Tronox is the world’s fifth-largest producer of titanium dioxide pigment (Tio2). Tio2 is a white pigment that is mainly used in paint, plastics and paper to provide opacity and brightness. The company has plants in the U.S., Australia and the Netherlands. On September 26, Tronox announced the acquisition of the mineral sands assets of Exarro, which are located in South Africa and Australia. Mineral sands are the source of the feedstock ores that are used to create Tio2 pigment. This acquisition makes Tronox the world’s only fully vertically integrated producer of Tio2 pigment.
Tronox was formed as a spin-off from Kerr-McGee in 2002. Kerr-McGee loaded the spin-off with a host of legacy environmental and tort liabilities related to discontinued chemical operations. The combination of a large debt load, legacy liabilities, and macro downturn led Tronox to file for bankruptcy in January 2009. Tronox emerged from bankruptcy in February 2011 recapitalized and having completely shed the legacy liabilities.
There was widespread speculation on the part of both shareholders and industry analysts that Tronox would be acquired by one of the other major Tio2 players after emerging from bankruptcy. Huntsman (HUN) had placed a bid on TROX in bankruptcy that was rejected, and Kronos (NYSE:KRO) had spoken publicly about their desire to acquire TROX assets. TROX stock ran up to 160 in July, I am guessing largely on the basis of buyout speculation (the only other pure play comp, KRO, was trading significantly lower at the time). But in September TROX announced they would be acquiring Exxaro’s mineral sands for 38.5% of TROX stock, which nixed any hopes for a buyout. The post-reorganization equity holders did not respond favorably to the deal, sending the stock from 130 to as low as 78. I think this reaction had more to do with the hopes for a quick liquidity event being dashed than the merits of the deal itself. The stock price has since largely recovered to pre deal announcement levels, but positive Tio2 industry trends and the enhanced industry position of TROX post deal have still not been reflected in the stock price.
Tio2 Pigment Industry Trends
I analyzed the tio2 pigment industry trends in my analysis of KRO, so please see this article for more detail. In brief, titanium dioxide is an oligopoly industry with very favorable supply/demand dynamics for the foreseeable future. A substantial amount of industry capacity was permanently shut down in the '08/'09 recession leading to a supply shortage. It takes three to five years to construct a greenfield plant, and prices are still not at levels which will spur significant new investment. Pigment pricing has risen about 40% in the past year, and is expected to keep rising in the fourth quarter and into 2012. The third quarter continued the current trends with producers selling everything they could make and prices continuing to rise. The fourth quarter typically sees some slight downward seasonality in volume (due to being off-season for paint sales), but all of the major producers have still announced price increases for fourth-quarter implementation.
I’d like to expand on two points more fully:
What happens to the Tio2 industry if we go into another recession? The Tio2 players did poorly in '08/'09, but I think they are much better positioned this time around. Tio2 pricing is much more supply than demand driven. Demand for Tio2 is fairly price inelastic. Looking at historical pricing, the industry has long cycles more heavily impacted by the relative production capacity of the industry than general macroeconomic factors. For example, pricing was actually up 7% from 2007-2009 even though volumes were off about 15% from peak to trough. The problem in 08 was that selling prices and profit margins were so low (and had been for almost 20 years) that the decline in volume and deleveraging of fixed costs wiped out the entire profit (and then some) of the pigment producers. With prices now having almost doubled, the pigment producers are much better equipped to deal with a decline in volume than they were in 08. They can give up some volume and fixed cost leveraging and still be at least minimally profitable. Additionally, customer inventories are still tight (HUN stated 45 days on their Q3 call) so the industry will probably not see the dramatic destocking we saw in 08. That means the volume drop off will also probably not be as great in another downturn of similar magnitude.
The specter of additional Chinese capacity
The only major additions to capacity over the next several years are expected to come from China (aside from DuPont’s 350 thousand metric ton brownfield expansion). Estimates range from 700 thousand to 1 million metric tons of announced capacity being added by 2016. But there are several mitigating factors:
It is doubtful that all of the announced capacity will be builtEven if the capacity is built Chinese plants do not typically run at “nameplate” capacityTio2 demand will grow in line with GDP in developed markets and much faster in emerging markets, which even in a minimal global growth scenario should add 100-200 thousand metric tons of demand a year.Most importantly, nearly all of the incremental capacity will be sulfate process Tio2. At the present time Chinese Tio2 does not compete with that of the major producers. This is clear in looking at the bifurcated market pricing trends in the Chinese market. In the last two months, domestic pricing for Chinese Tio2 has fallen almost 30%, while prices of imported Tio2 from the major Western producers continue to rise.[/list]
Tio2 Feedstock Industry Trends
Tio2 pigment is produced from either rutile or ilmenite ores, which are mainly mined in Australia, South Africa, and Canada. The ores differ in their Tio2 content, with rutile being 95% Tio2 and ilmenite 45-65% Tio2. The Tio2 content of ilmenite can be upgraded to produce either synthetic rutile or titanium slag, depending on the chemical process employed. The Tio2 feedstock producers are heavily concentrated like the pigment producers, with the top five producers accounting for 62% of production in 2010.
Pricing for rutile and ilmenite had been flat for years until 2011. In real terms pricing had declined even as input costs for energy and labor were rising. Additionally, the feedstock producers had typically sold their products under “cap and collar” contracts that limited their ability to maintain margins.
Low market pricing led to a long period of minimal investment in mine capacity even as demand for Tio2 enjoyed strong growth with the emergence of Asian and Latin American developing markets. Supply for titanium ores has continuously tightened as legacy mine assets have been depleted and not replaced. The feedstock producers are currently just meeting Tio2 demand with 6.3 million mt of titanium minerals mined in 2010. New mine projects that could yield 2.5 million mt of Tio2 have been announced. But those projects have lead times from 3-7 years and many of those projects will likely never even be started. Industry consensus is that feedstock supply tightness will not abate until 2014 at the earliest.
Another boon for feedstock producers is that their cap and collar contracts have largely rolled off over the past year, giving the producers more flexibility to adjust pricing. Additionally, the industry has moved to quarterly instead of yearly price adjustments on long term supply contracts to pigment makers, which will allow ore producers to take advantage of strong pricing momentum.
Pricing for high grade rutile and synthetic rutile feedstocks has doubled in the past year to roughly $1,000/mt, and industry consultant TZMI forecasts that it will double yet again in 2012.
Zircon is another mineral mined from titanium mineral sands that is used mainly as an opacifier for ceramics. The story is similar there, with supply already having tightened significantly in 2010, and strong demand being driven by emerging markets. The structural supply deficit is predicted to last for several more years, which should ensure higher pricing for producers.
The Deal and Rationale
Tronox will be buying two mineral sands mines in South Africa and Exxaro’s 50% stake in the mine and pigment operations of their Tiwest JV. (Tronox has been operating a JV of an integrated mineral sands and pigment operation in Australia with Exxaro for over 20 years.) Upon closing, TROX will have 465 thousand mt of pigment capacity, 600 thousand mt of Tio2 feedstock capacity, and 205 thousand mt of zircon capacity.
Tronox will issue 10 million shares in New Tronox to Exxaro. Exxaro will retain a 26% ownership in the South African mineral sands assets to comply with South African black ownership (BEE) regulations. There is a put/call provision for 1.4 million shares to be issued in exchange for this stake if Exxaro’s retention of the ownership stake is no longer necessary to comply with regulations. The deal gives Exxaro a 38.5% stake in the newly formed entity (42% assuming an exercise of the put/call shares). Tronox shareholders will exchange their shares for shares in the new entity as well as $12.50 per share to return cash generated by Tronox between the deal announcement and the anticipated closing in the first half of next year.
Tronox outlines their rationale for the transaction in detail in their slide presentation on the deal:
Securing raw materials for the pigment operations in a time of potential supply shortage and ensuring a platform for potential capacity expansionProviding a more stable earnings model by not being subject to raw material price fluctuationsEnhancing earnings by being able to capture value at multiple parts of the supply chainAdding value through the synergies created by an integrated operation[/list]All of these rationales are logical at any point in the cycle (and industry analysts have wondered why we had not seen extensive vertical integration until now), but they are especially timely given the tightness of ore supply, increasing ore cost, and the positive market fundamentals in place for both the feedstock and pigment operations.
Earnings and Cash Flow
“New Tronox” is somewhat difficult to model right now because prior to the deal Tronox and Exxaro did a substantial amount of business together through ore contracts and the Tiwest JV. Additionally, it is not yet clear to what extent Tronox will utilize the vertical integration to source feedstock internally. Tronox has said that in 2012 Exxaro’s feedstock supply is already contracted out so there will be no incremental sourcing of feedstock to Tronox beyond what had already been contracted with Exxaro. They have said that by 2014 supply contracts will have run off and Tronox could potentially source all of its ore internally.
In my model (click here to download) I have attempted to look at 2012-2014 as well as a steady state case under stable long term market pricing for feedstocks and pigment. I used the TZMI pricing forecasts for feedstocks and pigment, which Tronox also utilized in their presentation. In this scenario pigment pricing peaks in 2012 and then starts slowly falling before reaching a steady state in 2015. Ore pricing peaks in 2013 and then drops sharply before stabilizing in 2016. However, the steady states for both pigment and feedstock are a “new normal” that are well above the pricing seen prior to 2010. As Tronox management noted, this is a more bearish scenario than many industries analysts predict. Other analysts predict pricing rising somewhat more slowly but not peaking until well beyond 2013. Either way, what I think is most important for investors is that nobody is predicting a return to the depressed pricing environment of the 2000s. Both the feedstock and pigment producers have operated in long suffering oligopoly markets so they should be very conservative and rational in their capacity expansion investments.
To ease the analysis and be conservative I will analyze the mineral sands and pigment operations independently, and assign no value to integration synergies in the manufacturing operations. I will include the $30 million in initial cost savings from SG&A and logistics synergies that the company has earmarked.
Here are the major assumptions in my model:
Ore and pigment market pricing- per TZMI96% capacity utilization- based on Tronox management comment on practical full utilizationEBITDA margins: I assume 35% peak EBITDA margins on the pigment side, declining to 30% as ore prices catch up with pigment pricing in 13 and 14. Tronox is the only major pigment producer with 100% chloride production, which is regarded as more cost efficient than the sulfate process. Industry analyst Ti Insight has said that Tronox coming out of bankruptcy has the second lowest cost curve in the industry behind Du Pont (which utilizes a proprietary production process). Right now Tronox’s EBITDA margins for 1H 11 are 25%, which compares slightly unfavorably to the 28% 1H 11 margins of the Tio2 pure play KRO. But I think the current numbers are clouded by Tronox’s revenues from selling Exxaro’s share of Tio2 pigment from the JV, which Tronox books as revenue but only makes a small commission on. I assume the pigment side will at least hit the 35% peak margins I predict for KRO in 2012 and then fall back to the 30% range that KRO has done in Q3 11. On the mineral sands side, Exxaro had 31% EBITDA margins in the 1H 11, but that includes intercompany transactions. The closest comp Iluka had 50% EBITDA margins in 1H 11 for their mineral sands operations. For the miners, price raises are totally incremental margin so margins can expand very quickly as selling prices rise. I assume 45% peak margins for the mineral side in 13, and 40% margins in 12 and 14. For steady state margins I assume 27.5% for pigments and 35% for feedstock.[/list]Here are the EBITDA and cash flow numbers:
These assumptions yield EBITDA of around $1.5 billion for 2012-2014 and $1.05 billion in a steady state. In their deal presentation Tronox has guided for EBITDA of $1.4-1.5 billion in 12, $1.7-2.0 billion at the peak, and $1 billion in mid-cycle. I think the model confirms those numbers are reasonable. And they might even be conservative given that my scenarios assign no value to manufacturing synergies and the pricing scenario could very well be more bullish in the mid-cycle.
Tronox should be a significant generator of free cash flow. TROX is not taking on any additional debt in the transaction, and currently only has $470 million of debt and ~$30 million a year in cash interest expense. They also have a significant amount of deductions available to reduce cash taxes:
$160 million in pre-bankruptcy NOLsAn approximately $1.1 billion deduction to income in 2011 for value in excess of claims provided to creditors in the restructuring, which will create an NOL going forward.Tronox paid ~$400 million in cash and other assets into remediation trusts to resolve their legacy environmental and tort liabilities. They did not take a tax deduction on this payment at the time, but rather will receive the deduction as the money is disbursed from the trusts.[/list]The use of the NOLs will be subject to change of control limitations, but the company still estimates the tax assets have an NPV of $300 million. Combined maintenance capex is estimated at $90 million. The company is expanding one of the South African mines at a cost of $200 million between now and 2014. If we assume only a $50 million a year benefit due to the tax assets then FCF before working capital changes would be in the range of $850-900 million in 2012-14. The company has guided for $700 in levered FCF in 2012, so that is in the ballpark if we assume an uptick in working capital given the increasing sales.
TROX plans to institute a dividend in line with chemical industry peers, and the CEO on the recent conference call said it will be in the 2-3% range. They have also talked about return of cash to shareholders beyond the dividend, which I presume would be a share buyback. In the next few years they have the aforementioned $200 million project to expand mine capacity, and they will probably consider a project to expand pigment production capacity. The company is largely unlevered (less than .5x debt/steady state EBITDA) leaving them with significant balance sheet flexibility.
This is the pro forma capital structure:
The Exxaro assets are being contributed cash and debt free. I assume an exercise of the put/call shares in the share count.
Chemical comps vary, but generally trade in the 5-8X EBITDA range. I think 6X is a conservative multiple here. Whether to value off of 2012-2014 EBITDA or something more like the steady state scenario will obviously depend on what you think is sustainable over the longer term. But even if we take 6X steady state EBITDA that gives us a $225 stock. Taking a 10% FCF yield on steady state FCF gives us a $235 share price. And neither of those targets include the $11 per share NPV of the tax assets. Those are quite attractive targets for a $125 stock, and there is even more upside potential if the stock gets rated off of something closer to peak earnings and cash flow.
The key risk is that pricing momentum fails to materialize. However, the company is guiding for pro forma EBITDA of $780-800 million in 2011 (they have done $336 for the pro forma entity in 1H 11), and we are through three quarters of 2011 with strong pricing momentum for both feedstocks and pigment. I think there is almost no chance prices for feedstock and pigment drop over the next few years. So even without any further price increases beyond this year TROX is trading at 4.8X EBITDA, which is incredibly a below comp multiple. And it is important to note that through their vertical integration TROX has mitigated the ore cost inflation risk facing the other Tio2 producers.
Cyclical downturn: As discussed above, in a macro downturn selling volumes will get hit some but I expect pricing will remain flat and Tronox would be able to stay well above breakeven.
Deal closing: The deal announcement clearly disappointed those in the existing shareholder base that were looking for a quick exit post restructuring. Some have also questioned why Tronox would do an all-stock transaction before listing on a major exchange. But I don’t think Tronox got ripped off here and I could understand why they do not want to relever the company right after coming out of bankruptcy. The deal makes sense from both a strategic and financial perspective, and as management has explained their rationale more fully in two presentations and conference calls I don’t think getting shareholder approval will be a problem.
Integration with Exxaro: The usual merger integration concerns seem largely inapplicable here given the logical fit of the vertical integration and the fact that Tronox and Exxaro have already been working together for two decades with the exact same business model.
The major catalyst will be the quickly growing earnings and cash flows as ore and pigment prices continue to rise. We will also have the post-reorganization catalysts: major exchange listing following the deal close in 1H 12, increased sell side coverage, increased marketability of the stock etc. Management has indicated a dividend will be instated, which will draw income investors to the stock.
TROX versus KRO
If you buy into the Tio2 thesis then both TROX and KRO are attractive stocks at current valuations. I currently own both but favor TROX for several reasons:
Vertical integration will mitigate the risk of rising ore costs eating into pigment margins and allow TROX to also capture value on the ore sidePost reorg catalysts should provide an additional boost to the stock beyond rising cash flowsWhether valid or not, KRO has traded with sentiment on Europe (even more so than the general market) due to their heavily European customer base.[/list]
Disclosure: I own shares of TROX and KRO
Elie Rosenberg runs a value investing research website at valueslant.com. Sign up here to get his free value investing ideas and analysis by email and get his free ebook, "16 Ways to Find Undervalued Stocks."