Arriscraft International write-up

If a micro-cap Canadian income trust suspended its distribution, would anyone want to own it? And if it then reinstated its distribution a year later, would anyone notice? Those are the kinds of questions surrounding Arriscraft International Income Fund, a Canadian company engaged in the sexy growth business of manufacturing fake rock. Trading at C$4.50 per unit today, for a valuation of C$36MM, the company is yielding over 9% on current distributions, distributions that are amply covered by current cash flows. The trailing twelve month free cash flow yield is 20%. More importantly, margins hit historical lows last year and have only begun trending up; assuming traditional margins and taking into account an upcoming growth initiative that will receive a go/no go decision by the end of the year, there is reason to believe that distributable cash flow will double and cash distributions will triple within the next 36 months, hoisting the stock’s price up close to its intrinsic value of greater than $10. Combined with the distribution income, this would produce a triple digit gain and very attractive IRR for equity holders purchasing at today’s prices.

 

**all dollar amounts are CAD unless otherwise noted**

 

 

Arriscraft was founded in Ontario in 1956 to commercialize a process for manufacturing synthetic stone to use as masonry, wall cladding, and other building material applications. Today, the company operates manufacturing facilities at three locations in North America. At the oldest plant, in Cambridge, Ontario, the company operates four press units, making it the company’s largest and most efficient plant. In Saint-Etienne-des-Gres, Quebec and in Fort Valley, Georgia, the company operates one press unit at each location. Both of the Canadian facilities supply customers on either side of the border; Georgia, clearly, is targeted exclusively at the US market. The Georgia facility was completed only in 2004, and spent much of 2005 working through process improvements that limited its productivity. Those startup hiccups are now in the past, as management indicates that the plant is operating near the levels of productivity of its Canadian operations. What’s more, there are plans to expand Georgia by one or more press units, a topic which we’ll cover later in this write-up.

 

The company’s products, which are unique, occupy a niche in the building material market. Wall cladding and masonry can be classified as either: (i) “full bed” masonry products, which are typically 3” to 4” thick and are laid by a brick layer or stone mason in a mortar bed, and (ii) “thin set” or “adhered” masonry products, which are thin brick or stone units (or tiles) that are adhered to a supporting wall. Architects must specify which technique to use in the design phase. Full bed masonry construction is generally thought of as higher quality construction that thin set masonry construction, just as hardwood flooring is considered superior to laminate flooring.

 

Arriscraft produces exclusively full bed manufactured masonry. Cheaper than natural stone but more expensive than thin set “painted” cement, manufactured stone offers the aesthetic appeal, durability, and identical installation methods of natural stone, at a cost that is typically one third to one half that of the natural material. Arriscraft products are also available in infinite color and pattern varieties, regardless of geographic location – an attribute that natural stone cannot match, due to transportation costs and the locations of quarries. The following tables summarize the company’s product positioning:

 

 

 

 

 

At first glance, it’s easy to dismiss the company’s competitive positioning. However, it’s worthwhile to note that its products, which are chemically identical to natural stone (they are made from the same sources: coarse and fine silica sands, lime, and a dollop of tender loving care in the form of heat and compression) cost just one third to one half the cost of the natural stone products they are designed to mimic; the variation in the discounts has to do with the costs for different types of stone. Arriscraft management tells me that they often win designs in new projects after their customers have incurred cost overruns. When the value engineers come in to try to take out costs, one of the easiest ways for them to do so – without compromising the structural or aesthetic integrity of the building project – is to simply substitute inexpensive Arriscraft Natural Edge cladding for the expensive real stone product.

 

The company has no shortage of competitors. Builder Magazine, which concerns itself with the affairs of the homebuilding industry, reports that its readers regard Cultured Stone (a unit of Owens Corning), Eldorado Stone, and Coronado Stone as the premier brands in the United States. As indicated above, each of these companies offers a product inferior to Arriscraft, because each offers a cement-based, low density, thin set product, which is nothing more than a texture painted on a thin slab of cement. If by chance, a stray bullet or a speeding eight-year-old were to nick the stone, the cement-based products would reveal grey cement underneath, but the Arriscraft products would look the same, because their texture and color are properties of the material itself, not a thin coat of paint.

 

While acknowledging the existence of competitors like Cultured Stone and Eldorado, the company maintains that they are the only supplier of full bed, chemically identical manufactured stone, a claim I have been unable to disprove. They are also the largest supplier of full bed manufactured masonry, of any type, in North America. The closest competitor may be Century Stone, a small private Canadian company that offers full bed cement-based stone substitutes; there are bound to be similar small private companies in the US that I haven’t turned up yet. Nonetheless, Arriscraft’s product is unpatented and protected only by trade secret and know-how, a protection that is self-evidently meaningful, as the company has now sold its product for 50 years without imitation.

 

 

Arriscraft didn’t come public with the pre-IPO clumsiness of Google or the post-IPO existential absurdity of Vonage, but it came pretty close. The company IPO’d in December 2004, and promptly set upon a distribution schedule of C$0.30 per quarter. As for timing, the housing market in Eastern Canada began rolling over (Ontario housing starts down 8% and Quebec down 13%) and the Canadian loonie began to take off at around the same time, hurting Arriscraft’s earnings. These problems were simultaneous with production problems at the Georgia plant during its first year of operations. Hindsight renders the effects of these factors obvious – here’s a table of distributable cash flow per unit and actual distributions in 2005.

 

 

It does not take a VIC member to spot the problem here, especially for a company with approximately $0.70 per share of cash on December 31, 2004. Accordingly, management suspended the distributions in the summer of 2005, causing the stock to plummet from $10 or so to $4 or so. At the same time, the company replaced a credit facility with outside lenders with a new one from the previous owners of the company. The new credit facility is a $16MM subordinated unsecured debenture that carries interest at 4% for the first two years (through July 2007) and then at prime + 2% thereafter. The debenture is convertible into Arriscraft units at $9.50 per share, though it is redeemable before conversion at the company’s option.

 

In the fall of 2005 – as the stock hovered in the $3.50 to $4.00 range – a gentleman by the name of Saul Koschitzky used his company, IKO Enterprises, to acquire 16.7% of the outstanding units of Arriscraft. IKO is a leading roofing supply company based in Canada; evidently, Koschitzky saw an obvious bargain in an industry he knew well, and decided to buy the stock. Fearing a takeover below intrinsic value, the board approved a unitholder rights plan to prevent Koschitzky from acquiring any more stock. There have been no filings since December 13, 2005 to indicate that Koschitzky has bought or sold any stock since then. Incidentally, the Koschitzky family appears on the Canadian equivalent to the Forbes 400 list.

 

 

In December 2005, the board of Arriscraft recruited David Boles to be the company’s new CEO. Boles was formerly COO of the North American division of Qeubecor World, a large commercial printer. Boles’s division had revenue of more than $5 billion, which makes him more than qualified to run an $80 million company like Arriscraft. With only six months on the job, his handiwork is already impressive:

 

 

Factors affecting the year over year comparison for the 1^st half of 2006:

 

 

In the MD&A, the company provides a table to illustrate the calculation of distributable cash from net income. Here’s a reproduction of the table:

 

		Six months ended
		June 30 2005	30-Jun-06
Net income/loss		-2,459	482
Add:
Interest and accretion		984	1,015
Income taxes		0	43
Inventory FMV increase		964	0
Deprec & Amort		3,988	3,762
EBITDA		3,477	5,302
Add:
Mgmt bonus		488	0
Non-controlling interest		-271	53
Unrealized forex (gains) losses		98	-852
Adjusted EBITDA		3,792	4,503
Deduct
Mgmt bonus		-488	0
Maintenance capex		-897	-433
Income taxes		0	-43
Interest		-984	-821
Distributable cash		1,423	3,206

 

A similar table is provided to reconcile the cash flow statement to distributable cash, but I won’t reproduce it here to save on space. See the MD&A for details if you’re interested.

 

 

In August, the company announced that they would be resuming distributions at the rate of $0.42 per year. With recently commenced payout of $0.42 per year, the units are yielding 9.3% today, a payout that is well covered, considering that distributable cash per unit totaled $0.37 in the first half of the year and $0.94 in the trailing twelve months. As for cash flows and margins, the following tables should provide some color on the opportunity in front of Arriscraft:

 

 

	1H2006	1H2005	TTM	2005	2004	2003	2002	2001	2000
Sales	38	33	74.8	69.8	77.3	70.5	71.6	66.9	62.7
Gross profit	18.6	14.4	36.5	32.3	35.6
EBITDA	5.3	3.5	12.2	10.4	0.9
Adjusted EBITDA	4.5	3.8	9.5	8.8	10.9	13.6	20.2	17.1	17.4
Net income	-2.4	0.4	-9.9	-7.1	0.9	-0.9	7.6	8.1	8.4
Product volume	147	137	302	292	301	281	269	266	266
thousands of tons per year
Adjusted EBITDA margin	11.84%	11.52%	12.70%	12.61%	14.10%	19.29%	28.21%	25.56%	27.75%

 

At the annual meeting in May 2006, CEO Boles indicated that the company foresaw an operational break point at a revenue run rate of $75 - $80 million per year. According to Boles, “sales above this level will have a meaningful impact on distributable cash flow.” Based on 2006 YTD results, the company is already nearing this crucial level of annualized sales, and the preliminary results are encouraging. Gross and adjusted EBITDA margins are up, though the margin improvement has been partially masked by the investments in the US residential sales force. There should be further leverage as these new costs are absorbed by further sales gains.

 

It should be noted that the company builds inventory during the first half of the year, and then experiences their most robust selling season during the third quarter. This is driven by the need of Canadian builders to complete and seal the interiors of their construction projects before the onset of winter. Seasonality has been reduced in recent years as the company has increased sales in the United States. Raw material costs are primarily natural gas, electricity, silica sands, and lime. Each item represents less than 10% of COGS, allowing for some protection if any one were to spike in price.

 

As mentioned above, the company has a $16MM subordinated exchangeable debenture as its main credit facility, of which the entire amount is outstanding (though it is carried at only $14.3MM on the balance sheet – it is being accreted upwards to the full value of $16MM over its term). It carries interest at 4%, though it will reset to prime + 2% in the middle of 2007. Additionally, there is $5.9 in industrial development revenue bonds associated with the Georgia facility, which can be treated like a lease. In total, interest expense is running at $1.6MM per year, giving the company approximately 6 X interest coverage, using 2005 EBITDA.

 

The subordinated debenture is convertible, at the option of the holder at any time, into regular units of the Fund at a price of $9.50 per unit. Additionally, the units are redeemable by the fund at any time at 101% of principal value and unpaid interest.

 

The prior owner of Arriscraft was issued, in conjunction with the IPO, 948,833 exchangeable LP units. These units are eligible for distributions, provided that the annual distributions to the common units exceed $1.20 per year, almost three times the current distributions. They are convertible on a one for one basis into common units starting in December 2006.

 

There are 1.62 million subordinated management units that are convertible into regular units of the Fund in December 2006, if distributions at that time are in excess of $1.20 per year, which they are not currently. There a number of other terms, which you can find in the annual report, but the critical variable is the $1.20 per annum in cash distributions. Until that target is achieved, the subordination features of the different securities ensure that virtually all of the cash flow growth will benefit holders of the common units.

 

 

In the second quarter MD&A, and in the company’s annual meeting (a recording is available on the company’s web site), Boles spoke about an upcoming decision: whether to add another production press to the Georgia facility. They are contemplating this move because of a peculiar breakdown in the business mix, as illustrated in this table:

 

Sales in C$ (millions)	1H 2005	1H 2006
Canada
Residential	8.3	9.4
Non-residential	2.3	1.9
Other	7.9	9.1
United States
Residential	1.6	2.4
Non-residential	11.5	14.1
Other	1.3	1.9

 

Non-residential products sales are primarily to the commercial, office, and landmark (stadiums, courthouses, hedge fund offices) building sectors. “Other” includes (i) accessories assembled from manufactured masonry, like windowsills, doorways, quoins, and other products, (ii) brick manufactured on the company’s stone production presses and (iii) natural limestone products from one quarry the company operates.

 

Until the opening of the Georgia facility, the company had supplied the US market from its plants in Quebec and Ontario. As these facilities are located in areas with more commercial than residential construction, and as the company found itself selling out of its production capacity most years, they simply did not focus on the US residential market to any large degree. The opening of the Georgia plant, in the middle of a high-growth residential market, allowed the company to target residential sales in the US for the first time, and the new CEO has built up the sales force expressly to attack this new market. What’s more, the company is considering adding one or two presses at the Georgia facilities to greatly expand its presence in the US residential market. The company will announce their decision sometime in the next month or so, as they had been waiting to see third quarter business levels before making their decision.

 

The company indicates that the PP&E carried on the balance sheet is fairly indicative of the costs to construct new manufacturing facility. Note 3 of the 2^nd quarter MD&A indicates that the cost of the company’s buildings and machinery and plant equipment totals $33.2 million. Divide by the six production presses in use today, and we can estimate that the company would incur costs of $5.5 million to build each new production press. This estimate is corroborated by the fact that the fourth production press at the Ontario plant, completed in 2004, cost $5.2 million. They are contemplating whether to add zero, one, or two presses in Georgia, according to the recording of the annual meeting.

 

If the company were to add two production presses, it would cost them $11 million. Last year, the company generated $7.37 million in operating cash flow and spent $1.462 million in capital expenditures, all of which was classified as maintenance capex. This year, operating cash flow could be up anywhere from 50% to 100%, and maintenance capex will be lower, thanks to some of the cost-cutting initiatives put in place by Boles. Less than $2 million will be spent on distributions. Thus, with the free cash flow that will be generated in 2006 and 2007, the company will have more than enough cash to self-fund the 33% increase in production capacity represented by two additional production presses, without putting the distribution at risk. In an optimistic scenario, they could generate sufficient cash flow for this expansion in a single year.

 

The return on capital for an expansion should be excellent. Once the two new production presses achieve operating performance comparable to that of the older lines, they should add $25MM in annual revenues ($75MM TTM revenues, divided by six current production presses, gets you $12.5MM in annual revenues per press). If the company achieves adjusted EBITDA margins of 13% on those sales, the $3.25MM of adjusted EBITDA would be a return on capital of 30% ($3.25 divided by $11 of construction costs). If the company achieves its more normal 20% adjusted EBITDA margins on the new lines, the adjusted EBITDA return on capital approaches 45%. Of course, adjusted EBITDA may not the best number to use when calculating return on investment, but whatever earnings figure you choose to use in your analysis, expansion looks like a very sound use of capital.

 

One final note about the US housing market: Janis Joplin once said that freedom’s just another word for nothing left to lose. This perfectly describes the situation Arriscraft finds itself in with regard to sales in the US residential sector. With just $5 million in annualized sales into this market, they have very little to lose from a sharp downturn in housing. In fact, it makes wonderful sense for them to be expanding aggressively right now, while other building materials companies are likely to be suffering. If they make a “go” decision within the next few months, production in the new Georgia facilities would commence by the middle of 2008. Given that we are about a year into the housing bust, even a bad downturn should be over by then. Arriscraft will be well-positioned as the market bottoms, and by the time their new facilities are operating at standard efficiencies in 2009, US homebuilders should be in a strong position to appreciate the company’s high quality, low cost product.

 

 

Arriscraft presently sports a safe, well-covered 9% yield. In fact, they are paying out less than the distributable cash generated in 2005, a terrible year. On trailing twelve month distributable cash, the payout ratio is less than 50%. If they choose to expand, the distribution level is unlikely to change for some time, but will eventually rise by over 100%, perhaps even more than 200%. This would be driven by an increase in earnings and an increase in the payout ratio, up from less than 50% today to closer to 100%. Assuming that the company is priced to yield 500 basis points more than the US Treasury (which would imply 9.6% today), a $1.00 annual distribution would lead to a 2009 price target of $10.40. A $1.50 annual distribution would imply a target price of $15.62. This distribution target ($1.50) is not outlandish considering the current distributable cash flow run rate, the potential for margin expansion, and the potential for increased sales from capacity expansion.

 

If the company chooses not to expand, they will almost certainly raise the cash distributions up to close to 100% of distributable cash flow. Given their history in 2005, they should do this slowly, but even in this scenario, the distribution should double within two to three years.

 

With a 9% current cash yield, no matter what choice the company makes, current investors will be paid handsomely to wait for either scenario to play out.

 

Catalysts: