Description

SUMMARY

Carpathian (CPN) is one of the most convex risk/reward opportunities in the gold mining space.  Downside should be fairly limited at current prices with upside potential of >5x over the medium term.  

• Current enterprise value: US$135 million
• Two assets:

1. RDM (Brazil):  gold mine currently in construction => production of ~100koz p.a.; 8 year mine life (based on reserves only); cash costs: $700-750/oz; in production from Q2/Q3’13; funding in place to complete construction of the project.  Conservative valuation for this asset alone comfortably exceeds CPN’s current EV (more detail below)
2. RVP (Romania):  Gold/copper deposit with >7Moz gold only resource (~11Moz gold equivalent resource).   Robust economics based on Preliminary Economic Assessment (NPV 10% >$1bn).  CPN targeting production by 2016.  Reasonable valuation is several multiples of current EV (more detail below)

• Barrick (largest gold miner in the world, >$50bn EV) acquired 9% of Carpathian in July 2011 at $0.52/share (68% premium to today’s price).  This deal was structured to provide Barrick a strategic foothold on CPN’s Romanian asset which is of interest to Barrick longer term (Barrick are assisting with exploration and technical activities at the project as part of the deal that was agreed – more details on the deal below)
• CPN has put in place some gold and FX hedging for the Brazilian asset which partially mitigates the downside risk of a sharp fall in the metal price or a sharp rise in the BRL
• Thesis: the undemanding valuation and near term cash flow at RDM (Brazil) comfortably underpin CPN’s current valuation.  Upside is provided by the much larger RVP project (Romania) which contains a significant gold resource, good project economics and is a project of interest for other gold miners (e.g. Barrick interest).  At CPN’s current valuation you are essentially getting RVP “for free” and it could be worth many multiples of CPN’s current market value
• Catalysts: there are a number of catalysts over the next 12 months

OVERVIEW

Carpathian is a junior gold miner with two 100% owned projects, Riacho dos Machados (RDM) gold project in Brazil and Rovina Valley (RVP) gold and copper project located in Romania.  The RDM mine is currently under construction and scheduled to commence production in Q2/Q3’13.  RVP is an earlier stage (although much larger) project.  Carpathian is currently completing a pre-feasibility study on RVP, which is due to be released in Q4’12, with a longer term intention of bringing the project into production in 2016.  Company wide Carpathian has gold resources of >8 million ounces and copper resources of >1.5 billion pounds. 

(1)     Riacho dos Machados (RDM) (Brazil)

For those interested there is a significant amount of descriptive and financial information about the project contained in the RDM feasibility study, a copy of which was filed on SEDAR in June 2011 (http://www.sedar.com/CheckCode.do;jsessionid=00002RVzEojL5TICXt1CxxmOlVK:-1)

In summary, the project is located in South-eastern Brazil, approximately 145km by road north-east of the city of Montes Claros in Minas Gerais State.  The operation is fairly simple – mining will take place via an open-pit mine and processing will be through a conventional Carbon In Leach (CIL) processing plant.   In addition to the open-pit reserve and resource the RDM property currently has a small underground resource and further potential to expand both the open-pit and underground resources.   Neither the feasibility study, nor my own valuation (to be conservative), ascribes any value to anything other than the current reserve (i.e. anything beyond the current known reserve would be upside to the valuation). 

Below is a summary of the material economic points of the project based on the feasibility study:

• Mineral Reserve:  830 koz (reserve was calculated based on a $950/oz gold price)
• Production: avg. ~100 koz of gold production p.a. for the life of mine
• Pre-Production Capex:  $160m (includes $11m contingency)
• Sustaining Capex: $20m in aggregate for the life of mine (avg. of ~$2.5m p.a.)
• Cash Costs: $558/oz (pre-royalty*)

* The amount of royalties paid will depend on the gold price.  Based on the gold price assumption used to calculate project economics for the feasibility study ($1,150/oz) cash costs including the royalty payment was $593/oz.

Importantly, RDM is fully funded for project construction and development.  Many pre-production gold miners trade at discounts to project value in part because project funding has not yet been secured (i.e. equity holders will be diluted to fund construction and development).  This should not be the case for CPN as funding is in place.  RDM’s funding is comprised of:

1. Debt Project Financing: $90m Debt Facility with Macquarie Bank

• This facility is secured against the Brazilian assets but has no recourse to CPN’s Romanian assets.  I believe this is important in both an upside and downside scenario
  • Downside: if something goes wrong in Brazil equity holders can extract full value from Romania without Brazil’s debt providers having any claim on this asset
  • Upside: allows maximum flexibility in funding the Romanian project when this is necessary
  • Carpathian was required to enter to enter into both gold hedging and FX hedging (USD:BRL) as part of the debt financing agreement
  • FX Hedging:
    • Capex Hedging: US$117m @ USD:BRL 1.90 (all matures before construction completion)
    • Opex Hedging: US$194m @ USD:BRL 1.98 (maturities commence in 2013 and are spread over approximately 5 years, which is the term of the debt agreement)
  • Gold Hedging: 216koz forward sold at $1,600/oz (26% of Reserves)

• In return for an upfront payment of $30m CPN is required to deliver 12.5% of total production from the RDM open pit to Macquarie.  Macquarie will pay $400 for each ounce of gold delivered (in addition to the upfront payment)
• Macquarie has the option to have the gold stream apply to underground operations (if any) at RDM by contributing 12.5% of the capital required to develop the underground mine

• CPN’s most recent press release stated that $75m of equity would be contributed to the project.  This amount includes funds from the gold stream ($30m) (see here: http://www.carpathiangold.com/news/press-releases/press-release-details/2012/Carpathian-Provides-Update-on-Mine-Construction-and-Development-Progress-on-the-RDM-Project-Brazil-Remains-on-track-with-2013/default.aspx)
• Unrestricted cash balance at 30 June 2012 was ~$35m.  Approximately $25m had been spent on RDM at that date and $15m of gold stream proceeds were still to be received by CPN (i.e. $35m + $25m +$15m = $75m)

The funding sources above ($165m) provide a slight buffer versus the total forecast capital expenditure requirement of $160m (which includes an $11m contingency).

There are a few points / observations to make about RDM before discussing valuation:

• Reserves: RDM will very likely mine more than the current open-pit Reserve given both the existence of the underground resource and the exploration potential at the property.  As mentioned above I am ascribing zero value to this to be conservative
• Production: the crushing and grinding equipment being installed (9,000tpd) can handle more production than assumed in the feasibility study (7,000tpd).   As such, an increased production scenario can be easily accommodated with minimal additional capital expenditure (this is particularly relevant if there is exploration success).  I am not assuming any additional production in my valuation.
• Budget & Timetable: Based on the last update provided by the company (Oct’12) RDM remains on budget and on schedule.  The project is 45% complete and essentially all major contract items have been locked in (i.e. risk of material construction capex overrun is being mitigated) (see here: http://www.carpathiangold.com/news/press-releases/press-release-details/2012/Carpathian-Provides-Update-on-Mine-Construction-and-Development-Progress-on-the-RDM-Project-Brazil-Remains-on-track-with-2013/default.aspx)
• Financing: The financing package secured will result in a lower NPV than the “project” NPV shown in the feasibility study.  The feasibility study did not contemplate financing so, all else being equal, the following will impact “project” NPV:
  • Gold Stream Agreement: 12.5% of gold production will be sold at $400/oz (vs. the higher realized price assumed in the feasibility study)
  • Hedging: the FX rate hedged is slight lower than the rate assumed in the feasibility study (1.98 for opex hedging vs. 2.00 in the feasibility).  Gold hedging is at a price higher than assumed in the feasibility study base case ($1,600/oz vs. $1,150/oz in the feasibility).  Note both FX hedging and gold hedging are currently underwater vs. spot prices. However, the value of the hedge book will fluctuate over time
• Cash Costs:  There has been ongoing cost inflation in Brazil (and in general mining costs are higher now than they were at the time the feasibility study was completed).  As such, I believe actual cash costs at RDM will likely be higher than shown in the FS

Valuation

DCF valuation (7.5% discount rate) shown below.  The DCF models known reserves only, is based on the feasibility study assumptions but assumes ~30% higher cash costs (average of approximately $725/oz cash costs pre-royalty over the life of mine vs. $558/oz in the feasibility study).  The DCF valuation also explicitly models the cost/benefit of RDM’s financing structure and the hedge book (per above).  

After Tax DCF Valuation (7.5% Discount Rate)

Note:  the model uses estimated remaining pre-production capital expenditure at 30 June 2012 of approximately $135 million ($25 million already spent).  This is done to calculate the value of RDM at the last balance sheet date (i.e. to compare this value to the current enterprise value which uses the most recently available net debt at 30 June 2012)

Given RDM is nearing production I also think trading multiples of comparables are useful, especially companies with producing assets in Brazil (i.e. what would RDM be worth if it was separately listed as a single asset company).  I know of two listed gold mining producers with assets only in Brazil:

(1)     Jaguar Mining (JAG): 3 producing assets in South Brazil and a pre-development project in north Brazil.  JAG has a very high debt load and has suffered from recent operational problems at its producing assets (one of the three assets has been placed  on care & maintenance).  The CEO was pushed out, the Board has been “refreshed” and the company is now a turnaround story.   JAG currently trades at ~4.7x 2013 EBITDA (Bloomberg consensus)

(2)     Luna Gold (LGC): Single producing mine in North Brazil (731koz Reserves; 3.2Moz M&I Resources).  ~8-10yr mine life based on Reserves and current production plan.  Like RDM, LGC’s Aurizona mine is subject to a gold stream agreement (Luna’s is with Sandstorm Gold).   LGC currently trades at ~6.4x 2013 EBITDA (Bloomberg consensus)

I would argue that RDM is still pre-production and so should trade at a discount to a producer (successfully ramping up the mine should lead to a de-risking / re-rating to a producer multiple).   As such, I would suggest RDM should trade at a discount to LGC.  JAG’s valuation has compressed materially and is currently suffering from both its operating history and financial stress from its stretched balance sheet so it’s not the best comparable (although still relevant).    

Using $1,600/oz gold (which I think is lower than what analysts have in their models for 2013 – i.e. what underpins the consensus multiples above) and using a multiple range of 3.5x-4.5x EBITDA (a material discount to LGC and also a discount to JAG) yields an EV of $225-290m for RDM.

My conclusion is that a conservative case – gold below current spot prices, costs higher than feasibility study, modelling known reserves only (which were calculated using $950/oz gold assumption) – still yields a valuation for the RDM asset alone that justifies Carpathian’s entire current market valuation.  Less conservative assumptions yield upside to current valuation levels.  Valuation multiples of peers with producing assets in Brazil support this conclusion.

(2)     Rovina Valley Project (RVP) (Romania)

For those interested there is a significant amount of descriptive information about the RVP property is contained in the mineral resource update filed on SEDAR in August 2012 (http://www.sedar.com/GetFile.do?lang=EN&docClass=24&issuerNo=00018939&fileName=/csfsprod/data134/filings/01957472/00000001/v%3A%5CCarpathianGold%5CTechnicalReport%5CCarpathian-TR-Aug31-2012.pdf)

A Preliminary Economic Assessment (PEA) for the project, which contains descriptive information as well as indicative project economics, was filed on SEDAR in April 2010.  This might also be of interest (http://www.sedar.com/GetFile.do?lang=EN&docClass=24&issuerNo=00018939&fileName=/csfsprod/data107/filings/01585312/00000001/s%3A%5Ccpn521tr.pdf)

In summary, RVP is situated in the Metalliferi Mountains, near the Western boarder of Romania.  RVP consists of three proximal gold-copper porphyries along a north-northeast trend over a distance of 7.5km.  RVP is situated with considerable infrastructure already in place as water access is ~2km away and grid power is within 5km.  The Resource at RVP was updated in July 2012.  The Current Resource is >7 million ounces of gold and >1.5 billion pounds of copper.  Approximately 95% of the Resource is contained in the Measured and Indicated category.  A pre-feasibility study is currently being completed on the project and is due to be released in Q4’12.  I expect a large portion of the Resources to be converted to Reserves when the pre-feasibility is released.

RVP is located within what is known as the Golden Quadrilateral.  This is an area with a significant mineral endowment.  In addition to RVP there are two other significant properties in this area.  The first is Rosia Montana, a >15 million ounce gold deposit located approximately 25km north-east of RVP.  Rosia Montana is owned by Gabriel Resources (TSX: GBU).  The second is Certej, a 3 million ounce gold deposit located approximately 17km south-east of RVP.  Certej is currently owned by Eldorado Gold (TSX: ELD) and was acquired when ELD acquired European Goldfields in 2011. 

The 2010 Preliminary Economic Assessment for RVP (which was based on a Resource released in 2008) highlighted a 19 year mine life with an average annual production of ~200koz of gold for a total of 3.72 million ounces of recoverable gold.  Average annual copper production was estimated to be approximately 50Mlbs, for a total of 938 million pounds of recoverable copper.  Using $1,450/oz gold and $3.00/lb copper the PEA returns a pre-tax NPV10% of >$1bn and an IRR >35%.  Cash costs using these metal prices were estimated to be $81/oz of gold (with copper as a by-product credit).

Since the 2010 PEA was released additional drilling was completed and the mineral resource was updated.   The update resulted in an increase in the total resource and a significant increase in the resource contained in the measured and indicated category (i.e. the conversion of Inferred Resources into the Measured and Indicated category was very high – see table below).   The feasibility study (due to be released in Q4’12) will be based on the updated (larger) resource.  My expectation is that the pre-feasibility will show a larger project (higher annual production and higher aggregate life of mine production) but a larger upfront capital cost (in part to accommodate larger production).  Overall, I expect the economics of the project to remain robust (second quartile or better average cash costs, NPV and IRR similar to PEA).

RVP – Comparison of Resource: September 2008 vs. July 2012

Investment by Barrick Gold

The project size is material by global standards and has attracted attention from larger gold mining companies as the size of the resource has grown and the economics of the project have proven to be robust. 

In July 2011 CPN announced that Barrick Gold (the world’s largest gold mining company, EV >$50bn) had acquired a 9% shareholding in CPN via a primary placement of shares.  The share placement was completed at a share price of $0.52/share (+68% higher the current share price).  The placement was part of a broader agreement whereby: (i) the proceeds from the placement were to be applied exclusively to RVP; (ii) a technical committee (3 CPN, 2 Barrick representatives) was to be established for RVP; (iii) two Barrick employees were seconded to the RVP project; (iv) Barrick has a right to participate in future CPN equity placements to avoid dilution; (v) Barrick was to have a right of first refusal in relation to any disposal of the Romanian assets by CPN.

Valuation

RVP is a much earlier stage project than RDM. 

As a starting point the pre-tax NPV10% valuation using the PEA assumptions and $1,450/oz gold and $3.00/lb copper is ~$1.1bn.   However, the PEA is based on a dated (smaller) resource and will be updated with a pre-feasibility study that is scheduled to be released in Q4’12.  I expect the pre-feasibility to return robust project economics and a similar NPV to the PEA.

In addition to RVP’s modelled project economics a value for the project can be imputed from comparables.  Given the proximity of RVP to Gabriel Resources’ Rosia Montana project (~25km) and the fact that GBU is a single asset company, GBU is the best comparable for RVP. 

A side-by-side comparison of RVP and the Rosia Montana project is shown in the table below.

Comparison – Rovina Valley Project and Rosia Montana

(1) Based on current spot prices for gold, silver and copper

(2) RVP capital expenditure and cash costs based on 2010 PEA base case ($900/oz gold and $2.25/lb copper).  RM capital expenditure and cash costs based on 2009 NI 43-101 compliant technical report base case

(3) Cash costs reported on a by-product basis.  RVP costs receive by-product credits from copper production. RM costs receive by-product credits from silver production

Gabriel Resources’ current enterprise value is approximately $700 million, or approximately $55 per attributable ounce of measured and indicated resources (note: GBU only owns 80.7% of Rosia Montana, the remainder is owned by the Romanian government).  

I believe that Rosia Montana should achieve a higher value than RVP for a number of reasons including: (i) it is a larger ore body; (ii) it is higher grade; (iii) it is a gold/silver deposit vs. a gold/copper deposit for RVP (precious metals typically have lower volatility than base metals which impacts risked project economics in my opinion); (iv) It is further along in the permitting process (~2yrs ahead of RVP in my estimate, although GBU has had repeated permitting problems/delays).  

 However, as a starting point applying GBU’s valuation implies a value of approximately $400m for RVP (~3x CPN’s current EV).  This value needs to be discounted for the reasons mentioned above but even cutting it in half implies RVP is ~1.5x CPN’s entire current enterprise value (and this is ascribing zero value to RDM in Brazil).   The table below shows a number of potential valuations

(1) based on 2010 PEA using $1,450/oz gold and $3.00/lb copper

Eldorado’s Certej project (~17km south-east of RVP) is also a relevant, albeit more difficult to extrapolate, valuation data-point for RVP.  Eldorado has over a dozen projects globally and is a much larger company than either CPN or GBU (ELD has a >$10bn EV) so direct comparisons based on current market valuation metrics is difficult (and somewhat meaningless).  ELD acquired the Certej project as part of a $2bn takeover of European Goldfields which was announced in late 2011.  European Goldfields had three major projects all of which were pre-development (two in Greece and Certej in Romania) of which Certej was the smallest.  The acquisition price paid by ELD for European Goldfields equated to $166/oz Resources, although given Certej was the smallest of the three assets acquired it is difficult to extrapolate directly from this datapoint.   Certej is further along the permitting process than either RVP or Rosia Montana (reducing project risk and arguing for a higher valuation) and Eldorado has said publicly it plans to commence construction of this mine in 2013.

While Certej’s valuation is difficult to extrapolate it does provide evidence of two important points: (i) value should be enhanced as the project de-risks; and (ii) there are acquirers for assets in this district, including very large gold miners.

CONCLUSION AND VALUATION SUMMARY

For the reasons described above I think each of the two projects individually could justify CPN’s current market valuation (i.e. at the current valuation investors are getting one of the two projects “for free”).  Additionally, given its size and potential relative to CPN’s current market valuation, RVP has the potential to deliver many multiples of CPN’s  current value to shareholders. 

In addition to the current valuation providing investors a margin of safety, the fact that funding has been secured for RDM (limiting dilution risk) and Barrick’s interest in the company (which was secured at a much higher price than the current price) provide further downside protection for investors.

The table below shows a matrix of valuations for RDM and RVP, expressed as a multiple of current enterprise value.  What should be evident from the table below is: (i) the market is implying fairly draconian valuation assumptions for both assets at the current valuation; and (ii) there is material convexity in the return profile from current valuation levels.

Carpathian Valuation: Intrinsic Value Estimate vs. Current Market Value

Catalyst