Description
Note that all figures are in Canadian dollars.
REASONS FOR OPPORTUNITY:
Fortress' share price is down around 50% since February highs. The main reasons are:
- Generally weak recent financial results, although mainly driven by near-term transitory issues (planned facility overhauls, an ongoing facility conversion project)
- Market concern over productive capacity additions in the dissolving pulp space
- Possible profit taking given strong stock price performance over the last year
- Recent lawsuit filed by a competitor claiming breach of unspecified confidential information by former employee
- Recent $57.5mm dilutive equity offering, although it alleviated intermediate-term funding needs
- Japanese earthquake / tsunami, which impacted a variety of industries, including dissolving pulp
- All probably exacerbated by limited trading liquidity (average volume around 60k)
The only potential structural concern above relates to capacity additions. Fair enough, spot prices will likely come down from recent elevated levels. But as will be discussed below, the company has its production locked up for 5 to 10 years at prearranged prices that imply substantial EBITDA generation. The bottom lines is that as a result of the dislocation, at current levels, you can buy a contracted and visible cash flow stream for around 3x. Or said another way, the current reward-to-risk relationship is very asymmetric at 3x1 under what I consider highly conservative assumptions, with a vary identifiable catalyst on the horizon.
BRIEF DESCRIPTION OF BUSINESS:
Fortress is an international producer of security and other specialty papers and products, incorporated in May 2006 after being spun out of a company called Mercer International. They are currently listed on the TSX under ticker symbol FTP.
Side note: more detail on the historical business can be found in the company's financial reports. I haven't spent a ton of time here because, as will be discussed below, the largest driver is still unfolding and isn't reflected in the historicals. This no doubt presents some risks to this story (notably execution), although to my mind the recently negotiated long-term production contracts serve to de-risk the "bet on growth" aspect here...
The main contributor to the company's value is the Thurso pulp mill, which is described below.
Thurso Pulp Mill
- Located in Quebec and in the process of being converted from northern bleached hardwood kraft pulp to dissolving wood pulp
- Also known as specialty cellulose, dissolving pulp is a refined pulp used in products such as cellophane, film, computer chips, etc
- The primary use is in rayon, a cotton substitute and silk-like product growing in popularity within the textile industry
- Cotton supplies are currently and expected to be tight due to cyclical weather issues and structural lack of dedicated land
- Yet demand from emerging countries like India and China remain strong as populations continue to enter the middle class
- In light of the supply/demand disconnect, producers are currently adding capacity which some fear will alter supply/demand
- Demand has historically been determined by the level of economic growth and has been closely tied to overall business activity
The Thurso pulp mill is the core of the story here. Here's the 30,000 foot background. In mid-2010, the company purchased an idled pulp mill for next to nothing and concurrently secured roughly $125mm of mostly low-cost project financing (@ 5.0% to 5.5% annually for 10 years) to fund the conversion of the acquired mill to dissolving pulp production given the above-noted strong industry dynamics. The converted mill will have annual design capacity of roughly 200,000 metric tons and is expected to be completed in Q3-2011. Not surprisingly, the initial market reaction was favorable and the stock price rocketed. But after a string of recent headlines, some of which were noted above, the market has given back some of the gains. Yet despite all those "setbacks", most of which appear to be transitory, the company has announced positive developments related to the conversion project, which the market seems to be either ignoring or drastically discounting.
In late 2010, the company successfully locked up close to 80% of the converted mill's capacity with three rayon producers out of Asia. The contracts range between 5 and 10 years and are subject to floor and ceiling prices of $1,200 and $1,600 per metric ton. This pricing is admittedly below current spot pricing (which have recently reached upwards of $2,000 to 3,000 per metric ton) but nonetheless supportive of EBITDA generation in the $100-200mm range depending on the price and expected cash operating costs. For perspective, the company did $22mm of EBITDA in the LTM period, so you can see how game-changing the ramp up of this asset will be to financial performance...In the valuation section, I'll highlight the fact that this asset alone is arguably worth in meaningfully excess of the company's current stock price, implying that you're easily getting the other assets for free.
Outside of Thurso, the company has two legacy paper mills, as described below.
Dresden Paper Mill
- Located in Germany and produces mainly non-woven wallpapers
- The non-woven wallpaper market is fairly niche with not many pure play public comparables or competitors
- Global demand is generally declining, but the non-woven segment of the market is materially increasing driven by emerging markets
- Generally, non-woven wallpaper is more easy to remove, dries more efficiently, is generally more consumer-friendly, etc
- Wallpaper sales are clearly correlated to new construction (residential and commercial), so its a non-developed market story, where the company focuses its efforts anyway
- Fortress controls 50%+ of the non-woven wallpaper market (up from 27% in 2006) and is looking to expand further
The Dresden mill has been the lone bright spot for the company in recent quarters. As of March 2011, the company was doing roughly $28mm of operating income on a run-rate basis on 15% higher shipments year-over-year. As far as 2011 is concerned, the Dresden mill will likely offset expected weakness at the Landqart mill, as discussed below, and obsorb some of the costs related to the Thurso mill conversion.
Landqart Paper Mill
- Located in Switzerland and produces security papers such as banknotes, passports, and other brand protection and other security papers
- This is generally a lower volume but higher margin product given the complexities involved in production
- Most players are private controlled by central banks; the remaining players compete for the remaining currency contracts
- Analysts estimate the size of the global banknote market to be 150-160k metric tons annually with approximately 50% open for bid
- The largest non-central bank player is De La Rue with production in fiscal 2010 of approximately 15.5k metric tons
- With recently expanded capacity, Fortress should have running room to become a significant global player behind De La Rue
- Long-term fundamentals in the banknote industry are generally positive:
- Use of paper money grows 2-3% annually
- For security reasons, banknotes have been getting more complex, a higher margin opportunity
- Credibility and reputation is established over many years making it difficult for new entrants
The Landqart mill is currently a drag on cash flow. For one, the company has shuttered one and has been upgrading another paper machine to gear the operation toward higher margin banknote products, admittedly an investment in the future but an impact to near-term cash flow. So for the last few quarters, the plant and both paper machines have essentially been operating off-line with close to full operating costs. The company has shuttered several employees as a result of the consolidation. As of Q4-2010, the upgrade has more or less been completed. Another reason for the drag is the global banknote market, which is oversupplied as several major currency introductions have recently been postponed. Management expects operations at Landqart to be tough for 2011.
VARIOUS PERSPECTIVES ON VALUATION:
There are several ways to look at valuation here, but all illustrate that the company is materially undervalued. On an absolute basis, the company looks very inexpensive. If you believe consensus analyst estimates, the company is trading hands for 2.7x 2012 EBITDA and 1.7x 2013 EBITDA. I personally don't give much credence to forward looking estimates, especially when such estimates are coming from research firms whose ulterior motive is to pull-through investment banking business. But that said, in this instance, I do believe it's important to look at valuation a few years out to normalize cash flow given the importance the Thurso mill conversion project will have to financials, and as noted, that's all on the come. Ideally, you'd want to look at the normalized EBITDA contribution from the business on a fully-ramped or mid-cycle basis. The following table highlights my high level estimates.
|
|
EBITDA ($mm)
|
|
|
Asset
|
Low
|
High
|
Notes
|
|
|
|
|
|
|
Thurso
|
$100
|
$190
|
Capacity sold at either floor or ceiling at mgmt operating costs shocked 15%. Realistically, uncontracted supply will be sold at higher spot prices
|
|
Dresden
|
$20
|
$40
|
Run-rate operating performance at mid-point
|
|
Landqart
|
$0
|
$18
|
Low-end assumes asset breaks even on a mid-cycle basis. High-end uses peer operating metrics to derive a reasonable normalized mid-cycle EBITDA
|
|
Corporate
|
($15)
|
($20)
|
Generally run-rate
|
|
Total EBITDA
|
$105
|
$228
|
|
|
Enterprise Value
|
$450
|
|
|
EV / EBITDA
|
4.3x
|
2.0x
|
|
|
Midpoint Valuation
|
2.7x
|
|
So on an absolute basis, the company is pretty cheap. And to me the above assumptions are pretty conservative. As noted, the company will more likely than not offload the uncontracted dissolving pulp supply at higher relative spot prices (the above analysis assumes all product sold at either floor or ceiling). Also worth noting that if dissolving pulp spot prices remain high, there's little chance that the company sells its product for the floor price, biasing the EBITDA potential to the high-end and thus valuation lower...
Another way to look at valuation is on a sum-of-the-parts or break-up basis, valuing each of the discrete business units relative to public peers or recent transactions. Looking at valuation this way corroborates the absolute valuation outlined above.
|
Asset
|
Mid-Point Mid-Cycle EBITDA
($mm)
|
Discounted
EBITDA
Multiple
|
Implied
Price
|
Notes
|
|
|
|
|
|
|
|
Thurso Alternative 1
|
$145
|
5x
|
$48.25
|
Peers trade hands for 6x forward EBITDA
|
|
Thurso Alternative 2
|
N/M
|
N/M
|
$22.46
|
A recent transaction priced at $1,600 per metric ton of capacity, but it's misleading as the asset was much older and has higher operating costs
|
|
Weighted Average
|
|
|
$35.36
|
Equal weighted the two approaches, but arguably more weight should be given to the EBITDA-based method
|
|
Dresden
|
$30
|
5x
|
$9.98
|
There are no public comps, but the company has reportedly received overtures at 7.0x EBITDA for this asset
|
|
Landqart
|
$9
|
5x
|
$3.00
|
The company's closest peer has historically traded at a trimmed mean of 7-8x EBITDA, and recently rejected a 11x EBITDA offer
|
|
Corporate
|
($18)
|
|
($1.20)
|
Generally run-rate
|
|
Enterprise Value
|
|
|
$47.14
|
|
|
Plus: Net Cash
|
|
|
$1.16
|
$75mm of cash and $58mm of debt. Note that I have used Q1 figures, but the company has an outstanding project financing line so there may be movement quarter-to-quarter
|
|
Equity Value
|
|
|
$48.29
|
|
|
Premium to Current
|
|
|
55%
|
|
I also think the above analysis is conservative. If you apply the multiples that the market is assigning to peers, and bias the Thurso valuation to the EBITDA-based approach, you could easily make the case for 100%+ upside from here. It should be noted that such high-side case isn't out of the realm of reasonableness, since the company was trading at those levels just a few short months ago. Which also puts my base case into perspective as very achievable.
Thinking about or justifying valuation another way, at the current market price of $31.12 per share and assuming a valuation for the legacy assets of $10-15 per share as implied above, you get to around $20.00 per share of implied value for the Thurso mill. Using the company's cash operating cost structure of roughly $600 per metric ton (which I shocked by 15%), I back into a net sale price for dissolving pulp in the $800-900 per metric ton range, which is a huge discount from recent spot prices AND the company's contracted pricing of $1,200 to $1,600...Just doesn't make sense...
Okay, that's all fantastic but what's the downside look like? On a very worst case scenario, let's bite the bullet and assign a value of $20-25 per share for the Thurso mill, which is consistent with the valuation implied by the above-mentioned recent M&A transaction, and may be what the market is assigning. (Again, that approach is very misleading given the higher relative cost structure and older age of that acquired asset, but for argument sake, we'll go with it.) The Dresden mill is currently doing run-rate operating income of $28mm or so in its growing market. Let's say it doesn't ever grow from here and assign a no-growth multiple of around 3x EBIT. That'll get you to around $6 per share for Dresden. As noted above, the Landqart mill is currently a drag on cash flow given recent upgrading initiatives and certain delays in the banknote market. Let's say it remains a drag forever and partially offsets the value contribution from Dresden. After accounting for corporate costs and capital structure, I could argue a low-case valuation of something like $25-30 per share, maybe something slightly lower than that. Using the base case value above of roughly $50.00 per share, depending on where you come out, one could easily argue for a 3x1 reward-to-risk ratio on what I would characterize as very draconian assumptions in all cases. And I would argue that the reward-to-risk is even higher.
I can hear you now: if the company is so cheap, than what is management doing about it? Well in mid June, they announced a new $10mm share buyback program, which represents 3% of the current shares outstanding. While not a huge number on the face, it represents the majority of net cash on the balance sheet as of March 2011, and more importantly illustrates that management is aware of the valuation disconnect and is willing to take measures to close the gap.
CAPITAL STRUCTURE AND LIQUIDITY:
The company's capital structure is adequate. They have $75mm of cash, most of which was recently sourced in a Q1-2010 bought deal equity offering, that fully covers $58mm of debt at a weighted average cost of around 5-6%. Most of the debt is revolving with intermediate-term maturities. Importantly, the company should be fully-funded to finalize conversion of the Thurso mill to dissolving pulp. The bulk of the funding will come from roughly $100mm of project financing from Investissement Quebec, only $9mm of which has been drawn as of March 2011, and cash on-hand.
MANAGEMENT BACKGROUND:
The company's CEO, Chadwick Wasilenkoff, is a well regarded entrepreneur, and has a background that's constructive to capital discipline. He has worked in industry, investment banking, and private equity. He also owns 16-17% of the company's stock, so his interests are clearly aligned with investors. He was recently awarded a $5mm cash bonus and $10mm in additional long-term incentive awards in conjunction with a renewed contract, which seems a little excessive given recent financial performance. However, he was pretty instrumental in locking up the supply contracts that essentially guarantees strong future financial performance, so I'll cut him some slack.
RISKS TO THESIS:
Main risks that I see are as follows:
- Probably most important is execution / cost overruns, delays, and implementation issues with the conversion project
- Dependence on a few key, large customer contracts (namely the dissolving pulp supply contracts)
- Pulp, cotton, energy, chemicals, synthetic fibre, and other input cost inflation
- New entrants in the non-woven wallpaper market and possible erosion of market share
- Foreign currency exchange risk
- Unwise acquisitions - the company has been clear that they are looking for further opportunities to acquire mills to convert to dissolving pulp. Assuming the implied returns on capital remain attractive, this could prove to be a sound strategy, but it increases execution risks nonetheless
Catalyst
-
The most obvious catalyst is the ramp up of the Thurso mill upon full conversion. It's expected to be online in late Q3-2011, and management believes it will be 4 months or so until it's fully ramped up to 100% deliverability. So it's a 2012 event, but a pretty clear catalyst nonetheless.
-
As noted above, there has also been some noise at the Landqart mill, which the company has been upgrading. The market has also been somewhat soft, so a re-ramp of production capacity at Landqart should also be a catalyst for the shares.
-
An additional catalyst will be further share repurchases.