South Africa is well known for its abundant commodity endowment with massive reserves of chrome ore, platinum group metals (PGMs), diamonds and, most prominently, gold. Over a century of mining gold in South Africa has resulted in many mines reaching depths of over four kilometres, where natural rock temperatures frequent 50°C and rock pressure is high – increasing susceptibility to seismic activity.

These structural issues and consequent risks have led to declining gold production as the cost of mining has increased significantly, relative to the gold price. South Africa has now become one of the most expensive gold mining regions in the world.

A rich history
In 1886, the discovery of the Witwatersrand Basin initiated the biggest gold rush in world history, along with the establishment of “Egoli” (‘City of Gold’), or Johannesburg as we know it today.

This elliptical basin stretches 400km across the Free State, North West and Gauteng provinces, and is the largest gold ore deposit ever discovered. It has produced over 1.5 billion ounces of gold to date, representing over a third of cumulative global gold supply. An additional 1 billion ounces of ore (11% of total world reserves) still lies in the ground as at the end of 2019 (chart below).

Gold mining in South Africa is tarnished
The deep-level labour-intensive nature of conventional South African mining results in structurally higher mining costs than the rest of the world. Over 70% of costs are related to labour (costs have risen faster than inflation) and electricity (prices have trebled over the last seven years). In contrast, large Australian gold mines operate at lower costs as they are mechanised and shallow, therefore attracting additional investment and further exploration. Australia currently has the largest reserves of gold in the world.

The steep cost of mining in South Africa has been exacerbated by declining grades, meaning that for every tonne of ore mined, fewer ounces of gold are extracted. Furthermore, seismic activity is more prevalent in deeper mines where increased ground falls pose greater safety risks for workers. South Africa continues to have some of the highest mining-related injury and fatality rates in the world, which together with poor safety conditions, impacts overall productivity and adds to the already high structural costs.

At the present run rate of production, it is estimated that the local gold mining industry will see the last ounce of gold mined in approximately 11 years’ time. Interventions that could prevent or decelerate this decline include:
• research and development around more efficient ways to mine at great depths;
• the introduction of mechanisation to improve productivity and safety; and
• stabilising electricity pricing and ensuring consistent long-term supply by using sustainable, cheaper alternatives.
These options, however, require upfront investment that the industry cannot afford at present and which are not assured of success.

A sunset industry carries a cost
South Africa’s life-of-mine forecasts are dependent on various assumptions including the future gold price and mining costs. If these assumptions prove too optimistic (higher than expected costs and/or lower than expected gold prices), mines could face earlier closure as they become unprofitable. The act of closing a mine, though, is complex and costly.

To close a mine, employers need to retrench and pay all employees. With the gold industry currently employing over 95 000 people, this prospect carries a staggering price tag. Once a mine has stopped operating, the site must be decommissioned and safely returned to its original, natural state. This involves the removal of waste and hazardous materials, the reinstatement of indigenous vegetation and ensuring that underground water quality is normalised.

In South Africa, a further complication is the underground interconnectedness of a multitude of gold mines and the challenges posed by the abundant underground water reserves found in the western region of the Witwatersrand Basin, near the town of Carletonville. The chart below indicates the various mining companies operating in this region. Some mines have stopped operating and been placed on care and maintenance (limited operations where mostly rehabilitation and decommissioning is conducted) while awaiting closure.

Working mines ensure that the mine is dewatered (removal of groundwater) so that employees are safe when working underground, with the water used elsewhere in operations. Closed mines can be safely flooded due to the absence of workers on site. However, if a neighbouring mine believes that this may impact the safety of their employees, they can interdict the closure of the other mine for safety reasons. A successful interdict means that the mine owner must continue dewatering the mine for the benefit of, and until the closure of, the neighbouring mine.

With each mine operating at different cost levels this presents a significant consideration for mining companies. It could potentially necessitate merger and acquisition activity of neighbouring mines for the purpose of reducing closure liabilities.

South African mining companies are required to prepare for closure by building up funds for the decommissioning and rehabilitation of the mine. These funds are only available if a closure certificate is granted by the government, in which case the mining company passes the dewatering charges on to the government if the rehabilitation funds are insufficient to meet the requirements. Recent amendments to mine closure legislation state that companies will be liable for environmental rehabilitation costs regardless of having being issued a closure certificate.

Mines placed on care and maintenance do not generate cashflow, yet still need to cover hefty rehabilitation and decommissioning costs that must be funded by other cash-generating mining operations. Furthermore, the social impact of mine closures in South Africa is hugely negative as every mine employee typically supports approximately ten dependents.

Gold prices are now moved by sentiment
Though South African mines structurally now operate with such high costs, it is still possible that they can make great cashflow-generating investments if the rand gold price can remain high enough for the foreseeable future – in order for them to produce strong profits for shareholders. In formulating a long-term gold price outlook to interrogate this possibility, we look to assess the fundamentals of the gold market: annual primary demand (eg for jewellery or industrial uses) relative to primary mined supply into the future.

Presently dominating these fundamental, somewhat quantifiable economic forces, is the build-up and release of above ground stocks of gold. This primarily takes the form of movements in private gold investment (ie bars and exchange traded funds) and official gold holdings of central banks – investment demand. The degree of producer hedging (selling in advance) also affects gold prices, sometimes materially.

Gold does not tarnish or decay and almost all the gold ever mined still exists today. A large proportion of this stock does come back to the market on an annual basis. At the right price, above ground stock is incentivised to be sold to meet annual demand. As shown below, around 26% of annual gold supply currently comes from recycling. Above ground stocks are capable of meeting 58 years of primary gold demand, rendering primary mine supply unnecessary.

Industrial demand for gold is reasonably price-inelastic but is currently relatively insignificant and has somewhat cyclical drivers and some substitutes. Jewellery demand is price-elastic, as seen in the 10 years between 2002 and 2012. The rising gold price environment saw jewellery as a proportion of demand, decline from over 70% to less than 50%. This was due to high levels of investment demand over this period, spurred on by the prospect of an increasing gold price, which displaced much of the physical jewellery demand.

We find investment demand almost impossible to predict other than to posit that it has some momentum to it, ie higher prices from increased demand create more demand and may lead to higher prices still. “Conventional wisdom” suggests that gold is a useful inflation hedge, but there is little evidence for this other than in the unique period of the late 1970’s. Gold bulls, without any inflation around, currently argue that gold outperforms when real interest rates are low. Another argument is that it is an essential hedge against the large monetary stimulus experiment being undertaken by developed market central banks. However, we find little credence to these arguments and see far better hedges against those scenarios elsewhere.

We believe that a more sustainable and predictable gold market is where the physical jewellery demand is the dominant driver and, given the price sensitive nature of this market segment, this could well mean a market with lower gold prices that incentivize jewellery demand growth again.

Based on an unpredictable and seemingly elevated gold price, and the current high costs of running a gold mine in South Africa, coupled with future closure cost realities, we find South African gold shares unattractive. Our clients, therefore, currently have very limited exposure to South African gold mines in our portfolios.