HSBC 2012 Annual Report Download - page 260

HSBC HOLDINGS PLC
Report of the Directors: Operating and Financial Review (continued)
Risk > Appendix – Risk policies and practices > Credit risk
258
Impairment assessment
(Audited)
It is our policy that each operating company in HSBC creates impairment allowances for impaired loans promptly
and appropriately.
For details of our impairment policies on loans and advances and financial investments, see Notes 2g and 2j on the
Financial Statements.
Impairment and credit risk mitigation
The existence of collateral has an impact when calculating impairment on individually assessed impaired loans.
When we no longer expect to recover the principal and interest due on a loan in full or in accordance with the
original terms and conditions, it is assessed for impairment. If exposures are secured, the current net realisable value
of the collateral will be taken into account when assessing the need for an impairment allowance. No impairment
allowance is recognised in cases where all amounts due are expected to be settled in full on realisation of the security.
Personal lending portfolios are generally assessed for impairment on a collective basis as the portfolios typically
consist of large groups of homogeneous loans. Two methods are used to calculate allowances on a collective basis:
a roll rate methodology or a more basic formulaic approach based on historical losses.
The historical loss methodology is typically used to calculate collective impairment allowances for secured, or low
default portfolios such as mortgages, until the point at which they are individually identified and assessed as
impaired. For loans which are collectively assessed using historical loss methodology, the historical loss rate is
derived from the average contractual write-off net of recoveries over a defined period. The net contractual write-off
rate is the actual amount of loss experienced after the realisation of collateral and receipt of recoveries.
A roll rate methodology is more commonly adopted for unsecured portfolios when there are sufficient volumes of
empirical data to develop robust statistical models. In certain circumstances mortgage portfolios have a statistically
significant number of defaults and losses available, enabling reliable roll rates to be generated. In these cases a roll
rate methodology is applied until the point at which the loans are individually identified and assessed as impaired,
and the average loss rate for each delinquency bucket is adjusted to reflect the average loss expected following
realisation of security and receipt of recoveries. The average loss expected is derived from average historical
collateral realisation values.
The nature of the collective allowance assessment prevents individual collateral values or loan-to-value (‘LTV’)
ratios from being included within the calculation. However, the loss rates used in the collective assessment are
adjusted for the collateral realisation experiences which will vary depending on the LTV composition of the
portfolio. For example mortgage portfolios under a historical loss rate methodology with lower LTV ratios will
typically experience lower loss history and consequently a lower net contractual write-off rate.
For wholesale collectively assessed loans historical loss methodologies are applied to measure loss event
impairments which have been incurred but not reported. Loss rates are derived from the observed contractual write-
off net of recoveries over a defined period, typically 60 months. The net contractual write-off rate is the actual
amount of loss experienced after realisation of collateral and receipt of recoveries. These historical loss rates are
adjusted by an economic factor which adjusts the historical averages to better represent current economic conditions
affecting the portfolio. In order to reflect the likelihood of a loss event not being identified and assessed an
emergence period assumption is applied. This reflects the period between a loss occurring and its identification. The
emergence period is estimated by local management for each identified portfolio. The factors that may influence this
estimation include economic and market conditions, customer behaviour, portfolio management information, credit
management techniques and collection and recovery experiences in the market. A fixed range for the period between
a loss occurring and its identification is not defined across the Group and as it is assessed empirically on a periodic
basis, it may vary over time as these factors change. Given that credit management policies require all customers to
be reviewed at least annually, we expect this estimated period would be at most 12 months.