North American Commodity Chemicals – Shale Gale Maybe, Windfall More Definitely
SEE LAST PAGE OF THIS REPORT FOR IMPORTANT DISCLOSURES
April 11, 2012
North American Commodity Chemicals – Shale Gale Maybe, Windfall More Definitely
- Higher natural gas production and low pricing in North America is driving down the price of ethane – a key feedstock for commodity chemicals. This should drive a sustained peak in profits over the next few years. The benefit accrues because the high price of crude oil sets product prices globally, and there is limited capacity to consume ethane. The conclusion is insensitive to the demand outlook, suggesting it will be unaffected by lower GDP expectations
- Commodity Chemicals as a sub-group is fairly valued, but appears undervalued relative to this extreme tailwind. We would expect the group to trade towards the top of the historic relative range, suggesting as much as 30%-40% relative upside as the story plays out.
- Ethane based profit margins are high and rising: North American producers could generate $4bn of operating profit per quarter through 2015! Return on Invested Capital in the polyethylene space should exceed 35% through the period and beyond if we believe the forward curves for oil and natural gas.
Source: IHS, SSR Analysis
- Today, North American ethane based ethylene producers have an integrated margin in excess of $1100 per metric ton (50 cents per pound) above the break-even cost of the marginal global producer – which will provide a pricing floor. In the last real super-cycle of the late 1980s, this margin was less than $280 per metric ton, around 12.5 cent per pound. In that cycle, pricing was well above break-even costs for the marginal supplier – we do not need a peak pricing cycle for our bullish thesis to play out in this case.
- Demand will try to spoil the party, but we do not think that it will be successful. So much of global ethylene demand is satisfied by higher cost producers (using a crude oil derived feedstock), that demand would need to fall by 35-40% before we hit a real point of inflection in global pricing. This would take global demand down to levels last seen in 1997. A demand decline of 20% should only impact US margins by $100 per metric ton.
- Risks are two-fold: first, investors focus on the macro trade only and underweight the entire sector as GDP and consumer spending forecasts fall, despite the earnings momentum; and second, the high cash flows result in value destructive acquisitions, a move at which the sector is well practiced. These are clearly concerns and could overwhelm the earnings momentum story. While there is the possibility of relative downside, given the fairly valued nature of the group today we think it unlikely in the face of the expected earnings tailwind.
- If natural gas as a transport fuel were to develop in the US as some are suggesting, this would increase the demand for and production of natural gas and by association the production of ethane. This would be a further boost for the commodity chemical industry given the potential for increased ethane supply and a relatively fixed capacity to consume.
As production of natural gas in the US has risen it has increased not just the volume of methane available, but also the amount of associated gases; ethane, propane, butane and natural gasoline. With the exception of ethane, these products have values that are derived from crude oil and as a consequence greatly enhance the value of the gas being developed – this type of gas (wet gas) is being preferentially developed today as a result. There is a limit to how much ethane can be left in the methane gas stream that we burn every day, and when that limit is reached it has one valuable alternative use – as a feedstock for the production of ethylene, one of the major building blocks of the plastics and chemicals industry. Without this use, the ethane would need to be injected into storage or, more likely, flared.
We are at a tipping point in North America, as we have reached the limit of consumption by the chemical industry and ethane is about to slip into a surplus that could last for many years. Ethane prices have been by far the most attractive feedstock for the North American chemical industry since the price of oil increased materially, despite the fact that ethane prices have also risen, generating a very good margin for those extracting the ethane from natural gas. With oil prices expected to stay high and natural gas prices in North America expected to stay low, the very strong margin that US ethylene producers have already seen will expand as ethane prices fall to break-even extraction values – a territory they occupied for most of the period from 1980 to 2005. Ethane availability in the US Gulf is rising steadily today because of developments in the Eagle Ford shale basin in West Texas, but will step change in 2013 when pipelines are completed to bring supplies to the Gulf Coast from the Western Marcellus. By end 2013, ethane supply in the US Gulf could be 20-25% higher than it is today.
Our analysis shows that there is a real risk that in the near-term we are flooded with ethane – possibly through 2016/2017 (and perhaps beyond). Today we have a surplus, but it could get much larger. Ethylene expansions are underway, but the big projects all seem to be “just around the corner” in terms of putting a shovel in the ground or ordering long lead time equipment, almost as if there is a big game of chicken going on. Hardly surprising given that there is no domestic outlet for this new capacity and to drop $1bn for a new plant that takes 4-5 years to build and that could have no market as soon as it starts-up is understandably something you want to think long and hard about. This is not unlike the LNG export investment thesis; it is a nice idea but the sums of money involved are so large and the lead times so long, that capacity will be limited and slow to arrive.
There is a longer-term risk that the US over-builds capacity, and creates more consuming capacity than there is ethane availability. Additionally, on a 7-8 year horizon, it’s a brave man who forecasts that the oil/US natural gas spread will remain at all time high levels. Consequently, actions could align to create significant US surpluses 8-10 years from now and an export market that is much less attractive than it is today. However, in the near to medium term, cash flows will be very high.
The stocks do not reflect this earnings tailwind – the commodity chemical companies in the US are in aggregate fairly valued on an historic normalized basis, whereas, at the beginning of a significant profit super-cycle we would expect the sector to trade well above mid-cycle values. This suggests as much as 30-40% relative upside as the earnings momentum becomes more obvious.
Outside the commodity space the rest of chemicals look less interesting when focusing specifically on this issue. Many are large consumers of fuel and the lower natural gas environment has already played its part. Heavy energy users like PPG, DuPont, Air Products, Praxair and the smaller market cap names are either in industries where costs get passed through to customers, both on the upside or downside, or in end markets where a demand slowdown is more significant than feedstock cost or energy cost advantages.
The Feedstock Supply/Demand Logic is Hard to Fault
The price differential between natural gas and crude is hard to ignore and is more extreme than at any point in history – Exhibit 2. What is most interesting is that the logistics of moving gas out of the US and the logistics of changing consumption patterns in the US are complicated and not something that can be altered overnight. For example – for very obvious reasons, the chemical industry is consuming as much natural gas based feedstock (ethane) as it can – slightly more than a million barrels of ethane a day. To consume more it has to spend money to convert or expand facilities – a 12-24 months process, depending on the project – or spend a lot more money to build capacity from scratch – a 5 year process.
More generally, the US is seeing some restarts of natural gas based businesses that shut down in the early part of the last decade as gas prices peaked – most notably methanol and aluminum facilities (methanol consumes methane as a feedstock and the aluminum industry is all about the cost of electricity – which is very competitive when based on natural gas).
Source: IHS and SSR Analysis
The ethylene industry is not keeping up with the potential feedstock supply. Pipelines scheduled to come on line through 2013 could add as much as 25% to the available supply of ethane in the US Gulf while near-term capacity expansions will increase the ability to consume ethane by around 10% by 2014 and 18% by 2016, if we are to believe announced plans for expansion (by which time we would expect ethane availability to be higher still). While ethane is cheap today, relative to alternatives, it could get much cheaper, as today there is a reasonable margin in the ethane extraction process and as Exhibit 3 shows, we have seen periods of ethane oversupply when this margin has trended to zero and in some cases fallen below zero. The margin can fall below zero and stay there, as the alternative for ethane is to leave it in the natural gas stream, but once the upper limit is reached the options are re-injection (into wells or storage) and/or flaring, and both of these have tangible and intangible costs. As the chart shows, from 1982 through 2005, ethane was extracted from natural gas mostly because it was necessary, not because it was profitable.
Source: IHS and SSR Analysis
So What Exactly is a Lot of Money
We have done a very simple exercise to try to frame the opportunity. The price of polyethylene, in an oversupplied global market, is the cost of manufacture of the marginal supplier and today is set by the cost of manufacture using a crude oil fraction (naphtha) as a feedstock. Then we assume that the cost in the US is the integrated cost of starting with natural gas, extracting the ethane and converting through to the same polyethylene. The difference between the two is the margin available to a US manufacturer. Because we are not energy forecasters, we use the forward curve for crude and the forward curve for natural gas – as shown in exhibit 4
At this point we make some broad assumptions around production costs, but as we have one of the more robust historical databases on this subject, we are qualified to do so. The analysis is summarized below in Exhibit 5, and our broad assumptions are outlined below the exhibit.
Source: Bloomberg, IHS and SSR Analysis
- (1) The price of naphtha in c/gallon is 2.5x the price of Brent crude in $/bbl. This is slightly below the 40 year average, which has been fairly stable.
- (2) The co-products on the Naphtha plant are valued at about 80% of the cost of Naphtha, which is above the historic average and well above the average in high crude oil price environments– so we are being generous to the naphtha process. We discuss this assumption later.
- (2) The analysis is based on a capacity of 1.2bn pounds of polyethylene operating at 95% of capacity
- We are assuming that we are making High Density Polyethylene
- We are assuming that the clearing price for US polyethylene is set in the US and equivalent to the price setters cost, regardless of that location. (In the cost curve analysis that follows, the price setter, i.e. the marginal supplier is in Europe). This assumption is conservative based on historic fact, but we will challenge this assumption shortly as it is possible that as US exports rise, prices in the US could fall to reflect the price in export markets less the cost of freight from the US (particularly for the more commodity grades of plastics)
- The integrated capital cost of building capacity to separate ethane, make ethylene and then make polyethylene is $1.5bn – this is around 10-15% higher than most recent estimates from manufacturers and consultants.
To be clear – the price of polyethylene in the US does not reflect this theoretical global break-even price today, it sits well above that level. High Density Polyethylene pricing in the US is at around 95 cents per pound today, versus the break-even number of 68 cents in Exhibit 5. The price is higher because ethane pricing in the US has not yet fallen to break-even, which currently adds around 15 cents per pound to costs and historically the US price has sat a premium to international pricing.
The cash flows and returns derived from this analysis (even with a lower theoretical polyethylene price) are very strong, but have existed in the past. We saw a 36 month period of slightly higher margins in 1988, 1989 and 1990, where returns peaked at a higher level, because investment costs were lower. One of the differences this time is that the length of the period of high margins is expected to be longer. Of course back in 1988 the period was expected to be longer as well, but the expectation was based on a prolonged cyclical demand peak.
The other and more important difference this time relates to demand. Demand should not screw this up!
Demand may impact investor sentiment but it should not change the story for US gas based producers.
Following 1989 and 1990, conditions were very different, there was an economic slowdown that impacted most of the world and commodity chemical demand globally grew well below trend. In addition, we were in a period when the world was seeing its first wave of new capacity in a decade. Operating rates slid from around 95% in 1988 to 85% by 1993.
We are terrible forecasters of demand – not SSR, but everyone; SSR included. So under no circumstances are we going to suggest that we are in any way relying on demand. We are not.
The other difference this time is the steepness of the cost curve, and where US ethane based producers sit on that curve relative to the high cost players that ultimately set the price. In the late 1980s early 1990s oil was cheap, and the cost curve was relatively flat, the only significant advantage accruing to the new producer in the Middle East, running on close to free gas. The cost difference between the high cost producer – setting global pricing – and the average US producer was small. The US industry made very little money in a globally oversupplied market. Profits were high in 1988, 1989 and early 1990 because the market was short of material and prices rose well above costs – everybody’s costs. Exhibit 6 shows the global ethylene cost curve as it was in 1989 (crude oil was around $18 per barrel and US natural gas was at $1.60/MMBTU. The price of ethylene was so far above the cost of the break-even producer that no-one really cared about the shape of the cost curve (they really cared 3 years later!)
Source: IHS and SSR Analysis
Contrast Exhibit 6 with Exhibit 7, which shows where we believe the cost curve to be today. There is a lot of information in this analysis so we should walk through the conclusions one by one.
First, the shape of the curve: the low cost producer in 1989 was the same guy it is today – the original facilities in Saudi Arabia, based on a natural gas price where the alternative was to flare the gas. There were three of these facilities in 1989. The curve has moved to the left as more low cost (gas advantaged) plants have been built, mainly in the Middle East, and it has pivoted to the top right as the costs of naphtha feedstock and natural gas outside the Middle East and North America have risen relative to the Middle East and North America.
About 80 million tons of capacity has been added over the 23 year period – mostly in developing economies or where costs are low.
Second, how we arrive at a price: in 2012 and beyond, we are assuming that global pricing is set by the break-even price, not by some demand driven level set by the marginal consumer – as it was in 1989.
Third, the curve drives the margin: as the curve is much steeper than it was in 1989, the gap from the US natural gas based cost to the break-even level in 2012 is around $1100 per metric ton, whereas it was around $280 per metric ton in 1989. (50 cents per pound versus 12.5).
Source: IHS and SSR Analysis
Back to demand. In the analysis in Exhibit 7, we assume that the marginal producer has an operating margin of zero, so we are modeling an oversupplied market. The benefit accruing to the US producer is driven 100% by the steepness in the cost curve caused by the lower cost of natural gas in the US. Demand has to fall dramatically before a natural gas based producer is setting the price.
Looking at the same issue from a different perspective, as shown in Exhibit 8, in its most recent global analysis, IHS estimates that global ethylene produced from ethane accounted for around 42 million metric tons out of a global total production of close to 125 million metric tons in 2011. We would have to close a very large amount of capacity for the marginal price of ethylene to be set by a feedstock that was not a crude oil fraction (naphtha and gas oil in the exhibit). If oil stays where it is, gas remains cheap and ethane supply in the US remains above the ability to consume, the margin benefit is reasonably assured, regardless of the demand environment. In its worst ever year, global ethylene demand fell by around 12% (this was back in the early 70s after the first oil shock). Based on the analysis in exhibit 7, if demand fell by 20% in 2012, pricing (and therefore margins for US producers) would only fall by around $100 per ton (4.5 cents per pound).
Source: IHS and SSR Analysis
To be clear, we expect demand to disappoint. If you look at the long-term trend in ethylene demand growth, it has a negative slope – Exhibit 9. It has the slope partly because of the law of big numbers, the larger the market the harder it is to add absolute increments of demand, but also because there is a huge learning curve in terms of plastic use and substitution at play in the background, which makes significant growth in a period of high prices very unlikely. Consumers of plastics, particularly in the packaging space continually learn to use less for any given application and, in addition, build flexibility into their packaging options so that they can move to cheaper materials as they become available. In an environment where we expect plastic and chemical prices to be set by $100 plus crude oil it is intellectually inconsistent to think about strong demand growth. The bull case on demand is always wrong because it is based on multiplying current demand by some expected expansion of the global middle class urban consumer, without taking the learning curve or substitution into account.
Source: IHS and SSR Analysis
With this earnings tailwind Commodity Chemicals has upside
We use discount from normal (mid-cycle) relative value as our investment selection tool – as introduced in our first piece of research last week “Redemption May Come, But Only To Those Who Have Truly Suffered”
This methodology shows that the commodity chemical industry is volatile from a relative valuation perspective, as illustrated in Exhibit 10. Stocks swing from very expensive to very cheap, sometimes over a long period and sometimes quickly – more recently, quickly. The prolonged period of high relative value in the late 80s was in anticipation of the high and prolonged profitability peak that followed the fall in oil prices in 1986. The prolonged period of cheapness from 1995 to 2002 had much more to do with investors’ interest in growth and technology than it did with anything going on in chemicals.
It is important to note that when the industry has a wind to its back valuations can move dramatically. A one standard deviation event for the commodity group is a 21% movement in relative value – the sector is fairly valued today and the tailwind we foresee could well produce at least a two standard deviation event, as it has in the past. For the individual companies a one standard deviation event is more extreme (as much as 40% in some cases) as the group aggregation process tends to mute the volatility. The commodity group has underperformed month to date and is now slightly below normal value.
Source: Capital IQ and SSR Analysis
Ethane based production capacity in North America is summarized by publicly traded company in Exhibit 11. We add perspective by showing capacity in pounds per share. Dow Chemical and LyondellBasell have capacity in other parts of the world but we have chosen to highlight only the advantaged North American capacity (Dow has significant feedstock advantaged capacity in the Middle East). Where a plant has the ability to consume ethane as well as other feedstock we have shown the capacity in its entirety and as a consequence we are overstating for Dow and LyondellBasell, but as both are maximizing their ethane consumption today we are not overstating the numbers by more than 30%.
Source: IHS Company Reports and SSR Analysis
Risks To The Idea – Macro Apathy and Walking Around Money!
There are two risks to the idea that we have outlined in this piece – the first is that we are right and nobody cares and the second is that we are right and several of the industry players commit the business equivalent of self mutilation, scaring investors away quite quickly. Note that we believe that our fundamental conclusion is quite robust.
Basic Materials and many of the Industrial sectors are macro plays – buy when the economy is in the pits, wait for a recovery and sell before the recovery peaks. Essentially this is a demand surprise play. Economic peaks and troughs are very aligned with consumer spending highs and lows and companies in the value chain will also play around with inventory in anticipation of the changes, effectively magnifying the amplitude of the swing at the industry and company level.
So the key question is whether any of the Basic Materials sectors outperform in a period of negative economic and consumer spending revisions?
History tells us that for the most part – no! The exception would be when a very industry specific tailwind was in place. In almost every case this has been a unique supply/demand dynamic that has selectively impacted pricing – gold would be a great example, as would copper.
We believe that the tailwind for US commodity chemicals could be strong enough to do this and a catalyst will be Q1 earnings where producers admit to how much money they have made in the first quarter and then admit that half or more of this was in March, and that April looks as good or better. We would expect already high estimates to move higher.
The second risk: “I have to have that shiny thing” – Acquisitions
The second risk is far more interesting as the commodity chemical industry does have a habit of spending peak cash flows unwisely, either on new capacity which over estimates demand growth, or on expensive accessories. Private equity and producers alike have made regrettable decisions to buy businesses for cash at cyclical peaks, partly because no-one will sell at the trough and no-one has the money to buy at the trough.
One query which followed our first piece of work last week questioned how much of the declining story behind aggregate returns was driven by acquisitions which then resulted in large write-downs. On an aggregate basis, acquisitions do play a part in declining returns as they have increased coincident with returns falling, but they are not a big part of the picture and the chart in Exhibit 12 shows aggregate acquisition spending as a percent of revenues for the overall Industrial and Basics group.
Where the analysis is a bit more interesting is when you look specifically at chemicals, as the peaks in the exhibit are mostly explained by acquisitions in the chemical space (for example the 1981 peak is DuPont’s acquisition of Conoco). Dow Chemical, DuPont, Ecolab, Ashland, Cytec, Monsanto and Georgia Gulf have all made significant acquisitions (relative to their own size) over the last 15 years, while others, like PPG and Air Gas have been serial acquirers of smaller businesses. While we are not going to comment on the wisdom of any individual deal, acquisition spending in the chemical industry has likely had a more persistently negative impact on return on capital and for the shareholder than in many other sectors.
Moreover, there is already deal activity with Eastman acquiring Solutia in an agreed deal and Westlake making a hostile approach to Georgia Gulf. If we were to see a significant hostile move at values that clearly represent cyclical peaks, we could see a general cooling of interest in the group, particularly the larger cap names which will be assumed to be the potential buyers.
Source: Capital IQ, Company Reports and SSR Analysis
Sensitivities to our cash flow forecasts: US pricing relative to the rest of the world and co-product values.
In the analysis in Exhibit 5 we have assumed that US polyethylene pricing will not fall below international levels – this is despite significant US exports of polyethylene. We are comfortable with this conclusion for the next few years because what we see happening today is not inconsistent with history. The US has been a net exporter of polyethylene for decades and yet has seen domestic prices well above international clearing prices. If we were to see a wave of new US capacity dedicated to the export market there is a risk that domestic prices to fall to a level that reflected a netback from the major markets consuming US polyethylene, but as yet that capacity is not happening to the degree necessary and with the speed necessary to impact our views for the next 3-4 years.
The risk would be that the price in the US fell to the break-even value that we have established, less some sort of freight adjustment – in a worst case this could be 5 cents per pound ($110 per metric ton). If we look at the economics in exhibit 5 this could negatively impact our 25 cent per pound margin by around 20% in the out-years. This would drop return on invested capital to around 28%.
We have assumed that the co-products produced on a naphtha based unit (propylene, butadiene, etc.) are valued at roughly 80% of the naphtha feed – i.e. they offset 80% of the feedstock cost. This is close to the historic average, and higher than the levels we have averaged since oil have increased. This is partly because some of the lighter co-products have a natural gas based value and so you are essentially turning some of the “crude oil” to natural gas, which is value destructive.
Our analysis would be hurt if co-product values increased further, reducing the net cost of making ethylene from naphtha and history shows us that they have peaked in value in excess of the cost of naphtha, though again never in periods of high oil prices. Today the ratio is around 70%.
If we assumed that co-product values covered the cost of naphtha, our returns in Exhibit 5 would fall to zero as naphtha would be competitive with ethane as a feedstock. While such an occurrence would undermine the opportunity we have outlined in this report we believe that the risk is small. The primary drivers of the credit are propylene and butadiene prices. Higher propylene prices drive up the price of polypropylene, and today it sells for more than some grades of polyethylene. Polypropylene’s appeal has been its attractive price and it has grown in use through substitution of other polymers and paper/card. At current pricing demand is moving the other way – this will limit the ability to raise pricing and keep a limit on propylene pricing. Butadiene’s primary use is in synthetic rubber and here the marginal competition is with natural rubber.
©2012, 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.