Olefins II: Big Oil & China, the cycle disruptors
How the search for markets for a post-gasoline world disrupts the olefins cycle.
The Chemical industry is undergoing its longest and most vicious downcycle in recent history. Margins and utilization have rarely been so low for so long, and company prices reflect this.
At the center of this downcycle is the olefins and polyolefins industry, the largest chemical segment by far, and the heart of the petrochemicals and plastics industry. Many large chemical companies have olefin and polyolefin operations, and their economics make or break downstream industries. I provided a 15-page Intro to the industry some weeks ago, covering economics, geographies, and trends.
This article is a double-click on perhaps the two most important trends mentioned in that Intro. On the one hand, the increasing role oil majors are playing in the petrochemical industry, and on the other, the capacity boom happening in China. Both trends are very related.
5 out of the 6 largest polyolefin companies by capacity are either oil majors or controlled by them. Out of these, the two largest are Chinese.
Globally, oil majors worry that in the future, vehicle fuel demand will fall, and want to search for new markets for oil, of which the largest is clearly chemicals, with olefins being the gateway to chemicals. In no country is this more true than in China, where gasoline demand is already falling, and where the economic directions clearly point towards a ‘shift to chemicals’ (减油增化).
This is not a new situation in the oil industry. The image with which this article starts is a boat advertising the Meifu kerosene lamps (美孚灯) that Standard Oil sold in China, allegedly at a fraction of their cost. The goal of selling the lamp cheaply was to then sell kerosene (razor model). The example illustrates the early efforts by oil majors (Standard Oil, of all companies) to enlarge the demand market for oil (in China, of all places). History really rhymes.
Today, the search for oil-to-chemical markets leads majors to invest more than the economics of the industry would warrant. The Meifu lamps of yesteryear are the olefin plants of today. Contrary to the exhausted pure-play petrochemical companies, the oil majors are awash with cash and are planning to continue investing in the sector. If that is the case, investment, overcapacity, and the downcycle in petrochems might last way beyond what traditional industrial economics would predict.
Without further ado, 钻探！宝贝！钻探！
Disclaimer: The opinions expressed in the Blog are for general informational purposes only and are not intended to provide specific advice or recommendations for any individual or on any specific security or investment product.
Industry dominated by oil majors
When we observe the list of the largest olefin and polyolefin companies in the world, by nameplate capacity, the majority of them are oil majors or are owned and controlled by oil majors.
In the list below, we have Sinopec and CNPC (aka Petrochina), two state-owned Chinese majors, Exxon Mobil, from the US, SABIC, owned by Saudi Aramco, the state-owned oil monopoly from Saudi Arabia, and Borouge Group, ex-Borealis, soon to be controlled by ADNOC, the O&G state-owned operation from Abu Dhabi. Other important oil players in chemicals include Shell (UK/Netherlands), ChevronPhillips Chemical (US), and Reliance Industries (India). Formosa Petrochemical, from Taiwan, has refining but no oil extraction operations. The rest are ‘pure-play’ chemicals.
Peak light-vehicle fuel
Oil majors playing a large role in petrochems is not a new reality, especially in olefins. However, in recent years, some trends have increased their interest in the market.
First, oil producers are naturally long olefin feedstocks from the well because they produce crude oil mixed with gases (methane, ethane, propane, butane). In addition, they add long-feedstock exposure because they own refining capacity, and therefore produce a lot of naphtha, the most important global olefin feedstock.
Although feedstocks can be transported (naphtha is a liquid at ambient temperature and pressure, propane and butane at added pressure (LPGs), and ethane cryogenized only), there are advantages to connecting the oil-gas separation or refining streams directly to olefin crackers. Even when not owned by oil majors, large olefin plants are generally located close to refineries.
These factors have led to the oil players historically playing a large role in petrochem. However, for a long time (and still today), olefins and their feedstocks were unimportant, third-order considerations for oil majors, focused on the production of fuels.
As we can see below from IEA, petrochemical feedstocks make up only about 15% of oil consumption today (notice the chart’s Y axis starts at 70 million barrels per day). This does not include the large portion of petrochem feedstocks derived from gas molecules (ethane, propane, butane). The true business of oil majors is fuels. Gasoline, the most important of those fuels, and used primarily for regular cars, represents 50% of oil demand today.
The problem for oil majors is that most projections agree on peak light vehicle fuel consumption within this decade. Mind the wording there: not peak oil, not peak fuels, but peak light vehicle fuel (basically passenger cars). The driver of that trend is EVs.
In no country is this more true than in China, the largest EV market in the world. Last year, China’s light vehicle fuel consumption already fell, and it is still falling in 2025, despite the economy expanding. To aggravate the problem, China has the second-largest refinery capacity in the world. It has a lot of refineries, for which it has no (perspective of) vehicles.
Of course, we are talking of projections, which depend on a plethora of factors, all of which are highly uncertain. That is, we might never see peak vehicle fuels or peak oil. We don’t know really, but the projections above are what companies are using to make capital allocation decisions.
The quest for oil to chemical refining
In this context, petrochem feedstocks become one of the very few potential drivers of oil consumption growth ahead.
There is no peak projection for petrochemical consumption. Petrochems are a recurrent defensive consumption in developed economies (all kinds of disposables, plastics, coatings, and industrial processes), and their consumption grows almost linearly with GDP in developing economies (urbanization, construction, growth of the middle class, etc). Further, new materials and applications are introduced regularly (healthcare, renewables and electrification, performance materials, etc.).
Of course, today, chemicals can only accommodate so much oil production. It represents only 15% of oil consumption today. Still, the math adds up for it becoming much larger over time.
A ballpark estimate of all current petrochem feedstock consumption (olefins + aromatics) would be around 500 million tons per year. Part of this is produced from gas molecules, not oil. However, assuming a metric ton of naphtha becomes 800 kg of feedstocks (simplification), and that naphtha fully displaces gas feedstocks, then today all petrochems would represent about 1 billion metric tons of naphtha, or basically 7.3 billion barrels of oil per year (~20 million daily, or ~20% of current production) if all oil could be converted to naphtha, which as we will see below, is not the case.
However, as seen above, we also need to account for petrochem growth. Assuming the population continues to grow and it continues to get rich (biggest driver of petrochem consumption growth is wealth), and petrochems consumption grows at a 5% growth rate over 25 years, then by 2050 we would need 67 million daily barrels for chemicals, or ~67% of today’s production (again, assuming you can turn one ton of oil into one ton of naphtha). Not bad, really.
But there’s a big catch. In current refinery configurations, depending on the type of oil, only between 5/15% of a barrel becomes chemical feedstocks. That means we cannot serve all the chemical demand from oil today. At a 10% yield, you would need 200 million barrels instead of 20 million, or 2x the current production. Further, we would still have a lot of useless derivatives (gasoline mainly). However, these yields can be changed over time, with the right refinery-olefin cracker configuration and technologies. This last point is where oil-to-chemicals refining technologies come to the fore.
The ‘simplest’ solutions require retrofitting, replacing refining cracker equipment or catalysts in order to yield more LPGs and naphtha and less gasoline. Integrated plants can be designed to use that naphtha directly as feedstock. In addition, today, some naphtha goes into gasoline as reformate, and instead, this could be used as feedstock too. Some claim that with retrofitting and reconfiguration alone, you can get to 40% petrochem feedstock yields, and some integrated refineries (mostly Chinese locations like Shandong Yulong, or Hengli in Dalian) already claim these figures.
Then there are more experimental technologies that have not been as tested commercially, but that will probably have large world-scale examples in 5/10 years. With direct oil cracking (Sinopec, SABIC), there’s a promise of 70%+ petrochem feedstock yields.
That is, assuming population and wealth growth over the next three decades, plus the continuation of already fairly established technological changes, we could see a world where most oil goes to chemicals before 2050, and O&G majors want a part of that world.
Economic and institutional drivers of the ‘move to chemicals’
In their preference for concise, four-character, slogan-like parlance, the Chinese have characterized the chemicals strategy as 减油增化 (reduce fuels, add chemicals) and 减油增特 (reduce fuels, add specialties). I believe this is applicable to players from all over the world.
We know the expectation is for fuel-driven demand for oil to decline, and for a large part of growth to be concentrated on petrochem feedstocks, but that does not completely explain the investment in olefins, and especially downstream into polyolefins. At the end, it is not like O&G companies were manufacturing cars in the 20th century. They could simply sell oil or feedstocks to chemical manufacturers.
Does adding chemical capacity make economic sense for an O&G company? And if not, then why is it happening?
Economic drivers: marginal, sunk, and accounting costs
Let us first analyze the situation from a purely economic standpoint. That is, does investing in downstream chemicals add value to the aggregate oil+refinery portfolio of oil majors? My short answer after some thought is not really, except for two edge cases.
Every O&G asset (basins or refineries) has a breakeven cash cost level. Producing above the breakeven cash cost is economically sound if capital is already deployed in an asset.
For example, if Exxon has an offshore rig outside of Guyana with an all-in operational cost of $35/barrel (including maintenance CAPEX, plus all overhead functions, but no structural CAPEX nor returns on capital), then producing with oil above that price generates a cash-profit to the company, and should be carried out. If the price is permanently (or expected to remain) below $35, then Exxon loses money by producing and should simply abandon the operation.
A similar example can be applied for a refinery, but instead of the price of oil, we need to talk about refinery margins, which in turn depend on refinery capacity versus demand.
This calculation does not change because of downstream integration with chemicals, assuming a market for oil and feedstocks continues to exist (more on this last point below). If a company only has oil fields and no refinery, then it is economical to sell to the market above breakeven, and cut production if not. If the company has a refinery and uses its own oil below breakeven instead of buying from the market, it is losing money in comparative terms (and it is indeed decreasing the competitiveness of the refinery). In a similar vein, if a company has an olefin/polyolefin asset, then it pays to use its own feedstocks so long as they can be obtained at a similar or lower cost than in the market. If the breakeven costs of an asset (field or refinery) are above the (long-term expected) market’s price for the products of that asset, it is uneconomical to continue producing, independently of downstream integration.
The capital costs for exploring or acquiring the resource and building the asset are not economically relevant because they are sunk. They cannot be recovered by producing below breakeven (in fact, you lose more capital in relative terms), and they are not ‘lost’ if the plant is shut, because they were already ‘lost’ (or consumed) in the past.
What can change is the accounting cost of the asset, because a shutdown would trigger an impairment. By integrating with downstream operations, one might shift the profitability of one operation (say, chemical) upstream, and therefore subsidize the uneconomical operation of upstream assets. However, an impairment is only the accounting recognition of an economic reality, and you can ‘cook the books’ to avoid it, but that doesn’t change the reality.
A similar argument for integration, but that is also uneconomical, is diversification. Some companies might argue that their chemical or refinery divisions can sustain the oil production when it is not doing well in a cyclical fashion (that is, not permanently as considered above), and therefore smooth out returns over the different cycles. This argument does not make sense because investors can do their own diversification. Participating in more than one industry only makes economic sense for a company if it can gain competitive advantages by integrating, or if the managers are good capital allocators (in which case part of the work of the investor shifts to the managers of the company).
In any case, investing in chemical assets (or in chemical-oriented feedstock refineries) should be evaluated independently for capital allocation, as a segmented business. They cannot improve the economics of upstream units.
Edge case 1: small cost of retrofitting
For the specific case of refineries, if the cost of reconfiguring the operation to produce more feedstocks or to incorporate more downstream chemicals were low, then doing so may be economical. In this case, a small capital addition could leverage the previous sunken cost on the refinery assets. However, from the examples we will see below, this does not seem to be the case in most refineries, where a major overhaul is needed.
Edge case 2: market dislocation
Another case where downstream integration can protect from the loss of an otherwise economically viable asset is a scenario in which most oil production is integrated and there’s little oil (or refined products) trade.
In this scenario, oil demand decreases so much, and integration in the oil value chain increases so much, that companies decide to use their own oil, and therefore some unintegrated producers with cost-competitive products are left without a market.
It is hard to think of this scenario in purely economic terms, because companies should always be willing to trade for oil or feedstocks if someone is offering them at a discount to their own marginal costs. However, if, for the institutional reasons considered below, oil buyers prefer to use their own oil, then integration might be the only way to make use of cheap reserves.
Although this is an edge case in theoretical terms, its probabilities are not small (above all, in China).
Institutional drivers: strategic industries, and company survival
Given that it doesn’t make a lot of economic sense to build chemical operations downstream, but most oil companies are doing it, I think institutional factors explain the situation better.
At the company level, there is a big vested interest in keeping the company operating at a larger scale by subsidizing the oil and refining operations with chemicals. By shifting from an oil-producing company with chemical operations to a chemical company with oil operations, companies guarantee their own survival.
That is, an oil company might make more money for shareholders by producing until it becomes uneconomical, and then winding up, than by elongating its life via chemicals, but that is not in the interest of employees or managers. This includes employees, managers, and even the boards if they are not thinking from a pure shareholder view (now completely allowed under the stakeholder framework). There’s also a big interest in avoiding large accounting impairments on the part of managers and bonus-making employees.
In addition, there is a life-extension argument, because companies that are expected to wind up in the near term might have trouble attracting resources, especially talent and capital. For example, not a lot of people would want to work for a company that expects to wind up in 2035, and lenders are very unlikely to lend to this company into 2045.
A country-level factor, related to the market dislocation case above, is that countries might prefer to keep their (relatively more expensive) oil and refining operations running instead of buying from the market at lower prices, for strategic or political reasons. In this context, adding chemicals is a way to keep assets working. This can explain the behavior of state-owned oil companies, like Sinopec, PetroChina, Sabic-Aramco, or ADNOC. It definitely explains the Chinese strategy.
The situation in China
From the perspective of China, I believe the institutional driver for oil-to-chems is particularly evident. I’m not telling anything new, given that China applies industrial policy extensively. China has stated its goals to increase chemical self-sufficiency while shifting its refinery capacities to chemical feedstocks.
The country has already set max refinery capacity at 1 billion tons (~20 million barrels/day). This capacity is close to or has already been reached, but the country is still allowing large, chemical-integrated new or chemical-retrofitting projects to be constructed. Eventually, this growth will need to be offset by smaller refinery closures.
In the first table below, we can observe current new or expansion projects in China that have integrated 10M+ ton refineries, plus 1M+ tons of ethylene and downstream operations. They will eventually represent 47 million tons of ethylene capacity, or 60% of the expected 2030 capacity. Further, they represent 20 million tons in ethylene capacity expansion versus the current configuration, or 2/3 of the expected capacity expansion between 2023 and 2030.
The message seems to be ‘if you want to refine in China, you need to generate your own refinery demand, in chemicals’.
Of course, the next problem is that, by addressing the future gasoline/refinery glut, China has created or exacerbated both its chemical and refinery glut.
In refining, it continues to add capacity despite having reached its own established limit and demand already falling in some fuels. Refining capacity utilization is 70% or less, and in ethylene, capacity utilization might be only 60%. In all sorts of downstream markets (PVC, EPS, PO, EO, SM, ABS), capacities are 30/50% above domestic demand. In polyolefins, the less oversupplied market, demand was still only 80% of supply in the most commoditized lines in 2024. Further, as seen partially above, capacity additions will continue. In ethylene, capacity will expand by 50% from 2023 levels by 2030.
That has obviously led to thin or negative margins. The CCPI (China Chemicals Products Price Index) is currently at 3,800. A level not seen since the worst of the pandemic, and before that, since the 2015 market crash. Both Sinopec and PetroChina have negative chemical+refining margins.
The strategy, as I see it, is to deal with this in stages, by removing older capacity and waiting for end-market demand to grow organically.
The country is already closing older refinery capacity (non-integrated, below 2 million tons), plus announcements around considering equipment up to 20 years (30% of refining capacity) to be too old for operations. In chemicals, the comments are less stringent (‘rationally determine’ instead of ‘strictly control’), but given the degree of the glut, I would not be surprised by more aggressive measures. The calls for a specific Chinese government document suddenly triggering a wave of closures seem to repeat every few years, though, so I would not jump on that bandwagon fast.
In addition, the country seems to be solving a glut upstream by continuing to push it downstream. Some oversupplied products were already mentioned (EO, PVC, SM, ABS, etc.). I do not think the country will keep adding capacity here, given that the market growth is slower.
Instead, China wants to expand capacity in other, more advanced markets (high-value polyolefins, polyolefin elastomers, EVA, UHMWPE, and others). Here, the market is expected to grow much faster, and the most complex value chains play well with the overall goal of increasing value added, moving value chains higher in the pecking order, adding R&D and IP to value, etc.
The degree of additions is large here as well. From one account, in elastomers (PEO) alone, the country’s projects (for 3 million tons in yearly capacity) already exceed current demand by 3x. This means, eventually we will also have a glut in polyolefin specialty markets.
Capital allocation implications
In such a globalized environment as olefins, capital allocation decisions in one place affect the whole world. This is particularly true for one of the (if not the) largest chemical consumers and producers, like China.
When capacity increases in China, it directly competes with imports, that is, mainly with export-oriented assets in North America and the Middle East, and to some degree with export-competitive specialty production in Europe (probably to worsen as China pushes for specialty self-sufficiency as well).
Chinese capacity is not competitive with most exports, so eventually, capacity expansions inside China should lead to capacity reductions in China as well, in less competitive production (non-integrated, older, smaller). The problem is the speed of the adjustment, which is unlikely to be swift enough given the magnitude of the additions expected.
While uncompetitive production remains, it depresses margins for everyone, and it pushes China-focused imports towards other markets (mainly Europe, North East Asia, and Emerging).
At the same time, China will remain the largest chemical (and therefore oil) demand growth market, given that the scale of South East Asia is not as large, and that India is still well behind in terms of plastics consumption. That is why global oil majors continue to invest there to guarantee access.
We can also observe in the table below that most of the projects of the large oil majors are in China. This ties the two trends together (China capacity additions and oil-to-chems).
My conclusion is bearish for the olefins cycle, in particular for export-oriented locations (NA, ME), and wary for export-oriented specialties. The cycle will eventually turn, but we are not close to the bottom in terms of time, in my opinion.
Idk why your posts don’t get more attention. Truly a diamond in the rough.
Well written on China. It’s worth adding that the ME guys are also sweating over the decline in oil-as-a-fuel demand, and therefore also desperately trying to move downstream into petchem. The biggest loser(s) have been the European capacity which is old and bereft of cheap Russian feedstock is falling like flies.