Description

These debentures are money good (46% IRR thru year-end 2016) for a few simple reasons: Fortress already has the cash to repay them and cash levels should grow, plus they’re basically first in line to get repaid.  While this idea is clearly for small funds only, note that liquidity is probably better than it appears.  I was able to get $1.7 million face over a couple weeks without much impact on the price.  Other investors I know had similar experiences, but with larger amounts.

(All $ amounts in CAD unless noted.  Last traded at 73% of par.)

The Fortress Paper saga is a long one, some of which is detailed on past VIC writeups on the equity.  I’m happy to give my synopsis, but at this point all you really need to know regarding the 2016 debentures (the “2016s”) is the following:

1. The company’s biggest business – a dissolving pulp (“DP”) mill in Thurso, Quebec – has recently turned FCF positive and will likely continue to improve thru the maturity of the 2016s.  The USD/CAD rate is a risk, but you can hedge that easily.

2. Even if the Thurso mill returned to losing money, Fortress can likely borrow against (or sell outright) its other business – a mill in Landqart, Switzerland that makes security paper on which banknotes are printed.

3. Fortress had $63.2M in cash and securities at 6/30/15, well north of the $40.25M principal on the 2016s.

4. The guy in charge (Chad Wasilenkoff) is smart, hard-working, and owns 17% of the equity.  He has his flaws, but the last few years have taught him humility.

Deep dive on the Thurso DP mill:

DP is basically purified cellulose derived from plants – in this case hardwood trees from Quebec.  It’s used to make viscose, which in turn is used to make rayon for clothing and other textiles.  Demand grows ~7% a year for secular reasons (cotton is supply-constrained) and this should continue for quite a while.  

In recent years, however, supply has grown even faster.  Remember that spike in cotton prices about 5 years ago?  Well, DP prices spiked along with it, and pretty soon everyone with a dormant paper pulp mill was converting it to make DP.  At last this supply wave has receded, and there are no mills of any consequence scheduled to come online in 2016.  Both viscose and DP inventories have tightened significantly and as a result, DP prices bottomed in 1Q16 around USD 800/metric tonne.  Today they’re at USD 870/mt, and since there is a lag between the index price and the price Fortress receives, we should see revenues rise from here even if the index stays flat (but I assume flat pricing anyway).  

(Note: the benchmark price for DP is calculated by a consultancy called the China Chemical Fiber Group (CCF Group).  It represents the landed price in China (where most of the world’s DP goes) and is priced in USD, as are all DP transactions.)

Just as importantly, the CAD/USD rate has also weakened, boosting Fortress’ revenues.  Finally, a new manager at Thurso seems to finally be getting its cost structure fixed (the previous guy was terrible).  All three of these factors combined to bring Thurso EBITDA back to the black in 2Q15.  I’m highly confident that 3Q15 will be even better.

Fortunately, we have good visibility on DP supply thru year-end 2016.  DP mills are huge projects that don’t just pop-up overnight.  The outlook gets fuzzier as one looks further out, but thus far it appears that four large new mills could go online in 2017, and that’s the reason I’m not really interested in the 2019 debentures or the equity.  Note that Fortress also has $25.0M of debentures due to the Quebec government on 12/31/17 and $69M of publicly-traded debentures due 12/31/19.  Along with the 2016s, those are the only debts with recourse to the holdco.  The Thurso subsidiary owes ~$105M to the Quebec government, but there is no interest or principal due on this until 2017, and again, this debt is nonrecourse.

Notes:

1. Realized DP < index price due mainly to transport cost (~USD 100/mt), also to China tariffs, variations in cellulose purity, etc.

2. Thurso is a "swing" mill, capable of producing both NBHK (Northern Bleached Hardwood Kraft - i.e. paper-grade pulp).  All production should be DP from here on.

3. 2H15 estimates based on mgmt guidance of 80-82% utilization in this period.

Notes:

4. I've kept 2H15 COGS/mt steady at 2Q15 levels to be conservative.  Realistically, higher utilization will probably bring this number down.

5. I assume no more SG&A cost saves to be conservative.

6. Per mgmt guidance, cogen (cogeneration) revenues should be in line with 2Q15 at 22 MWh.  Like many paper mills, Thurso uses the byproducts of pulp production to generate eletricity which is then sold back into the grid.

7. I've included cogen revenues as an offset to costs because this is the way mgmt presents the numbers.  When they give guidance on costs, it's in terms of net opex/mt.  They're targeting CAD 725/mt over the next 18-24 months.  The increase in 1Q15 was due to a crack in the lime kiln, which will be permanently fixed during the annual shutdown in October.  Capacity estimates incorporate this shutdown.  While this fix will likely reduce costs, I've not factored in any improvement just to be conservative.

8. Excludes $2.0M of severance costs for a layoff in 2Q14.

Notes:

9. Wherever possible, I've assumed things stay at current levels in 4Q15 and all of 2016.  So, forex rates, CCF DP Index price, corporate overhead, cogen output, etc. all stay at today's rates/levels.  Note that I'm assuming no further improvement in production costs at Thurso, despite mgmt guidance.  The only variable that change are capex (using mgmt guidance) and Thurso capacity utilization in 2016, which I estimate will increase to 85% per mgmt guidance.

10. Thurso maintenance capex is $5.0M/year.  2H15 will see an additional $4.3M of discretionary capex intended to lower production costs.

11. Landqart maintenance capex is $3.0M/year.

12. I've kept corporate overhead at 2Q15 levels, except for two ridiculous payments of $1.0M each to be paid to Chad in 4Q15 and 4Q16.  These payments were triggered by his voluntary job change from CEO to Chairman.  Like I said, he has his flaws - being overpaid is one of them.

13. Keep in mind that $4.0M of this interest expense will go to the 2016s.

14. This expense could go away if LSQ is sold (more on this below)

15. When I said the 2016s are "basically" first in line, this is what I meant by basically.  This small amount of debt repayment will reduce the $5.5M of non-recourse debt that is owed by Landqart to the Swiss National Bank.

16. Note that ~$21M of this cash is currently restricted (tied to contracts at Landqart), but according to mgmt, substantially all of it should become unrestricted by year-end 2016.

Landqart security paper mill

This is a fairly stable and modestly growing business.  They have an innovative new product called Durasafe that could boost margins tremendously if it ever really takes off, but given the lengthy sales cycles in this industry, that’s unlikely to happen before the 2016s mature.   The mill’s land by itself has been appraised at CHF 69M (= CAD 95.4M).  If necessary, Fortress could do sale leaseback of the property, or perhaps even a sale or partial IPO of the business.  At a fire-sale 5x runrate EBITDA, Landqart should be worth at least $60M.

LSQ mill

This is a dormant pulp mill in Lebec-Sur-Quevillion (LSQ), Quebec that Fortress might be able to sell.  Fortress had originally hoped to convert it to DP, but at current prices and tariffs the math just doesn’t work.   And it’s contractually prohibited from making paper pulp for years to come per the agreement they made when they bought it from Domtar.  So it’s basically good for only two things: producing electricity with its cogen plant, or for scrap.  Mgmt seems confident that a sale is possible and that the proceeds would exceed of the ~$7M of non-recourse environmental liabilities.  Note also that a sale would eliminate the care and maintenance spending shown in the table above.

Finally, I should mention that there’s a risk that Chad could spend a small, but still meaningful amount of cash acquiring new distressed assets.  Call it ~$2M.  Chad just loves finding extreme bargains.  It’s his raison d’etre.  That said, I don’t think he’s reckless or stupid, and as mentioned he owns shares worth $11.5M.  Other 2016 holders and I have told him loud-and-clear that Fortress ought to deal with upcoming maturities before spending money for anything but debt repurchases, debt redemptions, or cost improvements at Thurso.

Odd & ends:

You won’t find the CCF Index price for DP on Bloomberg or anywhere on the web (that isn’t behind a paywall).  You could get it for a subscription fee from the CCF Group (ccfgroup.com), or just call Fortress or any other public company that makes DP (e.g. Tembec, Rayonier Advanced Materials, et al).

There are many research outfits that track supply & demand for DP. I’ve been told that Brian McClay & Associates are the best (pulpmarket.ca).

That sharp spike in Fortress shares in May was due to an unfounded rumor that Greece was indeed leaving the Euro and that the Landqart mill had a contract to make paper for the new drachmas.  The reason it hasn’t returned to prior levels is probably due to the improved fundamentals at Thurso.

You can buy 2016s through any of the major I-Banks in Canada, or by calling GMP Securities in NYC.  You can also get them through Charles Schwab by calling Mike Magee at 800-803-9232 ext 33620  (note that Schwab’s execution sucks … not Mike’s fault!).

Catalyst

Bonds repaid on-time and in cash at 12/31/16.  How’s that for a hard catalyst?