Industrial Gases – APD Must Focus on Costs, But PX the Better Investment

Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0


Graham Copley / Nick Lipinski



July 28th, 2014

Industrial Gases – APD Must Focus on Costs, But PX the Better Investment

  • APD is trading at a 3x multiple premium to PX today – a reversal of a recent history of PX premiums. Investors must be convinced that APD can cut costs aggressively and quickly to bridge that gap, as there is nothing in the fundamentals of the business or the operations of each company that can justify the spread. We think that APD needs at least a 10% headcount reduction.
  • The current multiple premium equates to an NPV of around $230 million of pre-tax cost reductions at APD falling fully to the bottom line. If we assume that the valuation relationship prior to Pershing’s involvement is still appropriate, APD would need to cut costs by more than $450 million to justify the current price.
  • Even if APD can materially change the returns on investment on future capital spending the company will take years (probably 6, best case) to get to the average level of returns that PX earns today. The risk is that the projects in the pipeline today – capital already allocated – show flat or declining returns ahead of any improvements – delaying the path to improvement.
  • M&A is not the answer. Accretive sales are hard to imagine given the high current earnings multiple for APD and asset swaps would likely benefit PX as much as APD.
  • We see a 12% relative move to get to multiple parity today and we think that multiple parity better reflects the relative opportunity at APD. The last 3 years have been disappointing for both PX and APD, but APD has consistently seen more significant negative revisions than PX. Once the euphoria of the new leadership is over – revisions will matter.
  • Near term there is a risk that APD estimates rise further (making the task harder) as the consensus target price is below the current price. Analysts will likely want to avoid having to change their recommendation to a sell.

Exhibit 1

Source: Capital IQ, SSR Analysis

The Case for PX relative to APD

To understand our relative view on PX versus APD it is necessary to study the chart below (Exhibit 2) in detail. The chart shows the deterioration of Praxair’s multiple relative to Air Products after it was spun out from Union Carbide – effectively through late 2000 – followed by a slow but steady improvement. This improvement effectively lasted until the activist interest in APD began in Q2 2013.

  • The spikes in PX’s values in the 1990s were the result of: an unjustified premium at the time of the spin; the CBI spin out in 1995; and the APD BOC acquisition debacle in late 1999.
  • The peak in 1999, quickly corrected but the last small peak of that period was coincident with Dennis Reilly being appointed. The market gave Dennis a little bit of the benefit of the doubt for around 10 minutes.

Exhibit 2

Source: Capital IQ, SSR Analysis

The trajectory from late 2000 to early 2013 was driven in our view by superior execution, and focus on costs and project returns. This is best demonstrated in Exhibit 3 which shows net income margins from January 2001 on a rolling 12 month basis. We picked 2001 because the APD data for 2000 was distorted by the currency hedge loss on the failed BOC deal.

Note that our model uses 12 month forward sales and earnings and from August 2013 is consequently picking up increasing levels of forward estimates, and the step change in APD’s line at the right end of the chart is almost entirely driven by consensus forecasts. As shown in Exhibit 4, APD has seen some fairly significant negative revisions/misses over the last few years, and while PX has also seen negative revisions/misses, they have been less severe and are shrinking – Exhibit 5.

Exhibit 3

Source: Capital IQ, SSR Analysis

Exhibit 4

Source: Capital IQ, SSR Analysis

Exhibit 5

Source: Capital IQ, SSR Analysis

The faster net income margin growth at PX has, for the last 7 years, been accompanied by higher revenue growth – Exhibit 6.

Exhibit 6

Source: Capital IQ, SSR Analysis

Note: Divestments caused the step change at APD in 2009, but growth has lagged PX since. Some of the 2008 to 2010 revenue decline was a function of declining natural gas prices and their impact on “pass through” contracts

Our view, which we think is supported by the analysis above is that PX has earned its superior multiple relative to APD through consistently superior revenue growth and net income margin growth, driving a better trajectory in return on capital (Exhibit 7) and clearly superior shareholder returns. Note that with the return on capital analysis in Exhibit 7 we again have an estimate bias in the most recent data.

Exhibit 7

Source: Capital IQ, SSR Analysis

So How Does A New CEO at APD Justify a Relative Multiple Swing from 1.15 to 0.80?

One can argue that APD was inexpensive prior to the Pershing Square action, and indeed we had highlighted the company as unusually inexpensive in early 2013, but does a change of leadership support removing essentially all of the relative gains that PX has made over the last 13 years? Unlikely.

Like PX, APD’s capital base has been established over many years, mostly with fixed term contracts and fixed prices or margins, and it is very difficult to do anything with this capital base except for some incremental tweaking. As we think about APD relative to PX there are a few things that the company cannot control or cannot change dramatically in the near or medium term, which are important as we think about shareholder returns.

  • The Macro Environment – no influence, but no real relative difference between the two companies save a few regional biases which are well known
    • What will benefit APD sales will likely equally benefit PX sales
    • A new CEO is not going to suddenly increase demand for APD products
  • Construction and operating efficiencies – here PX is dominant and there is an opportunity for APD to play catch-up but on future capital; there are fewer options for capital in the ground
    • ADP has around $15bn of capital employed today, earning an average return of approximately 9%. As a best case, assume that the company can immediately move to PX’s average on new projects (currently 12%) – unlikely given that much of the 2015/16 pipeline is in place, but let’s use this for illustration purposes.
    • If we also assume that the company has 10 year straight line depreciation, we get the return on capital profile shown in Exhibit 8 – improving but still not coming close to the PX number until 2020 – this is very much a best case.
    • This is a path that APD must follow, but we think it is unlikely that they could do as well as quickly as we are suggesting. What we show in Exhibit 8 assumes that PX stands still, which is equally unlikely. This does not justify the current premium in APD.

Exhibit 8

Source: Capital IQ, SSR Analysis

Things that can be done near term include M&A and Cost Reduction.


Barring something totally off the wall, we think this is a dead end from a value and relative value justification perspective. Almost nothing in APD’s portfolio would sell accretively today given the very high multiple for the overall company. Asset swaps in the industrial gas business might improve market structure in some parts of the US and in Europe, but PX is as likely to be a beneficiary of such moves as APD, and while APD might see better margins and returns on restructured assets, it is very likely that PX would also.

Cost Reduction

APD’s current multiple premium over PX implies that the company can quickly realize roughly $230 million of pre-tax cost reduction – assuming multiple parity is appropriate. If we assume that PX’s recent multiple premium is appropriate, APD would need to cut roughly $450 million of pre-tax costs, which is close to what would be needed to get to net income margin parity.

APD has a very similar SG&A ratio to sales as PX but a much higher cost of goods sold ratio to sales, which explains where PX’s higher margins come from. We can debate whether the SG&A costs are too high at Air Products based on business mix, but optically, looking at the size of the Allentown campus versus the minimal footprint in Danbury for PX, one might instinctively feel that they are. However, the bigger target may be operating costs. If APD could lower operating costs as a percentage of sales to the same level as PX they would be able to find $800 million.

This is, of course, a naïve assumption as part of the story here is business mix which explains a great deal of the difference. On-site/tonnage gas cost of goods sold is mostly the pass through of natural gas and electricity and the business tends to have much lower margins – but lower SG&A needs as a consequence. By contrast a packaged gas business has much higher margins – lower cost of goods sold – but higher SG&A.

PX has much more packaged gas and APD has more tonnage – as shown in Exhibit 9. Simply looking at the packaged gas component we can probably conclude that SG&A costs are high at APD relative to PX and there is also probably a cost of goods sold lever or two that could be pulled. In conclusion, it is likely that APD can find significant cost reductions, and they could total what is priced into the stock today – the question is how quickly they could be realized, and how much business disruption and lost focus might accompany these reductions. We suspect that the implied timing in the share price is ambitious.

Exhibit 9

Source: Capital IQ, SSR Analysis

Most of what we would be looking at here are people costs. Praxair has around $440,000 of sales per employee, while APD has $480,000 of sales per employee. One would reasonably expect PX to have lower sales per employee than APD because the packaged gas business is more people intensive, even though the price point is higher. For perspective, Airgas has around $320,000 or revenues per employee. If anything should encourage APD to look at headcount it is the chart in Exhibit 10, which shows PX playing catch up over the last ten years despite a business mix that is, in theory, significantly more people intensive.

Exhibit 10

Source: Capital IQ, SSR Analysis

If APD is going to come close to the cost savings needed to justify the current relative price the new CEO is going have come up with at least a 10% headcount reduction – and it needs to be done quickly. Assuming that the average employee costs around $100,000 per year, this would get you half way to the larger numbers suggested above (2,100 to 2,200 employees) – $220 million.

Where We Go From Here

Most likely, earnings at ADP continue to disappoint, made worse by the need for analysts to raise estimates simply so that they can maintain a positive view at current high prices. Note that, according to Capital IQ, consensus is positive, but not excessively so, despite an aggregate target price that is $7 per share lower than the current price!

By contrast, PX has a slightly more positive average rating, but at least it has an aggregate target price that is above the current level. APD has seen positive revisions for 2016 EPS while PX has not, perhaps reflecting an expectation of lower costs at APD, but the improved estimates do not justify the price differential.

Our normalized value charts for both APD and PX are shown in Exhibits 11 and 12. To be clear, neither stock is overly expensive, but PX does not deserve the relative discount; moreover there is likely no arrow in APD’s quiver than can justify where we are today.

Exhibit 11

Source: Capital IQ, SSR Analysis

Exhibit 12

Source: Capital IQ, SSR Analysis

©2014, SSR LLC, 1055 Washington Blvd, Stamford, CT 06901. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein.  The views and other information provided are subject to change without notice.  This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results.

Print Friendly