Description

Jupiter Mines                               

Summary

Jupiter owns a 49.9% stake in a major open-cut manganese mine in South Africa. The shares are trading at an historic dividend yield of 21%, EV to EBITDA of 2.1 and a pe ratio of 3.3 (9m basis). Given further strong cash flows since the interims, there’s a possibility of a further dividend of 2.5c that would take the yield for the year to 30%.

While the manganese price appears to have peaked, this high quality long life mine is trading at a fraction of the valuation of its peers and we expect corporate action or buy-backs in the near future that could see investors being well rewarded for buying into the current weakness. With Jupiter and the underlying mine debt free, investors should continue to earn handsome dividends, while waiting for the corporate action to unfold.

About Jupiter

Jupiter is an ASX listed holdco, owning a 49.9% stake in the Tshipi Borwa manganese mine in the Kalahari manganese field in South Africa. The other 50.1% of the mine is owned by a black empowerment consortium and OM Holdings. The mine produces more than 6% of the world production of manganese from a single, shallow open pit with a geologically straight forward ore body. While the manganese ore market is notorious for dramatic swings between feast and famine, the mine had a ROCE of 96% in the financial year to Feb 2018 and the ROCE for the current financial year should be well over 100%.

The JORC Ore Reserve of the open pit equates to a life of 29 years at the current run rate, with additional shallow underground resources for a possible further 20 years. The ore reserve is calculated at a price of $4 per dry metric ton unit of 37% Mn content lumpy ore FOB Port Elizabeth (spot is currently $5.50 / dmtu). The ore reserve is also calculated based on a pit shell revenue factor of only 0.67, which effectively means that 33% of nominally profitable ore (at $4/dmtu) is excluded from the mine plan by choice, to maximize short-term profitability and NPV.

Tshipi’s ore grade of 36.5% lumpy ore, is regarded as a medium ore grade, between South 32’s Gemco mine in Australia (46%) and Kazzinc’s Zhairem polymetallic mine (24% Mn). China has several manganese mines with even lower ore grades in the 15% - 20% range, but it’s difficult to find reliable information on them. The advent of Tshipi and other new entrants has displaced many of these old, deep and marginal Chinese mines in a shift reminiscent of the displacement of Chinese low grade iron ore mines by the likes of Fortescue.

The Tshipi mine was built from scratch in 2010, exported first ore in late 2012 and repaid its shareholder loans in late 2017. The ore body sub-crops at 70m below surface and is 15m – 19.5m wide. The ore body dips at 7 degrees to the northwest, yielding a Life of Mine strip ratio of 10.4. The ore only requires a simple dry crushing and screening to produce the saleable product.

Tshipi is contiguous to South 32’s Mamatwan mine and the companies have agreed to extract the boundary pillar between them starting in 2019 up to 2031. This will effectively merge the 2 pits at ore level. This area is both shallow and high grade and will monetize the significant block of ground that would otherwise have remained sterilized due to the angle of the 2 adjacent pit walls.  

The Manganese market

90% of Manganese is used in the making of carbon steel, with the balance used in specialty steels, alkaline and NMC batteries and chemicals. Only 10kg of Manganese is required in the making of 1 ton of steel, driving off unwanted phosphorous and sulfur and improving the ductility and hardness of the steel. That represents a cost of less than $20 relative to a ton of steel worth ~$500/ton (spot rebar steel).

Manganese is not traded on any exchange, with the mines selling their product directly to customers or via intermediaries such as Glencore. The market has therefore seen extreme price volatility viz. the spike from $3 to $18/dmtu between 2006 and 2008 and the subsequent return to $3 in 2009. More recently, the price fell to $1.30 in late 2015, before recovering to $8.00 in late 2016. (See graph below for the period April 2015 – Jan 2019).

Source : Jupiter Mines website

Exacerbating this price volatility, is the ebb and flow of demand for volume. In other words high prices generally indicate a shortage of material, allowing producers to sell more tonnage, subject to production and infrastructure constraints. Adding further to this formula, Tshipi produces a medium grade product that is easier to sell when demand is buoyant, meaning that it can gain market share when times are good. The opposite holds true in equal measure.

As the following graph from Ferroglobe (Nasdaq : GSM) shows, the Manganese ore cost only comprises 27% of the overall cost of producing manganese based alloys – the actual feedstock used in steel. (Ferroglobe Business Update Sep 2018 slide 11)

With Chinese smelters adding significant new manganese alloy capacity, the margins of the alloy producers have been squeezed, while demand for the ore has remained strong. This is the same process that has played out in the ferrochrome and other markets. This explains why the recent diversion in price between manganese alloys and the ore has been slow to correct.

Tshipi’s flexible business model

Given the extreme feast and famine cycles described above, it’s comforting to know that Tshipi has been able to adopt a flexible business model that allows it to limit the damage into the down swings and maximize the upswings.

The model starts by having an open pit mine with sound geology and a conservative pit shell design to cherry pick the ore body rather than maximizing mine life. Added to this the mining is outsourced to a contractor (the reputable Moolman Brothers), the crushing and screening is outsourced, the railways and harbors belong to Transnet and the trucking volume is outsourced as well.

This means that the entire operation can be shut down at 3 months notice, as happened in early 2016. On the flipside, the mine has the possibility to ramp up volumes from the current 3.6mt p.a. run rate to 5.0mt p.a. although that presumes either a further switch to expensive trucking by road or additional capacity from Transnet, which is unlikely before 2021. The current transport split is 60% rail and 40% road, with road being US$0.50/dmtu more expensive than rail. At the current 60/40, the mine’s cash cost of ore at the Port Elizabeth harbor (FOB) is $2.30/dmtu, while spot is trading at $5.50/dmtu.

So why is Jupiter so cheap?

The Tshipi shareholders tried to sell the mine in late 2017 and found 10 bidders. (International Manganese Institute Annual Review 2017, p.10) However, the shareholders (Jupiter and Ntsimbintle) couldn’t agree on price and the process was abandoned. Pallinghurst then decided to re-list Jupiter on the ASX. The listing took place by way of a placing of 600m existing shares at A$0.40 each by the major shareholders out of a total of 1.95 bn shares in April 2018.

After the listing, Investec sold their entire holding of 13% in a book build via UBS at 34c in early August. Investec was an original backer of Brian Gilbertson’s strategy to build the Tshipi mine in partnership with Ntsimbintle in the early 2000’s. Investec must therefore have an exceptional grasp of the asset. While US$60m might be small beer for Investec, it’s still curious that they sold the stock less than 2 months before they could have banked a 5c dividend. The Jupiter’s announcement indicated that the sale came as a surprise to management.

The stock price has not recovered from Investor’s departure. Niggling operational setbacks have also popped up since then. First the CEO was booked off with pancreatitis, which appears to have set a whole series of events in motion. There was a minor failure of the pit wall. A breakdown of the contractor’s excavators at the pit caused a loss of 100,000 bcm (350,000 tons) of ore (5 weeks’ worth of production). A Section 54 stoppage was enforced due to injuries but was quickly lifted. Tshipi’s small but valuable rail allocation to Durban was “reallocated” (revoked) without explanation, entailing higher cost road transport. And last but not least, a crane collapsed in high winds at Port Elizabeth harbor delaying ore shipments by a month.

So the good news of a 5c dividend (compared to 3.5c forecast), strong manganese price, weak Rand exchange rate, recent weakening in the Diesel price, approval of the Mamatwan Boundary Pillar project, and the inclusion of Jupiter in the S&P ASX 300 index and the eventual return of the CEO from hospital couldn’t entice buyers back into the stock.

Jupiter’s other warts include the minimalist disclosure on Tshipi itself, the purposeful cross-disclosure of FOB and CIF benchmark prices (seemingly for competitive reasons), vague disclosure of ore tons between normal lumpy qualities and once-off sales of erzats fines, alternate reporting of figures in Australian dollars, Rands and US Dollars, the curious February year-end that complicates comparative valuations, the recent correction of a US$17.1m misstatement of the interim results, the haphazard dividend policy, the lack of clarity on future dividends vs. buybacks and lastly the risk that management are hinting at the possible purchase of shares in its BEE partner’s holdco with cash that could best have been applied for dividends or Jupiter buy-backs.

The Gilbertson factor must also be mentioned. Jupiter’s chairman – Brian Gilbertson – made a fortune for South African shareholders in the 1990’s by cleverly merging the erstwhile Gencor into Billiton and then merging Billiton with BHP. While he was on the up, investors were amused by his abrasive personality and he could do no wrong. Since then, BHP shareholders have counted the cost. The demerger of South 32 is an almost exact undoing of the ill-conceived acquisition of Billiton.  

At his Pallinghurst vehicle, Gilbertson recently forced through the acquisition of the minorities in Gemfields on the London market in an all scrip transaction in a way that raised eyebrows. His Platmin mine is seemingly adrift with little hope of a quick turnaround and the extent of his fees and carries at every level has not gone unnoticed. Even at Jupiter, the resolution on directors’ remuneration at last year’s AGM was barely carried at 54% vs 45%. The passing of a poison pill defense structure at the AGM could also not have been a highlight for minorities.

The route to unlocking value

With no single dominant shareholder, we believe that Jupiter offers compelling value for South 32 who controls the contiguous Mamatwan mine via Samancor. Jupiter’s 50.1% BEE partner in Tshipi – Ntsimbintle – is also Samancor’s 8% BEE partner in Mamatwan. There is therefore scope for a discussion.

Assore could also view Jupiter as an attractive bolt-on acquisition. Assore owns its manganese and iron ore mines in a 50/50 structure with ARM, so is culturally comfortable with JV’s. Given Jupiter’s sound shareholder’s agreement with Ntsimbintle, Jupiter has the right to market its half of Tshipi’s ore, which is Assore’s forte. Likewise any sale of Jupiter would force Ntsimbintle to accept such an offer on similar terms or remain the new owner’s BEE partner in Tshipi (for free).

Catalysts

The declaration of a final dividend in February of more than 2c.

A further weakening in the Rand in the run-up to the general elections in April.

Catalyst

The declaration of a final dividend in February of more than 2c.

A further weakening in the Rand in the run-up to the general elections in April.