Pan African Resources — Tails wagging

Pan African is a South Africa based gold mining group which has dual primary JSE and AIM listings and operates seven major assets in South Africa: Barberton (target output 95koz Au pa), the Barberton Tailings Retreatment Project (20koz), Evander (95koz), the Evander Tailings Retreatment Project (10koz), Elikhulu (53koz), Phoenix Platinum (12koz) and Uitkomst (400kt saleable coal pa).

We have updated our forecasts for FY17, which now include Uitkomst and reflect the effect of H117 section 54 regulatory stoppage notices (21 days lost production) as well as the current Evander production halt. Our group wide forecasts are for production of 183.8koz of gold in FY17 compared to management guidance of 181koz of gold. This, along with our current price assumptions, results in EPS of 1.36p (1.35p normalised), rising to 2.41p in FY18. In the longer term (and in FY17 had it not been for the current production challenges), we estimate that Pan African is capable of generating average earnings of 2.65p/share and paying average (maximum potential) dividends of 2.31p (including Elikhulu). At currently prevailing forex rates, the net present value of such a dividend stream is 21.38p (discounted as 10% pa). This valuation assumes that the grade profile at Evander averages 6.43g/t from FY17-30 and that the gold price will average US$1,283/oz (real). In the meantime, it is cheaper than its immediate peers on at least 70% of a range of valuation measures. It also has the third highest (consensus) forecast dividend yield of any dividend-paying gold company, globally

Pan African had £33.2m in net debt on its balance sheet as at 31 December 2016 after the payment of a £17.1m final dividend in late December (cf net debt of £22.8m as at 30 June 2016, £16.2m as at 31 December 2015 and £18.0m as at 30 June 2015). Net debt currently equates to a gearing (net debt/equity) ratio of 16.7% and a leverage (net debt/[net debt + equity]) ratio of 14.3%. We have now incorporated our estimated capex requirement for Elikhulu of £154m over six years from FY17 to FY22. Consequently, whereas we had previously been expecting Pan African to be net debt free within the current year, we now expect this to occur in FY21.

The valuations above were conducted at a long-term gold price of US$1,283/oz, changes in which will (inevitably) affect our valuation. Other risks include the timing of the refurbishment programme and the consequent production hiatus at Evander and timely progress of the Elikhulu project. Exogenous factors include exchange rate movements (in particular the rand/dollar and rand/sterling rates and, by extension, the sterling/dollar rate), the regulatory framework for energy costs and labour agreements.

Pan African’s interim figures were consistent with both its trading statement of 27 January as well as our prior full-year expectations. Compared to the prior year period (ie H116) gold produced and sold decreased 10.0% to 91,613oz owing to a combination of operational constraints coupled with a sharp increase in Department of Mineral Resources’ (DMR) Section 54 stoppage notices, while cash costs of production increased by 25.8%, to US$932/oz (ostensibly the same level as in H215). On a headline basis, earnings per share, in particular, were higher than at any time since H114 (ie the second half of calendar 2013), when the gold price averaged US$1,311/oz. Moreover, this coincided with a period in which the South African rand has been unusually strong for the first time in a reporting period since H111, averaging ZAR17.8771/£ (vs ZAR22.0942/£ in H216) and ZAR13.9875/US$ (vs ZAR15.4132/US$ in H216).

In rand terms, PAF’s gold cost of production (excluding realisation costs) increased 14.4%, to ZAR1,165.6m compared with H116; including realisation costs, it increased by 16.4%, to ZAR1,193.3m. This could be predominantly attributed to:

Compared with H216 however, PAF’s gold cost of production (including realisation costs) increased by only 5.5% in local currency terms; relative to the aggregate number of tonnes processed, it increased by only 1.4%, to ZAR643/t (vs ZAR634/t).

While overall gold production fell 10.0% cf H116, there were sharp increases at both the BTRP (+14.0%) and the ETRP (+77.3%), while production at both of Pan African’s traditional, underground operations fell, with output at Barberton down 21.0% and at Evander down 27.2%. As a result, the BTRP and ETRP were Pan African’s second and third most profitable business units overall, behind Barberton (underground), but ahead of Evander (underground). As a result, for the first time the company recorded greater profits, in the form of adjusted EBITDA, from its tailings retreatment projects than from its underground operations.

We had anyway expected a decline in tonnes milled from underground at Evander to 349kt for FY17 (174.5kt pro rata for H117) versus c 400ktpa in prior years. In the event, this decline was slightly greater than expected, being augmented by Section 54 DMR stoppage notices, in particular.

During the six month period under review, Evander was issued with four Section 54 regulatory notices by the DMR, which resulted in 13 lost production days, compared with three notices and two days in the prior year period.

While cash costs of production increased by 58.8% in US dollar terms, to US$1,457/oz, most of the increase could be attributed to the lower number of ounces produced and the adverse moves in the rand:dollar exchange rate. In rand terms, aggregate costs increased by a much more respectable 2.8%, indicating a continued focus on cost control, notwithstanding the production challenges experienced during the half year.

Operations at Evander have now advanced from 25 to 26 level, which has simultaneously improved access to more high-grade panels as well as allowing management to manage and blend the ore mined towards achieving its target of 110,000oz of gold per annum. All other things being equal therefore, the head grade is expected to remain in the range 6-7g/t for the next three to five years. Note that, had an underground head grade of 7.2g/t prevailed in H117 (and applying the group’s H117 marginal tax rate to the incremental profits), we estimate that it would have added c 0.41p to both Pan African’s EPS and headline EPS in H117. Operations will remain focused at 25 level for eight years, before progressing to 26 level. Excluding Elikhulu, management’s target is to achieve a consistent all-in sustaining cost of production of US$1,100/oz at Evander (vs US$1,310/oz in the period under review).

In fact, immediately prior to the February announcement of a fatality at seven shaft, Evander was reported to be mining material at a diluted grade of 8g/t from underground and at a mine call factor in excess of 80% (vs 65% in H116, 59% in H115 and a target of 73%) as a result of a focus on blasting and cleaning practices, in particular, including the use of blasting barricades. In the wake of the fatality, there will now be a 55-day underground mining hiatus at Evander, while steelwork at the 7A, 7 and 8 shafts is refurbished. During this period, the pump column at 8 shaft will be the only one at the mine that is operational, which represents a technical engineering risk in the event of a similar failure as at 7 shaft. During this period, management is also planning to drill a second exploration drill hole into the Kinross 2010 pay channel as a precursor to potential future exploitation.

On 10 March, PAF announced that repairs at 7 shaft were “progressing on schedule” and that they were “still expected to be completed within the 55 day period previously communicated” (ie around the middle of April). At the same time, in order to ameliorate Evander’s underground fixed cost base once mining recommences, management also announced a retrenchment programme (with the agreement of the National Union of Mineworkers and the appropriate South African government agency) whereby approximately 30% of Evander Mines’ employees will be retrenched at an estimated cost of ZAR54m (US$4.2m or £3.4m). In order to minimise the number of job losses overall, Evander has agreed to re-engage a number of retrenched employees once site activities commence at Elikhulu. In the meantime, Edison estimates that this initiative will save Evander approximately 15% of its cost base in the longer term. Pro-rata, this implies a net additional cost in the order of ZAR25m in H217. However, this should be substantially offset by management’s initiative to utilise available plant capacity to continue processing tailings and additional surface sources.

As at Evander, Barberton was served with six Section 54 regulatory notices in H117, which resulted in eight days of lost production (cf one notice costing three days of production in H116). In aggregate therefore, Barberton and Evander were served with a combined ten Section 54 regulatory stoppage notices during the 184 day period, which resulted in 21 days of lost production (cf four notices and five days lost in the prior year period). In addition however, Barberton also experienced a degree of community unrest, which was designed to target government service delivery, but which also had the collateral effect of costing Barberton six days of lost production, as a result of protests preventing employees from reporting to work. Finally, Barberton experienced flexibility constraints at its Fairview mine and specifically at its high grade 11-block, which resulted in lower grades being mined. As with Evander however, while cash costs of production were reported to have increased by 42.0% compared with the prior year period, to US$967/oz (vs US$681/oz), in fact aggregate cash costs increased by only a much more modest 3.9% in local currency terms compared to H216:

Barberton is commencing multi-year wage negotiations with its labour-force later this year (note that Barberton traditionally negotiates its agreements outside the auspices of the South African Chamber of Mines’ collective wage bargaining process). In the meantime however, exploration drilling has confirmed at least a 200m down-dip extension of the high-grade 11-block of the MRC orebody. Consequently, work is underway to develop additional production platforms to expose additional high grades panels in order to increase mining grades and flexibility, with the result that a much improved performance is expected from BGMO operations’ in the second half of PAF’s financial year.

In contrast to BGMO and Evander, PAF’s retreatment operations typically outperformed our expectations. In the case of the BTRP, although throughput declined, the head grade was maintained at a very high level, which resulted in materially higher metallurgical recoveries, while aggregate costs actually fell in local currency terms. Consequently, the BTRP produced 14,741oz of gold during the period, compared to a long-term, sustainable expectation of 10,000oz per semi-annual period (ie representing production outperformance of 47.4%):

Nameplate capacity at the BTRP is 100ktpm and management is working towards achieving this level in H217. For the moment, we continue to anticipate a moderation in grade and metallurgical recoveries. However, management has also recently been investigating the potential to reduce retention times, while maintaining metallurgical recoveries at relatively high levels. To the extent that it is successful, our forecasts may therefore prove to be relatively conservative.

In the longer term, approximately two-thirds of the tailings being treated at the BTRP originated from the Fairview concentrator at a grade of c 1.6g/t. The remaining third are from the same origin, but are supplemented with material from the Biox plant at a grade of 3-10g/t. As a result, a substantial portion of the resources being treated by BTRP exists at a grade in excess of 1.6g/t and there is therefore ample opportunity for management to pursue higher-grade strategies in order to increase financial returns.

Albeit achieved at the expense of higher costs in local currency terms, throughput, grade and metallurgical recovery at the ETRP all increased materially during the period compared to H216. Prima facie, this could be attributed to the processing of higher-grade surface material. In addition however, the plant also operated at 98.4% capacity utilisation.

The grade of the dam being re-mined at the ETRP is 0.3g/t. However, the operation is commercially viable given its ability to fill unutilised capacity in the Kinross plant such that it therefore attracts only incremental operating costs (eg 14 additional employees). Re-mining is being conducted without breaching the dam wall and the tailings are being redeposited at Winkelhaak, thereby simplifying the environmental requirements (ie the Environmental Impact Assessment) with respect to re-filling the Kinross dam at a later date.

Nameplate capacity at the ETRP is 200ktpm and management’s target is to achieve 150-160ktpm (75-80% capacity utilisation on a sustainable, long-term basis). At ambient head grades (0.32g/t) and metallurgical recoveries (45%), this should result in the production of 10,000oz per annum at a total cash cost of US$1,012/oz (Edison calculation), based on prevailing forex rates and the ETRP’s current cost base of c ZAR130m per annum. In the meantime however, management will continue to source toll-treatment material with higher grades than the ETRP’s reserve and resource grades. Note that, in this respect, it is at a commercial advantage in that it is the only retreatment operator in the area and is therefore (effectively) the buyer of choice (or even the buyer of last resort) for tailings assets destined for retreatment in the region.

Effectively, the ETRP represents a substantial pilot plant, designed to prove recovery and cost parameters, before the development of the much larger Elikhulu project, which is currently poised to break ground imminently, subject to the conclusion of the final financing package. Compared with the ETRP’s 2.4Mtpa processing plant capacity, Elikhulu will process 12Mtpa to produce c 50koz per annum, at a cost of c ZAR1.7bn in initial capex and ongoing costs of c US$398-504/oz in opex.

Phoenix returned to profitability at the EBTIDA level during H117, despite a material decrease in head grade owing to re-mining from the lower grade Elandskraal/Kroondal tailings facility, as opposed to Samancor’s Buffelsfontein facility, and the fact that re-mining was limited by the recent drought in South Africa’s North West province. Compared with H216 however, metallurgical recoveries were materially higher following the installation of high energy agitation cells in the plant, as were sales of platinum group metals, while aggregate costs increased by only 3.6% in local currency terms.

Effective from 1 July 2016, the life of operations at Phoenix has decreased to nine years as a result of the cessation of mining operations at Lesedi, following IFM’s business rescue plan. Note that the right to the PGEs contained within the Lesedi resource continues to reside with Phoenix. Hence, the life of operations is likely to return to c 28 years in the event that the mine is re-opened once again. Nevertheless, it creates the risk of an impairment to PAF’s carrying value for Phoenix at the financial year end (albeit, this is clearly a non-cash cost). It also imposes on Phoenix a requirement for a new, permanent tailings storage facility (at an estimated cost of c ZAR30-40m). Nevertheless, Phoenix remains a ‘strategic’ investment for Pan African, providing it with an entry into the PGE market. It will also benefit in future periods as a result of a new offtake contract, signed with Northam Platinum, which lowers chrome penalties associated with the refining of Phoenix material.

Having been purchased effective from 1 April 2016, there are no meaningful comparative measures for the performance of the Uitkomst colliery under Pan African management. During H117 however, it was reported to have sold 327kt of coal from own and third-party sources at an average price of US$49/t after incurring production costs of US$41/t. The colliery’s contribution to Pan African’s EBITDA was reported to be ZAR38.0m, on which basis we estimate that it made a contribution to PAF’s EPS in the order of 0.1p/share during the period.

Pan African’s strategy is not to hedge its gold production forward, except in specific circumstances and to provide protection against specific risks. In July 2015 however, in order to protect its operational revenue, Barberton Mines entered into a short-medium zero cost collar via the following instruments:

As a consequence, the decline in the price of gold, from ZAR625,886/kg on 30 June 2016 to ZAR506,917/kg on 31 December 2016, resulted in a mark-to-market fair value profit of ZAR90.0m (£5.3m) in the period under review, compared with a loss of ZAR40.6m (£1.8m) in the prior year period, which is included in ‘other’ expenses in PAF’s income statement. Based purely on intrinsic value, a profit of this magnitude implies a net open position in the order of 24,000oz (which is as expected, given that some of the contracts will have been delivered into).

At the time of writing the rand price of gold is ZAR521,547/kg and we expect this to increase modestly to US$1,275/oz, or 535,457/kg, by 30 June 2017, resulting in a small equivalent loss in H217. Note that, for the purposes of our financial forecasting, we calculate mark-to-market fair value adjustments based on our gold price forecasts for the year in question and the intrinsic value of the remaining open option contracts (ie no time value is included). These are included in our forecasts of ‘other’ expenses in Pan African’s income statements.

In addition to its hedge position, PAF has a gold loan, valued at ZAR53.9m (cf ZAR110.4m previously) – equivalent to c 3,419oz of gold (cf 6,706oz of gold previously).

We have updated our forecasts for FY17 based on the estimates made for each of the operations highlighted above (NB FY17 forecasts now include Uitkomst). Note that, as per the forecasts detailed above for the individual operations, our group-wide forecasts for PAF are for production of 183.8koz of gold in FY17 compared to management’s guidance of 181koz. Within this context, our detailed financial forecasts for PAF for H217 and FY17 are as follows.

Note that our FY17 EPS forecast of 1.36p per share (above) compares with a mean consensus estimate of 1.94p, within the range 1.40-2.66p (source: Bloomberg, 31 March 2017). Our forecast of 2.34p for FY18 assumes a gold price for the year of US$1,248/oz and compares with a mean consensus of 2.44p within the range 2.00-3.05p.

Updating our long-term forecasts to reflect these changes, our absolute value of PAF decreases from 23.25p/share in December to 21.38p/share currently (not least as a result of the recent, preternatural strength in the value of the rand against both sterling and the US dollar). This is based on the present value of our estimated maximum potential stream of dividends payable to shareholders over the life of its mining operations (applying a 10% discount rate).

Pan African is the third-best performer of the London-listed gold miners, with its share price having risen by 111.1% in US dollar terms, since March 2010 (NB: more than one standard deviation above the mean, which has been a decline of 22.8% over the same period) and outperforming the gold price by 86.5%:

Of as much significance, PAF remains cheaper than its South African peers on at least 70% of valuation measures (ie 21 out of 30 measures in the table below on an individual company basis) regardless of whether consensus or Edison forecasts are used:

Within the global context, meanwhile, it has the third-highest dividend yield of the 39 ostensibly gold counters paying dividends to shareholders (including royalty companies):

Pan African had £33.2m in net debt on its balance sheet as at 31 December 2016 after the payment of a £17.1m final dividend in late December (cf £22.8m as at 30 June 2016, £16.2m as at 31 December 2015 and £18.0m as at 30 June 2015). Net debt currently equates to a gearing (net debt/equity) ratio of 16.7% and a leverage (net debt/[net debt + equity]) ratio of 14.3%.

Our forecasts for Pan African’s immediate capital expenditure commitments related to Elikhulu by financial year are as follows: