As contrarian investors we are always on the lookout for assets which we can buy for less than our estimate of their intrinsic value. As a result, much of our time is spent assessing what we regard to be fair value. Anglo American has recently entered our top 10. Rory Kutisker-Jacobson discusses the investment case.
Given the surge in commodity prices since the turn of the century, one might be surprised to learn that an investment in Anglo American has underperformed the FTSE/JSE All Share Index (ALSI) over this same period. R100 invested in the ALSI on 31 December 1999, with all dividends reinvested, would be worth R545 today, but just R392 had one invested it in Anglo American. A poor relative performance for South African investors, but then, like Anglo, aggregate ALSI earnings have benefited immensely from the commodities’ boom. For a global investor, R100 invested in the MSCI World Index over this same period would be worth just R144 today (see Graph 1). Determining whether Anglo American has been a good investment depends on where you sit. What is disappointing for some may be jubilation for others.
Of course, such comparisons also depend on when you start and end. In the build up to the global financial crisis, Anglo was the Cinderella of the market. It had significantly outperformed the ALSI and by June 2008 shareholders were sitting pretty with the share price around R550 per share (US$70). Anglo could seemingly do no wrong. Unfortunately, management was not immune to this optimism, and over the preceding two years had increased the debt on the balance sheet significantly by making what now appear – with the benefit of hindsight – to be overpriced acquisitions. When the clock struck 12 in late 2008, the subsequent deceleration was fierce. Anglo found itself in an over-indebted position, its dividend was cut, and its share price plummeted to R134 per share (US$13). What had once appeared to be a golden carriage, now looked decidedly like a pumpkin.
Since then commodity prices have recovered strongly, noncore assets have been sold, the dividend has been reinstated and the balance sheet is much stronger. Accordingly, the share price has recovered to R296 per share (US$37), but it is still some 48% below its 2008 dollar peak and trading close to its long-term relative low versus the ALSI.
Is Anglo American cheap? In our minds the investment case is mixed.
Valuing Anglo American
Anglo American can be thought of as the sum of three parts:
- A 65.3% stake in Kumba Iron Ore (Kumba), which in turn has a 74% stake in the Sishen Iron Ore Company (SIOC).
- A 79.7% stake in Anglo American Platinum (Amplats).
- The remaining assets (known as the ‘rump’ or ‘stub’) of Anglo American, which include its stake in De Beers, its coal mines in South Africa, Australia and Colombia, its copper and nickel mines in South America, the Minas Rio iron ore project in Brazil, and other non-core assets.
As Amplats and Kumba are both listed, one can think of buying one share in Anglo American for R296 as the equivalent of investing R84 in Kumba, R89 in Amplats and R123 in the ‘rump’.
Kumba was formed in November 2006, when Kumba Resources was split into Exxaro and Kumba. Had one invested R100 in Kumba Resources when it first listed in November 2001, retained one’s stakes in Exxaro and Kumba in the 2006 unbundling, and reinvested the respective dividends in each entity, one would now have an investment worth approximately R4 221. This equates to a return of 45% per year and a tenfold outperformance over the ALSI over this same period. With such spectacular growth, it may seem surprising that Kumba is trading on a historical P/E multiple of just 9.6 and a dividend yield of 8.5%, both of which appear attractive at first glance. However, we are concerned that Kumba’s current profitability is unsustainably high.
Primarily due to rapid growth in Chinese industrialisation, the iron ore price has increased fivefold over the last 10 years, leading to a massive increase in industry profitability. In 2001, the big four diversified miners generated operating profits of US$2bn from iron ore. For 2011 this is estimated to have been around US$70bn. Similarly, in 2001, iron ore accounted for 18.5% of the diversified miners’ operating profits; in 2011 it is estimated to have accounted for 67% (see Graph 2). With the industry investing record amounts of capital, iron ore projects already under construction are forecast to add one-third to existing supply. At the same time, question marks remain over the sustainability of Chinese consumption. While we cannot predict the future, knowing where we are in the cycle gives us reason to be cautious and we believe the risks are now to the downside.
In contrast to iron ore, current platinum group metal prices are below those required to generate a fair return on capital.
As a result, the platinum miners have all underperformed the market significantly. While we view this as unsustainable, we regard Impala Platinum and some of the juniors as more attractive long-term investments than Amplats at current prices. Despite excellent cost control over the past three years, Amplats’ unit costs remain above those of Impala. We have our doubts as to whether any further savings are achievable to change Amplats’ relative position. We further have doubts about whether Amplats is investing sufficient capital to maintain and indeed grow production over the long term. In contrast, Impala is reinvesting a considerable portion of current profits in sinking new shafts, which should entrench its position as a low-cost producer over the longer term. Impala is a top 10 holding.
In valuing companies on a relative basis, it can be illuminating to look not only at historical performance, current earnings and share price multiples, but also at the total absolute value that the market is attributing to them. At present, the market values Kumba at R161bn, which implies a market value of R218bn for SIOC. At present SIOC produces around 44 million tons per annum of iron ore, which is around 3% of global production. The vast majority of this comes from its Sishen mine in the Northern Cape. In contrast, the market is attributing a value of R140bn to Anglo Platinum, a company which controls roughly 40% of global platinum production.
Finally, we regard R123 as an attractive price to pay for Anglo’s remaining ‘rump’ assets. While the ‘rump’ accounts for 64% of current earnings, it makes up just 43% of Anglo’s current market cap (see Graphs 3 and 4). Amongst others, the ‘rump’ has excellent copper assets in Chile, metallurgical coal assets in Australia and a soon-to-be majority stake in De Beers, which accounts for a third of global diamond production. On a copper equivalent basis, the ‘rump’ accounts for over 80% of the group’s volume growth over the next five years. Using commodity prices well below those presently prevailing, the ‘rump’ is trading on less than 10 times our estimate of normal earnings.
Recent corporate action has underpinned the value inherent in the ‘rump’: In November, Anglo sold 24.5% of its Anglo Sur copper division in Chile for US$5.4bn, implying a value of US$23bn for 100% of Anglo Sur, including debt. Adjusting for future growth, Anglo Sur accounts for roughly 60% of Anglo’s total copper production. At current share prices, the implied market value of the entire ‘rump’ is just US$20bn. While we view the Anglo Sur price as high and unlikely to be repeated, one could then in theory buy the ‘rump’, sell 60% of the copper assets and effectively get all the remaining assets in the ‘rump’ for free.
The investment case is mixed
In buying Anglo American, one is acquiring Kumba, which we consider to be expensive, Amplats, whose prospects are good but in our view inferior to its peers, and the ‘rump’, which we find attractive. Alternatively, one is acquiring exposure to Kumba and Amplats at a considerable discount to their independent market values.
Where possible we have skewed our exposure in Anglo American towards the ‘rump’ by investing a portion of the total position in Anglos ‘stub’ certificates, which are listed on the JSE and priced daily. Issued by Deutsche Bank, the ‘stub’ certificates had a quarter end price of R123. While Anglo American is now a top 10 holding, we remain underweight relative to the benchmark.
In his personal capacity, Rory has a short position in both Anglo American Platinum and Kumba Iron Ore.