Description

We had previously submitted a long recommendation for Air Products (APD), and we are now recommending Praxair.  Praxair is the quintessential Warren Buffett-type business with very high barriers-to-entry, low business obsolesce risk, good pricing power, and an A+ management team.

Praxair is the largest industrial gas company in North and South America.  We think that, overall, Praxair has the best geographic footprint of the four major industrial gas companies with good exposure to the US, Brazil, Canada, Mexico, China, India, and Thailand.  We are much less enthused about the economies of Japan and Western Europe, and fortunately Praxair is “underweight” these countries with 0% exposure to Japan and 12% exposure to Western Europe.

Major end markets for Praxair are manufacturing (24%), metals (17%), energy (13%), chemicals (10%), healthcare (8%), electronics (8%), food and beverage (8%), and “other” (9%).

The history of the industrial gas business and its current structure are important, so please allow a digression.  The industry began in 1895 with the invention of atmospheric air separation independently by a German engineer and a French engineer.  Simplifying the process which was invented, air is cooled to a liquid and then boiled-off.  The gases of nitrogen, oxygen, argon and rarer gases are separated due to their different boiling temperatures.  It is worth noting that three (Praxair, Linde, and Air Liquide) of the four global industrial gas companies began as the initial business ventures by these two engineers, Air Products being the sole exception less than 100 years old.  The industry has not been kind to second movers.

Through the world wars and into the 1960s, the volume of demand for industrial gases grew cyclically but consistently over time.  The initial demand was predominantly oxygen and relatively low technology, such as welding and production of steel.  The industry structures which emerged were near monopolies or strong oligopolies within national/regional areas.  In the 1970s, the industry became more globalized, as previous gentleman's agreements against entry broke down and the large firms globalized.  The four large global players have presences world-wide, but tend to have density in their traditional geographies.

About the same time, the usage of gases diversified and the technology requirements became more demanding.  Demand for nitrogen became more significant (food processing), as well as numerous rare gases for semiconductor production.  A related business of producing hydrogen from natural gas also developed, used for refining crude oil.  Finally, in the 1980s, the outsourcing of large tonnage industrial gas production for steel production, oil refining, and chemical production took hold, so that the world-wide chemical and refining concerns largely exited the business of producing gases for their own use.

The business is unusual, with the production of multiple gases, delivered by multiple modalities (almost always local/regional), primarily by four key firms (the big 4 have 69% of world-wide market share).  Depending on the modality by which the gas is delivered, the business terms with customers and dynamics change. But the salient dynamic is that the firm which has local density and delivers many gases through many modalities is advantaged.  

A hypothetical example is illustrative.  A steel producer decides to build a greenfield mill, and contracts with Praxair to provide oxygen.  Praxair builds an air separation unit adjacent to the mill and pipes the oxygen gas "over the fence," in exchange for a 10-year take-or-pay contract with cost pass-throughs for key inputs, mainly electricity.  The gas is sold for 5¢ for 100  cubic feet and the plant has a 12% cash-on-cash return after seven years.  Given the terms of the contract, the primary risk Praxair takes is credit risk of the customer.  If the customer is solvent or the plant operates, Praxair should have a reasonable rate of return on its investment (as an example of the leverage with customers during stress, see how Wilbur Ross and Mittal fared when they tried to "cram down" Praxair during ISG's restructuring - punch line is it didn't work).  

But in this example, Praxair looks at the 200-300 miles around its new facility where it can economically deliver oxygen and other gases and decides there are a lot of other potential customers.  So it overbuilds the capacity somewhat and adds a liquefier.  Praxair enters into 5-year contracts with regional manufacturers, hospitals, food and beverage companies, and other customers to provide them their industrial gas.  It delivers liquefied gas by truck to its own storage vessels, built at the customer site. It also attempts to match the local demand to the different gases being produced at its facility, as gas which is not sold is vented back into the atmosphere. This balance of selling multiple gases produced simultaneously is functionally a form of capacity utilization, and takes time to achieve.  Oxygen gas in this "merchant" example sells for 30¢ for a 100 cubic feet delivered and generate 15%+ all-in cash-on-cash returns for the overall project.  

Finally, Praxair sells gas and other "hard goods" to small users who consume the gas using cylinders, such as welders and small manufacturers.  In doing so, it rents the cylinder and provides delivery.  It also sets up "store fronts" to sell the related items.  Oxygen gas sold in this example sells for 50¢ for 100 cubic feet, 10x the price of the pipeline gas, as the significant cost is the cylinder and the delivery, not the gas itself.  The combination of pipeline, merchant and cylinder business sourced from a single facility increases the all-in returns of the project to near 20%.

While this hypothetical plant is a good investment, in my view the best plants are the ones built 10-20 years ago, such as in the US.  Without many greenfield steel mills or other industrial production facilities being built nearby, there is limited opportunity for a competitor to build a new competitive facility from scratch.  But as the plant is being fully depreciated and pricing power is increasing, existing customers are increasing their consumption through "creep" and additional customers are popping-up around the facility, as new uses for gases emerge.  

Current customers tend to increase gas consumption over time, as their own efficiency gains drive more throughput.  Industrial gases also often offset energy inputs, so energy inflation tends to drive increased demand.  Industrial gases are often used in the manufacturing of higher quality products or to reduce environmental impact, which is why usage will grow in developing countries as their standards improve.  Lastly, new products and production methods seem to sometimes involve gases.  Ten years ago, I would not have foreseen the gas demand from LCD and thin-film solar production, or oxygen for coal gasification.  Praxair's management believes that the combination of "creep" in existing customer demand and new customers around an existing facility generate 3-5% in revenue growth per year, before any significant growth capex.   

Taking a step back from a hypothetical and broad discussion, I want to enumerate the key industry dynamics.  The business of local production and distribution has significant barriers to entry, including: 1) need for local production due to transportation costs, 2) high customer switching costs due to long-term contracts and inertia, 3) high capital costs of initial production facility, 4) need for local density to minimize transportation and selling costs, and 5) capacity utilization being a key determinant of profitability, as capital costs represent up to half of production costs.  At the same time, however, economies of scale are also important, which is why this is not a Balkanized industry.  Economies of scale include: 1) cheaper capital costs, 2) engineering know-how and intellectual property (both for production and customer application), 3) supply-chain know-how, 4) benefit of multiple production facilities for redundancy, maintenance and efficient capacity utilization and 5) importance of reputation (industrial gases are often an essential element of production for customers, but a small relative cost of production).  I have come across few growth industrial businesses with these types of favorable competitive dynamics. 

We anticipate Praxair earning $8+ in 2016.  If the average company (S&P 500) has traded historically at 15.5-16.0x earnings, we would expect PX to trade at least above 18x (but we wouldn’t be surprised if the p/e multiple were in the low-twenties).  The share’s dividend yields just over 2%.

With the recent management changes at Air Products (Pershing Square was successful in getting the board to change management, and John McGlade was replaced by Seifi Ghasemi, who was very successful as CEO of Rockwood Holdings), we expect a very benign competitive environment going forward. The previous management was more growth-oriented than returns-oriented, but Seifi Ghasemi appears to be a very rational competitor and we believe this will lead to a good pricing environment going forward. Risks are primarily macro, Fx (Brazilian Real, etc.), an acquistion of Airgas by Air Liquide.

Catalyst