Pan African Resources — Everything falling into place

Pan African is a South Africa-based gold mining group, which has dual primary JSE and AIM listings and operates four major assets in South Africa: Barberton (target output 95koz Au pa), the Barberton Tailings Retreatment Project (20koz pa), Elikhulu incorporating the Evander Tailings Retreatment Project (55koz + 10koz pa) and Evander underground incorporating the Egoli project (currently 36koz rising to >100koz pa). Within these, it has a plethora of short- to medium-term organic growth projects and an abundance of ESG initiatives (see results presentation for details).

Pan African is cheap relative to both its historical trading record and its peers. Our core valuation of the company is 41.79c/share (30.42p/share) cum-div. Stated alternatively, we calculate that investors buying Pan African’s shares, cum-div, at a price of 17.50p per share, will enjoy an internal rate of return on their investment of 21.8% per year over the next 18 years, until FY39, in US dollar terms. In addition, our core valuation rises by an additional 13.5%, to 47.43c/share (34.53p/share) once next level growth projects – and in particular, the Mintails/Mogale assets – are taken into account. To this must then be added the value of c 19.2m underground Witwatersrand ounces, which we estimate could lie anywhere in the range of 0.22–5.24c to take the total to 47.65–52.67c/share (34.69–38.34p/share). Alternatively, if PAF’s historical average price to normalised EPS ratio of 8.9x in the period FY10–21 is applied to our FY22 and FY23 forecasts, it implies a share price of 32.15p in FY22 followed by 32.99p in FY23. In the meantime, PAF remains cheaper than its South Africa- and London-listed gold mining peers on 86.1% of comparable common valuation measures (31 out of 36 individual measures) if our forecasts are applied over the next two years or 80.6% if consensus forecasts are applied (see Exhibit 13 on page 12). Thus, applying PAF’s peers’ average year 1 P/E ratio of 9.2x to our forecast normalised HEPS forecast of 4.99c/share for FY22 implies a share price for the company of 33.3p at prevailing forex rates, while applying its peers’ average year 2 P/E ratio of 8.6x to our forecast normalised headline earnings per share (HEPS) forecast of 5.12c/share implies a share price of 32.0p in FY23.

Pan African reported total net debt of US$39.0m at end-FY21, which equates to a gearing ratio (net debt/equity) of 13.8% and a leverage ratio (net debt/[net debt+equity]) of 12.1%. Of this, net debt to financial institutions amounted to only US$23.6m, which compares to two and a half years ago, when it was completing capex on Elikhulu and net debt amounted to £102.7m (31 December 2018) and gearing and leverage to 82.2% and 45.1%, respectively. Despite capex commitments, we are continuing to forecast the company will be free of net debt to financial institutions by the end of FY22. In the meantime, the group remains comfortably within its debt covenants to the point at which they are now, to all intents and purposes, irrelevant.

Our valuation of Pan African (above) has been conducted at a real gold price declining from US$1,819/oz to US$1,524/oz over the next six years to CY28 and then flattening off. Apart from gold price risk, other risks include the sovereign risk relating to South Africa and normal geological and engineering risks (albeit more than a third of production in the immediate future is expected to be derived from tailings retreatment operations). Otherwise, normal commercial risks include exchange rate movements (in particular the rand/dollar exchange rate), regulatory risks and workforce risks.

Pan African announced its FY21 financial results within the context of known production results, which were released to the market on 13 July and only very slightly adjusted in the final results announcement on 15 September (see Exhibit 1). Overall, production in FY21 was 6,777oz (3.5%) higher than PAF’s own guidance of 195,000oz from as recently as May 2021 and as a result, the group’s production profile grew, rather than declined, in H221 cf H121.

In general, all of the outperformance could be attributed to operations at Evander underground. Output at Evander underground was restrained by a ventilation shaft lining fracture in H121 as well as technical difficulties relating to the pseudo packs used for ground support (now overcome) and lower than usual metallurgical recoveries (although this was, to some extent, counterbalanced by an elevated head grade, which, we estimated at 8.51g/t). In the event, however, Evander’s performance in H221 not only returned output levels to those expected for the six-month period, but more than made up for the shortfall in the first half, such that production for the full year was 36,016oz cf an original Edison forecast of 35,667oz (see our note, The sun rises over Egoli’s city of gold, published on 14 October 2020). Over the same period, production at Barberton underground benefited from increased mining footprints on the 256, 257, 258 and 358 platforms, thereby engendering greater mining flexibility. Throughput at Elikhulu was restricted on account of remedial work on its tailing storage facility’s lower compartment. In addition, unexpected concentrations of carbonaceous material in the lower benches of the Kinross dam negatively affected gold recoveries, which we estimate must have been in the order of 42.6% in H221 – in line with our prior expectation and also H121 and H220, but notably lower than H219 and H120 (both >50%).

The table overleaf presents Pan African’s H221 and FY21 results relative to both historic results and our prior expectations. In general, both earnings and headline earnings were within 1% of our forecasts for the full year (post-tax). Within that, the main variances were a negative US$3.3m variance in the depreciation charge, which increased more than capex (+31.2% H221 cf H121) and production (+5.1% H221 cf H121) and a negative US$6.1m variance in ‘other’ expenses, which included US$7.3m in costs incurred on the increased value of employee incentive schemes’ liabilities, consistent with the increase in the group’s share price. These were almost exactly balanced by a US$10.4m positive variance in the tax charge, of which the majority (US$7.2m) could be attributed to deferred taxes.

While normalised earnings and profit after tax were within 1% of our prior expectations, however, Pan African outperformed our underlying expectations (as measured by normalised HEPS, which excludes other income and expenses, in particular), by 6.9%, as shown below.

In the event, deferred taxes accounted for the majority (52.6%) of the total tax charge, with cash taxes amounting to only the balance of 47.4%.

While modestly outperforming our expectations for the full year, normalised HEPS of 4.54c/share was nevertheless almost exactly in the middle of the range of analysts’ expectations of 3.01–3.72p/share, or 4.16–5.14c/share (source: Refinitiv, 21 July 2021).

Barberton reaped the rewards to increased mining flexibility in H221, owing to the establishment of four platforms from which to cycle high-grade production on the MRC orebody (cf three in FY20), with tonnes milled remaining at high levels and the underground grade remaining close to its average for the last 10 years. As a result, production recorded its highest level since H116, while unit costs increased by only 3.1% (H221 cf H121) in rand terms, which was a creditable performance given that year-on-year CPI inflation in South Africa was 4.9% in June after recording a 30-month high of 5.2% in May and that the number of tonnes milled from surface sources decreased by 5.9 percentage points, from 24.3% of the total in H121 to 18.4% in H221. As a result, Barberton recorded its second highest level of adjusted EBITDA in recent history in H221 (after H121) and otherwise the highest since at least H115 and easily covered capex of ZAR230.4m (see Exhibit 3).

Full details of Barberton’s operating performance in FY21 are provided in Pan African’s results announcement. In summary, the worst depredations of the coronavirus were mitigated by: increased reserve delineation drilling on the 256 platform of the high-grade MRC orebody at Fairview to increase confidence in, and predictability of, management’s geological models; Barberton’s ability to mill ore from surface sources (requiring a lower complement of workers); and its ability to focus on higher-grade areas of the orebody after the establishment of the 257 and 258 platforms, thereby allowing four platforms to cycle (flexible) production on the MRC. The improved flexibility, resulting from accelerated underground development programmes, has now increased the face length available for mining to over 200m (cf 130m at end-FY20). At the 257 platform alone, geological mapping and reserve delineation drilling have identified mineralised widths in excess of 15m (cf the usual 7m ordinarily encountered on the upper platforms).

Since the year end, Pan African announced a three-year wage agreement with the National Union of Mineworkers (NUM) and a five-year wage agreement with the United Association of South Africa union (UASA). While there are a number of conditions to the agreements, the substantive points are that the NUM deal provides for an average annual wage increase of c 5.6% for three years ending 30 June 2024, while the UASA deal provides for 5.0% increases for the five years to end June 2026, but with the potential to be higher or lower depending on the rate of South African consumer price inflation. The blended average annual increase is expected to be 5.4% compound annually for the initial three-year period of the agreements. While price increases in South Africa (as measured by the CPI) were 3.0–3.3% between July 2020 and March 2021, they recently accelerated to 5.2% in May and 4.9% in June. Over the past six years, they have been 3.1–6.3% (simple average 4.7%). With little apparent political will to rein prices in, however, in the face of the country’s unemployment situation, inflation is expected to remain at elevated levels at or around 4.5% into the foreseeable and, as such, Pan African’s agreements represent a real-term increase of less than 1%, while at the same time conferring upon Barberton the security of operating in an environment free from the prospect of old fashioned industrial action.

Elikhulu’s performance in H221 was broadly similar to both that in H121 and Edison’s prior expectations, with the exception of the fact that unit cash costs (as measured in ZAR/t) did not decline relative to H121 as we had expected. Remedial and optimisation work on the Elikhulu tailings storage facility’s lower compartment restricted tonnage throughputs, in mitigation of which the group was required to install elevated drains on the south-western edge of the compartment in order to facilitate the removal of excess water from the facility (TSF) and to ensure sustainable operations. At the same time, the lower benches of the Kinross TSF were found to contain higher than expected concentrations of historically processed fine carbon, which adversely affected metallurgical recoveries, compounded by the mining of the coarser but high-grade outer wall of the Kinross TSF, which also acted to reduce recoveries.

As a result, whereas we calculated that Elikhulu accounted for 47% of group-wide H120 adjusted EBITDA, 91% of H220 adjusted EBITDA and 68% of FY20 adjusted EBITDA, in FY21 we estimate it accounted for 39% of H121 adjusted EBITDA, 27% of H221 adjusted EBITDA and 33% of FY21 adjusted EBITDA. Nevertheless, while adjusted EBITDA at Elikhulu was comparable to that of Barberton in FY21 (ZAR786.0m cf ZAR965.6m), its capex was still an order of magnitude lower (ZAR64.2m cf ZAR418.3m).

Capex at Elikhulu will continue to increase in FY22 as it transitions from Phase 1 of its operations (the re-mining of Kinross tailings) to Phase 2 (the re-mining of Leslie and Bracken tailings), necessitating the installation of approximately 6km of piping and a pump station between the plant and the areas to be mined. Thereafter, it is expected to produce approximately 60,000oz gold pa until FY26 as re-mining progresses from the Kinross to the Leslie and Bracken TSFs. For the final seven years of operation, while processing the Winkelhaak TSF, it is expected to produce c 50,000oz gold pa (excluding c 102,000oz inferred mineral resources delineated in the soil material beneath the existing TSFs).

In the meantime, however, Elikhulu is expected to produce 55,000oz gold in FY22, with improved throughput and higher recoveries from the planned re-mining of areas on the upper benches of the number three Kinross TSF dam (in line with its mine plan from the start of production).

Relative to prior periods, underground operations at Evander in H221 recorded a vastly improved performance, with both tonnes processed and grades increasing materially at the same time that unit costs (measured in ZAR/t) decreased materially, to result in a near seven-fold increase in adjusted EBITDA (cf H121) to a level that is almost certainly a record for underground operations at Evander since they came under PAF ownership and management in H213.

After a difficult H121, in which output was below expectations, among other things, as a result of a ventilation shaft lining fracture, successful remedial work on the shaft barrel allowed production in H221 to more than make up for the shortfalls in H121, with average production of c 5,134oz per month for each of the last three months of FY21 (ie an annualised production rate of 61,608oz pa).

The 8 Shaft pillar now has a remaining life in excess of two years and is expected to produce approximately 79,160oz of gold during this period, at approximately 39,000oz pa, after which production at Evander underground will be redirected towards Phase 1 of the 8 Shaft decline, 24 Level project and potentially Phase 2 thereafter (extending mining down to 25 and 26 Levels).

Pan African has recently added two major new projects to its existing portfolio of growth projects (already including Royal Sheba, the Fairview sub-vertical shaft, Rolspruit, Poplar, Evander South etc) in the form of the Mintails/Mogale project and the 8 Shaft decline 24 Level project.

One of the assets with the most immediate optionality in the company’s portfolio is Mintails/Mogale, which could yet prove very similar in nature to Elikhulu, and into which PAF is conducting due diligence with a view to acquiring it.

On 6 November 2020, PAF announced it had entered into a conditional agreement with the liquidator of Mintails’ assets for the purchase of the total share capital and associated loans of Mogale Gold and Mintails SA Soweto Cluster. Due diligence has been extended until January 2022. In the meantime, Pan African has successfully concluded both a fatal flaw analysis and a high-level financial evaluation of the project (which would be similar in nature to Pan African’s flagship Elikhulu project). In July, it subsequently completed a pre-feasibility study (PFS) on the Mogale Gold assets. Key outcomes of the PFS (cf the financial evaluation) are as follows:

While not explicitly reported in the results of the PFS, metallurgical recoveries were estimated to be c 53% in the initial financial evaluation (cf Elikhulu’s 48%). In the meantime, by way of comparison, investors should note that Mintails’ and Mogale’s aggregate resource of 2.36Moz compares favourably to Elikhulu’s original resource of 1.7Moz and its initial reserve of 1.5Moz, but at a fractionally higher grade of 0.30g/t (cf Elikhulu’s 0.29g/t). PAF announced the results of an independent definitive feasibility study (DFS) on Elikhulu on 5 December 2016, which demonstrated an NPV9 of US$75.9m (or, then, 5.0c/share, or US$40.95 per resource ounce) at a gold price of US$1,180/oz and a forex rate of ZAR14.50/US$. At the time, we estimated Elikhulu to be worth US$69.9m (or 4.6c/share) at a 10% discount rate and to be capable of adding 1.33p to EPS in the first eight years of its operation (albeit there are now 28.0% more shares in issue). Now, however, with capex having been expended (albeit with not all associated debt having quite been repaid), we estimate a valuation for Elikhulu of c US$140.98 per initial resource ounce or US$185.19 per remaining resource ounce. As such, and albeit with suitable caveats such as the Mintails/Mogale assets developing in a similar fashion to Elikhulu, PAF could acquire for US$1.31/oz an asset that should be worth US$9.88/oz as an in-situ resource (see Gold stars and black holes, published in January 2019), could be worth US$23.98/oz (pre-development) and may be worth up to US$208.49 per remaining ounce (or US$126.23 per initial ounce), post-initial capex and debt repayment.

Pan African has continued to maintain the integrity of the underground infrastructure at Evander even after the end of high-volume, deep-level underground mining in May 2018. While limited vamping operations have continued since then, PAF has now concluded an internal technical and economic study into the merits of mining the number 2 decline on 24 Level at 8 Shaft (Phase 1 of the project) with an option to extend mining to levels 25 and 26 at a later date (Phase 2).

An integral component of the Phase 1 study was the identification of pre-May 2018 problems at Evander underground and appropriate mitigation of the major challenges encountered during the mining of the Kinross orebody. These included:

To date, the study has yielded the following results for the project:

While not large by the standards of some of Pan African’s other recent projects, the 8 Shaft no. 2 decline 24 Level project will extend the 8 Shaft Pillar project’s 2.5-year life by a minimum of another 2.5 years at approximately the same level of production of c 35,000oz per year and has now been officially brought into PAF’s life of mine plan for 8 Shaft (and therefore our estimates for the purposes of our valuation of Pan African, below).

As a consequence of the positive concept study on Mintails/Mogale and the group’s assessment of the opportunity provided by the 24 Level project at Evander, Pan African has now reprioritised its capital expenditure programmes as follows:

As a result of the above changes to our assumptions (particularly the deferral of the development of Egoli), whereas we had forecast that group production at PAF would reach 274.5koz in FY25 previously, we now forecast that it will reach 263.8koz in FY27 – approximately two years later than originally forecast, but with intermediate production augmented by the 24 Level project (and potentially the Phase 2 25 and 26 Level projects) and without the need for third-party debt funding for Egoli.

In the light of Pan African’s operational update for FY21 and fractionally revised external factors such as the gold price and forex rates, our absolute value of the company (based on its existing four producing assets plus Egoli) has increased by 14.4% to 41.79c/share (cf 36.52c previously), on the basis of 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):

A summary of the major components of the change in our valuation is provided in the ‘bridge’ chart below:

Including its other potential growth projects and assets (namely the residual Evander underground resource and its shareholding in MC Mining), our updated total valuation of PAF is as follows:

Stated alternatively, excluding its other assets (ie the Fairview Sub-Vertical Shaft, Royal Sheba, Mintails/Mogale, MC Mining shares and EGM underground resource), we estimate that investors buying Pan African’s shares, cum-div, at a price of 17.50p per share, will earn an internal rate of return on their investment of 21.8% per year (over the next 18 years, until FY39, in US dollar terms.

Exhibit 12, below, depicts PAF’s average share price in each of its financial years from FY10 to FY21 and compares this with normalised HEPS in the same year. For FY22 and FY23, the current share price (17.50p) is compared with our forecast normalised HEPS for FY22 to FY23. As is apparent from the graph, Pan African’s price to normalised HEPS ratio of 4.8x and 4.7x for FY22 and FY23, respectively (based on our forecasts, see Exhibit 18) is close to the bottom of its recent historical range of 4.1–14.8x for the period FY10–21:

Stated alternatively, if PAF’s average Year 1 price to normalised EPS ratio of 8.9x for FY10–21 is applied to our normalised earnings forecasts, then it implies a share price for PAF of c 32.15p in FY22 followed by 32.99p in FY23.

Over the next two years, Pan African remains cheaper than its South Africa- and London-listed gold mining peers on 86.1% of comparable common valuation measures (31 out of 36 individual measures in the table below) if our forecasts are applied or 80.6% if consensus forecasts are applied.

Alternatively, applying PAF’s peers’ average year 1 P/E ratio of 9.2x to our forecast normalised HEPS forecast of 4.99c/share for FY22 implies a share price for the company of 33.3p at prevailing forex rates. Applying its peers’ average year 2 P/E ratio of 8.6x to our forecast normalised HEPS forecast of 5.12c/share implies a share price of 32.0p.

Pan African reported net debt to financial institutions of only US$23.6m at end-June 2021 (see Exhibit 16), which equates to a gearing ratio (net debt/equity) of only 8.3% (cf 24.5% at the interim stage) and a leverage ratio (net debt/[net debt+equity]) of 7.7% (cf 19.7% at the interim stage). Despite capex commitments, we are continuing to forecast the company will be free of net debt to financial institutions by the end of FY22.

Including all other components, total net debt amounted to US$39.0m, made up of the following components:

This US$39.0m level of net debt equates to a gearing ratio (net debt/equity) of 13.8% and a leverage ratio (net debt/[net debt+equity]) of 12.1%.

Relative to H121, lease liabilities remain almost unchanged, at US$5.3m (cf US$5.0m). In addition, during FY21, PAF entered into a loan with Redink Rentals (RF) to fund the solar photovoltaic renewable energy plant located at Evander Mines. For the purposes of our financial modelling, debts to non-financial institutions, such as lease liabilities, are included in ‘Other long-term liabilities’ in Exhibit 18, below.

In the meantime, the group remains very comfortably within its revolving credit facility debt covenants to the point at which they are now, to all intents and purposes, irrelevant: