|Shares Out. (in M):||15||P/E||0.0x||0.0x|
|Market Cap (in $M):||108||P/FCF||0.0x||0.0x|
|Net Debt (in $M):||173||EBIT||0||0|
Note: This write-up uses U.S. and Canadian dollars interchangeably given how close the two currencies are to parity.
|Share Price||$ 7.20|
FTP shares, down 70% YTD, trade for approximately 1x a reasonable estimate of mid-cycle earnings per share. The company has been left for dead after a string of operational disappointments, increasing competition, falling prices, and indiscriminate tax-loss selling going into the end of the year. The current FTP share price ascribes a significant probability to bankruptcy, but with $84 million of cash in the bank, no material debt maturities until 2015, and a debt structure with minimal recourse to the parent company, Fortress is in no danger of near-term financial distress. An investment in FTP is not without risk, but if Fortress can get back on track then those who buy in at this price stand to make many multiples of their original investment.
Fortress Paper was founded in 2006 by its current CEO, Chad Wasilenkoff, as a vehicle for contrarian investments in beaten-down forestry assets. The company has been written up previously on VIC and elsewhere, and I suggest those write-ups for company and industry background. The recent initiation by CIBC contains an abundance of background and industry data as well.
Fortress has four primary assets:
1) A dissolving pulp mill in Thurso, Quebec, that Fortress recently converted from an idled kraft pulp (NBSK) mill. This mill should reach full production capacity in Q1 or Q2 of 2013
2) A nonwoven wallpaper mill in Dresden, Germany
3) A currency paper mill in Landquart, Switzerland
4) An second idled kraft pulp mill, this one in Lebel-sur-Quevillion (“LSQ”), Quebec, which Fortress just purchased, and which Fortress plans to convert into another dissolving pulp mill beginning in 2013
Why Fortress Isn’t Going Bankrupt
I will discuss each of the Fortress assets individually, but first I want to jump directly to the company’s financial situation. As of 9/31/12 Fortress had $84 million of cash on the balance sheet versus $257 million of debt (yielding a blended cash rate of 5.8%). $104 million of the debt was provided to Fortress by the Canadian government in order to finance the Thurso mill conversion. This debt pays 5.5%, is non-recourse to Fortress, and does not begin amortizing until 2015. Here is a schedule of Fortress’s future debt maturities, before including the debt Fortress will draw to fund the LSQ conversion:
|Non-Recourse -- Thurso||38.4||12.8||12.8||12.8||12.8||12.8|
As you can see, Fortress does not face a material debt maturity until 2015, and does not face one with recourse to the parent until 2016. There are no important covenants endangering the company either. Conversely, it is possible Fortress will be cash flow positive in 2013 (before discretionary capital investments in the LSQ mill conversion, which I will discuss in a bit).
So while the share price suggests in serious distress, the reality is different.
Asset #1 – Thurso Dissolving Pulp Mill
In 2010 Fortress negotiated a deal with the government of Thurso, Quebec, to take over an idled kraft pulp mill there and convert it into a dissolving pulp mill. Fortress paid a $1.2 million net purchase price and promised to invest $150 million for the conversion of the mill. In return, Fortress was able to eliminate the mill’s defined benefit pension liability, gain indemnity from its legacy environmental liabilities, and negotiate a 7 year union deal at a 15% discount to national rates. Fortress was also able to enter into a 15 year fiber supply agreement, as well as a 15 year agreement to sell its excess power to a Hydro Quebec at subsidized “green” energy rates.
The now-converted Thurso mill has nameplate capacity of 200,000 metric tons. The conversion was completed earlier this year and the mill is now working through optimization and de-bottlenecking. In the last few weeks the mill has been operating above 90% of capacity, and hopefully should be able to operate at that level consistently by sometime in early 2013. In the company’s slideshow you can find a timeline graph of Thurso’s capacity utilization.
In Q1 2013 Fortress will also complete the construction of a $100 million cogeneration recovery boiler at Thurso. This boiler will produce power as a byproduct of the dissolving pulp production process. The power will sold under a contract that will generate $20 million of EBITDA annually, reducing the mill’s cash cost per ton by approximately $100. All in, Thurso’s cash production cost will be in the high $700s, making it lower-cost than approximately two thirds of the DP industry.
Now seems like a good time to discuss dissolving pulp more broadly. Dissolving pulp is used primarily to make rayon. Growth of the middle classes in emerging economies is driving increased demand for textiles, and rayon is gaining popularity as a substitute for cotton. From 2000 to 2010, global dissolving pulp production grew from approximately 3 million tons to 5.3 million, a 6% CAGR. The pace has quickened in the last few years, as DP consumption has grown about 12% annually since the financial crisis.
The dissolving pulp industry is growing but it is volatile. When the price of cotton went through the roof in 2010, textile manufacturers upped their usage of rayon, and the price of dissolving pulp followed cotton upwards—DP prices were above $1,500 for most of that year, and even hit $2,500 at one point. Cut to today, and the price of dissolving pulp has fallen to somewhere between $900 and $950 per ton (you can find a price graph for dissolving pulp in the company’s slideshow).
Falling cotton prices have contributed to the DP price plunge, but increasing industry capacity has been a problem as well. The industry has just gone through something of a land grab. The bout of high cotton prices and the rapid turnaround in rayon acceptance made it evident to everyone that dissolving pulp has a bright future. What ensued was a rush of kraft mill conversions in North America as well as a number of greenfield mill constructions in China.
New capacity that was planned in 2010 and 2011 will continue to come online for a few years, but I believe there will be some mitigating factors. First and most importantly, at today’s DP price, between one third and one half the industry is losing money. Further, much of the new capacity coming online is located in China, where the lack of native fiber sources makes Chinese mills the highest-cost producers in the industry, at a cash cost of roughly $1,200 per ton.
The equilibrium price for dissolving pulp seems to be approximately $1,200 per ton—this is the price at which the Chinese mills can cover all of their costs. The Chinese are the industry’s marginal producers, and right now all of them are losing money. Some of the mills will tolerate losses for a while longer because they are start-up mode, but I believe a lot of this capacity will be forced to idle if the price of pulp doesn’t recover soon.
Returning to Fortress’ Thurso mill, to date the mill has lost money because production has been ramping up. But the mill should swing to profitability sometime in early 2013 as production reaches full 200k ton capacity and the cogeneration recovery boiler comes online. At that point, the mill should generate approximately $30 million of EBITDA even at today’s very low $900-950/ton price for dissolving pulp. At the marginal DP price of $1,200, Thurso EBITDA should be in excess of $75 million.
$30-75 million of EBITDA compare favorably to the mill’s maintenance capital requirements of less than $15 million and cash interest of approximately $5 million. The Thurso mill has taken longer to bring online than the company originally anticipated, and FTP shares have paid the price for it, but once the mill hits its stride, it will contribute roughly $15 million of cash flow before interest, even at today’s pulp price —not insignificant in light of Fortress’ current $100 million market capitalization. At a higher pulp price the mill will generate a lot more than $15 million of cash flow, as the costs of running the mill are largely fixed (this is true for each of the Fortress businesses).
The Thurso mill conversion was partly financed by the Quebec government via a $102 million, ten year, non-recourse loan paying 5.0-5.5% interest. The friendliness of this financing source cannot be overstated. The Quebec government’s #1 concern is jobs—namely the preservation of the jobs at the Thurso mill. In the extremely unlikely scenario that Fortress is unable to meet the Thurso debt amortization payments that begin in 2015, I strongly believe the ultimate outcome would be a renegotiation of terms rather than any material financial remedy or recourse. The Quebec government has no desire to repossess the Thurso mill or push Fortress into insolvency.
Asset #2: Nonwoven Wallpaper Mill in Dresden, Germany
In 2006 Fortress acquired this mill from Mercer International. The mill is the largest nonwoven mill in the world, accounting for almost 60% of worldwide nonwoven wallpaper supply. Nonwoven wallpaper is easily removable, flame-retardant, and potentially heat-efficient and antimicrobial as well, depending on the treatments applied. The nonwoven market is small, but because nonwoven wallpaper is superior to traditional wallpaper in so many ways, it is taking market share, and those share gains are driving total nonwoven volume growth of 5-10% annually.
The Dresden mill’s paper machine is the biggest and fastest in the industry. Nobody else can produce nonwoven wallpaper as profitably. The mill’s management has been able to squeeze out additional capacity year after year, and on the most recent conference call the company said it has a new capital plan to increase capacity from the just-reached 60k tons all the way to 70k tons. Because the industry is so small, building a new mill at this point in the industry’s lifecycle is uneconomical. Doing so would increase industry supply by 20-60% and punish everyone’s margins. The result is constrained supply, and therefore high margins—25% at the EBITDA level—for Fortress.
The mill is a cash cow. When Fortress bought the mill, it was producing $5 million of EBITDA. Today the mill produces $40 million, and I believe EBITDA in 2013 will approach $45 million, against CapEx of about $8 million. 95% of the mill’s sales are in Euros, so there is no FX mismatch risk.
Valuing the mill is a bit of a puzzle. A few years ago there were rumors in the market that Fortress was close to a deal for the mill at 7x. If you were to show the mill’s financial statements to a group of analysts without divulging the nature of the underlying business, I think most of them would say the business merits a 9-12x EBITDA multiple. Fear about potential industry capacity expansion—and therefore margin erosion—probably drives some of the gap in multiple. But it is worth noting that nobody in the industry has announced any large-scale capacity expansion plans. Maybe more importantly, Ahlstrom, a major wallpaper competitor, recently canned a failed 20k ton nonwoven wallpaper machine project. Ahlstrom was attempting to build and operate a replica of Fortress’ Dresden machine, but the company couldn’t get the technology or production process right. After two years, millions of dollars, and not a single saleable ton of nonwoven wallpaper, Ahlstrom gave up and shut the machine down.
Not surprisingly, Fortress receives regular indications of interest in the Dresden mill from both strategic and financial buyers. Obviously Fortress’ technology and knowhow would be quite valuable to Ahlstrom, or most other wallpaper competitors for that matter, and the financial profile of the mill should make it attractive to private equity buyers as well.
Eventually this mill will be sold or spun out. What it is worth will depend on the conditions under which it is divested, but assuming a 7x multiple and $40 million of LTM EBITDA, it would be worth $280 million. A more conservative 6x would be $240 million.
Asset #3: Landquart Security Papers Mill
This mill, located in Landquart, Switzerland, produces banknote paper for the Swiss Franc as well as other undisclosed currencies. The mill makes money, but right now it’s losing a lot of money. I don’t want to spend too much time discussing it because it isn’t very important in the grand scheme of things.
From 2004 to 2009 the Landquart mill generated average annual EBITDA of $6 million. Beginning in 2010 Fortress invested $65 million of fresh capital into the mill to expand capacity and install cutting-edge security feature technology on its printing presses. As part of this process, the mill became a dedicated currency paper mill, and stopped producing cheque paper and other security documents. The loss of revenue from these ancillary products, combined with a few other lost contracts, has left quite a bit of the mill’s capacity unused. Meanwhile, the sales cycle for new currency orders has been longer than originally anticipated, and Fortress has had to deal with rising production costs as a result of the strong Swiss Franc (ironic?). In 2011 Fortress announced a new banknote contract that will absorb most of the Landquart mill’s remaining capacity, but the contract has been slow to ramp.
Losses have been the result. The company has not provided any profit guidance related to the new contract, but the people who run Fortress are not stupid and presumably they would not be willing commit to a significant long-term contract unless the contract could restore the mill to profitability. But we will see. The contract is supposed to commence at the end of this year. All in all I expect the Landquart mill will lose money in 2013 (but less than in 2012), but hopefully become profitable thereafter as a result of new contracts and increasing interest in DuraSafe, a patented security technology that is the best in the industry.
Ultimately this mill will be sold, most likely to a strategic buyer, or it will be liquidated. Both scenarios would involve a positive cash recovery for Fortress. The mill’s assets and technology are valuable to a strategic buyer, and in liquidation the real estate is valuable and the severance and closure liabilities would not be large (perhaps 10 million CHF) thanks to the Swiss government’s business-friendly orientation. So while Fortress is unlikely to recoup its full investment in this mill, I believe the company will still be able to realize some value from it.
Asset #4: Dissolving Pulp Mill in Lebel-Sur-Quevillion
In June of 2012, Fortress closed its second deal for a kraft mill conversion, this time for a mill in Lebel-Sur-Quevillon (“LSQ”), Quebec. The mill was idled by Domtar in 2005 and permanently closed in 2008. This deal is substantially similar to the Thurso deal. Fortress is buying the mill for a nominal price, and in return it is committing to convert the mill, this time for $222 million. Upon completion, the LSQ mill will produce 250,000 tons annually at an even lower cash production cost than Thurso. Fortress estimates that the LSQ mill will have a cash cost in the low $700s per ton—lower than 75% of global DP capacity.
The LSQ deal also is being financed in the same way Thurso was. The Quebec government is lending Fortress $132 million in the form of loans that pay 5-5.5%, are non-recourse to the parent, and do not begin amortizing until 2016. This debt is not yet drawn. Fortress financed the remainder of the conversion cost through a July convertible issuance at a $31 conversion price.
Fortress plans to spend $30-40 million in 2013 to bring the mill back online as an NBSK mill. Then it will spend the remaining $180 million or so over the next twelve months in order to convert the mill to dissolving pulp production, putting the mill on course for profitable DP production sometime in 2015.
How Will It All Play Out?
On a consolidated basis, here is my rough estimate of how the financials will look IF the price of dissolving pulp doesn’t budge from its current low level:
|Dissolving Pulp - Thurso||4||30||30||30||30|
|Dissolving Pulp - LSQ||0||0||0||25||50|
|Dissolving Pulp - Thurso||-29||-14||-14||-14||-14|
|Dissolving Pulp - LSQ||0||-40||-150||-46||-15|
|LSQ Debt Drawdown||132|
Again, I think it worth emphasizing that at current DP prices, one third to one half of the industry is losing money. I will let you decide for yourself whether you think it is that the industry will continue to operate in a lossmaking position for three straight years. It is also worth noting that Fortress has flexibility in its LSQ conversion timeline—if cash starts to get tight, Fortress can delay some of the LSQ capital investments.
In a scenario in which DP prices gradually increase to $1,200 by 2015, this is how I think the financials might look:
|Dissolving Pulp - Thurso||4||45||63||81||81|
|Dissolving Pulp - LSQ||0||0||0||57||114|
|Dissolving Pulp - Thurso||-29||-14||-14||-14||-14|
|Dissolving Pulp - LSQ||0||-40||-150||-46||-15|
|LSQ Debt Drawdown||132|
These 2016 estimates represent my guess for a “steady state” of mid-cycle pricing and maintenance-level capital expenditures.
Either way, I think Fortress will be fine as long as DP prices don’t fall materially further. What if they do, and then they remain low indefinitely? This would undoubtedly be the worst-case scenario, as it would entail four years of prices low enough to wipe out most of a growing, capital-intensive industry. Fortress would have decisions to make. The company could walk away from the LSQ conversion completely, sell Dresden for a material but discounted sum, and/or jettison Landquart for a positive but nominal sum. None of these options would be ideal, obviously, but pursuing them would enable the company to survive. And even if Fortress gave up on LSQ totally, it would still be in position to earn ~$3+ per share on a mid-cycle basis, as I’ll discuss in a moment.
What’s It Worth?
I’ll start with a valuation based on earnings. Here is my estimate of mid-cycle earnings, based on all of the foregoing:
I’m not sure what kind of multiple to put on these earnings, but the answer is not 1x. Clearly there is room for a return of many multiples of your investment if Fortress can get through the current rough stretch.
Asset value is another way to look at FTP. In 2011, when DP prices got as high as $2,500 per ton, there were three non-Fortress DP mill transactions. In total, these transactions valued the mills at roughly $1,600 per ton of converted capacity, including both the mill purchase price and the cost to covert the mill.
|Prevailing DP Price||Date||Mill Location||Buyer||Seller||Value ($M)||Tons (K)||Value/ Ton|
|$ 2,000||2/2/2011||Port Alice||Fulida Group||Wellspring||285||160||$ 1,781|
|$ 2,500||4/18/2011||Sweden||Aditya Birla||Domsjo Fabriker||340||210||$ 1,619|
|$ 2,200||5/3/2011||Prince Albert||Paper Excellence||Domtar||225||145||$ 1,552|
A year later, in 2012, the Aditya Birla Group bought another idled kraft mill for conversion, this time in Terrace Bay, Ontario. There are conflicting reports on the purchase price, but it equated to something between $1,200 and $2,000 per ton depending on which purchase price you use. Even though the Terrace Bay is substantially higher-cost than Thurso or LSQ (rumored to be ~$1,000/ton cash cost), let’s assume that the $1,200/ton purchase price is correct and applicable to Fortress’ dissolving pulp assets. Here is what a sum-of-the-parts might look like for Fortress under that assumption.
|Dissolving Pulp - Thurso||Capacity (kmt)||200||$1,200||240|
|Dissolving Pulp - LSQ||Capacity (kmt)||250||$1,200||300|
|Wallpaper - Dresden||EBITDA ($M)||$40||7.0x||280|
|Security Papers - Landquart||10|
|Gross Value of Assets||740|
|Net Debt (incl. LSQ debt)||305|
|Net Value of Assets||435|
|Net Value per Share||$29|
Finally, Fortress has three convertible debentures outstanding, totaling $114M of debt. They mature between 2016 and 2019 at a weighted-average conversion price of $33. For the purposes of this write-up I’ve assumed they are NOT converted because the FTP share price would need to rise 450% for them to be in the money. Obviously, if the FTP share price gets above $33 then they will be converted. In that scenario the company would be free of $114M of debt, but at the cost of 25% dilution (4 million new shares). For those interested, here is the sharecount math:This is NOT replacement value—the cost to build a new mill is probably $2,000/ton in China, and above $3,000 per ton in Brazil. But those values are not pertinent here because we will not see a greenfield DP mill in North America any time soon.
|Share Type||Quantity||Strike/Conversion||Proceeds ($M)|
|Dec 2016 Convert||1.073333||$37.50||40.2|
|June 2017 Convert||0.774473||$32.28||25.0|
|Dec 2019 Convert||2.225806||$31.00||69.0|
Fortress is founded, run, and 12% owned by its CEO Chad Wasilenkoff. Chad has been a serial entrepreneur since childhood. He has a knack for making contrarian investments in beaten-down resource sectors. He spent five post-college years at Cannacord as an investment adviser, then quit in 2002 to form a “blank check” company that became Dynasty Metals & Mining, a gold miner. After cashing out of that deal, he took control of a distressed uranium producer in 2004 for $200,000. He took that company public in 2005 at C$0.55, and the stock quadrupled by the time he passed the CEO reins over to the President a year later at the peak of the uranium bubble. Forestry was his next contrarian resource bet. In 2006 he founded Fortress Paper and bought two assets: a distressed wallpaper mill in Germany and a distressed currency papers mill in Switzerland. In 2007 he took Fortress public at C$8 per share.
At the height of the DP frenzy, FTP shares hit $60 and Chad was a darling of analysts and investors. Today, however, many of those same people are angry with him. He has been accused of being overly promotional and of failing to execute. He is guilty to some extent—his targets have consistently been too aggressive, and he has not done a good job setting market expectations conservatively. He is naturally optimistic and aggressive, and too often he has let those traits seep through into his public comments and into the market expectations for the company.
But Chad and Fortress have also suffered from some bad luck. The Thurso plant conversion has been delayed by everything from national mill worker strikes to lightning strikes. Fortress is accused of being “unable” to execute, but the NBSK-to-DP mill conversion is a new and complex process, one that has only been completed a few times. The dissolving pulp operations manager that Fortress hired, Peter Vinall, probably has more experience with these conversions than any other human being on the planet.
Personally I think Chad is a brilliant capital allocator. He has turned $180 million of paid-in capital into an enterprise that I believe is worth closer to $1 billion than $180 million. The carefully constructed debt structure shows how much thought has gone into the company. What’s more, Chad is still young and this is not his last rodeo. His reputational (as well as financial) investment in Fortress is huge and he will work as hard as he can in order to get things back on track. I have received e-mails from him at 11pm on Saturday night and phone calls at 9am on Sunday morning.
The shareholders who have lost 80% of their investment in FTP will probably never have a positive view of the stock again, but that is part of what creates the opportunity. I have seen this same cycle of shareholder elation and revulsion play out quite favorably with some other left-for-dead small caps, including AIPC, SWM, and MIC.
A Final Note, on Cotton
I originally came to Fortress through a broad examination of agricultural companies. The #1 driver of the price of dissolving pulp is the price of cotton. Cotton is one of the few agricultural commodities that is not currently at or near an all-time high. Cotton already had its moment in the sun in 2010, when prices spiked to their highest levels since the American Civil War. Those high prices spurred massive plantings that filled world inventories. Now cotton is in the doldrums and every farmer in the world with a choice is planting corn rather than cotton, because the profit differential is astronomical. Eventually this lack of cotton plantings will whittle away the cotton supply and drive cotton prices higher again. This cycle is inescapable.
The long-term outlook for the cotton price is up, up, up. Emerging middle classes grow their cotton consumption even more than their protein consumption. Yet worldwide cotton acreage has been in decline for decades as a result of urbanization, soil erosion, and better prices for protein-related crops. Cotton farming requires a lot of water and fertilizer. Chinese cotton farming is already a large-scale environmental disaster (Google “Aral Sea disaster” for a precedent to what is happening in China) and the government there is trying to shift textile production toward rayon. The simple fact is that rayon can be produced more cheaply than cotton, and in a way that is far friendlier to the environment. Over time, demand for cotton will only increase, but cotton will only become more expensive, and higher cotton prices will mean higher demand for cotton substitutes, and therefore higher dissolving pulp prices.