Industrial Gas Firms Pull Economic Moats Out of Thin Air
Two players look undervalued at today's prices.
Industrial gas producers pull economic moats literally out of thin air by separating this abundant resource into its specific components: oxygen, nitrogen, and other specialty gases. The firms, via their on-site distribution model, receive handsome remuneration because of the substantial explicit and implicit costs of a plant shutdown. Establishing an on-site plant creates a regional monopoly, and firms maximize their returns on invested capital by selling excess product via bulk and cylinder methods.
Not all players are alike, however, so our economic moat ratings differ. Praxair (PX) earns a wide moat for its best-in-breed portfolio and industry-leading returns. Air Products (APD) earns a narrow moat because of its on-site exposure, offset by its outsize nongas exposure and large concentrations in electronics and hydrogen (energy) that weaken its returns on invested capital. Airgas (ARG) specializes in the least desirable distribution method (cylinders), but earns a narrow moat because it generates enough gas internally to deliver positive returns on invested capital.
Industry's Moat Emanates From Substantial Customer Opportunity Costs
The industrial gas industry sells atmospheric and process gases via three distribution methods: on-site, bulk, and cylinder. For atmospheric gases, the industry separates air into its main properties: oxygen, nitrogen, and argon. For process gases (carbon dioxide, hydrogen, helium, and other specialty gases) industrial gas companies purchase the raw byproduct typically from a refiner or a chemical company and process it into its refined state.
Broad end market and geographic reach combined with multiple customers enhances the industry's bargaining power. Multiple end markets depend on industrial gases for various applications. For instance, the steel industry uses oxygen in blast furnaces, and the food and beverage industry relies on carbon dioxide for fountain drinks and cooling and freezing applications. Further, no one customer is material to a firm's bottom line. Airgas' largest customer accounts for 0.5% of annual sales, Air Products' largest customer accounts for at most 10%, and Praxair, which doesn't provide a specific percentage, says it's not dependent upon a single customer or a few customers. As a result, we believe the industrial gas firms hold the bargaining power.
The on-site distribution method exists because of the substantial explicit and implicit costs of an operational stoppage or shutdown. Refiners require hydrogen to purify crude oil of organic sulfur to meet environmental requirements, making hydrogen a critical element. Refiners must run at greater than 50% capacity to maximize operating efficiency or they are forced to shut down production. Shutdowns are costly because output becomes off-specification (unsellable) and is diverted to tankage. A startup usually takes 12-24 hours before a refining complex can reprocess all of the off-specification products to specification, compounding the explicit costs.
The implicit costs are equally as great. A reasonably sized refiner accommodates 100,000-150,000 barrels of oil per day and typically operates with a utilization rate upward of 88%. Using approximate oil prices and crack spreads, we estimate that a typical refiner generates $1.4 million-$2.1 million in gross refining profits per day. This compares with typical daily operating costs of hydrogen of around $91,000, or roughly 1% of a refiner's typical cost deck.
This lopsided dependency favors the industrial gas companies and solidifies the industry's bargaining power, in our opinion. Although we focused our example on refining, similar dynamics hold for large customers in other end markets, such as steel manufacturing and chemicals.
Given the substantial costs of a shutdown, a customer commits to a contract with minimum take-or-pay levels and cost escalators. Large customers generally have good visibility into their long-term prospects. Although companies can shutter inefficient plants in an economic downturn or secular decline, these anchor customers produce fundamental products and are likely to survive technological substitutes for at least a decade. As a result, large industrial gas users have the confidence to enter a 10- to 15-year contract. To minimize transportation costs and ensure reliability of supply, an industrial gas company constructs an on-site plant adjacent to its customer's plant, creating massive customer switching costs. In return, industrial gas companies index their largest input costs (electricity and energy) to inflation, helping preserve operating profit dollars.
Since typically only one customer exists in each 250-mile region, the first company to secure the large customer essentially gains a regional monopoly. Onerous transportation costs prohibit gas deliveries beyond 250 miles, making the initial contract win extremely important. Further, with renewal rates for the on-site contracts hovering around 95%, winning a contract can create a customer for life. Aside from the large customer, there are numerous smaller customers in the surrounding area, so industrial gas companies build excess capacity for cross-selling opportunities. For cross-selling gases such as oxygen, the anchor tenant usually consumes about 70% of the annual output, enabling the industrial gas firm to sell the remaining product via two different, albeit less desirable, distribution models: cylinders (packaged) gases and bulk (merchant).
Vertical Integration of Distribution Methods Strengthens Returns
Competitive dynamics and cost structure differ across distribution methods. As a stand-alone business, the cylinder distribution method contains qualities of a no-moat enterprise. Cylinder contracts depend heavily on logistics, which add 10% to the cost profile. Another detractor is the competitive environment. Unlike the global gas industry, where the top four players represent 80% of the $74 billion global market, the $13 billion domestic cylinder gas business contains more than 900 companies with limited scale to offset the 40% markup required to purchase gas from the larger players. Combined with shorter contract durations, the cylinder business is less desirable. Bulk--in which liquid product is delivered via trucks and filled into a large storage tank--is a better mode than cylinder gases thanks to longer contracts and higher operating margins, but is still less desirable than on-site.
However, this stand-alone analysis misses the incremental revenue opportunities from owning all methods of the distribution channel. There is typically one large customer within a geographic region, but numerous small customers--such as welders, restaurants, and medical facilities--will require smaller quantities of product. Therefore, we believe creating sufficient spare capacity achieves the highest net present value for each company compared with stand-alone on-site plants. In fact, we believe this potential value is the reason that Air Products, after selling its packaged gas business to Airgas in 2002, attempted to acquire Airgas in 2010.
Praxair Has Carved Out a Wide Economic Moat
Praxair has best harnessed the industrial gas business model. Praxair derives 85%-90% of its annual revenue from selling industrial gases, compared with roughly 80% for Air Products. Praxair's remaining revenue is a mix of no-moat operations such as selling equipment and service applications. Praxair has less exposure to the electronics end market (which is more commodified than gases) and selling hydrogen gases (fewer cross-selling opportunities) than Air Products. As a result, Praxair's operating margins and returns on invested capital have outperformed its peers' over the past decade.
Air Products Earns a Narrow Moat
Like Praxair, Air Products benefits from the industrial gas business. However, its narrow moat stems from operating with a less desirable mix than the average industrial gas company (selling hydrogen to the energy industry and electronics) and selling more nongas products (like equipment).
Increased hydrogen demand in energy applications offers strong promise, but it carries lower operating margins and fewer incremental selling opportunities. Air Products holds the industry's leading energy position because of its large share in hydrogen. The firm commands roughly 40% market share (double its closest competitor) in the 6.1 billion square cubic feet per day industry. Air Products expects the confluence of heavier crude oil and increased transportation fuels will cause hydrogen demand to increase approximately 80% (6% compound annual growth rate) by 2022. This will help buoy long-term growth prospects for the company.
However, the cross-selling economics of a hydrogen plant aren't as desirable as a nonhydrogen on-site liquid plant. Cross-selling for hydrogen only works if there are other refiners in the same geographic region and if they all operate on the same pipeline, reducing the opportunity set compared with a traditional on-site liquid plant. As a result, a refiner receives roughly 80% of the output, compared with 70% in an oxygen plant. While this contract is still profitable, it doesn't secure Air Products' ability to cross-sell the remaining product, lowering the incremental net present value. Second, the use of natural gas increases the cost to produce hydrogen, resulting in lower reported operating margins compared with atmospheric gases.
Air Products' electronics exposure tilts more toward commodity products. The firm serves the electronics industry in two general fashions: on-site contracts for general plant operations and specific uses as part of manufacturing computer chips. For general plant operations, a chip manufacturer requires nitrogen as a purge gas to purify the operating chamber of imperfections. Computer chip manufacturers cannot operate without nitrogen, and these contracts provide Air Products with favorable economics. The on-site business, including bulk gases, represents about 43% its electronics business, or about 7% of overall sales.
However, we view the specific usage items, which include gases and enabling equipment, as commodified products with poorer economics. For starters, there are numerous competitors in this arena versus the global industrial gas industry. Second, these items lack the long-term contracts associated with traditional gas products.
We highlight nitrogen trifluoride as the poster child for this commodification, with pricing power declining substantially over the past decade. Including the enabling equipment, this exposure represents about 57% of Air Products' electronics business, or about 10% of overall sales, compared with 4% of Praxair's revenue.
Airgas Deserves Narrow Moat as Largest U.S. Industrial Gas Distributor
Airgas' transformation into a narrow-moat firm began with the 2006 acquisition of Linde's U.S. bulk gas assets. Airgas used a combination of organic growth and tuck-in acquisitions to become the largest player in the U.S. packaged gas (cylinder) industry. However, even with 20% share of the 2006 $10 billion packaged gas and welding hardgoods market--twice the share of its second-largest competitor and substantially larger than the nearly 1,000 independent suppliers--Airgas wasn't able to produce returns on invested capital typifying a narrow-moat enterprise. However, with the 2006 purchase of eight air separation units, Airgas tripled its internal gas generation to 30% of annual demand compared with 10% preacquisition.
We estimate increased vertical integration boosted Airgas' operating margins and returns on invested capital by 400 basis points and 250 basis points, respectively. To arrive at our savings estimate in the year of the acquisition, we drill down to Airgas' annual gas sales and remove the markup that Airgas pays its suppliers to deliver it gases. Since Airgas produces the gas for internal consumption, it will not record a profit on any intercompany sales in its financial statements.
Airgas generates about 65% of its annual sales from gas and rent, with hardgoods representing the balance. Of this gas and rent, the gas portion represents 80%, implying that gas revenue represents 52% of annual sales, or $1.7 billion in 2007. Before the air separation units acquisition, Airgas purchased 90% of its gases externally from suppliers like Air Products or Linde--we use a blended gross margin rate of 38% for the average markup Airgas paid its suppliers for industrial gases. Afterward, Airgas increased its internal gas capabilities, reducing its purchase requirements to 70% of its annual needs, and eliminated the gross product markup on the remaining 20%.
With its market-leading position and enhanced cost profile, Airgas should generate strong returns on invested capital over the coming decade. Today, Airgas holds a commanding lead over the 900 independent players. Number-two player Praxair has effectively ceded the title to Airgas, and with Air Products' 2002 exit of the packaged gas industry, we expect Airgas' dominance will remain unchallenged. The smaller geographic reach and lack of internal gas generation offset the administrative costs of a public company, and we estimate Airgas' size advantage compared with independent peers is worth approximately 10 percentage points of gross margin. As the current crop of independent operators retires, we expect Airgas will pursue further acquisitions to consolidate the industry.
Airgas Mostly Fairly Valued; Praxair and Air Products Offer Better Opportunity
Our fair value estimate for Airgas is $95 per share. We arrive at our revenue forecast by estimating the market share growth in the domestic packaged gases and welding hardgoods markets. Traditionally, industrial gas usage grows at a 1.5-2.0 times multiple of domestic GDP, which translates into 4%-5% market growth at the high end of our analysis. Outside of market growth, we forecast that Airgas can capture an additional 1 percentage point of share due to its broader product offering.
We think Airgas could achieve its 15%-16% long-term operating margin target if it increased its internal generation to 40% of total consumption versus 30% today. Another 10-percentage-point increase would boost operating margins roughly 200 basis points, enabling Airgas to achieve 15% operating margin based on current levels of around 13%. Airgas intends to add an additional air separation plant per year, suggesting that cost savings from increased internal gas generation is a realistic assumption.
We think Airgas is mostly fairly valued and require greater confidence in the firm's ability to sustain further operating margin expansion before we could justify a higher valuation.
Our $128 fair value estimate for Praxair provides investors with 12% upside from today's levels. We estimate average revenue growth of 6% per year, which is driven by slower U.S. and European growth (63% of sales) and higher emerging-market prospects. We project long-run operating margins of 21%.
Praxair generates 85%- 90% of annual sales from gases, with the remaining revenue from its surface technologies segment, materials science, and other service businesses. We think Praxair's lean operating practices and desirable product portfolio support operating margins above the typical project levels. Praxair ended 2012 with 22.3% operating margins, suggesting 21% midcycle operating margins are obtainable.
Our $110 fair value estimate for Air Products is about 22% above today's price. We estimate average revenue growth of 6% per year for the firm, driven by slower European growth offset by greater emerging-market opportunities. We project long-run operating margins of 17%.
The profitability on hydrogen is lower than with a traditional oxygen on-site plant. In part, the reason is the cosmetic nature of the natural gas pass-through. A traditional on-site plant that generates approximately $1 in revenue can incur as much as an additional $1.50 in natural gas pass-through costs. Although the contracts enable the supplier to generate the same operating profit dollars, the reported operating margin is artificially low: 12% compared with 30%.
Air Products has an outsize exposure to hydrogen, which accounts for about two thirds of its on-site (tonnage) business. Assuming this weighting persists, Air Products must generate 30% operating margins in its remaining tonnage business to achieve an 18% midcycle operating margin in its on-site business. This 30% operating margin falls within the range of a typical on-site contract, suggesting that our 18% long-term margin in the on-site distribution is realistic.
Air Products generates roughly 80% of annual sales from gases, with the remaining revenue from other sources like equipment and specialty materials. We believe our 6% long-term growth assumptions are reasonable, given the typical 1.5-2.0 times multiple of industrial gases usage to GDP growth. While the company has stumbled recently with operational setbacks, we don't think our fair value estimate implies heroic margin assumptions, providing investors with an opportunity to invest in a narrow-moat firm trading at an 18% discount to its fair value estimate.
Industrial Gas Industry Should Maintain Its Edge
Our economic moat ratings for the industrial gas players emanate from the substantial opportunity costs for large on-site customers. The companies then smartly exploit this advantage into incremental albeit less desirable stand-alone opportunities that bolster their return profiles.
We believe the industrial gas industry will retain its competitive advantage over time. Foremost, we think it's highly unlikely that the industry will migrate to a commodity pricing structure. Transportation costs far outweigh production costs, making it nearly impossible for outside supply to enter a local market--there is very little exporting of industrial gases outside country borders. Second, we don't think there is a high risk that new entrants will emerge, as the industrial consolidation during the past decade has fostered better capacity discipline in the developed markets.
Accordingly, we think the industrial gas industry is worth consideration by investors, especially given some of the compelling discounts to our fair value estimates at this time.
Basili Alukos does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.