Goodwill and intangible assets

12 Months Ended

Dec. 31, 2018



21	Goodwill and intangible assets



		2018	2017
	Footnotes	$m	$m
Goodwill		12,986	13,588
Present value of in-force long-term insurance business		7,149	6,610
Other intangible assets	1	4,222	3,255
At 31 Dec	2	24,357	23,453


1	Included within other intangible assets is internally generated software with a net carrying value of $3,632m (2017: $2,641m). During the year, capitalisation of internally generated software was $1,781m (2017: $1,157m) and amortisation was $687m (2017: $570m).


2	Information regarding the effects of adoption of IFRS 9 can be found in Note 37.



Movement analysis of goodwill
	2018		2017
	$m		$m
Gross amount
At 1 Jan	22,902		21,445
Exchange differences	(617	)	1,490
Other	(105	)	(33	)
At 31 Dec	22,180		22,902
Accumulated impairment losses
At 1 Jan	(9,314	)	(9,115	)
Exchange differences	120		(327	)
Other	—		128
At 31 Dec	(9,194	)	(9,314	)
Net carrying amount at 31 Dec	12,986		13,588

Impairment testing

The Group’s impairment test in respect of goodwill allocated to each cash-generating unit (‘CGU’) is performed as at 1 July each year. A review for indicators of impairment is undertaken at each subsequent quarter-end and as at 31 December 2018. No indicators of impairment were identified as part of these reviews.

Basis of the recoverable amount

The recoverable amount of all CGUs to which goodwill has been allocated was equal to its value in use (‘VIU’) at each respective testing date for 2017 and 2018. For each CGU, the VIU is calculated by discounting management’s cash flow projections for the CGU. The key assumptions used in the VIU calculation for each significant CGU are discussed below.



Key assumptions in VIU calculation
	Goodwill at 1 Jul 2018	Discount rate	Nominal growth rate beyond initial cash flow projections	Goodwill at 1 Jul 2017	Discount rate	Nominal growth rate beyond initial cash flow projections
	$m	%	%	$m	%	%
Cash-generating unit
Europe
RBWM	3,565	8.1	3.8	3,508	8.9	3.7
CMB	2,626	9.4	3.7	2,570	9.9	3.6
Global
GB&M	4,045	9.8	5.6	4,000	10.6	5.8

At 1 July 2018, aggregate goodwill of $3,061m (1 July 2017: $3,059m) had been allocated to CGUs that were not considered individually significant. The Group’s CGUs do not carry on their balance sheets any significant intangible assets with indefinite useful lives, other than goodwill.

Management’s judgement in estimating the cash flows of a CGU

The cash flow projections for each CGU are based on plans approved by the GMB. For the goodwill impairment test conducted at 1 July 2018, management’s cash flow projections until the end of 2022 were used.

Discount rate

The rate used to discount the cash flows is based on the cost of capital assigned to each CGU, which is derived using a capital asset pricing model (‘CAPM’). CAPM depends on a number of inputs reflecting financial and economic variables, including the risk-free rate and a premium to reflect the inherent risk of the business being evaluated. These variables are based on the market’s assessment of the economic variables and management’s judgement. The discount rates for each CGU are refined to reflect the rates of inflation for the countries within which the CGU operate. In addition, for the purposes of testing goodwill for impairment, management supplements this process by comparing the discount rates derived using the internally generated CAPM, with the cost of capital rates produced by external sources for businesses operating in similar markets.

Nominal long-term growth rate

The long-term growth rate is used to extrapolate the cash flows in perpetuity because of the long-term perspective within the Group of business units making up the CGUs. These growth rates reflect GDP and inflation for the countries within which the CGU operates or from which it derives revenue.

Sensitivities of key assumptions in calculating VIU

At 1 July 2018, none of the CGUs were sensitive to reasonably possible adverse changes in key assumptions supporting the recoverable amount. In making an estimate of reasonably possible changes to assumptions, management considers the available evidence in respect of each input to the model, such as the external range of discount rates observable, historical performance against forecast and risks attaching to the key assumptions underlying cash flow projections.

Present value of in-force long-term insurance business

When calculating the present value of in-force long-term (‘PVIF’) insurance business, expected cash flows are projected after adjusting for a variety of assumptions made by each insurance operation to reflect local market conditions and management’s judgement of future trends, and uncertainty in the underlying assumptions is reflected by applying margins (as opposed to a cost of capital methodology). Variations in actual experience and changes to assumptions can contribute to volatility in the results of the insurance business.

Actuarial Control Committees of each key insurance entity meet on a quarterly basis to review and approve PVIF assumptions. All changes to non-economic assumptions, economic assumptions that are not observable and model methodologies must be approved by the Actuarial Control Committee.



Movements in PVIF
		2018		2017
	Footnotes	$m		$m
As at 31 Dec 2017		6,610		6,502
Impact on transition to IFRS 9		(78	)	N/A
At 1 Jan		6,532		6,502
Change in PVIF of long-term insurance business		673		24
– value of new business written during the year		1,117		919
– expected return	1	(719	)	(599	)
– assumption changes and experience variances (see below)		292		(280	)
– other adjustments		(17	)	(16	)
Exchange differences and other movements		(56	)	84
At 31 Dec		7,149		6,610


1	‘Expected return’ represents the unwinding of the discount rate and reversal of expected cash flows for the period.

Assumption changes and experience variances

Included within this line item are:


•	$(56)m (2017: $(98)m), directly offsetting regulatory-driven changes to the valuation of liabilities under insurance contracts.


•	$455m (2017: $(141)m), reflecting the future expected sharing of returns with policyholders on contracts with discretionary participation features (‘DPF’), to the extent this sharing is not already included in liabilities under insurance contracts.


•	$(107)m (2017: $(41)m), driven by other assumptions changes and experience variances.

Key assumptions used in the computation of PVIF for main life insurance operations

Economic assumptions are set in a way that is consistent with observable market values. The valuation of PVIF is sensitive to observed market movements and the impact of such changes is included in the sensitivities presented below.



	2018		2017
	Hong Kong	France¹	Hong Kong	France¹
	%	%	%	%
Weighted average risk-free rate	2.29	1.52	2.02	1.50
Weighted average risk discount rate	5.90	2.35	6.20	2.20
Expense inflation	3.00	1.70	3.00	1.48


1	For 2018, the calculation of France’s PVIF assumes a risk discount rate of 2.35% (2017: 2.20%) plus a risk margin of $109m (2017: $80m).

Sensitivity to changes in economic assumptions

The Group sets the risk discount rate applied to the PVIF calculation by starting from a risk-free rate curve and adding explicit allowances for risks not reflected in the best-estimate cash flow modelling. Where the insurance operations provide options and guarantees to policyholders the cost of these options and guarantees is an explicit reduction to PVIF, unless it is already allowed for as an explicit addition to the technical provisions required by regulators. See page 190 for further details of these guarantees and the impact of changes in economic assumptions on our insurance manufacturing subsidiaries.

Sensitivity to changes in non-economic assumptions

Policyholder liabilities and PVIF are determined by reference to non-economic assumptions, including mortality and/or morbidity, lapse rates and expense rates. See page 191 for further details on the impact of changes in non-economic assumptions on our insurance manufacturing operations.