Despite Trading At Multi-Year Lows, Cliffs Has A Long Way To Fall

seekingalpha.com
Anuj Kumar
Sep. 23, 2014

Summary

• Tepid Chinese demand and increasing global supply has caused iron ore prices to fall off a Cliff which has muted the profitability of all segments.
• Asset sales are likely to fall below market expectations based on current iron ore prices / forward curve.
• Making the decision to buyback shares while facing a liquidity crisis appears to be the tactic of an activist investor with a short term investment horizon.
• MS values their Chromite project at original cost + all additional capital for a total of $500 million. Given that the project is currently shut down, undepreciated book value is likely an aggressive number. Using a combination of my analysis and MS's estimate of take or pay contracts, I am assuming the value of the entire Canadian operation is zero (if they put that division in bankruptcy for example).

Cliffs Natural Resources (NYSE: CLF) is a mining company primarily involved in iron ore with a small division in coal. They have recently faced many headwinds with falling global iron ore and coal prices. Iron ore prices have plunged following a large increase in production, particularly from their Australian competitors, as well as reduced demand expectations out of China. Activist investor Casablanca Capital (~5% ownership) recently won a majority of the board seats and has replaced the CEO. Despite a large debt load and negative cash flow for the foreseeable future, Cliffs has made the surprising announcement to buy back $200mm of their shares. This comes just 18 months after an offering for mandatory convertible preferred shares and ordinary common shares and I will not be surprised to see them have to reverse course and end up offering shares again in the future (likely at a lower price than the buyback). Cliffs has ~180 million fully diluted shares outstanding pro forma (to give a sense for how much the sensitivities affect the end thesis).

They report four distinct segments:

U.S. Iron Ore
Eastern Canadian Iron Ore
Asia Pacific Iron Ore
North American Coal

Source for divisional data: Production numbers and rev / cost guidance are from Q2 2014 earnings release and reserves are from CLF 10-K

North American Coal

Cliffs owns and operates four metallurgical coal operations (used in steel making) in West Virginia and Alabama and one thermal coal (power plant) mine in West Virginia. Met coal is sold at a premium typically to thermal. As of 12/31/2013, Cliffs had proven or probable coal reserves of 187.7 million short tons. They estimate as of their Q2 earnings release to produce approximately 7 million tons in 2014. Estimated full year revenue per ton is $75-80 and cash costs are $85-90 (with depreciation, depletion and amortization of $16 / ton).

With coal falling out of favor in the United States as power plants switch over to cleaner, cheaper US natural gas, coal prices have fallen putting this division in the red. On September 3, the Wall Street Journal reported that Cliffs has hired Deutsche Bank to sell its coal assets. After this news came out, Wells Fargo estimated that the coal assets could be worth $350 million based on coal reserves. Based on estimated 2014 ending reserves (187.7-7 produced = 180.7 at year end), this would value the coal division approximately at $1.94 / proven or probable ton. Given the coal assets currently available or recently available in the US following some sizable bankruptcies (James River, Patriot), I think this could be an aggressive number. Although the disclosure is limited, James River is selling their Triad mines in Indiana and Hampden mines in West Virginia (in addition to the assets of the Logan & Kanawha Coal Co.) to Blackhawk for $20 million in cash plus the assumption of $32 million in liabilities. Based on James River's 2012 10-K, Hampden and Triad alone had ~90.8 million tons of proven and probable reserves. Adjusting for two years of production (to 83.6 million tons) and assuming zero value for Logan & Kanawha, this would value their mines at 0.62 / ton. To be conservative (in favor of longs), I'll use $1.50 / ton which values coal at $271 million.

Asia Pacific Iron Ore

The Asia Pacific Iron Ore division consists only of the wholly owned Koolyanobbing complex, which includes the deposits at Koolyanobbing, Mount Jackson and Windarling in Western Australia.

As of 12/31/2013, Cliffs had proven and probable iron ore reserves of 64.5 million metric tons. They estimate from their Q2 '14 earnings release that they will produce 11 million tons in 2014. Revenue per ton is estimated at $85 - 90 however, this assumes 62% Fe seaborne iron ore fines price (C.F.R. China) of $112 per ton. For every $10 / ton that this base price varies from $112, the revenue will be reduced by $5 (defined as a +/- $5 / ton sensitivity per the Q2 release. I'll use this methodology going forward). The average 62% Fe price for the first 6 months of 2014 was $112 and the back half of the year is looking like it will average around $85 / ton; this would put the full year average for the iron ore "index" at about $98.5 / ton. This would lower the Cliffs Asia Pacific revenue range per ton to: 78.25-83.25 ($112 - 98.5 = 13.75; 13.75 / 10 * 5 = $6.75 / ton reduction). Cash costs for the year are estimated at $55 - 60 / ton (with $14 / ton of DD&A).

With activist Casablanca Capital in control, they are pushing for the sale of these Asia Pacific assets (and coal) to allow the company to focus on their core North American iron ore positions. Per the Wall Street Journal article mentioned previously, Jefferies has been hired to sell the Asia Pacific assets and a source thinks they could be worth up to $1 billion. Wells Fargo came out shortly after and refuted this analysis and estimated a value of $300 - 400 million. I wanted to understand WF's analysis further so I put together a mini projection model on Australia only. With only approximately five years of remaining production and a tepid forecast for iron ore prices (especially in Australia given the large increases in production from their competitors), the net present value of the Australian assets are significantly less than the $1 billion reported in the WSJ article and are in line with Wells Fargo estimates. Interestingly, you can see that this implies only a 2.7x EBITDA multiple which helps frame the true value of declining assets like these. I assumed ~$40mm of capex per year to maintain and operate the mines. Given they are estimating $300 million this year in overall capex and most of that is just maintenance capex, I thought this was a conservative estimate. If there was zero capex required for the next 5 years, the NPV would be $470 million to get a sense for how important this assumption is. The following model output shows how I derive the revenue per ton based on the 62% Fe forward curve, a 3% growth rate beyond 2016 for the "index" and how this affects the realized pricing (sensitivity) for Asia Pacific as explained above.




U.S. Iron Ore

Cliffs U.S. Iron Ore business manages and operates five iron ore mines in Michigan and Minnesota with annual rated capacity of 32.9 million tons - representing 59% of total US pellet production capacity. Production is generally sold via long-term supply agreements with price adjustment provisions. This is their largest segment and drives a significant amount of their revenue and gross profit. (Note: the long-term nature of the supply agreements is a positive in the short term as USIO isn't as affected by changes in the base commodity price but as these contracts roll, the US mines could be hit substantially. My model is based on currently achieved pricing and I do not assume any changes to the contracts to be conservative in favor of longs).

As of 12/31/2013, Cliffs had approximately 836.9 million tons of proven and probable iron ore in the US. They estimate (based on their Q2 '14 earnings report) producing 22 million tons this year (so they have many years of production available). Estimated revenues per ton are $100 - 105 with a $1 sensitivity (explained in Asia Pacific above). They estimate cash costs per ton of $65 - 70 (with DD&A per ton of $5). With production expected for the next 40 years, I did not need to do a production analysis (as above with Asia Pacific) and instead am valuing the US based on a very simplistic model:




The assumptions are based on Cliffs' Q2 guidance for the full year 2014 in terms of cash costs and adjusted revenue / ton. I have assumed $6 / ton of capex ($132 million). With 2014 capex (primarily if not all maintenance) of ~$300 million and US Iron Ore doing 50%+ of total company revenue, I believe this is a fair, if not conservative, estimate. Historically their capex has been significantly higher and I think that in actuality I am underreporting realistic maintenance capex in an effort to be conservative. NPV is calculated using a simple present value of a growth annuity:

Cash flow * (1 + growth) / (Discount rate - growth rate)

Eastern Canadian Iron Ore

The Eastern Canada segment consists of three mines with production sold primarily into Asia: Wabush mine, Bloom Lake Mine and Pointe-Noire. Pointe Noire was idled in Q2 2013 due to high production costs and low pricing. In February 2014, they announced plans to idle Wabush as well.

Production is sold through a series of short-term pricing arrangements linked to the spot market, which has caused weakness in this segment as well.



They plan on producing 7 million tons per year for 2014 with revenue guidance of $85 - 90 / ton with sensitivity of +/- $4 per ton. Their cash costs are estimated at $80-85 / ton with $23 of DD&A.

Without positive earnings, it is hard to evaluate Eastern Canadian Iron Ore. As of this time, the new CEO Lourenco Gonclaves (Casablanca appointee) has not made a decision on Canada in terms of exit, shut down or some other alternative. They would need over $1 billion to expand the 2nd phase at this point. If they are to shut down, they have significant take or pay rail and processing contracts that they would have to settle which would weigh on any potential shut down. A Phase II completion however would apparently drop the cash costs towards the $65 / ton range.

Casablanca Capital estimates $1.25 billion for Phase II however, I am assuming only $1 billion. If they were able to complete Phase II, the left hand side financials / valuation would be possible, albeit I believe aggressive particularly around capex and cash costs. I have assumed 14 million of annual tons in this scenario based on Phase II expectations. I have only assumed $7 / ton of capex, which would equate to about $100 mm for maintenance capex per year (again a view, which I think is conservative).

Based on the above financials and returns, no partner would be willing to finance this project. Their current partner (Wuhan Iron and Steel) has chosen not to put any more money into this venture and thus their ownership has reduced from 25% to about 17% today so it substantiates my claim as well that a new partner is unlikely to join. Morgan Stanley (July 31, 2014 US Steel Monthly Chartbook research report) estimates that the present value of the take or pay contracts + shutdown costs in Canada are ~776 million (Cliffs Canada has a series of rail and other take or pay contracts that they would have to settle or continue to pay even if they shut down or do not proceed with Phase II). MS values their Chromite project at original cost + all additional capital for a total of $500 million. Given that the project is currently shut down, undepreciated book value is likely an aggressive number. Using a combination of my analysis and MS's estimate of take or pay contracts, I am assuming the value of the entire Canadian operation is zero (if they put that division in bankruptcy for example).

Source for Phase II cash costs and tonnage expectations

Source for Phase II investment requirement

Summary

Adding up a sum of the above parts ($ millions):




Even with what I deem to be aggressive numbers (in favor of longs) across the board, I am only able to get to an equity value per share of $6.59. On Bloom Lake, I am not even discounting the fact that there will a high negative cash flow in the interim until it is shutdown. For the US Iron Ore division, I am assuming ~$100 / ton as the realized pricing, but I think as the longer-term contracts roll off that this number will come down significantly but for conservatism, I am not adjusting for that fact. For discussion purposes, if the US Iron Ore realized pricing falls by $10 / ton long term, this drops its value by almost $1.9 billion. With 180 shares outstanding, I think you get the picture.

The biggest risk to the short thesis here is an upward movement in iron ore pricing. Given the current forward curve, production from Australian producers, and tepid demand from China, the outlook is poor. If base Iron Ore prices were to move up $10 / ton (to $92 / ton on average for 2015), the equity value per share based on my model is still only $9 / share. Goldman has just released an updated report and forecast for iron ore and it is very bearish: $80 iron ore next year and lowers its estimates for 2016 and 2017 to $79 and $78 ("The End of an Iron Age," September 10, 2014). Since my original writing of this article, iron ore prices have fallen another few dollars as Chinese demand outlook remains tepid.

Additional disclosure: Personally have no position in the stock.

Despite Trading At Multi-Year Lows, Cliffs Has A Long Way To Fall

seekingalpha.com
Anuj Kumar
Sep. 23, 2014