U S SILICA HOLDINGS INC SLCA S
October 13, 2016 - 2:25am EST by
roc924
2016 2017
Price: 48.39 EPS -0.79 0
Shares Out. (in M): 65 P/E nm 0
Market Cap (in $M): 1,552 P/FCF nm 0
Net Debt (in $M): 100 EBIT -50 0
TEV ($): 1,652 TEV/EBIT nm 0
Borrow Cost: General Collateral

Sign up for free guest access to view investment idea with a 45 days delay.

Description

U.S. Silica (SLCA- $48.39) has been written up twice before, long and short, as has SLCA’s competitor HCLP. This is an idea to short SLCA. SLCA is a frac sand and industrial sand mining company. The frac sand piece is what’s driving the stock so I’m not going to write about the industrial specialty products division, which accounts for only 15% of my 2018E operating income.

 

SLCA has now retraced 61% of the 80% crash that occurred in 2014/2015 and is up 250% off its low just eight months ago, putting the stock far ahead of fundamentals and priced for perfection at 7x tangible book, 6x EV/sales, and 50x the average 4 year trailing EBIT (which includes the last up-cycle). This is a cyclical commodity business, that hasn’t generated free cash flow in the last five years, yet it is trading at a growth stock valuation.

 

The prevailing view and bull case today is that SLCA has significant cost and contribution margin leverage and when demand rebounds, SLCA will be raking in enough dough to more than compensate for today’s valuation. There is also significant hype around increasing sand/well, but I think this could be disappointing too. In addition, there is a view that capacity won’t come on line fast enough to meet demand. This is not dissimilar to the story that drove frac sand stocks up in 2013/14.

 

While demand will certainly increase, I think the bull case is almost certainly wrong about cost leverage and contribution margin leverage. I also think sand/well could disappoint high expectations and capacity will come on faster than expected, meaning the ramp in selling prices may not be as steep as expected. There are also trends in the industry toward cheaper brown sand and in-basin sand, which if they continue, could hurt SLCA significantly as 80% of its frac sand business does not fit in these categories. As a result of all these factors, I think earnings will fall short of the lofty expectations reflected in the stock.

 

Insiders appear to agree with me: there was a large cluster of insider selling totalling $6m since August at prices ranging from $38-46. While insider selling for many companies is not a reliable signal, this cluster of selling is very unusual for this company. Also, there were no purchases this year, despite the stock hitting $14. Management did a significant (10m shares, 20% of outstanding), unforced secondary earlier this year around $18/share, signalling they thought that $18 wasn’t a low valuation.

 

There’s no doubt in my mind this stock is significantly overvalued, caught up in momentum and hype about the next cycle, which hasn’t even started. The question is how overvalued and I don’t have a precise answer, but I think the stock has very roughly 50% downside to $24. At $24 this stock would trade at 25x the average of the last four years of EBIT (this includes the upcycle years of 2013/14), 9x the highest EBIT the company has ever achieved (2014), and 7x (almost a turn higher if you capitalize $400m in operating leases) very optimistic (I would say unrealistic) EBIT of $248m in 2018. A $24 stock is 9.6x EPS in 2018 of $2.50, again using optimistic assumptions.

 

Catalysts: New supply from expansions and restarts happening already even in depressed volume and price market. SLCA does not realize cost leverage and earnings that investors expect.

There’s No Cost and Contribution Margin Leverage

 

If there is leverage on operating costs, then proppant COGS should not have a linear relationship with proppant tons. SLCA has had no leverage as you can see below. The scatter plot below shows this clearly. If there was leverage, there would be a logarithmic shaped curve instead, with incremental COGS decreasing as tons increased.

 

 

SLCA breaks out its COGS in its 10-K:

 

COGS ($millions)

2015

Transportation

258

Operating labor

80

Electricity and drying fuel (mostly nat gas)

28

Maintenance and repair

38

Other

91

Total

495



We can see that most or all of its COGS are variable. Interestingly (shockingly?) the company noted in its March 2016 presentation that it takes 100 rail cars and 430 trucks to complete one well, so it’s no surprise transportation is over half of COGS. Depreciation, depletion and amortization is not included above and is obviously a leveragable fixed cost, but the company doesn’t break out proppant vs ISP DD&A. So we can regress total COGS + DD&A vs proppant tons and ISP tons. We get the same results with COGS/ton of 71 (slope) and R square is 94%, with only the slope of proppant tons statistically significant. I’m sure ISP tons matter, but swings in proppant tons swamp ISP. .



 

Coefficients

Standard Error

t Stat

P-value

Intercept

-8297

21383

-0.388

0.70345

Proppant tons

71.0873

4.61629

15.3992

1.3E-10

ISP tons

-14.094

19.0783

-0.7387

0.47147



If there was leverage in proppant contribution margin, i.e. higher proppant contribution margin / ton as tons sold increased, we would get an upward slope in the following graph. We don’t, it’s negative, meaning negative leverage. The slope isn’t statistically different from zero, so we can use the average as our estimate (it’s $30/ton).

 

Here I exclude the big downturn quarters of 2Q15 through 2Q16. Again a negative slope, but only significant at the 10% level, so assume its zero. Average contribution margin for these quarters was $38, which I use in my model for 2018.

 

 

One more chart if you aren’t convinced.



What’s going on here? Obviously from the above, most of the costs are variable. I believe a big factor is that when demand is high and the supply chain is tight, the railroads are capturing the bulk of the value. There’s no leverage in SG&A: for example 4Q14, the best volume quarter, saw SG&A spike as a % of sales, a big part of which was bonuses and bad debt expense.

 

 

Industry

 

I don’t do a detailed description of the industry and business, but instead focus on the areas I think are important for the stock now. The Smart Sand S-1, company filings (FMSA, HCLP, EMES, SLCA, etc.) and industry reports/journals are good sources.

 

Capacity and Demand

 

There is significant frac sand capacity coming on-line again. In Wisconsin alone, companies have plans to add an additional 10m tons/year of capacity, equivalent to 20% of 2017E industry proppant demand of 45m tons and the industry is only running at 35%ish utilization now.

 

http://lacrossetribune.com/jacksoncochronicle/news/local/frac-sand-mines-invest-for-efficiency-new-markets/article_814e5490-07c0-5362-b6e3-17feaf35889f.html

 

2Q16 FMSA conference call: We can further ramp up our overall capacity with limited time and investment in our other currently idled facility.

 

Credit Suisse estimates that there is 110m tons per year of frac sand capacity today with 75m “ready to work.” I have a hard time reconciling industry utilization of 30-40% (demand of 33m tons this year / 75-110m of capacity) and demand of 63m tons with CS’s bullish outlook.

 

 

Annualized 1H16 sales and capacity for top public (or soon to be) companies. If we included Unimin, this group would represent about half of industry capacity.

 

 

1H16 sales anlzd

Capacity today

Notes

SLCA

11.0

15.0

estimated capacity

EMES

1.7

6.9

 

HCLP

  3.6

  10.4

2.9 idled, not included in 10.4

Smart Sand

0.4

3.3

can add 1m tons in 6-9 mos; ultimately expandable to 9m

FMSA

5.6

 16.8

 

Total

  22.3

    52.4

 





Frac Sand Demand (millions of tons)

 

2013

2014

2015

2016

2017

2018

Spears

55

73

53

30

45

59

CSFB

     

33

49

63

Avg

55

73

53

32

47

61

growth

 

33%

-27%

-40%

49%

29%



 



Proppant per well is going to increase, but there is a lot of hype about this leading to skyrocketing demand. Some counterpoints:

 

“At some point, increased sand no longer translates into higher production, however. Murphy, along with Whiting and other companies with major fracing operations like EOG Resources (ticker: EOG), has been trying to find where that fine line in the proverbial sand is drawn. Adding too much sand into fracs runs the risk of obstructing the flow of oil. “You get to the point where you’re jamming the fractures and your conductivity goes down,” said Jonathan Garrett, a Houston-based analyst for Wood Mackenzie.”

IHS Markit’s associate director for Plays and Basins, Reed Olmstead said at the September 2016 meeting of the National Oil-Equipment Manufacturers and Delegates Society in Houston:


“In a move pioneered by EOG during 2014, proppant intensities (measured by lbs of sand per lateral ft) have been increasing across most major oil plays, said the analyst. “Through much of 2015 and early 2016, operators were pushing the limits of proppant loading and seeing a corresponding increase in productivity. However, the most recent data have given indications that this may be plateauing.”

 

Brown Sand and In-basin Sand

 

There is a trend in the industry (for basins in and near Texas, e.g. Eagle Ford and Permian basins) toward toward lower cost brown sand / in-basin supply and away from more expensive northern white sand from out-of-basin locations like Wisconsin and Illinois. This is important because only about 20% of SLCA’s production and proven reserves are in-basin or lower cost brown (aka “Brady”) sand (see table at end of report).

 

“The move towards brown sand in oil and gas extraction marks the culmination of a trend which has seen E&P firms opt for increasingly cheaper proppants to enhance efficiency. This has resulted in a shift from ceramic proppants to resin coated, then to white sand and now to brown.

As in the case of each shift before it, the key driver behind the higher demand for brown sand is cost savings, with brown selling at cheaper price levels than white.”

 

SLCA’s Wisconsin and Illinois mines are between 700 and 1,300 away from the point of use. Over half of SLCA’s COGS are transportation related (rail, truck, transload, etc and in the range of $25-30/ton). For out of basin mines, the industry spends $30-60 per ton transporting sand from out-of-basin locations like Wisconsin to the oil and gas basin (i.e. in-basin mines have a $30-60 cost advantage). An ongoing shift to in-basin sand could put SLCA at a severe cost disadvantage. I think it’s early to say how this plays out, but this trend was discussed at a September 2106 industry conference.

 

http://www.indmin.com/Article/3585733/Frac-Sand-16-Shift-to-brown-sand.html

http://www.indmin.com/Article/3585277/Frac-Sand-16-In-basin-supply-key-to-company-survival.html

 

Management Compensation and Targets

 

Management has missed significant targets

 

In its September 2014 investor presentation, management set an EBITDA target of $550-600m by 2017. They have since pulled the target from their presentations. I think the company might hit only half of that in 2017 and 2016 EBITDA is going to be about $10m.

 

SLCA 2013 investor presentation shows industry frac sand volumes doubling by 2016, instead volumes are going to be about flat. The same presentation showed sand/well going up 300% from the 2013 level of 2,000 tons/well.  Instead, sand/well has increased an estimated 32% from 2013 to 2016 according to the 2Q16 Spears and Associates report. Eagle Ford, Appalachia, Permain, and Baken sand per well increased an estimated average of 74% from 2013 to 2016 according to the 2016 EMES investor presentation.

 

Management compensation increased 162% from 2013 to 2015 (source: Morningstar), while operating income plummeted 76% during the same period.



No Free Cash Flow

 

USD in millions

2011

2012

2013

2014

2015

1H16

Total

CFO

      43

    101

          46

    171

      61

        1

    423

PP&E

     (67)

   (106)

         (60)

     (93)

     (54)

     (23)

   (403)

FCF

     (24)

       (5)

         (14)

      78

        7

     (22)

      20

Acquisitions, net

      -  

      -  

          -  

     (98)

      -  

      -  

        (98)

FCF w/ acquisitions

     (24)

       (5)

         (14)

     (20)

        7

     (22)

     (78)



Miscellaneous

 

Wisconsin industrial sand mines and facilities as of May, 2016

Total Number of Industrial Sand Facilities (Mines, Processing & Rail Loading)

128

Number of Active Facilities

92

Number of Inactive Facilities

32

Number of Facilities Reclaimed/In Process of Final Reclamation

4

http://dnr.wi.gov/topic/Mines/ISMMap.html

 

 

In 4Q15, the company spent $36 million repurchasing shares at $19.67 (average) that employees received through the exercise of stock options.

 

Sales to our two largest customers, Schlumberger N.V. and Halliburton Company, accounted for 13% and 12% of our total revenues, respectively, during the year ended December 31, 2015.

 

Mine/Plant Location

Proven reserves (000s tons)

Probable reserves (000s tons)

P&P reserves

2015 tons mined (000s)

Primary end markets

Ottawa, IL

63,394

40,468

103,862

4,171

Oil and gas proppants, glass, chemicals, foundry

Voca, TX

34,467

41,900

76,367

2,033

Oil and gas proppants

Mill Creek, OK

 

15,009

15,009

1,408

Oil and gas proppants, glass, foundry, building products

Sparta, WI

27,179

2,740

29,919

1,134

Oil and gas proppants

Utica, IL

10,895

0

10,895

842

Oil and gas proppants

Mapleton, PA

3,954

5,000

8,954

766

Glass, building products

Pacific, MO

17,377

7,994

25,371

617

Oil gas proppants, glass, foundry, fillers and extenders

Kosse

10,898

10,898

519

Glass, recreational products

Berkeley

1,431

1,431

473

Glass, building products, fillers and extenders

Columbia

5,581

5,581

418

Glass, building products, fillers and extenders

Dubberly

3,918

3,918

256

Glass, foundry, building products

Montpelier(1)

13,481

13,481

227

Glass, building products

Hurtsboro

736

736

154

Foundry, building products

Jackson

97

725

822

144

Fiberglass, building products

Mauricetown

12,211

12,211

138

Filtration, foundry, building products

Rockwood (2)

6,563

6,563

 

Glass, building products

Fairchild

38,975

38,975

 

Batesville

34,732

34,732

 

Total

237,676

162,049

399,725

13,300

 
         

Notes

Northern white and other

101,468

43,208

144,676

6,147

Estimate

Brown and other

34,467

41,900

76,367

2,033

Estimate

Other

17,377

23,003

40,380

2,025

Estimate

Proppant and other

153,312

108,111

261,423

10,205

Estimate

Brown and other % of proppant and other / i.e. in-basin

22%

39%

29%

20%

Estimate