Harmony’s current CEO, Peter Steenkamp, was appointed in January 2016. His previous
experience includes an unfortunate episode as CEO of the JSE listed Pamodzi Gold. Under
his leadership Pamodzi acquired several marginal gold mines in South Africa at elevated
prices – mostly against debt. The group was finally liquidated in 2009, after investors
showed little appetite to invest in Pamodzi shares. This was in the post-Lehman’s phase,
with gold breaking through $1000/oz, but the Rand exchange rate recovering strongly and
the mines battling with electricity supplies and cost pressures.
The decision to acquire Moab is Steenkamp’s third important strategic move since taking
office at Harmony. The first was to high grade the Kusasalethu (Elandsrand) and
Masimong (Erfdeel) mines, sacrificing life for cash flows, which we applaud.
The second was the controversial decision to acquire the other half of the Hidden Valley
gold and silver mine in Papua New Guinea from Newcrest for $1. Hidden Valley has a
history of weak management, unpredictable output and of consuming a never ending
stream of cash. Steenkamp’s new strategy at Hidden Valley is to invest in a major push
back of the open pit, reconfigure the ore belts, invest in larger yellow equipment and
dramatically increase production to achieve a lower unit cost. For a small group such as
Harmony, this is a do or die strategy with significant risk. The risk of failure also extends to
Harmony and Newcrest’s reputation in PNG, where they hope to build the large Wafi
Golpu underground Copper/Gold mine. (See below)
Our fundamental thesis is that Steenkamp could easily have walked away from the Moab
deal or bid a lower price, given the extent of the environmental liabilities and scarcity of
cash flush, likely buyers. The willingness to overpay with borrowed money for a
challenging asset, seems eerily reminiscent of the Pamodzi story.
Moab shaft (Middle Mine)
The Ore Reserve as postulated by SRK, basis 1 January 2018 is 1.37m ozs, with the budget
showing a significant improvement in ore grade, as tonnage declines over the next 5 years.
While we’re skeptical of this ability to raise the recovered ore grade from 7.5g/t to 9.2 g/t
(2017 and 2021), we used these in our model. Where we differ from SRK is in our estimate
of unit cost per ton of ore. We expect unit costs to already climb this year, as underground
volumes drop off and potentially if Kopanang’s ore is no longer treated at the Great
Noligwa metallurgical plant.
On that basis, the model delivers highly taxed earnings in 2018, 19 and 20, before
potentially showing losses in the last 2 years of operation. Any plans to treat the Mispah
slimes dam will therefore have to be executed rapidly, so as to avoid missing the tax shield
of the next 3 years. By spending capex on Mispah, net cash will be reduced in the short-
term.
Great Noligwa shaft pillar
On paper, the Great Noligwa (GN) shaft pillar is a great concept. Instead of mining bitty
blocks of ore far from the shaft, the project can be accessed in close proximity to the shaft