In our view, the best opportunities to be had today are in dominant, growing businesses being sold at below average prices.  One that has "bubbled-up" to near the top of our list is Air Products.

Air Products is a well above average business. First and foremost, the barriers to entry in the industrial gas business appear to be among the largest of any industrial business which I have studied, and are getting more formidable over time.  Second, the industrial gas industry has attractive growth prospects, in the short, medium and long-term.  I believe that Air Products will continue to grow revenue at 8%+ and earnings per share growth at a 10%+ annual rate.  Third, the combination of long-term take-or-pay contracts, significant and steady cash flows, the financing of half of the significant capital base with equity, and replacement value that far exceeds accounting value provide for a very strong financial position.

I believe that Air Products can earn $6.80 per share in a normal economy in calendar year 2012.  If an average company is worth 15x earnings, then Air Products is worth at least 17x earnings, or $115 per share in 2012.  This compares to the common stock price around $75 today.

History
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 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.  To get a sense of the breadth of uses today, review "applications and equipment" on Linde's US website:  http://www.lindeus.com/international/web/lg/us/likelgus30.nsf/docbyalias/nav_industry
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
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 Air Products to provide oxygen.  Air Products 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 Air Products takes is credit risk of the customer.  If the customer is solvent or the plant operates, Air Products 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, Air Products 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.  Air Products 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, Air Products 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 (unless this is the US, more on this later).  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 4-6% 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.

The balance sheet and cash flows (numbers are fiscal year-end, but they haven't changed materially)
Air Products has an accounting value of $7.5B for its physical plant, including long-term receivables which should be considered as such.  The gross plant, however, is $16.4B, which I believe is more representative of replacement value.  In comparison, the net debt is a little less than $4.0B.  Stated hard book is $3.6B, and there is a closed pension with $3.4B in gross liabilities which reduces the stated hard book by $1.1B.  Considering a) replacement value, b) long-term take-or-pay contracts and c) how the business performed in 2009, with world-wide industrial production down 9.2%, I believe this is a sleep-well-at-night financial position.

Cash flow from operations (which need to exclude growth in non-current capital leases, which is functionally capex) has been $1.5B or more for the last five years, including 2010.  The ~$6.5B of cash flow generated in 2006-2009 was spent as follows: capex and similar ~$5.1B, dividends ~$1.3B, and repurchases of ~$1.4B (acquisitions and divestments roughly net).  Spending exceeds cash flow as increased capital base allows for increased financial leverage.  

Earnings and valuation
Air Products expects to earn somewhere near $5 per share in fiscal year 2010 (ending in September), which is also what the company earned 2008, before Lehman Brothers.  So the shares trade at 15x the current year's expected earnings, which is slightly cheap given the quality of the business.  

But the earnings power of the business is higher than current results due to suppressed volumes and the benefit of costs which were cut during the recession.  In its 10-Ks and conference call slide decks, Air Products discloses the dollar-per-molecule-weighted changes in volume.  Based on this disclosure, I believe that volumes in 2010 will have roughly returned to 2007/2008 levels.  In the interim, however, Air Products has brought online capacity built for ~$3.0B, compared to 2009 EOY gross and net plant and similar of $16.4B and $7.5B.  And as discussed above, Air Products should have a small amount of "creep" in volume over time as its plants are run more efficiently, its existing customers use more gas, and as new customers emerge around an existing facility.  I estimate that volumes today are 10% below the "normal" activity levels of two-three years ago.  Management says that this "missing volume" would generate $1B in revenues and perhaps $300 million in incremental operating income, or $1 per share in earnings.  Incremental margins are high because many costs are fixed: the plant, the employees, the trucks, and the route costs.  This implies current year earnings would be near $6 per share.

Separately, Air Products appears to have cut costs significantly, some of which will not return, due to efficiencies.  At the nadir in Q3 2009, Air Products said that its quarterly earnings declined by 47¢ due to volume reductions, while 20¢ of cost reductions offset this.  Interestingly, in Q3 2010, a year later, returning volumes and new additions had added 36¢ in quarterly earnings, while there was only a 1¢ giveback in the previous year's costs.  Analysis of other quarterly disclosures thus far are likewise consistent with the costs being permanently reduced.  The net result is that management appears to be saying that they have increased earnings 50-80¢ per year by reducing costs.  While I am skeptical that all of the claimed cost reductions are permanent, I do believe it is possible that the earnings power of the business is stronger than it would have been if the recession had not occurred.  

A simpler way to estimate earnings power is to use a point of time before the recession and extrapolate forward.  In fiscal year 2008, Air Products earned $5.05 per share adjusted.  With plant additions, and some margin expansion consistent with management guidance and current performance, I believe the company should grow EPS 10%+ per annum (in comparison, Praxair says it will achieve 12-18% EPS growth).  If this is correct and assuming 2008 normal earnings were a little lower at $4.50, then earnings power should be at least $6.60 fiscal year 2012 and $6.80 for the calendar year.  Regardless of the approach used, $6.80 in normal EPS 2012 feels reasonable to me.  

Bid for Airgas
Air Product's current tender offer for Airgas's common stock provides enough fodder to keep an arbitrageur busy full-time. While I view the potential of value destruction due to overpayment unlikely, I cannot rule it out completely.  But because I believe Air Products business and financial position are so fundamentally sound, and the common stock can be purchased at such a discount, the price today is an attractive opportunity in spite of the uncertainty.  

Here are the salient elements, in my view.  Air Products has been trying to purchase Airgas for a little less than a year, which led to a public tender offer for the common stock in February.  But they have been considering buying the business for much longer.  Airgas is a strange bird.  Founded in 1982 by the current CEO Peter McCausland, it is the only example of sizable entry into the industrial gas business in the last 100 years, excluding Air Products and recent start-ups in China.  While much of the industrial gas industry involves high technology and large industrial concerns, the distribution of gas cylinders and equipment to small manufacturers and similar, particularly welders, is much more akin to traditional industrial distribution businesses, such as Grainger, with the added tie-in and revenue from renting cylinders.  The rapid growth of the small industrial base following World War II allowed for many local and regional mom & pop gas distributors to establish themselves in this portion of the supply chain.  Like other small business, cylinder and related distribution appears to have been a "lifestyle" business for many proprietors.  Peter's genius was seeing the roll-up opportunity as these businesses changed hands and needed to be professionalized, and economies of scale became more pronounced.  Airgas was even able to purchase sub-scale distribution businesses from some of the majors, including Air Products in the early 2000s.  Unlike in Europe and South America, which are more vertically integrated from gas production through cylinder distribution, Airgas represents a largely non-integrated enterprise, buying much of its gas from suppliers.  Airgas's share in US "packaged gases" is 25%, compared to 12% for Praxair, and with 50% independents.  Airgas does produce some of its own gas needs, operating ~10% of the air separation units in the US, mostly east of the Mississippi river.  Airgas has grown revenues and EPS faster than Air Products for the last half decade, due to the limited capital requirements and ample acquisition opportunities in the business.  

The current offer price is $63.50 per ARG share, which equates to $7.2B price including net debt, compared to peak revenue in fiscal year 2009 (mostly calendar year 2008) of $4.3B.  Airgas earned $3.12 in fiscal year 2009 ending March and management talked about achieving earnings per share of somewhere around $4.20+ in calendar year 2012. Note: these numbers include ~$25 million a year in amortization from acquisitions.  Also, accounting depreciation of around $200 million a year may exceed the economic costs.  All-in, the offer price looks fair to expensive before synergies.  

Air Products believes it can achieve $250 million in synergies from the combined entity.  Assuming $200 million generates $1.44 per ARG share, fully-taxed.  I note that there does appear to be an industrial logic to the combination, with Praxair's returns apparently benefiting from its own cylinder distribution business in the US and the vertical integration evident in other markets.  Consolidation in the industry appears beneficial to all players.

Air Products management has said the right things thus far regarding the hostile acquisition, such as having a top price which they are not willing to go above.  They have been mealy-mouthed about how they will finance the purchase long-term.  There is the potential that they will issue equity to finance part of the purchase, as they initially intended before their purchase offer went hostile (I view an equity issuance as likely and positive, given the proforma balance sheet).  My accretion-dilution model (which is similar to many street models) shows that an acquisition at $70 per ARG share would be accretive in 2012 assuming street estimates for ARG's earnings, $200 in pre-tax synergies, financed half by equity issuance at $70 per APD share.  One can make ones own assumptions.  The point is that barring significant overpayment, a low-priced equity issuance, or some impairment of the business during the interstitial acquisition/integration period, all of which I view as remote, I have difficulty imagining a scenario which impairs anything close to enough value to warrant APD's current cheap price.  

Conclusion
In summary, Air Products is clearly an above average business, being sold in the market at an average company price today, and significantly below average adjusting the earnings for a normal economy.  The Airgas hostile offer will play-out over time, but in all reasonable scenarios in my view, the common stock remains a compelling value.  

If you: 1) disagree with any logic or analysis, or 2) have in mind another industrial business which you believe has similar barriers to entry and higher growth potential, please expand in the discussion.

None.