Cost curve analysis critical in evaluating commodity chemicals

Integration back to the cracker must be the raison d’etre of the commodity polyethylene business. But does this make cost curves simple enough to act upon?

Paul Ray/London

IT’S A love/hate relationship we all have with the volatile petrochemicals market.

This event-driven business is rarely short of intriguing happenings that keep us all guessing. Even if we claim that we’ve seen this or that event before, and can justify its occurrence, we all usually feel (when being honest) that we’re passengers on a very large ship whose ultimate destination will always be somewhat unknown.

Ever wondered how people stay the course in the cyclical world of petrochemicals? How they justify builds, exits, mergers and acquisitions, or asset swaps? After dispensing with the many and varied human characteristic drivers of this business, by and large it all has some connection to cost curve rationale and a perceived place (or value) related to positioning on such curves.

POLYETHYLENE ANALYSIS

For commodity polyethylene (PE) there are three simple, broad types of manufacturing assets that participate in the market. These are illustrated in the conceptual cost curve picture.

Low-cost, large-scale integrated manufacturing assets typified by the rapidly emerging class of Middle East assets, where the advantages presented from feedstock costs normally outweigh lost benefits from not having the asset based in a large market center.

The advantages of this asset class are particularly pronounced in today’s high-cost oil and gas markets. By and large, they represent the fastest-growing segment of supply to the market despite many practical project delivery challenges, such as rising capital costs, constrained contractor resources and politics, to name a few.

Integrated (i.e. cracker plus polymer unit) manufacturing assets located in large regional markets such as the US, Europe and Asia: this integration does not always mean colocated assets but must reflect common equity across assets, or individual company material balances effectively providing equity integration.

In Europe and Asia, more than 90% of installed PE capacity is effectively equity-integrated back to the cracker. The business model is driven by cracker feedstock procurement and the sale of PE and coproducts.

Nonintegrated (polymer unit only). This is the simplest business model where one buys ethylene as a feedstock and sells PE. This is also the most exposed, as regardless of how efficient the polymer producing operation is, the procured price for ethylene feedstock is almost always greater than the cash cost of producing ethylene.

As a quid pro quo, the capital in a nonintegrated facility is lower than that required for either of the two preceding categories.

OBSERVATIONS FROM COST CURVE

The conceptual PE cost curve when considered in combination with the contribution margin assessment over raw material and key variable manufacturing costs, in the new ICIS pricing report “Weekly Margin – PE Europe” leads to several observations.

Examination of volatile week-by-week changes provides simple signals on the direction of business contributions, as dictated by the environment, informing market positioning by sellers, buyers and traders.

Nonintegrated producer margins fall short of adequate reward to justify any fresh investment, but do currently support continued operations.

Sustainability of marginal, nonintegrated operations is questionable as lower-cost alternative supplies grow rapidly and the currently increased global economic uncertainty may taper down market growth rates.

One or both of these factors will eventually contribute to some softening of the market and the likelihood that pricing will start to be driven from a lower (integrated) step on the cost curve. This scenario existed in Europe for much of 2002 and 2003 with integrated economics effectively setting a market price floor.

It is most unlikely that pricing would move down so far as to force the exit of efficient integrated producers in the market. This would imply an enormous product overhang. Such a scenario is possible for short periods, but not for months or years.

Restructuring moves do drive exits or integration from marginal positions. A nonintegrated asset closure is Basell‘s high density polyethylene (HDPE) unit at Tarragona at the end of 2005. A recent example of effective integration is SABIC‘s PE assets at Gelsenkirchen, Germany, following the acquisition of the Huntsman cracker complex at Wilton (selling ethylene out of the UK effectively provides an offset for purchasing ethylene in Germany).

A similar move by Basell, acquiring the Ruhr Oel cracker, has integrated its PE asset at Munchmunster, Germany. This does not necessarily overcome accounting or geographic hurdles, but does strategically group assets together.

Plant exits can be far from intuitive, due to value-added PE grade production slates, site exit costs exceeding the pain of continued operation, prohibitive contractual exit terms, cross-business support, exits tied to market backfill from new projects, or to plant obsolescence.

RETURNS AND INVESTMENT

Over a full business cycle, cost-efficient players should derive adequate returns to continue to justify investing in satisfying the growth of the markets.

The definition of an acceptable return or a hurdle rate is a perilous exercise, and to borrow a comment from Lee Raymond, the retired Chairman and CEO of ExxonMobil: “Hurdle rates are a self-fulfilling prophecy. It doesn’t take much of an analyst, and most of us early in our careers have been analysts, to show that if you know the hurdle rate, you can always find enough credit to just get a project going.”

In simple terms, those meriting investment should always be found toward the left of the cost curve. Those toward the right-hand side and only producing commodities are typically seeking opportunities to reposition their businesses.

In a commodity world, the winners are invariably those with the asset base most strongly positioned to the left-hand side of any cost curve picture. The rest will just become history on the continued volatile course of the business.

EXPLAINING THE COST CURVE PICTURE

• The cost curve model ranks all manufacturing plants in ascending cash cost order. Defining cash cost can be perilous in itself, but here it is simply considered to be the sum of feedstock, manufacturing variable and fixed cash costs for running the business.
  Any product price premium over this cash cost represents a cash margin available for supporting the business working capital, taxes, royalties, any corporate costs, debt service costs, capital costs and owner’s returns from the business.
• In a normal market, the market will only pay up to the last marginal tonne that is needed to balance demand against supply. Any tonnes above this cut point on an ascending cost supply curve should not be required and therefore not remunerated by the market. This is a concept of laggard-driven pricing.
• In a tighter market, where the market demand requirement is high on the cumulative PE supply curve, prices should rise in order to justify continued operation of higher-cost operations.
• In a longer market, where the market demand requirement is lower on the cumulative PE supply curve, prices should fall to only reward the more efficient participants that are needed in order to balance the market demand. Any higher-cost players will be driven to exit.
• From an investor’s and market sustainability perspective, the key issue is whether there are adequate returns from continued investment from the asset classes that are needed to balance future market growth. Once again, there is a wide range of possible definition for “adequate.”
  However, there should be no dispute that the recent year returns from advantaged feedstock players have been more than adequate (e.g. SABIC’s earnings before interest, tax, depreciation and amortization margins in excess of 30%/year).

ICIS PRICING WEEKLY MARGIN – PE EUROPE REPORT

The report provides a consistent, weekly view of the impact of costs and prices on an integrated PE business, with simple signals on the direction of business contributions, informing market positioning by sellers, buyers and traders. The analysis is driven by manufacturing yield data from Linde Engineering and from ICIS market price reporting. A key objective is to deliver clarity of the real impact from volatile feedstocks through to contracted petrochemical markets. It provides a contribution margin measure over feedstock and key variable manufacturing costs. For more information, contact csc@icis.com

Paul Ray joined ICIS early in 2007 as the head of data and analytics, to drive the development of new data services. He moved from INEOS, after supporting transition from BP. At BP, his varied roles focused on competitor intelligence, chemical business strategy and planning and performance management. Contact Paul at paul.ray@icis.com