years	More than 5 years	Undated	Total	RWAs
	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn
IRB advanced approach	729.1	382.5	429.8	52.4	1,593.8	581.6
Retail:
- secured by mortgages on immovable property SME1	0.1	0.2	2.8	-	3.1	0.6
- secured by mortgages on immovable property non-SME1	2.9	4.1	281.9	-	288.9	71.6
- qualifying revolving retail	66.2	-	-	-	66.2	15.3
- other SME1	3.3	7.0	3.6	-	13.9	6.2
- other non-SME	13.8	12.7	20.8	-	47.3	12.4

Total retail	86.3	24.0	309.1	-	419.4	106.1
Central governments and central banks	212.7	80.2	34.5	-	327.4	54.1
Institutions	100.9	25.4	4.1	-	130.4	38.7
Corporates2	318.6	247.1	60.1	-	625.8	328.5
Securitisation positions3	10.6	5.7	22.0	-	38.3	40.7
Non-credit obligation assets	-	0.1	-	52.4	52.5	13.5

IRB foundation approach	10.5	12.9	2.4	-	25.8	16.8
Central governments and central banks	-	0.1	-	-	0.1	-
Institutions	-	0.1	-	-	0.1	-
Corporates	10.5	12.7	2.4	-	25.6	16.8

Standardised approach	242.1	201.6	116.8	30.0	590.5	356.9
Central governments and central banks	123.5	37.7	20.5	7.6	189.3	19.7
Institutions	16.2	0.9	13.0	-	30.1	11.2
Corporates	70.2	142.6	27.2	0.1	240.1	224.7
Retail	17.1	12.8	18.0	-	47.9	35.2
Secured by mortgages on immovable property	1.9	3.0	33.7	-	38.6	13.8
Exposures in default	2.2	1.3	1.2	-	4.7	6.1
Regional governments or local authorities	0.4	0.3	0.4	-	1.1	0.6
Equity	-	-	-	13.2	13.2	26.9
Other4	10.6	3.0	2.8	9.1	25.5	18.7


At 31 December 2014	981.7	597.0	549.0	82.4	2,210.1	955.3

IRB advanced approach	642.5	405.0	421.3	-	1,468.8	521.2
Retail:
- secured on real estate property	2.8	5.0	302.9	-	310.7	105.4
- qualifying revolving retail	66.9	-	-	-	66.9	15.4
- SMEs1	3.8	8.7	6.1	-	18.6	8.9
- other retail	7.0	23.1	16.7	-	46.8	11.0

Total retail	80.5	36.8	325.7	-	443.0	140.7
Central governments and central banks	206.4	106.1	29.2	-	341.7	53.0
Institutions	99.1	29.9	1.0	-	130.0	28.0
Corporates2	223.1	230.6	55.0	-	508.7	279.7
Securitisation positions3	33.4	1.6	10.4	-	45.4	19.8

IRB foundation approach	10.6	11.5	1.5	-	23.6	13.6
Corporates	10.6	11.5	1.5	-	23.6	13.6

Standardised approach	248.0	233.5	101.2	85.0	667.7	329.5
Central governments and central banks	154.9	50.4	14.7	-	220.0	0.7
Institutions	17.9	4.3	13.0	-	35.2	12.1
Corporates	53.7	146.7	21.2	0.2	221.8	202.1
Retail	15.7	19.6	12.4	-	47.7	36.1
Secured on real estate property	2.7	9.2	38.5	-	50.4	28.4
Past due items	2.4	1.0	0.7	-	4.1	5.4
Regional governments or local authorities	0.3	0.1	0.4	-	0.8	0.8
Equity	-	-	-	3.3	3.3	3.5
Other items5	0.4	2.2	0.3	81.5	84.4	40.4


At 31 December 2013	901.1	650.0	524.0	85.0	2,160.1	864.3

For footnotes, see page 36

Key points

· The impact of foreign exchange movements on exposure value is discussed under table 22.

· The CRD IV requirement to report exposure gross of any cash collateral has resulted in an increase of exposure value primarily in the Less than 1 year and Between 1 and 5 years bandings.

· The impact of higher corporate lending, including term and trade-related lending, is reflected in the increase in exposure value, primarily in the Less than 1 year banding.

· The reduction in the Other Undated residual maturity banding is driven by the reclassification of Non Credit Obligation Assets to be reported separately under IRB approach. See table 9 for further information on CRD IV impacts.

· The reduction in deposits with central banks in Europe in Central governments and central banks under the standardised approach is reflected primarily in the Less than 1 year banding.

· Sale of government bonds in North America in Central governments and central banks under the IRB approach has resulted in a decrease in the Between 1 and 5 years banding.

· The decrease in retail IRB approach in the More than 5 years banding is primarily due to the continued run-off of the US CML retail mortgage portfolio in North America.

Application of the IRB approach

The narrative explanations that follow relate to the IRB approaches: advanced and foundation IRB for distinct customers and advanced IRB for the portfolio-managed retail business. Details of our use of the standardised approach can be found on page 70.

Our Group IRB credit risk rating framework incorporates obligor propensity to default expressed in PD, and loss severity in the event of default expressed in EAD and LGD. These measures are used to calculate regulatory expected loss ('EL') and capital requirements. They are also used with other inputs to inform rating assessments for the purpose of credit approval and many other management decisions.

Use of internal estimates

PDs, LGDs, and EAD applied in the calculation of regulatory capital requirements are also extensively used for other purposes, for example:

· credit approval and monitoring: IRB models are used in the assessment of customer and portfolio risk in lending decisions;

· risk appetite: IRB measures are an important element in identifying risk exposure at customer, sector, and portfolio level;

· pricing: IRB parameters are used in pricing tools for new transactions and reviews; and

· economic capital and portfolio management: IRB parameters are used in the economic capital model that has been implemented across HSBC.

Roll-out of the IRB approach

With PRA permission, we have adopted the Basel advanced approach for the majority of our business. At the end of 2014, portfolios in much of Europe, Asia and North America were on advanced IRB approaches. Others remain on the standardised or foundation approaches pending the definition of local regulations or model approval in line with our Basel IRB roll-out plans, or under exemptions or exclusion from IRB treatment. Additionally, in some instances, regulators have allowed us to transition from advanced to standardised approaches for a limited number of immaterial portfolios.

In June 2014 the EBA published a consultation on the thresholds for the application of the Standardised Approach for exposures treated under permanent partial use and the IRB roll-out plan. Subject to the publication of the finalised RTS it is expected that the roll-out plan will set target thresholds for IRB rather than the advanced IRB approach specifically.

Under the advanced IRB approach, banks are allowed to develop their own empirical models to quantify required capital for credit risk. All such models developed by us, and any material changes to those models, must be approved by the PRA, subject to de minimis exceptions. Material changes are those that individually have a high impact, or where a number of small changes in aggregate have a high impact. Quantitative and qualitative materiality thresholds for these model changes are determined by CRD IV which also requires us to obtain PRA approval before implementation where these thresholds are breached.

The effectiveness of this process is monitored by the PRA through an annual review of IRB usage, focusing on the proportion of total credit risk assets for which IRB approaches are used.

Banks have experienced difficulties in adopting advanced IRB in some cases, for example in portfolios which have very low levels of default, such that the PD, LGD and EAD cannot be assessed to a sufficiently high degree of confidence due to a lack of default or loss data. Difficulties also arise in countries where the rules and requirements of the local regulator's implementation of Basel requirements are different from those of the PRA, or where the regulators have introduced capital floors and overlays to mitigate perceived model deficiencies. Tables 27 and 31 below detail several material regulatory thresholds and overlays. Whilst recognising the complexity of adopting IRB in some situations, we remain committed to working constructively with our regulators to achieve acceptable roll-out plans.

The wholesale risk rating system

This section describes how we build and operate our credit risk analytical models, and use IRB metrics, in wholesale customer business.

PDs for wholesale customer segments, that is central governments and central banks, financial institutions and corporate customers, and for certain individually assessed personal customers, are estimated using a Customer Risk Rating ('CRR') master scale of 23 grades. Of these, 21 are non-default grades representing varying degrees of strength of financial condition, and two are default grades.

The score generated by a credit risk rating model for the obligor is mapped to a corresponding PD and master-scale CRR. The CRR is then reviewed by a credit approver who, taking into account all relevant information, such as most recent events and market data, where available, makes the final decision on the rating. The rating assigned therefore reflects the approver's overall view of the obligor's credit standing and propensity to default.

The finally assigned CRR determines the applicable master-scale PD range from which the reference PD, generally the arithmetical mid-point, is used in the regulatory capital calculation.

Reviewing the initial model score, relationship managers may propose a different CRR from that indicated, where they believe this more appropriate. Such amendments may only be made through an override process and must be approved by the Credit function. Overrides for each model are recorded, and override levels are reviewed, as part of the model management process.

The CRR is assigned at obligor level, which means that separate exposures to the same obligor are generally subject to a single, consistent rating. Where unfunded credit risk mitigants such as guarantees apply, these may also influence the final assignment of a CRR to an obligor. The impact of unfunded risk mitigants is considered for IRB approaches on page 69 and for the standardised approach on page 70.

If an obligor is in default on any material credit obligation to the Group, all of the obligor's facilities from the Group are considered to be in default.

Under the IRB approach, obligors are grouped into grades that have similar PD or anticipated default frequency. The anticipated default frequency may be estimated using all relevant information at the relevant date ('Point-in-time' or 'PIT' rating system), or be free of the effects of the credit cycle ('Through-the-cycle' or 'TTC' rating system).

We generally utilise a hybrid approach of PIT and TTC. That is, while models are calibrated to long-run default rates, obligor ratings are reviewed annually, or more frequently if necessary to reflect changes in their circumstances and/or their economic operating environment.

Thus, over the economic cycle, a cycle will also appear in CRR migration. The influence of longer-term economic cycle factors implied by the model's calibration, combined with the effect of ongoing credit review, will result in long-term PDs generally above the actual default frequency during benign economic periods, but not changing so fast in a downturn. In practice, under a hybrid approach, ratings tend to be more volatile than would be the case in a pure TTC system, but less volatile than in a pure PIT one.

Moreover, our policy requires approvers to downgrade ratings on expectations but to upgrade them only on performance. Therefore, ratings will typically migrate during a downturn in response to higher perceived risks, but be upgraded more slowly in an upswing. This leads to expected defaults overall typically exceeding actual defaults.

For EAD and LGD estimation, operating entities are permitted, subject to overview by Group Risk, to use their own modelling approaches for those parameters to suit conditions in their jurisdictions. Group Risk provides co-ordination, benchmarks, and the sharing and promotion of best practice on EAD and LGD estimation.

EAD is estimated to a 12-month forward time horizon and represents the current exposure plus an estimate for future increases in exposure, taking into account such factors as available but undrawn facilities, and the realisation of contingent exposures post-default.

LGD is based on the effects of facility and collateral structure on outcomes post-default. This includes such factors as the type of client, the facility seniority, the type and value of collateral, past recovery experience and priority under law. It is expressed as a percentage of EAD.

Wholesale models

To determine credit ratings for the different types of wholesale obligor, many different models and scorecards are used for PD, LGD, and EAD; there are over 100 wholesale IRB models in use or under development within HSBC. These models may be differentiated by region, customer segment and/or customer size. For example, PD models are differentiated for all of our key customer segments, including sovereigns, financial institutions, large, medium and small sized corporates.

Global PD models have been developed for asset classes or clearly identifiable segments of asset classes where the customer relationship is managed globally, for example sovereigns, financial institutions and the largest corporate clients, typically those which operate internationally.

Local PD models, specific to a particular country, region, or sector, are developed for other obligors. This includes corporate clients when they show distinct characteristics in common in a particular geography. The most material local Corporate PD models are the UK mid-market PD model, and the Hong Kong and Asia-Pacific mid-market models.

The two major drivers of model methodology are the nature of the portfolio and the availability of internal or external data on historical defaults and risk factors. For some historically low-default portfolios, e.g. sovereign and financial institutions, a model will rely more heavily on external data and/or the input of an expert panel. By contrast, where sufficient data is available, models are built on a statistical basis, although the input of expert judgement may still form an important part of the overall model development methodology.

Most LGD and EAD models are developed according to local circumstances taking into account legal and procedural differences in the recovery and workout processes. However, our approach to EAD and LGD also encompasses global models for central governments and central banks, and for institutions, as exposures to these customer types are managed centrally by Global Risk. In 2013 the PRA required all firms to apply an LGD floor of 45% for senior unsecured exposure to sovereign entities. This floor was applied to reflect the relative paucity of loss observations across all firms in relation to these obligors. This floor is applied for the purposes of regulatory capital reporting.

In addition, the PRA has published guidance on the appropriateness of LGD models for low default portfolios generally. The PRA has determined that there should be at least 20 defaults per country per collateral type for LGD models to be approved. Where there are insufficient defaults, an LGD floor will be applied. As a result, in 2014, we were required to apply LGD floors for our banks portfolio and some Asia corporate portfolios where there are insufficient loss observations.

In the same guidance, the PRA also indicated that it considered income producing real estate to be an asset

class that would be difficult to model. As a result, we have migrated to the supervisory slotting approach for our UK CRE portfolio and have migrated our US Income Producing CRE portfolio on to the standardised approach.

Local models for the corporate exposure class are developed using various data inputs, including collateral information and geography (for LGD) and product type (for EAD). The most material corporate models are the UK and Asia models, all of which are developed using more than 10 years' worth of data. The LGD models are calibrated to a period of credit stress or downturn in economic conditions. The global LGD models for sovereigns and for banks reflect the expected increase in observed losses during an economic downturn period.

None of the EAD models are calibrated for a downturn, as analysis shows that utilisation decreases during a downturn because credit stress is accompanied by more intensive limit monitoring and facility reduction.

Table 27 below sets out the key characteristics of the significant wholesale credit risk models that drive the capital calculation split by Basel wholesale asset class, with their associated RWAs, including the number of models for each component, the model method or approach and the number of years of loss data used.

Table 27: Wholesale IRB credit risk models

Basel asset classes measured	RWAs for associated asset class US$bn	Compo-nent	Number of significant models	Model description and methodology	Number of years loss data

Central governments and central banks	54.1	PD	1	A constrained expert judgement model using a combination of expert judgement and quantitative analysis. The model inputs include macro-economic and political factors.	7
		LGD	1	An unsecured model built on assessment of structural factors that influence country's long term economic performance. Floor of 45%, applied as required by the PRA.	7
		EAD	1	Because of limited internal default experience and sparse historical data on utilisations and limits, the model was developed based on a combination of expert judgement and similar exposure types.	7

Institutions	38.7	PD	1	The model is a combination of expert judgement and statistical analysis. The model inputs include balance sheet information, country risk factors and qualitative data.	9
		LGD	1	Regression model that produces a downturn LGD and expected LGD. Inputs include collateral and country risk data. Floor of 45%, applied as required by the PRA.	9
		EAD	1	Regression based model that predicts Credit Conversion Factors taking into account current utilisation, available headroom, product type, and committed/uncommitted indicator.	9

Corporates1	322.3

Global large corporates		PD	1	Even though the portfolio is low-default, the model is statistically built and calibrated on 15 years of data. The inputs include balance sheet information, market data, macroeconomic and country risk indicators and qualitative factors.	>10
Other corporates		PD	5	Corporates that fall below the Global large corporate threshold are rated through local PD models, which reflect regional circumstances. These models use balance sheet data, behavioural data and qualitative information to derive a statistically built PD.	>10
All corporates		LGD	3	Local statistical models covering all corporates including Global large corporates developed using various data inputs, including collateral information, recoveries and geography.	>7
		EAD	3	Local statistical models developed using various data inputs, including product type and geography.	>7

1 Excludes specialised lending exposures subject to supervisory slotting approach (see table 29).

Table 28 below sets out Basel metrics broken down by region, table 30 shows the same metrics, however these are broken down by CRR band. Table 29 sets out an

analysis of those exposures to which a supervisory slotting approach is applied.

Table 28: Wholesale IRB portfolio analysis

	Europe	Asia	MENA	North America	Latin America	Total
	%	%	%	%	%	%
At 31 December 2014
Exposure weighted average PD
IRB advanced approach
Central governments and central banks	0.09	0.09	1.23	0.01	0.57	0.17
Institutions	0.66	0.22	0.55	0.13	0.76	0.36
Corporates1	2.62	1.44	0.09	1.26	-	1.85

IRB foundation approach
Central governments and central banks	-	-	0.04	-	-	0.04
Institutions	0.13	-	0.03	-	-	0.10
Corporates1	1.36	-	2.86	-	-	1.74

Exposure weighted average LGD
IRB advanced approach
Central governments and central banks	45.0	45.0	45.0	45.4	45.0	45.1
Institutions	35.3	45.3	39.8	40.6	45.4	42.0
Corporates1	25.8	44.3	13.7	36.6	-	35.6

1 Excludes specialised lending exposures subject to supervisory slotting approach (see table 29).

Table 29: Wholesale IRB exposures under the Slotting Approach

	Remaining maturity less than 2.5 years		Remaining maturity greater than 2.5 years		Total
	Exposure value	RWAs	Exposure value	RWAs	Exposure value	RWAs
	US$bn	US$bn	US$bn	US$bn	US$bn	US$bn
Supervisory Category
Category 1 - Strong	7.0	3.4	9.7	6.7	16.7	10.1
Category 2 - Good	4.4	3.1	3.7	3.2	8.1	6.3
Category 3 - Satisfactory	1.4	1.7	1.5	1.7	2.9	3.4
Category 4 - Weak	0.9	2.4	0.3	0.8	1.2	3.2
Category 5 - Default	1.4	-	0.2	-	1.6	-

At 31 December 2014	15.1	10.6	15.4	12.4	30.5	23.0

Table 30 and the graphs below set out IRB exposures by obligor grade for central governments and central banks, institutions and corporates, all of which are assessed using our 23-grade CRR master scale. We benchmark the master scale against the ratings of external rating agencies. Each CRR band is associated with an external rating grade by reference to long-run default rates for that grade, represented by the average of issuer-weighted historical default rates.

The correspondence between the agency long-run default rates and the PD ranges of our master scale is

obtained by matching a smoothed curve based on those default rates with our master scale reference PDs. This association between internal and external ratings is indicative and may vary over time. In these tables, the ratings of Standard and Poor's ('S&P') are cited for illustration purposes, though we also benchmark against other agencies' ratings in an equivalent manner.

For further details of the Group's approach to credit quality classification, please see the definition of 'obligor grade' in the glossary, and also page 207 of the Annual Report and Accounts 2014.

Table 30a: Wholesale IRB exposure - by obligor grade1 - Central governments and central banks

	CRR		PD range		Exposure value2		Average PD3		Average LGD3		RWA density3		RWAs		Mapped external rating
			%		US$bn		%		%		%		US$bn
Default risk
Minimal	0.1	0.000 to 0.010		122.8		0.01		45.2		7		8.7		AAA
	1.1	0.011 to 0.028		60.3		0.02		45.0		7		4.4		AA+ to AA
	1.2	0.029 to 0.053		59.2		0.04		45.4		13		7.4		AA- to A+

Low	2.1	0.054 to 0.095		51.6		0.07		45.0		20		10.4		A
	2.2	0.096 to 0.169		6.0		0.13		45.2		25		1.5		A-

Satisfactory	3.1	0.170 to 0.285		11.3		0.22		45.0		43		4.9		BBB+
	3.2	0.286 to 0.483		3.6		0.37		45.0		53		1.9		BBB
	3.3	0.484 to 0.740		1.6		0.63		45.0		63		1.0		BBB-

Fair	4.1	0.741 to 1.022		1.7		0.87		45.0		81		1.4		BB+
	4.2	1.023 to 1.407		0.4		1.16		45.0		125		0.5		BB
	4.3	1.408 to 1.927		0.2		1.65		43.3		100		0.2		BB-

Moderate	5.1	1.928 to 2.620		0.9		2.25		45.0		111		1.0		BB-
	5.2	2.621 to 3.579		0.7		3.05		45.0		129		0.9		B+
	5.3	3.580 to 4.914		5.6		4.20		45.0		130		7.3		B

Significant	6.1	4.915 to 6.718		0.7		5.75		45.2		157		1.1		B
	6.2	6.719 to 8.860		0.1		7.85		45.0		200		0.2		B-

High	7.1	8.861 to 11.402		0.7		10.00		45.0		186		1.3		CCC+
	7.2	11.403 to 15.000		-		-		-		-		-		CCC+

Special management	8.1	15.001 to 22.000		-		-		-		-		-		CCC+
	8.2	22.001 to 50.000		-		-		-		-		-		CCC+
	8.3	50.001 to 99.999		-		-		-		-		-		CCC to C

Default4	9/10	100.000		-		-		-		-		-		Default

At 31 December 2014				327.4		0.17		45.1		17		54.1

Default risk
Minimal	0.1	0.000 to 0.010		132.4		0.01		45.1		7		9.3		AAA to AA+
	1.1	0.011 to 0.028		74.3		0.02		45.0		6		4.8		AA to AA-
	1.2	0.029 to 0.053		38.7		0.04		45.0		14		5.6		A+

Low	2.1	0.054 to 0.095		64.1		0.07		45.0		18		11.7		A
	2.2	0.096 to 0.169		11.4		0.13		45.0		29		3.3		A-

Satisfactory	3.1	0.170 to 0.285		5.3		0.22		45.0		42		2.2		BBB+
	3.2	0.286 to 0.483		3.7		0.37		45.0		49		1.8		BBB to BBB-
	3.3	0.484 to 0.740		2.4		0.63		45.0		67		1.6		BBB-

Fair	4.1	0.741 to 1.022		1.1		0.87		45.0		82		0.9		BB+
	4.2	1.023 to 1.407		0.2		1.20		45.0		100		0.2		BB
	4.3	1.408 to 1.927		0.3		1.65		45.2		-		-		BB-

Moderate	5.1	1.928 to 2.620		0.9		2.25		45.0		111		1.0		BB-
	5.2	2.621 to 3.579		1.4		3.05		45.0		121		1.7		B+
	5.3	3.580 to 4.914		1.1		4.20		45.0		136		1.5		B+

Significant	6.1	4.915 to 6.718		0.3		5.75		45.4		167		0.5		B
	6.2	6.719 to 8.860		3.7		7.85		45.0		168		6.2		B-

High	7.1	8.861 to 11.402		0.4		10.00		45.0		175		0.7		B-
	7.2	11.403 to 15.000		-		-		-		-		-		CCC+

Special management	8.1	15.001 to 22.000		-		-		-		-		-		CCC
	8.2	22.001 to 50.000		-		-		-		-		-		CCC-
	8.3	50.001 to 99.999		-		-		-		-		-		CC to C

Default4	9/10	100.000		-		-		-		-		-		Default

At 31 December 2013				341.7		0.17		45.0		16		53.0

For footnotes, see page 50.

Table 30b: Wholesale IRB exposure - by obligor grade1 - Institutions (continued)

	CRR		PD range		Exposure value2		Average PD3		Average LGD3		RWA density3		RWAs		Mapped external rating
		%		US$bn		%		%		%		US$bn
Default risk
Minimal	0.1	0.000 to 0.010		1.8		0.02		50.2		22		0.4		AAA
	1.1	0.011 to 0.028		15.3		0.03		41.0		12		1.8		AA+ to AA
	1.2	0.029 to 0.053		27.4		0.04		31.7		11		3.0		AA-