Value Investors Club / CLIFFS NATURAL RESOURCES (CLF Secured Debt)

Description

I’m having a lot of trouble finding anything to buy these days, but I think the secured debt from Cliffs Natural Resources is a compelling investment. After having been short the company’s stock for most of last year, I now own both the 8.25% Notes due 2020 and the 7.75% Notes due 2020. The 8.25% Notes are first lien obligations of the company and trade at 91% of par to yield 10.6% to maturity. The 7.75% Notes are second lien obligations of the company and trade at 61% of par to yield 20.7% to maturity.

Cliffs is the largest manufacturer of iron ore pellets in the United States and produces approximately 22 million tons per year of pellets. Substantially all of its production is sold to steel manufacturers that use blast furnaces to manufacture steel. Sales tend to be on multi-year contracts and are often at prices that have a fixed component and an adjustable component based on either seaborne iron ore prices, steel prices, or both. As a result, Cliffs’ iron ore business is a somewhat less commodity-price sensitive business than traditional iron ore mining, but it still has significant swings in realized pricing. The chart below shows the sensitivity of realized pricing for the company relative to seaborne iron ore prices:

The U.S. Iron Ore business is only partially tied to the seaborne market for two major reasons. First, the U.S. steel market is concentrated around the Great Lakes, which makes taking seaborne iron ore deliveries very tricky. In the seaborne market, prices are quoted delivered to the port in China. This means a miner such as Vale realizes net revenue equal to the seaborne iron ore price minus inland transportation to the exporting port, fees at the exporting port, and bulk shipping costs to China. The Chinese steel manufacturer purchases the ore delivered to the Chinese port and is then responsible for the port charges in China and the inland transportation to the steel mill. In contrast, the logistics for seaborne iron ore to reach the Great Lakes is significantly more challenged. Seaborne iron ore must be unloaded from large bulk carriers at the mouth of the St. Lawrence river and loaded onto barges, which must then travel through multiple locks (as many as 15), before reaching a port and being loaded onto inland transportation. We believe that this could make purchasing seaborne iron ore as much as $30-35 per ton more expensive for a Great Lakes-based purchaser than it is for a Chinese purchaser. The second reason that there is only a limited link between domestic iron ore prices and seaborne prices is because the domestic market is composed primarily of iron ore pellets whereas the seaborne market is composed mostly of iron ore fines. U.S. mills are not configured to produce pig iron directly from iron ore so they use pellets, which command a premium to unprocessed fines. Even at today’s iron ore prices, we believe that premiums are at least $25-35 per ton.

At current seaborne iron ore prices, we believe that an integrated miner which owns its inland transportation and ports, such as Vale, realizes about $40-50 after paying for shipping. For Vale to realize a similar amount and sell to the Great Lakes-based steel manufacturers, we believe that costs for the steel manufacturers would be $80-100, which is calculated as follows:

This supports an assertion that Cliffs made in a recent investor presentation that “any potential competition from outside the US has a breakeven price of $95 - $100 per ton”.

Furthermore, as recently as February 2015, Essar Steel Algoma stated “Essar’s only option for obtaining iron ore pellets in the Great Lakes is through Cliffs because Cliffs does not have a reliable competitor for the production and supply of iron ore pellets at this time. The reason Cliffs does not have a competitor that Essar could turn to is because the breakeven cost for any competitor to enter this market from Brazil or from Eastern Canada (for example) is prohibitive. Essar has no choice, therefore, but to work with Cliffs regarding its current iron ore pellet needs.”

In comparison to the $80 100 cost per ton for a Great Lakes-based steel manufacture to purchase seaborne iron ore, current prices of $75-80 per ton seem supportable. Obviously, if iron ore drops another $10-20, seaborne could become more competitive. However, a prolonged period of seaborne iron ore at today’s $50 price, much less 20-40% lower, seems unlikely in light of the industry cost curve as presented below:

The significant differences between the U.S. Iron ore market and the seaborne iron ore market as well as Cliffs’ strong market position had historically resulted in a negotiated price that was entirely separate from the seaborne price. This changed in mid-2000s when the company saw the seaborne iron ore market go bananas and decided to include a link between seaborne prices and its prices going forward. That decision looked brilliant for a while, but it has obviously come back to haunt the company. Going forward, we actually expect that the company will earn a fixed price per ton similar to what it previously did. The efforts by the steel manufacturers to backwards integrate into iron ore pellets have been disastrous, such as Magnetation where cash costs are over $100 per ton. Hence, it makes sense for the steel manufacturers to stay with Cliffs, but there are few reasons for them to take risks on seaborne prices like they did in the past.

In addition to the U.S. Iron Ore business, Cliffs also owns two ancillary businesses: a metallurgical coal mining business and an Australian iron ore business. Each of these businesses operates near EBITDA breakeven, but after accounting for SG&A and capex, we believe that annual cash losses from these businesses will be nearly $65 million in 2015. We believe these cash losses are approximately equally split between SG&A and capex but that the coal business accounts for about 60% of the losses and the Australian iron ore business accounts for about 40% of the losses. The company has already sold half of its coal assets in the last 6 months and has repeatedly stated that it is in the process of selling the other half. We believe that the company will realize about the same for its second set of coal assets as it did for its first ($175 million) and that the company will be able to reduce SG&A by $10 million and capex by $30 million following this sale. While we would be pleased if the company also sells its Australian assets, we do not believe that this is part of the current plan.

We believe that the company’s secured debt (both the first lien and the second lien debt) are well covered by the asset value of the company. Following the first lien offering, the company has approximately $475 million of cash on its balance sheet and. After taking into consideration a $180 million working capital reversal expected in 2H 2015, a $185 million income tax receivable expected in 2H 2015 and $175 million of proceeds from the sale of coal mines, we believe the Cliffs will have over $1 billion in pro forma cash. This means that the company has more cash than it has first lien debt outstanding:

The credit analysis for the second lien debt is a bit more nuanced since we should also look at the earnings of the ongoing business and not just the balance sheet. Based on the trading price of the second lien debt, we create the company at a value of $1,180 million or $164 million above the cash and non-core asset value. Including the value of U.S. Iron Ore would mean, in our opinion, that the bonds are well covered at par. We believe that Cliffs will be able to reduce its cash costs to about $55 per ton in 2017 due to the closure of the Empire mine and continued cost cutting by the new CEO. This would result in EBITDA of approximately $300 million and free cash flow of -$20 to +$40 million due to the company’s $200 million of annual interest expense. Hence, valuing the bonds at par would mean that investors create the company for about 1x EBITDA and less than 2x EBITDA-Capex:

No investments are without risk. We believe that the greatest risk to our investment comes from the Essar Steel Minnesota iron ore mine which is being built. We have already accounted for lost volumes of 3.5 million tons due to the closure of the Empire mine in conjunction with ArcelorMittal moving those tons to the new Essar project being built. In total, the Essar Steel Minnesota project is supposed to produce 7 million tons of pellets per year, which means that Cliffs’ production could theoretically drop to 15 million tons. At a $15 margin per ton and after incremental SG&A cuts, we think that EBITDA might get as low as $150 million in this scenario. In this case, our second lien bonds would still create the company for about 2.5x EBITDA and 5.0x EBITDA-Capex.

In the prospectus for a 2014 bond offering that went bust, Essar Steel Minnesota disclosed that its off-take contracts for pellets are priced at a discount to IODEX and will magnify moves in seaborne iron ore pricing. Extrapolating the pricing table from the prospectus (below), Essar Steel Minnesota could be contractually obligated to sell pellets for less than $30/ton in the current seaborne iron ore pricing environment, far below what we believe the project’s marginal and all-in sustaining cost to produce a pellet will be, if it ever starts up.

There are a few reasons to be relatively sanguine about this possibility. First, the Essar Steel Minnesota project has been delayed multiple times and is over budget. Second, the project is clearly uneconomic in the current iron ore price environment. Third, given the operational problems at Bloom Lake and Magnetation, it is entirely possible that the new Essar mine would not be competitive with Cliffs’ mines. Fourth and finally, we would note that Cliffs does occasional fly-overs of the Essar site and the pictures do not suggest much progress has been made. On the flip side, we don’t think it would be crazy to imagine Cliffs volumes remaining above 20 million tons with a shutdown of Magnetation, which would be an economically rational choice for AK Steel.

Below are Magnetation’s forecasts for its cost to produce pellets (assuming they remain operational) where the cost creep in cost per ton to produce a pellet has steadily crept upwards over the last year.

Most recent Magnetation forecasted 2015 cost to produce pellets:

Prior Magnetation forecasted 2015 cost to produce pellets:

Earlier Magnetation forecasted 2015 cost to produce pellets:

A second risk to our investment comes because the Company has the right to issue up to $750 million of 1.5 lien bonds in front of the second lien bonds we are recommending or up to $700 million of additional second lien debt. We believe that the most likely reason to use these baskets would be to repurchase existing senior notes at a discount or to exchange existing senior notes at a discount. One thing making this tricky today is that the bonds would probably need to be issued with a yield somewhere in the neighborhood of 13-14%. It’s not clear to us that it’s a great idea for the company to issue debt at these yields. Investor willingness to sell or participate in an exchange at current levels also seems limited because the uptake in the last exchange was significantly below the maximum levels. One of the reasons for the low participation exchange was that retail holders of unsecured notes which were issued under investment grade indentures were unable to exchange into the 144a second lien bond. It seems most likely to us that the company will try to wait for lower yields before it tries another exchange offer. The issuance of this debt would push the second lien bonds down in the capital structure and in a base case they would be 3.6x leveraged and 7.2x leveraged in the downside case:

A third risk to our investment could come from the company spending its cash and non-core asset sale proceeds on repurchasing the unsecured debt which is junior to us. In this case, we would be entirely dependent on the earnings of the U.S. Iron ore business to support the bonds. A benefit from this would be reduced interest expense – possibly as much as $80 million less per year. We would then expect free cash flow to be $80-130 million per year in a base case and for the company to be able to repay the first lien debt from accumulated cash at maturity (or to come close). We also think the second lien could be easily refinanced. Even so, the depletion of the cash and non-core asset sale proceeds would make the business appear significantly more leveraged:

A final risk comes from the Canadian Iron Ore operations. In 2011, Cliffs acquired Consolidated Thompson for nearly $5 billion. Consolidated Thompson’s primary asset was an Eastern Canada iron ore project called Bloom Lake. Bloom Lake suffered from stubbornly high operating costs and caused Cliffs significant recurring losses. Upon assuming control of Cliffs, Lourenco Goncalves promptly put the Bloom Lake asset up for sale. When no deal materialized in the near term, Cliffs caused Bloom Lake to file bankruptcy in Canada in January 2015, removing the Canadian operations’ cash drain from the rest of the company. On its 2014 4Q earnings call, Goncalves stated “the Bloom Lake Group has been effectively ring-fenced, therefore Cliffs Natural Resources no longer has any obligations to fund these pre-filing liabilities…the previous $650 million, $700 million figure now equals zero.” The $650-700 million figure referenced a November 2014 announcement regarding what the company believed its maximum Bloom Lake liability could be over the course of 5 years if Cliffs were to have closed the project outside of a bankruptcy filing. This follows a similar path that U.S. Steel took with one of its loss making Canadian operations.In September 2014, a U.S. Steel subsidiary, U.S. Steel Canada filed for bankruptcy protection in Canada, relieving the parent of $1.0 billion in underfunded pension and post-employment benefits.

In summary, we think that the first lien notes are pretty close to bullet-proof and we have a hard time seeing them impaired. We also think that the second lien notes are covered in most reasonable scenarios unless a weird confluence of events happens such as the company wasting all of its cash on repurchasing the unsecured debt and then losing all of the potential volumes it can to the new Essar mine and also seeing the Canadian liabilities flow back and somehow end up senior to the second liens. Even so, investors will get 38.75 points of coupons back over the next five years and the only maturity coming up first is $436 million of 2018s now trading at 70.

I hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise do not hold a material investment in the issuer's securities.

Catalyst

- Non-core asset sales

- Essar Steel Minnesota construction

- ArcelorMittal contract renegotiation

show sort by

			2015	2016
Price:	61.00	EPS	0	0
Shares Out. (in M):	540	P/E	0	0
Market Cap (in $M):	329	P/FCF	0	0
Net Debt (in $M):	0	EBIT	0	0
TEV (in $M):	329	TEV/EBIT	0	0