Index Member Comment - February 8, 2011: Arafura's Nolans rare earth project very sensitive to a rosy price future
Arafura Resources Ltd has been working on the Nolans project since its discovery in 1999 and 2011 marks the home stretch year that will determine whether or not Nolans should be developed as a rare earth mine. Arufura plans to open pit mine the Nolans deposit at a rate of 1,000,000 tonnes per year or about 2,700 tpd with a waste to ore stripping rate of about 1:1. The ore will be concentrated through crushing, gravity separation and flotation for a 30% mass reduction that upgrades the 3.2% head grade to a 4.6% concentrate grade. Arufura will annually transport 700,000 tonnes of concentrate by rail about 1,500 km to its processing facility at Whyalla on the coast of the state of South Australia. Arufura has hinted that beneficiation studies indicate a 6% concentrate is possible which would reduce the annual concentrate shipment to about 500,000 tonnes. The Whyalla Complex will be designed to produce 20,000 tonnes of separated rare earth oxides, 80,000 tonnes of 61% technical grade phosphoric acid, 500,000 tonnes of gypsum and 150 tonnes of uranium which must be removed as a contaminant. Arafura is in the midst of a bankable feasibility study initiated in early 2010 which it expects to complete by the end of 2011 at which point it believes it will also have approvals in hand for an open pit mining operation and concentration plant at the Nolans mine site in Northern Territory, and for the Whyalla chemical plant complex in South Australia. Arafura plans to raise the $950 million project financing in early 2012 and complete construction within 18 months to allow commissioning in H2 of 2013 in preparation for full scale production in 2014. The project output capacity is equivalent to that of Lynas' LAMP facility in Malaysia after its Phase 2 expansion which will process concentrates shipped abroad from Australia. However, Arufura's Whyalla Complex is much more complex because the concentrate will at best contain 6% TREO and will generate considerable "waste" which the company must convert into usable by-products in the form of phosphoric acid and gypsum. Assuming "mine footprint" approvals in separate jurisdictions (Arafura chose the longer route to Whyalla on the southern coast of Australia over the shorter route to Darwin on the northern coast because South Australia was more eager for the chemical plant jobs than Northern Territory) are received by the end of this year, it is something of a stretch to believe that project financing, procurement and construction will take only 18 months. A more realistic timeline would be for full production to start in 2015.
There is reason to be skeptical that Arafura will receive a positive feasibility study from its engineering consultants. As seems to be the norm for ASX listed juniors Arufura has done a very poor job disclosing technical details about the Nolans project, with the result that analysts must lift what they can from the compact summaries offered in corporate presentations. There is a stunning contrast between the 200-300 page technical reports Canadian listings such as Avalon, Quest and Rare Element have filed in support of their projects compared to the skimpy news releases that count as full disclosure for Australian listings. Oddly enough Arafura has developed a large following among German investors, which just goes to show that stereotypes such as the one of Germans as rigorous, detail oriented thinkers need not apply in the investment world where limited disclosure offers relief from the pain of straining the brain over details. German investors through JP Morgan Nominees own just under 25% of the stock; for their sake one hopes their faith will be rewarded, because as my sensitivity analysis below using Arafura's numbers shows, the stock has considerable upside despite its 384 million fully diluted capitalization if the feasibility study validates Arafura's cost and mine parameters, and current FOB spot prices emerge as the long term reality.
What worries me is the scant disclosure Arafura has made about the mineralogy of Nolans and the less than forthright manner in which it presents recoveries. The key REE bearing mineral is the phosphate cheralite, a member of the monazite group. On the plus side this means that Nolans has a much better distribution of rare earths in terms of the heavies, but on the down side it also means that Arufura will have to deal with a fair amount of thorium as well as a lesser but still nuisance amount of uranium. In addition Arafura will have to deal with the fluorine in the main ore mineral, fluorapatite, from which the phosphoric acid will be derived. Arafura generally predicts REO recovery at 86% but if you look close enough you will find it stated as 77.5%. And that is before Arufura decided in early 2010 to expand its business model from producing a mixed rare earth oxide product in the form of a rare earth carbonate to producing and selling separated rare earth oxides. The current feasibility study involves pilot plant studies of the processes Arafura plans to use to crack the rare earth minerals, strip out impurities such as thorium and uranium, and separate the oxides into individual oxides that can command the FOB spot price. I do not trust Arafura's capital and operating cost projections because these were generated before the company has the benefit of the current feasibility study. Normally in mine development it is the prefeasibility stage that sucks up the most dollars while feasibility involves more engineering and permit requirement adaptation measures than physical work. Rare earth mines are better described as "chemical plant development" where establishing the feasibility of scaling the process to a commercial level sucks up big dollars and time. For example, in December 2010 Arafura raised AUD $90 million at $1.20 to fund completion of the bankable feasibility study. Between now and the end of the year there is plenty of room for unpleasant surprises.
Rare Earth Oxide Prices US $/kg as of February 3, 2011
|Rare Earth Oxide||FOB 3 Year|
|FOB Spot||Domestic 3|
|1) FOB and Domestic average and spot prices as of January 27, 2011 - Metal-Pages|
|2) JP Morgan prices used in Sept 29, 2010 research report as basis for DCF valuation that set A $1.71 price target for Lynas, where no REO given, 3 year FOB averages were used|
The chart below shows the value breakdown of Arafura's 20,000 tpa projected REO production in terms of FOB and domestic spot prices prevailing at the end of January 2011 and applied to the recoverable distribution inherent to the deposit (Arafura does not project recovery of holmium, erbium, thulium, ytterbium and lutetium which represent only 0.21% of the Nolans REO composition). The chart makes it clear that Nolans will make only a modest contribution to the world's heavy rare earth needs, primarily in the form of yttrium. Nolans is very similar to Mt Weld in terms of output scale and composition, with the differences that Mt Weld has a four times higher grade and a much better concentration factor. Assuming that Arafura can deliver its 77.5% recovery while Lynas achieves its 52% recovery, Arafura will be a somewhat bigger supplier of neodymium, the key magnet rare earth. Both are similarly vulnerable to future REO prices in that there is a substantial difference between the revenue potential at FOB and domestic spot prices which is due to the fact Chinese export policy since mid 2010 has resulted in supply shortages outside of China for the cheaper light rare earths and yttrium. There is a wide divergence between domestic and FOB spot prices.
We have constructed an after tax cash flow sensitivity model for Nolans by assembling as best as we could from Arufura's disjointed and murky disclosures the mine and cost parameters outlined below. This involved backwards calculating output numbers, which is why in our model we ended up with a mining rate of 2,200 tpd for annual ore production of 803,000 tonnes instead of 1,000,000 tonnes. We have worked all cost numbers back into dollars per tonne of ore mined so that, if necessary, we can see what happens when grade, recovery or throughput changes, as well as stage specific processing costs. Because Arafura must dispose of all the waste material, which may not be so easy 1,500 km from the mine site if it stays as pure waste, we have assigned all processing costs to the production of rare earths, and accepted Arafura's numbers in terms of phosphoric acid, gypsum and uranium by-product credits. In Arafura's disclosures these by-products would generate $100 million in annual revenues, which works out to about $125/t in by-product credits. Subtracting this by-product credit results in a cash cost of $344 per tonne of ore processed, which works out to $13.87/kg of TREO produced. Since we cannot tell from Arafura's disclosures how much of the operating cost is attributable to producing marketable phosphoric acid and gypsum, it is difficult to tell how much higher the operating cost would be if the by-products proved non-saleable.
Cash Flow Metal Price Sensitivity Analysis
|Arafura Resources Ltd||Nolans||Australia|
|Parameter Source:||Arafura Presentations||Scenario Author:||JK||Primary Metal:||REO|
|Breakeven||Domestic Spot||Domestic 3 Year Avg||FOB 3 Year Avg||JP Morgan||FOB Spot|
|TREO Price $/kg :||$22.22||$19.39||$10.23||$20.29||$36.20||$79.81|
|Recoverable Rock Value $/t:||$551.06||$480.87||$253.70||$503.19||$897.76||$1,979.29|
|Average Annual Revenue:||$397,323,900||$346,719,641||$182,926,350||$362,812,868||$647,305,364||$1,427,111,632|
|Average Annual Pre-Tax Cash Flow:||$105,251,173||$54,646,914)||($109,146,377||$70,740,141||$355,232,636||$1,135,038,905|
|Pre Tax NPV:||$98,844,289||($327,751,208)||($1,708,533,806)||($192,084,796)||$2,206,195,896||$8,779,987,284|
|After Tax NPV:||$797,543||($338,882,383)||($1,708,533,806)||($226,060,159)||$1,553,339,274||$6,299,445,445|
|After Tax AUD Price/Share:||$0.00||($0.88)||($4.44)||($0.59)||$4.04||$16.39|
|Internal Rate of Return:||7.5%||-1.5%||0.0%||1.8%||37.6%||112.9%|
|Mine Parameters||Cost Parameters||Other Parameters|
|Tonnage||30,200,000||Capital Cost||$950,000,000||Fully Diluted||384,425,342|
|Grade||2.78%||Annual Sustaining Cost||$20,000,000||Net Interest||100.0%|
|Primary Metal||REO||Net Smelter Royalty||0.00%||CAPEX Funding||100% Equity|
|Primary Recovery||78%||Marketing Cost||0.0%||Years to Startup||1|
|Primary Production||19,914 tonnes
||Mining Cost $/t||$18.00||Mine Life||23|
|Mining Method||Open-Pit||Concentrating Cost $/t||$23.00||AUD:USD Exchange Rate||1.00|
|Processing Method||Cracking/Separation||Transportation Cost $/t||$48.00||Discount Rate||10%|
|Milling Rate tpd||2,200||Cracking/Separation Cost $/t||$362.00||Tax Rate||28%|
|Operating Days||365||G&A Cost $/t||$0.00||Payback Tax Holiday||Yes|
|Ore Mined Annually (t)||803,000||Reclamation Cost $/t||$0.00||Cost Inflator||0%|
|Waste to ore Ratio||1.00||By-Product Credit $/t||($125.00)|
|Concentrate||0.0%||Total Operating Cost $/t||$344.00||Currency||USD|
Analysis: Our sensitivity analysis shows Arafura's Nolans project is very sensitive to the future price of rare earth oxides. For example, assuming the prices JP Morgan used last fall as the basis for its Lynas and Molycorp buy recommendations become the long term FOB reality, and Arafura suffers no further dilution, the stock would be worth $4.04 per share. A mere doubling of the JP Morgan price set to the current FOB spot prices would translate into a fourfold increase to $16.39 per share. Sensitivity, however, cuts both ways. If we plug in 3 year average FOB prices Arafura has a negative $0.59 per share value, and if we plug in domestic China prices the value worsens to negative $0.88 per share, and at 3 year average domestic prices it collapses to negative $4.44 per share. In other words, for speculators to profit from Arafura at the current stock price, not only must Arafura's cost and recovery projections be confirmed by the feasibility study, but so must be the rosier projection that rare earth demand will grow robustly during the next five years, China will indeed limit its supply expansion to 100,000 tonnes per year as suggested by Dr Chen, and not every advanced rare earth mine will come on stream successfully. We do not think current FOB spot prices for lanthanum and cerium are sustainable, and, noting that Nolan's projected cerium and lanthanum output represents 55% of the total output value at FOB spot prices, we would be inclined to favor JP Morgan's price set as the long term reality, which makes $4 a reasonable price target for optimistic investors. However, as suggested above, we lack confidence in Arafura's cost and recovery disclosures, and given that the project on those numbers is sub-economic at trailing FOB prices and the current domestic spot price, we would recommend avoiding exposure to the high risk of a double disappointment in the form of bad technical news from the feasibility study and a possible retreat in FOB spot prices. While we are of the view that high FOB spot prices will prevail for several years, and that domestic China prices will rise and stabilize at levels well above the three year averages, we also acknowledge that the near to medium term supply situation is so dire that the risk emerges that the rest of the world simply gives up on rare earths and decides to do without the technologies they enable. We think this will happen unless there is a massive and collective push to finance and develop as quickly as possible the more advanced rare earth projects without the usual risk aversion concern that plagues the traditional mining industry. We think this sort of climax will happen this year, and because any bad news on the technical front is at least a year down the road for Nolans, Arafura is well positioned to trade as a proxy for the rare earth sector.