Barrack Gold Corporation – Annual report – 31 December 2015
20 Impairment of Goodwill and Non-Current Assets
In accordance with our accounting policy, goodwill is tested for impairment in the fourth quarter and also when there is an indicator of impairment. Non-current assets are tested for impairment when events or changes in circumstances suggest that the carrying amount may not be recoverable.
When there is an indicator of impairment of non-current assets within an operating segment consisting of a CGU that does not contain goodwill, we test the non-current assets for impairment and recognize any impairment loss on the non-current assets. When there is an indicator of impairment of non-current assets within an operating segment that contains goodwill, we test the non-current assets for impairment first and recognize any impairment loss on goodwill first and then any remaining impairment loss is applied against the non-current assets.
An impairment loss is recognized when the carrying amount exceeds the recoverable amount. The recoverable amount of each operating segment for goodwill testing purposes has been determined based on its estimated FVLCD, which has been determined to be greater than the VIU amounts. The recoverable amount for non-current asset testing is calculated using the same approach as for goodwill; however, the assessment is done at the CGU level, which is the lowest level for which identifiable cash flows are largely independent of the cash flows of other assets. A CGU is generally an individual operating mine or development project.
Summary of Impairments (Reversals)
For the year ended December 31, 2015, we recorded impairment losses of $1.7 billion (2014: $2.7 billion) for non-current assets and $2.2 billion (2014: $1.4 billion) for goodwill, as summarized in the following table:
2015 Indicators of Impairment
Fourth Quarter 2015
In fourth quarter 2015, as per our policy, we performed our annual goodwill impairment test. Primarily as a result of the lower gold price assumptions used in this year’s test which are consistent with current market conditions, we identified that the carrying values of our Pueblo Viejo, Goldstrike, Cortez and Lagunas Norte mines exceeded their FVLCD. At Pueblo Viejo, a goodwill impairment loss of $412 million and a non-current asset impairment loss of $1,101 million was recorded and the recoverable amount after the impairment, based on the mine’s FVLCD, was $3.2 billion (100% basis). At Goldstrike, a goodwill impairment loss of $730 million was recorded and the recoverable amount after the impairment, based on the mine’s FVLCD, was $2.7 billion. At Cortez, a goodwill impairment loss of $355 million was recorded and the recoverable amount after the impairment, based on the mine’s FVLCD, was $3.4 billion. At Lagunas Norte, a goodwill impairment loss of $247 million and a non-current asset impairment loss of $36 million was recorded and the recoverable amount after the impairment, based on the mine’s FVLCD, was $480 million. Refer to note 19 for our remaining goodwill balances.
As at December 31, 2015, all of the assets and liabilities of Bald Mountain and Round Mountain were classified as held-for-sale. As the agreed selling price is lower than previously recognized carrying values, we recorded a non-current asset impairment loss of $81 million.
Throughout fourth quarter 2015, the trading price of the Company’s shares declined such that the carrying value of our net assets exceeded our market capitalization. We have determined that this is an indicator of impairment and tested the remaining assets that were not included in the annual goodwill impairment test. As a result, we determined two additional impairments. At our Pascua-Lama project, we recorded an impairment loss of $404 million (net of a $46 million reversal related to a specific PP&E asset). The recoverable amount after the impairment, based on the project’s FVLCD, was $810 million. At our Buzwagi mine in Tanzania (part of our Acacia subsidiary), we recorded a non-current asset impairment loss of $37 million. The recoverable amount after the impairment, based on the mine’s FVLCD, was $81 million (100% basis).
We evaluated the FVLCD of an oil royalty that we received as part of the consideration for one of the Barrick Energy dispositions in 2013 and concluded that due to the significant decline in current oil prices in fourth quarter 2015 and the corresponding constraints on capital investment in the oil industry, its carrying value was not recoverable. We recorded an impairment of $36 million and reduced its carrying value to nil.
Third Quarter 2015
In July 2015, the Zambian government passed legislation that amended the country’s mining tax regime. This was an indicator of potential reversal of previous impairments recorded on our Lumwana mine in fourth quarter 2014. In third quarter 2015, we evaluated the FVLCD and concluded that, based on the current mine plan, lower short-term copper prices and a higher observable discount rate offset the lower royalty rate. Therefore no reversal of impairment was required at that time.
As at September 30, 2015, all of the assets and liabilities of Zaldívar were classified as held-for-sale as the transaction will result in a loss of control. The agreed selling price was lower than our previous assessment of FVLCD due to lower short-term copper prices, the impact of 10 months’ worth of production on the fair value and an increase in observable discount rates. For the year ended December 31, 2015 we recorded a goodwill impairment loss of $427 million as a result of this transaction.
In third quarter 2015, a net reversal of $16 million was recognized relating to the termination of contracts of certain leased assets at Pascua-Lama that had previously been impaired. They are now carried at their expected realizable value.
Second Quarter 2015
In second quarter 2015, we determined that we expect to sell the Monte Rio power asset at our Pueblo Viejo mine. Power supply to Pueblo Viejo is not impacted by this disposition. In third quarter 2015, we entered into an agreement to sell the asset and recorded a partial reversal of this impairment based on the agreed upon sales price. For the year ended December 31, 2015, we recorded an impairment loss of $11 million to reduce its carrying value down to its net realizable value.
2014 Indicators of Impairment
In second quarter 2014, our Jabal Sayid project in Saudi Arabia met the criteria as an asset held-for-sale. Accordingly, we were required to allocate goodwill from the Copper Operating Unit to Jabal Sayid and test the Jabal Sayid group of assets for impairment. We determined that the carrying value exceeded the FVLCD, and consequently recorded $514 million in impairment charges, including the full amount of goodwill allocated on a relative fair value basis, of $316 million. The recoverable amount after the impairment, based on FVLCD, was $560 million. In fourth quarter 2014, we closed a transaction to sell a 50 percent interest of Jabal Sayid for cash proceeds of $216 million.
We reached an agreement to sell a power-related asset at our Pueblo Viejo mine for proceeds that exceeded its carrying value. This asset had previously been impaired in fourth quarter 2012, and therefore we recognized an impairment reversal of $9 million. This transaction closed on September 30, 2014.
In fourth quarter 2014, as described in note 19, we reorganized our internal management reporting structure. As a result, the goodwill attributable to our former North America Portfolio, Australia Pacific and Copper segments was allocated to the individual CGUs within those operating segments on a relative fair value basis. The allocation of goodwill to the carrying value of our Bald Mountain and Round Mountain CGUs resulted in their carrying values exceeding their FVLCD and, as a result, we recorded goodwill impairment losses of $131 million and $36 million, respectively. The recoverable amounts after the impairment of Bald Mountain and Round Mountain, based on FVLCD, were $482 million and $131 million, respectively.
On December 18, 2014, the Zambian government passed changes to the country’s mining tax regime that would replace the current corporate income tax and variable profit tax with a 20 percent royalty which took effect on January 1, 2015. The application of a 20 percent royalty rate compared to the 6 percent royalty rate the Company was paying has a significant negative impact on the expected future cash flows of our Lumwana mine and was considered an indicator of impairment. As a result, we conducted an impairment test and as a result of the new royalty rate along with the decrease in our copper price assumptions, recorded $930 million in impairment charges, including the full amount of goodwill of $214 million allocated to Lumwana as a result of the change in segments (see note 19). The recoverable amount after the impairment, based on FVLCD, was $300 million.
Our Zaldívar mine experienced a significant decrease in the estimated FVLCD of the mine, primarily as a result of the decrease in fourth quarter 2014 of our forecast of the long-term copper price and, to a lesser extent, as a result of the final assessment of the tax rate increase in Chile. Accordingly, we recorded a goodwill impairment loss of $712 million on this CGU. The recoverable amount after the impairment, based on FVLCD, was $2,411 million.
In November 2014, we completed a strategy optimization study for our Cerro Casale project with the goal of identifying a development model that would improve the project economics and risk by reducing the upfront capital requirements in order to generate a higher return on our investment. The study was unable to identify an alternative that provided an overall rate of return above our hurdle rate for a project of this size and complexity. As a result, the budget for 2015 for the project was significantly reduced, with the 2015 budget focused on preserving the optionality of the project. We will continue activities to protect the asset and assess alternative ways to develop the project in a more economic manner; however, management’s expectation of achieving a suitable rate of return in the metal price environment has been diminished. The foregoing developments were deemed to be indicators of impairment, and as a result, we assessed the recoverable amount of the project and have recorded an impairment loss on the project of $1,467 million. The recoverable amount after the impairment, based on the project’s estimated FVLCD, was $500 million (100% basis).
In December 2014, the Chilean Supreme Court declined to consider Barrick’s appeal of the Environmental Court Decision on Pascua-Lama on procedural grounds (see note 35). As a result, the Superintendencia del Medio Ambiente (“SMA”) will now re-evaluate the resolution. Although we cannot reasonably predict the outcome of the resolution, this risk, in combination with the decrease in our long-term silver price assumption in fourth quarter 2014 due to declining market prices, and the continued uncertainty about the timing, cost and permitting of the project, were deemed to be indicators of impairment. As a result, we assessed the recoverable amount of the project and have recorded an impairment loss on Pascua-Lama of $382 million. The recoverable amount after the impairment, based on the project’s estimated FVLCD, was $1,200 million, which is equal to the project’s carrying value at the start of the year.
At our Porgera mine in Papua New Guinea, we have revised our LOM plan to include a portion of the open pit resources that were removed from the plan in the prior year. In 2013, we did not have a feasible plan to access the open pit reserves due to technical and financial issues with respect to the west wall of the open pit. In 2014, management resolved these technical issues and developed an optimized mine plan to sequence the west wall cutback in an economical manner. As a result, management was able to bring a significant portion of the ounces from the open pit back into the LOM plan. The new plan resulted in an increase in the estimated mine life from 8 to 12 years, and an increase in the estimated FVLCD of the mine, which has resulted in a partial reversal of a previous impairment loss of $160 million. The recoverable amount after the impairment reversal, based on FVLCD, was $600 million.
The annual update to the LOM plan at Cortez resulted in a cessation of mining in one of the open pits at the mine. This was identified as an indicator of impairment, resulting in the impairment of assets specifically related to this pit of $46 million.
The key assumptions and estimates used in determining the FVLCD are related to commodity prices, discount rates, NAV multiples for gold assets, operating costs, exchange rates, capital expenditures, the LOM production profile, continued license to operate, evidence of value from current year disposals and for our projects the expected start of production. In addition, assumptions are related to observable market evaluation metrics, including identification of comparable entities, and associated market values per ounce and per pound of reserves and/or resources, as well as the valuation of resources beyond what is included in LOM plans.
For the gold segments, excluding Pascua-Lama and Cerro Casale, FVLCD for each of the CGUs was determined by calculating the net present value (“NPV”) of the future cash flows expected to be generated by the mines and projects within the segments (level 3 of the fair value hierarchy). The estimates of future cash flows were derived from the most recent LOM plans and, where the LOM plans excludes a material portion of total reserves and resources, we assign value to reserves and resources not considered in these models. Based on observable market or publicly available data, including forward prices and equity sell-side analyst forecasts, we make an assumption of future gold and silver prices to estimate future revenues. The future cash flows for each gold mine are discounted using a real weighted average cost of capital (“WACC”), which reflects specific market risk factors for each mine. Some gold companies trade at a market capitalization greater than the NPV of their expected cash flows. Market participants describe this as a “NAV multiple”, which represents the multiple applied to the NPV to arrive at the trading price. The NAV multiple is generally understood to take account of a variety of additional value factors such as the exploration potential of the mineral property, namely the ability to find and produce more metal than what is currently included in the LOM plan or reserve and resource estimates, and the benefit of gold price optionality. As a result, we applied a specific NAV multiple to the NPV of each CGU within each gold segment based on the NAV multiples observed in the market in recent periods and that we judged to be appropriate to the CGU.
Pascua-Lama and Cerro Casale
The FVLCD for Pascua-Lama and Cerro Casale was determined by considering observable market values for comparable assets expressed as dollar per ounce and dollar per pound of proven and probable reserves (level 3 of the fair value hierarchy). We used the market approach as the LOMs for Pascua-Lama and Cerro Casale have significant uncertainty with respect to the estimated timeline for the project and the estimated remaining construction costs. The observable market values were adjusted, where appropriate, for country risk if the comparable asset was in a different country and any change in metal prices since the valuation date of the comparable asset.
For our copper operating segments, the FVLCD for each of the CGUs was determined based on the NPV of future cash flows expected to be generated using the most recent LOM plans (level 3 of the fair value hierarchy). Based on observable market or publicly available data including spot and forward prices and equity sell-side analyst consensus, we make an assumption of future copper prices to estimate future revenues. The future cash flows for each copper mine were discounted using a WACC depending on the location and market risk factors for each mine.
Our gold price assumptions used in our 2015 impairment testing are 2016: $1,000, 2017: $1,100 and 2018+: $1,200. In fourth quarter 2015, market consensus prices ranged from $1,000 to $1,250 in the short term and $800 to $1,300 in the long term. Consequently, our gold price assumptions are consistent with the assumptions a market participant would use to value a gold mining property. In 2014, impairment test gold prices were $1,250 for 2015–2016 and $1,300 for 2017 onwards.
The other key assumptions used in our impairment testing are summarized in the table below:
Should there be a significant decline in commodity prices, we would take actions to assess the implications on our life of mine plans, including the determination of reserves and resources, and the appropriate cost structure for the operating segments. The recoverable amount of the CGUs would also be impacted by other market factors such as changes in net asset value multiples and the value per ounce/pound of comparable market entities.
We performed a sensitivity analysis on the gold price, which is the key assumption that impacts the impairment calculations. We assumed a $100 per ounce change in our gold price assumptions, while holding all other assumptions constant, to determine the impact on impairment losses recorded, and whether any additional operating segments would be impacted. We note that this sensitivity identifies the key assets where the increase/decrease in the sales price, in isolation, could cause the carrying value of our operating segments to exceed its recoverable amount for the purposes of the goodwill impairment test or the carrying value of any of our CGUs to exceed its recoverable amount for the purposes of the non-current asset impairment test.
The results of this analysis are as follows:
We also performed a sensitivity analysis on our WACC, which is another key input that impacts the impairment calculations. We assumed a +/-10% change on the WACC, while holding all other assumptions constant, to determine the impact on impairment losses recorded, and whether any additional operating segments would be impacted. The results of this analysis are as follows:
In addition, for our Cerro Casale and Pascua-Lama projects, we have determined our valuation based on a market approach. The key assumption that impacts the impairment calculations, should there be an indication of impairment for these CGUs, is the value per ounce of gold and per pound of copper based on an analysis of comparable companies. We assumed a negative 10% change for the assumption of gold, silver and copper value per unit, while holding all other assumptions constant, and based on the results of the impairment testing performed in fourth quarter 2015 for Cerro Casale and Pascua-Lama, the fair value of the CGUs would have been reduced from $500 million to $450 million and $810 million to $730 million, respectively. We note that this sensitivity identifies the decrease in the value that, in isolation, would cause the carrying value of the CGU to exceed its recoverable amount. For Cerro Casale and Pascua-Lama, this value decrease is linear to the decrease in value per ounce/pound.
Based on the results of the impairment test performed in fourth quarter 2015, the carrying value of the CGUs that are most sensitive to the change in sales prices used in the annual test are: