Description

Summary

When a business with no contingent liabilities or other issues trades at 8x forward earnings, the market believes earnings to be unsustainable. It’s difficult to have an analytical edge in large cap stocks, but the huge dispersion in estimates for LYB tells us it is a good place to look. I believe LYB is mis-priced because the market is overly discounting the downside to margins that a combination of lower oil prices and more capacity will bring. In this write-up I’ll lay out my methodology for determining “normal” earnings, the path to it over the next few years, and why the stock is worth more in this scenario. In preview, I believe LYB is at 9.7x normal earnings with fair value of ~$100/share.

Read no further if you believe gas is headed much higher and oil is headed lower in the long term. The key variable in my long case for LYB is that this ratio stays where it is today or perhaps even improves. I can promise that this write-up is NOT based on the notion of great managers of “specialty” chemicals. The LYB management team, unlike basically every other chemicals company, fully admits and even embraces being a commodity chemicals company, and they manage costs and allocate capital accordingly.

LYB’s business can be split into two components: 1) chemicals that have steady cash flow characteristics based on low costs or an advantaged process; 2) chemicals like ethylene and polyethylene (“E-PE”) whose profitability is based on the spread between oil and natural gas. In this write-up I will concentrate on E-PE, which comprises the bulk of profitability and earnings volatility. My projections and mid-cycle case for the other businesses don’t differ much from their run rates. Below, you can see EBITDA by product. E-PE (Americas) and E-PE (Europe) are reported as the Olefins & Polyolefins segments. PO & Derivatives, Intermediates, and Oxyfuels are reported as the Intermediates & Derivatives segment. Refining and Technology are also reported separately.

Ethylene production

Ethylene can be produced by cracking NGLs including ethane, propane, and butane, or by cracking heavy liquids including naphtha, a refinery product, and condensates[2].

NGLs are present in the majority of gas production, and ethane is the most common NGL. The shale revolution made gas and NGLs cheaper relative to oil on an energy equivalent basis. US based producers of ethylene responded by switching from primarily using naphtha as a feedstock to using 90% NGLs (mostly ethane) as the price of NGLs followed gas downward. US producers of ethylene now sit on the low end of the cost curve. This is still true today even as oil (and naphtha) have become cheaper.

There are two key differences between these processes to keep in mind. First, domestic NGL prices track natural gas closely while naphtha prices track crude oil. Second, cracking ethane (the most widely available and widely cracked NGL) yields mostly ethylene (~80%), whereas cracking naphtha yields ethylene (~20%) and other co-products. This is one source of confusion that has contributed to the mis-pricing. As domestic producers of ethylene switched from cracking naphtha to cracking ethane, more ethylene was produced for the same amount of capacity. This lead to an oversupply of ethylene and a decline in ethylene prices relative to polyethylene. The upgrade of ethylene to polyethylene went from barely profitable to earning several hundred dollars per tonne. Additionally, the sudden increase in profitability for domestic E-PE has lead to the construction of new cracking capacity (all due in the next few years).

These factors have caused analysts to make two mistaken assumptions. First, they normalize margins for the upgrade of E into PE back to historical levels. Second, they assume that the additional ethylene capacity will cause further ethylene margin deterioration.It’s helpful to look at how how prices for polyethylene are set to see why these assumptions are wrong.

Ethylene-Polyethylene pricing

Ethylene is a local market because it is a gas and isn’t easily transported, but ethylene derivatives including PE are easily transported and thus a global market. If ethylene isn’t oversupplied in a geography, its price is set by PE less conversion costs and a small margin for PE producers. Asian naphtha based producers are the marginal producer of PE and other polyolefins.

Integrated (i.e. producers of both ethylene and PE) U.S. producers earn margins based on the price they can receive for PE (set by Asian naphtha based producers) less the cost of producing ethylene from advantaged feedstocks (mostly ethane). Because naphtha is priced from crude oil and NGLs are priced from gas, we can simply say that LYB’s E-PE profitability is based on the crude oil to gas spread.

An integrated US based producer of PE looks like this: Ethane -> Ethylene -> Polyethylene. The marginal producer in Asia is: Naphtha -> Ethylene -> Polyethylene. In normal times, low margins are captured in the Naphtha -> Ethylene step, and essentially no margin is captured in the Ethylene -> Polyethylene step. In a tight market more margin is realized in both steps.

The graphic below depicts the base margin earned from the crude oil to gas spread (E-PE), Asian ethylene producer margins, and Asian polyethylene producer margins as components of the E-PE margins of a US integrated producer (y-axis is $/metric ton). Notice that in 2014 and 2015 Asian producers (orange and yellow bars) earned exceptional margins. This had a variety of causes including a rapid fall in oil and naphtha prices (with lag effects on derivatives) and shortages.

Shortages are the most significant factor to be aware of. Prior to crude’s decline, some polyethylene supply came from other high cost processes including methanol-to-olefins and recycling. At lower crude prices this supply left the market as it became uneconomic. Utilization rates were already healthy and the reduced supply pushed them into shortage territory.

In my projection years and my normal case, I don’t give Asian producers any excess margin caused by shortages. Instead, I use the average margin earned over the past decade.

To understand the E-PE margin for a US producer, we need to look at each of these components:

A) The price of ethane

B) The cost and margin of cracking ethane to produce ethylene

C) The cost and margin of upgrading ethylene into polyethylene and other plastics

D) The realized price of polyethylene

Components (B) and (C) are less important when considering an integrated producer, because they capture margin at both steps to varying degrees.[3] Because LYB is (mostly) an integrated producer, (B) and (C) are less important for our purposes as well. Going back to my point above, we don't really have to care about whether excess margins will be earned in the ethylene step or the derivative (PE) step when we’re dealing with an integrated producer like LYB.

A) The price of ethane

The majority of gas wells[4] are considered “rich,” meaning NGLs are present. There are five NGLs – ethane, propane, butane, iso-butane, and natural gasoline – and typically ethane comprises 50% of the NGLs.

Ethane has little use other than for its heating value or as a petrochemical feedstock. If there isn’t demand for ethane as a feedstock, producers can either flare the ethane or send (“reject”) certain amounts[5] of it to utilities for heating uses. Because the shale revolution brought increased NGL production, the U.S. is oversupplied ethane. Ethane has traded at heat value since 2013.

Until the excess supply of ethane (seen by rejection) finds a home, we should see ethane prices in line with natural gas. There are concerns over the potential for a tight ethane market driven by recent reductions in drilling and also by planned ethane exports. These concerns are overblown because long term higher gas and oil prices will bring more NGL production, and planned ethylene cracker capacity additions aren’t enough to eat into all of the rejected ethane.  

In my projection years, I increase the price of natural gas in line with the forward curve, and I have ethane trade at a frac spread of 7 cents per gallon.[6]

D) The realized price of polyethylene

The price of polyethylene tightly tracks the price of crude oil over time. The relationship exists because the marginal producer of polyethylene uses ethylene derived from naphtha. To find the price of polyethylene in globally and in North America, we follow these steps:

Crude -> Naphtha (Asia) -> Ethylene (Asia) -> Polyethylene (Asia) -> Polyethylene (North America)

I’ll go through each of these four steps (highlighted in red) in turn.

1) Crude -> Naphtha (Asia) -> Ethylene (Asia) -> Polyethylene (Asia) -> Polyethylene (North America)

Because naphtha is a product of oil refining, naphtha prices closely correlate with Brent crude.

2) Crude -> Naphtha (Asia) -> Ethylene (Asia) -> Polyethylene (Asia) -> Polyethylene (North America)

Over the past decade (to March 2016), the cash margin of converting naphtha to ethylene has averaged $189/MT.  The net ethylene cost depends a lot on co-product credits because naphtha cracking produces less ethylene (and more co-products) vs. ethane cracking. Luckily, the prices of these co-products closely correlate with the price of naphtha over time so a simplifying step is to regress each of the co products to crude when building price projections.

3) Crude -> Naphtha (Asia) -> Ethylene (Asia) -> Polyethylene (Asia) -> Polyethylene (North America)

The conversion of ethylene to polyethylene should run at very low cash margins for the marginal producer. Over the past decade in Asia that has been the case.

4) Crude -> Naphtha (Asia) -> Ethylene (Asia) -> Polyethylene (Asia) -> Polyethylene (North America)

Polyethylene is easy to ship and is a global commodity. The difference in price between Asia, North America and other regions is negligible over time.

Going through these steps gives the components of E-PE margin. In the graphic below, I translate this margin into LYB’s EBITDA for E-PE produced in both North America and Europe. You can also see historic and projected EBITDA for the other segments as well as my mid-cycle “normal” year. I should point out that these projections quickly take any away any above normal margins like we've seen recently, which is why my 2016 and 2017 EBITDA numbers will be lower than many. I’m less concerned about being precisely wrong for those years than for being vaguely right about my estimates for mid-cycle.

It’s worth mentioning that there are a few other ways to produce ethylene, because it ties into my assertion that people are overly concerned about cracker capacity additions and low oil prices. These other processes include coal-to-olefins (“CTO”) and methane-to-olefins (“MTO”). A significant amount of PE also comes from recycled PE. All of these processes work well in a world with high oil prices, but as prices plummeted we saw reductions in supply from all three. In 2015, even as capacity utilization wasn’t at shortage levels, the world went into shortage because enough supply from these three sources came out just enough to get us there.

We could have world where oil stays lower for longer and where the loss of supply from CTO, MTO and recycled PE isn’t nearly made up for by ethane cracker additions (and where lower PE prices lead to more PE use). Or, we could have a world where higher oil prices lead to higher profitability for everyone (with the side benefit of making the production of rich gas more profitable and increasing the supply of NGLs). Meanwhile, the demand for PE has only had one down year in the past twenty. Either scenario is fine for LYB. This is why fears of oversupply are overblown.

Valuation

At $74.50, LYB trades at 9.7x my estimate of normal earnings, which uses the normal EBITDA shown above applied to the current capital structure. LYB is 5.6x consensus EBITDA and 7.9x consensus EPS for 2016 and has a 4.6% dividend yield.

I believe under 10x normal earnings is too cheap for a company where near term earnings should be above normal. Additionally, LYB is taking all excess cash flow and buying back stock aggressively. In the past two years, the company bought back 10% of the beginning share count in each year. My normal case contemplates a world where enough supply comes on such that the marginal guys don’t make much money. We have a while until we get to that world, so looking at a multiple of normal potentially understates fair value.

Commodity price projections

Because price projections for commodities including crude oil and U.S natural gas are crucial value drivers, I summarized my projection years and “normal” case below. Besides projecting these commodity prices, I’ve made assumptions about the normal margins earned by the high cost Asian producers as well as margins earned at other steps in the process.

Over the past decade, Asian ethylene producers have earned between $150-$200/metric tonne. I use $170 for the projection years. Note that recently these margins have been much higher, driven by shortages.

Another key assumption is the “frac spread’ for ethane. Note that in 2011 the frac spread was huge. This was a temporary disruption as the shale boom accelerated and infrastructure was unable to keep up. I normalize the frac spread at 7 cents / gallon, consistent with longer averages.

Earlier I detailed concerns about ethane prices rising faster than gas due to increased ethylene capacity and other uses like ethane exports. Ethane exports have gotten press recently, but it doesn’t make economic sense to export ethane to Europe except for a few specific scenarios.[7] Also, if prices really disconnected from gas, producers could switch to more propane and butane feedstock.[8]

Risks
- Oil languishes while natural gas prices rise

- NGL prices rise faster than gas prices

- Capacity additions far exceed announced/expected additions 

[1] UK tax resident after migrating tax residency from The Netherlands in 2013, and Houston TX headquartered.

[2] The terms “condensates”, “natural gasoline”, and “light naphtha” are often used interchangeably. Technically, naphtha should only refer to a refinery product. Generally, these terms refer to C5/C6 hydrocarbons.

[3] Typically, the bulk of margin is captured in cracking, not in the upgrade into derivatives. However, in the U.S., the upgrade of ethylene into polyethylene has enjoyed much higher margins than normal. When US crackers shifted feedstock to ethane, they produced more ethylene and less co-product for the same amount of capacity. Thus, the US ethylene supply increased and excess ethylene means a price based on lower value derivatives.

[4] NGLs are also present in oil wells.

[5] Limits are set by the utilities. Ethane has ~1,800 btu/ft^3 vs. ~1, 000 for methane but shares other properties, allowing it to be mixed with methane and used for heating.

[6] The ethane “frac” spread refers to the value of fractionating ethane vs. rejecting it, i.e. it is the vale of ethane above its heat value. Long term, ethane has traded at a 7c frac sread.

[7] For example, if a cracker were only able to use ethane.

[8] Propane trades around cracking arbitrage with ethane, but its price varies for commodity specific reasons. Its price rose in 2013 as export capacity opened up, but more recently traded lower as even more supply came on.

Catalyst

LYB enjoys higher cash flows as oil rises, or LYB enjoys higher cash flows as oil doesn't rise and high utilization and periodic shortages keep margins more than healthy.