Cradle Resources — Update 11 August 2016

Cradle Resources has a 50% interest in the Panda Hill niobium project in a known mining jurisdiction. Unlike other projects, its geology and metallurgy are almost entirely conventional and, as such, it is poised to become only the world’s fourth producing niobium mine.

At the project level, Edison calculates a value for the project of US$365m, based on the net present value of cash-flows, discounted at 10% per year. This compares with the DFS valuation of US$404m (the difference probably being attributable to pricing assumptions). On this basis therefore, Cradle’s 50% share of Edison’s Panda Hill valuation equates to US$1.11/share (before potential dilution). Assuming dilution at Cradle’s current share price in order to fund its share of joint venture capex however, this valuation declines to A$0.70/share (post-dilution).

In quantitative terms, Cradle’s sensitivity to a number of key parameters may be summarised as follows:

Edison’s base case scenario valuation of A$0.70/share assumes that Cradle will meet its equity funding obligations exclusively via the issue of its own equity. To the extent that it is able to fund its equity funding obligations via its own debt however, the value of its effective interest in the project may be geared as follows (albeit attended by increased financial risk):

In qualitative terms, the principal risk to which the Panda Hill project is immediately exposed may be described as execution risk. Once in production, however, this risk will abate and other risks (eg commercial, commodity price, foreign exchange and global economic) will become relatively more significant. In the meantime, a number of circumstances mitigate the immediate technical risks. In geological terms, the ore body is almost entirely conventional, while its near-vertical orientation means that it is amenable to low-cost, bulk mining techniques. In the meantime, the grade is comparable to an underground mine. Finally, geological risk mitigation also exists in the form of the relatively close-spaced 50m × 50m drilling. Metallurgical risk is mitigated by the fact that the majority of the Panda Hill niobium mineralisation is in the form of pyrochlore and lesser columbite within fresh to moderately weathered carbonatite lithologies. As a result, the process flow sheet is also almost entirely conventional.

The Panda Hill project is located on three mining licences, which were granted to Panda Hill Mines in November 2006 and cover a total area of approximately 22.1km2. Cradle became involved in early 2012 (in its incarnation as Verona Capital). After a corporate reorganisation, which involved transferring the mining licences from Panda Hill Mines to a company called RECB and finally Panda Hill Tanzania, Cradle became 100% owner of the project via its Panda Hill Mining Pty subsidiary in November 2015. In the meantime, in June 2014, Cradle concluded an agreement with Tremont Investments (backed by Denham Capital), whereby Tremont would fund the project to definitive feasibility study via a US$20m investment in return for a 50% interest in the project. The title to these licences (which allow the owner to search for, mine, dig, mill, process, transport, use and/or market niobium or other minerals found to occur in association with that mineral) has since been extended for a further 10 years to November 2026.

For a fuller background note on niobium, please see Niobium – The envy of the gods. The growing recognition of niobium as a vital new-age metal with limited supply was reflected in Anglo American’s recent sale of Catalão. The auction process was keenly competitive and resulted in Anglo selling Catalão for US$1.5bn, approximately 50% more than prior market forecasts. Two years earlier, in 2014, Magris Resources (50% owned by Cheung Kong and 50% by the Canadian Imperial Bank of Commerce) and Temasek acquired Niobec from IAMGOLD for c US$500m. In 2011, Posco and Nippon jointly invested c US$2.0bn for a 15% interest in CBMM (Companhia Brasileira de Metalurgia e Mineração), followed by a Chinese consortium comprising CITIC, Baosteel and Shougang on similar terms.

Niobium, formerly columbium, is a transition metal in the vanadium family (Group 5, according to the new IUPAC system of classifying elements) with symbol Nb, located between vanadium (above) and tantalum (below). Niobium metal is produced by the aluminothermic reduction of high purity niobium oxide with aluminium, lime and fluorspar. Ferro-niobium is produced from pyrochlore/columbite concentrates to which hematite (iron oxide) and aluminium are added to produce a 66% ferro-niobium matte.

Niobium’s first use was as filaments in incandescent lamps, although it was quickly rendered obsolete by tungsten, which has a higher melting point. In the 1920s, however, it was discovered that niobium improves the strength of steel, and this application remains its predominant use, with c 89% of the market for niobium accounted for by ferro-niobium.

As with most other metals, the major industrial nations represent the largest market for ferro-niobium (and therefore, by extension, niobium), with the exception China, which is the world’s largest consumer.

Supply: over 93% of official, global niobium resources occur in Brazil, with the balance being in Canada. Numerous other carbonatite-hosted niobium deposits are known across the world, the most significant being the Tomtor deposit in Siberia and the Morro dos Seis Lagos deposit in Brazil. Other countries with unquantified niobium resources include Egypt, Malawi and Greenland.

As with its resources, niobium production is concentrated in Brazil, which accounts for 92% of the world’s supply, the majority of which (c 85%) is controlled by one private company, CBMM. The balance of the world’s supply is almost exclusively produced in Canada (8%). As a result, Brazil is also the world’s leading producer of ferro-niobium. From 1995 until 2011, the USGS estimates that global output of ferro-niobium increased at an average compound growth rate of 8.2% per year, before stabilising in 2012.

Demand: since 2001, global apparent ferro-niobium demand has increased at a compound average growth rate of 6.9% per year, notwithstanding a material contraction at the time of the global financial crisis in 2009. While, currently, about 10% of the steel produced globally contains niobium, that share is expected to rise to as much as 20% in future.

Niobium materials are not openly traded on any metal exchange. Key characteristics of the niobium market therefore are the important role played by bilateral, long-term contracts between buyers and sellers, which now cover about 95% of total sales. Under these, FeNb is sold directly to steelmakers, which has led to a highly stable pricing environment.

Our (conservative) view is that niobium prices (and particularly ferro-niobium prices), are expected to remain at approximately the same level as in late 2008 and early 2009 (c US$35.16/kg) and will display little volatility in coming years, primarily due to CBMM’s dominance over supply (cf an average import price in excess of US$41/kg in the years 2010-15).

The Panda Hill carbonatite intrusion has been the subject of multiple phases of exploration work over the past 65 years, initially for phosphates, but latterly (and in the majority) for niobium:

Until 1982, a total of 96 diamond drill holes were completed over 6,419m (average 67m per hole). Cradle began work on the project in 2013, drilling a further 137 holes (both diamond and reverse circulation) over 20,724m (average 151m per hole), as well as undertaking an extensive geological mapping campaign and regional magnetic and radiometric surveys. The majority of the drilling was conducted on a 50m x 50m grid, although some spacing was broadened to 100m x 100m.

An initial scoping study was undertaken at Panda Hill in 2013. In addition to Class 5 capex and opex estimates, this included metallurgical test-work, pit optimisations, mining schedules and plant and infrastructure designs. The scoping study was followed a year later by a pre-feasibility study (PFS). This envisaged a 2Mtpa open pit operation to produce an average 6,800tpa ferro-niobium (FeNb) over a 30-year life for an initial capital expenditure of US$158m. In addition to a process plant including a pyrometallurgical converter, the PFS also included an on-site, heavy fuel oil (HFO) power plant.

In June 2014, Cradle concluded an agreement with Tremont Investments (backed by Denham Capital), whereby Tremont would fund the project to definitive feasibility study via a US$20m investment in return for a 50% interest in the project. Initially, a review of the PFS was carried out to consider the consequences of the discovery of the high grade Angel zone and to take into account the effect of production from Panda Hill on the global market. As a result of this review, the decision was made to proceed with a detailed feasibility study at an ore feed rate reduced from 2.0Mtpa to 1.3Mtpa in the first four years of the project, but then doubling to 2.6Mtpa in year five with a coincidental conversion from HFO power to grid power.

On 20 April 2016, Cradle announced the results of the Cradle DFS on the Panda Hill niobium project. Compiled by MDM in collaboration with Coffey, SRK, SGS, SLR and Roskill (among others), the DFS envisaged a 1.3Mtpa open pit operation, expanding by 100% in year five, mining an average 0.68% Nb2O5 grade in years 1-10 (and 0.54% over the 30-year life of mine) to produce an average 8,200t of ferro-niobium per year at an average stripping ratio of 1.5:1 and metallurgical recovery of 61%. Initial capex was estimated to be US$196m, with a further US$93m (to be funded from project cash flows) in year four. Average life-of-mine EBITDA is estimated to be US$112m pa (US$0.83/share undiluted at that time) and the payback period to be 4.75 years. Ultimately, the DFS calculated a pre-tax project IRR of 32% and a post-tax NPV10 of US$404m (equating to US$1.50/share, undiluted, at that time, for Cradle’s 50% share of the project) at an initial product price of US$37.65/kg Nb and average unit operating costs of US$48.04/t ore processed or US$21.34/kg Nb produced.

The Panda Hill niobium project is located near the town of Songwe in the southwestern corner of Tanzania, approximately 70km (47 miles) from the Zambian border at Tunduma. It is 5km from the main Dar es Salaam-Tunduma road (the A104), 2km from the TAZARA (aka Tanzam) railway line and 8km from Songwe airport (usually referred to in flight schedules as Mbeya), which is serviced twice daily from Dar es Salaam by Fastjet. Mbeya itself is 26km away and is the site of major power infrastructure. The TAZARA rail line is designed with a 1,067 mm gauge that permits through-traffic operations with the contiguous railway of Southern Africa and has a design capacity of 5Mtpa.

Primary access to the mine will nevertheless be via an 8km laterite road connecting with the southern side of the A104.

The Panda Hill carbonatite is a mid-Cretaceous volcanic intrusion into gneisses and amphibolites of a northeast-southwest trending mobile belt. It forms a steeply dipping, near-vertical plug of approximately 1.5km (about one mile) in diameter and is partly covered by a fenitised cap, which is, in its turn, overlain by residual and transported soils. Volcanic ash over part of the complex suggests a later stage of volcanic activity.

Evidence suggests three stages of carbonatite activity, with an early-stage (Sovite) calcite carbonatite forming the core, surrounded by late-stage carbonatites composed of more magnesium- and iron-rich carbonatites. Later-stage apatite-magnetite rich rocks and ferro-carbonatite dykes are also found in the complex.

The majority of the Panda Hill niobium mineralisation is in the form of pyrochlore and lesser columbite and is found within primary (ie, fresh to moderately weathered) carbonatite lithologies. Higher-grade material up to 3% Nb2O5 is related to magnetite rich bands and flow-banding (schlieren) within the carbonatite. However, the weathered carbonatite lithologies can also contain up to 3% Nb2O5 as well.

Cradle’s updated mineral resource at Panda Hill is based predominantly on new drilling undertaken in 2013 and 2014 and is estimated using multiple indicator Kriging on 2m composites with a 25m x 25m x 5m panel to generate a recoverable estimate simulating a selective mining unit (SMU) including mining dilution of 6.25m x 12.5m x 5m. Dated as at April 2015, the mineral resource is summarised in the table below by weathering type (at a cut-off grade of 0.3% Nb2O5).

Assaying for Nb2O5 was by Borate fusion XRF carried out by SGS in Johannesburg.

To date, only approximately 40% of the area of the carbonatite has been properly drill tested with the result that the operators have an ongoing exploration target of 200-400Mt of mineralised material at Panda Hill at a grade of 0.4-0.6% Nb2O5.

In the meantime, a reserve of 20.6Mt at 0.68% Nb2O5 has been delineated, which is sufficient to support operations for the first 10 years of the mine’s operational life:

Ore reserves were estimated only on the measured and indicated portion of the resources. The average cut-off applied was 0.46% Nb2O5, although the actual cut-off grade varied from year to year. The ore reserve was achieved by creating a MineSight model from the Multiple Indicator Kriging (MIK) resource model and was driven by the Whittle Optimisation work and the detailed mine design and scheduling. The mining schedule includes a 5% mining loss, with the mine dilution incorporated through the MIK model.

The mining study relating to Panda Hill was conducted by SRK based on conventional open cut mining based on drill, blast, load and haul using a typical medium fleet arrangement of two 120 tonne excavators and 6-18 ninety tonne haul trucks (eg CAT 777). The pit design incorporates six pushbacks – the first three occurring in the first 10 years of operations and the second three occurring in the subsequent 20 years. For the first 10 years of operations, run-of-mine material will be sourced exclusively from the measured and indicated categories of resources, with inferred resources and low-grade material being stockpiled together for potential subsequent processing. In addition, sub-optimal metallurgical material will be stockpiled along with waste material, albeit in a dedicated area, so as to be separately accessible if required. The effect of this strategy is to ensure the earlier mining and processing of higher-grade material, although it does so at the cost of a higher initial stripping ratio. After the first 10 years of operations, inferred material is also assumed to be directed towards the processing plant.

Initial crushing is effected by a primary crusher and two-stage SAG-ball mill combination. The milled product is then de-slimed and subjected to magnetic separation to remove magnetite before entering the staged flotation circuits. These comprise a pyrite float, a calcite float (with cleaning and de-watering) and then the niobium float to produce a c 40-45% concentrate. Impurities are removed by a two-stage leach process that removes phosphates and sulphur. The final leach residue is then dried and fed to a DC furnace for standard grade ferro-niobium production using aluminium as a reductant.

Significantly, the flotation process is similar to Niobec in Canada, which has a similar geology and mineralogy to Panda Hill primary material, while the leach process is based on Catalão and the converter process on some of the principles from the final stage of the CBMM pyrometallurgical circuit.

The project has excellent infrastructure. Apart from the TAZARA rail line, the Tunduma Highway and the airport, there is also a dry port and a major fuel depot, both located at Mbeya.

Currently, two power lines run from Mbeya to Songwe. One provides power to the town of Songwe; the other is a dedicated power line to Lafarge’s cement factory (6km away). Initially, the plant was designed to operate on power supplied by heavy fuel oil (HFO) generators (which will be leased). However, TANESCO is planning to build a new 400kV power line that will run past the licence area. As a result, management expects to connect to the national grid in the fifth year of operations, with the result that energy costs should decline from 18.8c (US) per kWh (based on current fuel prices) to 8.5c once the connection to the grid occurs. A capital amount of US$15m has been included in the capex schedules under Infrastructure (see below) to account for this contingency in the form of a sub-station and dedicated plant line.

Water demand for the project is anticipated to average 0.8Mm3 in the first four years of operations, rising to 1.3Mm3 from year 5 onwards, with the plant expansion to 2.6Mtpa. The project has three potential sources of water: water harvesting & storage, the Songwe River and boreholes. Surface water studies have indicated that the initial stage of operation will be self-sufficient in water if enough storage is provided to collect water during the wet season to feed into the process during the dry season. Once the expansion occurs, the balance is most likely to be supplied by the Songwe River (for which Cradle has a modest extraction permit). Any additional requirement will be supplied by a potentially large aquifer to the southeast of the tailings storage facility.

Notwithstanding the proximity of the TAZARA rail line, Cradle’s preferred logistics route is via road from Johannesburg and Botswana (for equipment and steel for construction) and via road to Dar es Salaam (for imported consumables and finished product). Panda Hill’s average output of 8,200t of FeNb per year equates to c 683t per month or c 23t per day, which will be transported in shipping containers on 20t trucks. By contrast, the volume of incoming and imported consumables will be much larger. Nevertheless, Cradle estimates this should be able to be accommodated by no more than 20-25 twenty tonne trucks running on a continuous basis per month, adding no more than 1-2% to the total number of trucks already using the road.

Security of the finished product is not perceived to be a significant risk given the absence of a ready local market for ferro-niobium.

As per our separate niobium report (also summarised on pages 3-4 of this note), the long-term ferro-niobium price is assumed to be US$35.15/kg (cf Roskill’s estimate of US$40-45/kg, depending on region).

Operational assumptions are based on the definitive feasibility study. Ore delivered to the mill for processing is assumed to start at a rate of 1.3Mtpa in CY18/FY19, rising to 2.6Mtpa in CY22/FY23. The head grade varies, but generally declines and averages 0.54% Nb over the 30-year life of the mine:

Metallurgical recovery is forecast to be 61%.

Capital costs estimates were prepared by MDM Engineering with input from SLR Consulting (for the tailings and water facilities) and SRK Consulting for mining. The plant is to start up on power supplied by heavy fuel oil (HFO) generators, which will be leased, with a capital amount included in the year-four expansion estimate of US$15m (included within Infrastructure) for connection to the national grid. The estimates were finalised in Q116 to an accuracy of -10% to +15%, as follows:

Initial capital expenditure was assumed to be phased over three years in the proportion one third, one half, one sixth. Expansion capital expenditure is all assumed to be expended in year four. Sustaining capital expenditure is estimated to be US$4.1m per year; the majority relates to tailings dam lifts.

In addition to the above, US$60m of closure costs related to the plant, tailings storage facility, waste rock dump and stockpiles has also been assumed.

Operating cost estimates were similarly prepared by MDM Engineering with input from SLR Consulting (for the tailings and water facilities) and SRK Consulting for mining. They were also finalised in Q116 to an accuracy of -10% to +15% (although without the inclusion of any contingency):

Initially (ie for Phase 1), the Panda Hill project is assumed to be funded 50:50 debt:equity, implying total equity funding of US$102.2m (including funding working capital), split equally between its two partners, Cradle and Tremont. Phase 2 is then presumed to be funded internally via project cash-flows.

Our base-case scenario assumes that Cradle will meet its equity funding obligations exclusively via the issue of its own equity to the same extent. Inasmuch as it is able to fund its obligations via its own debt, however, it may be able to increase its effective financial gearing to the project (this is considered in the Sensitivities section, below).

Additional assumptions include a 30% Tanzanian mining tax rate and a 3% government royalty and a 0.3% local government levy on gross revenues. Capital is depreciated over five years (in line with Tanzanian legislation) and peak working capital is assumed to be US$8.7m (which equates to approximately 24 days of initial turnover).

Cradle accounts for its 50% interest in the joint venture using the equity method of accounting. As such, Panda Hill will be reflected in just a single line in Cradle’s accounts in the income statement and balance sheet with three further lines in the cash flow statement (predominantly related to funding).

In the meantime, we calculate a value for the project of US$365m based on the net present value of cash-flows, discounted at 10% per year. This compares with the DFS valuation of US$404m – the difference in the valuations probably being attributable to different pricing assumptions (the DFS using pricing assumptions being provided by Roskill, which were not disclosed for reasons of commercial sensitivity). Note that at a flat, real niobium price of US$36.58/kg, Edison’s financial model dovetails with the DFS financial model such that both yield a project NPV10 of US$404m. This price compares with a known starting price (derived from Roskill) of US$37.56/kg used in the DFS – possibly implying that Roskill calculated a declining flat, real niobium price over time.

On this basis, therefore, Cradle’s 50% share of Edison’s Panda Hill valuation equates to US$1.11/share (before potential dilution) vs US$1.50 based on the DFS valuation at the time of Cradle’s DFS results’ announcement on 20 April. However, assuming dilution at Cradle’s current share price to fund its share of joint venture capex (ie via the issue of 254.7m shares at A$0.26 to raise A$66.2m, US$51.1m, or 1.5x the company’s existing market capitalisation), this valuation declines to A$0.70/sh (post-assumed dilution), as shown below:

Note that in the maximum case (when the highest grade ores are mined) EPS and maximum potential DPS projections approximate the current share price.

In qualitative terms, the Panda Hill project is immediately exposed to geographical/sovereign, geological, metallurgical, engineering, financing and management risks. In general terms, these may be summarised as execution risk – ie management’s ability to bring the project to account within its geographical jurisdiction and the required technical parameters. Once in production however, these risks will be perceived to have reduced and others, such as commercial, commodity price, foreign exchange and global economic risks will become relatively more significant. In the meantime, however, a number of circumstances mitigate the initial risks identified.

From a quantitative perspective, variations in Cradle’s discounted dividend net present valuation with metals prices, costs, the discount rate and foreign exchange rates, respectively, are shown in the tables below.

The 23.0% discount rate implied by the current share price compares to the universe of mining companies followed by Edison over time, analysed by stage of development, as follows:

Note that for the purposes of this analysis, 1 denotes companies at scoping study stage, 2 denotes companies at PFS stage, 3 denotes companies at BFS stage, 4 denotes companies that are financed, 5 denotes companies that are in development, 6 denotes companies which are in production via a subsidiary asset and 7 denotes companies that are in production via their main asset.

With respect to foreign exchange rates, the relationship between the Australian and US dollar is one of conversion between (effectively) US dollar-denominated earnings and an Australian dollar-denominated share price.

Finally, as noted above, Edison’s base-case scenario assumes that Cradle will meet its equity funding obligations exclusively via the issue of its own equity to the same extent. To the extent that it is able to fund its equity funding obligations via its own debt however, the value of its effective interest in the project may be geared as follows (albeit attended by increased financial risk):

The same analysis, performed with a sensitivity related to the share price at which new funds are raised is then as follows:

Cradle had A$2.865m in cash as at 31 March 2016. Since then it has raised an additional A$2.76m (before expenses), which should provide ample funds to advance it to a final investment decision in H117 (the second half of CY16), subject to the conclusion of offtake agreements and financing arrangements. In the meantime, Cradle intends to undertake a number of critical tasks for relatively minor costs, which could then reduce the execution schedule to c 18 months (vs 21 months) to around March 2018.