By John Richardson
WE DON’T HAVE THE DATA sets nor the data tools to work out what is going to happen next with any acceptable degree of reliability. Until or unless we develop the necessary data sets and tools we will remain, in my view, all at sea about the direction of petrochemicals demand.
The data sets we need are real-time sales figures from supermarkets, car showrooms, consumer-electronics stores and building-material merchants etc. There is a mass of data already out there, in the hands of the retailers and the wholesalers, that would be of huge use in creating better demand models.
Whether, of course, the retailers, wholesalers – and, also, the brand owners – will share their data is another question altogether. In a world driven by constant improvements in quarterly profits and competitive advantage, superior market visibility can give you a lot of leverage.
What is also unclear to whether the data tools can be built to process what would be an enormous quantity of numbers constantly flowing into petrochemical company planning departments. Major advances in digitalisation are necessary to process and make sense out of this flood of data.
As Moore’s Law slows down (it now takes closer to three years rather than the previous two years to double computing capacity on microchips), some data scientists I have spoken to wonder whether we are stretching the laws of physics to breaking point. They question whether the speed with which we can further sophisticate microchips can match the speed of our digitalisation ambitions.
And/or it could be that the huge cost of producing chips at 5 nanometres (billionths of a metre) and below will create bottlenecks in microchip development. We also need to consider, as I discussed last month, the geopolitical risks attached to either the US or China winning the 5 nanometre and below semiconductor race.
If, say, China won the race then the US and the countries in its geopolitical bloc may lack access to state-of-the-start of art microchips, the ones that are again 5 nanometres and below.
The next wave of global economic growth will likely be driven by electric vehicles, the internet of things, smart grids and digitalisation that enables to us to better monitor carbon emissions and plastic waste. If, for example, if the US bloc were the loser we would have to work out the implications for economic and so petrochemicals growth within the bloc.
You cannot run an information technology-based economy without rare earths. Geopolitics are again at play here. Last month, China said it was considering restricting exports rare earths. Whilst it controls four-fifths of the world’s rare earths refining capacity, China’s control over the supply of the rare earths themselves is far more fragile.
President Biden has made improving US semiconductor and rare earths supply one of his top priorities. But in a highly partisan political system with his power in the Senate severely restricted by the supermajority rules, it is questionable whether he can make much progress. The mid-terms are also only two years away when he could lose control of the Senate.
Democrats want to amend or even get rid of the supermajority or filibuster rules. But, as they, say, “good luck with that” as they don’t have the support of Joe Manchin, the Democratic Senator from West Virginia. In a delicately balanced Senate, his support is crucial.
A rapprochement between the US and China is possible, resulting in both sides being able to share access to sufficient semiconductors and rare earths. But such an outcome right now looks very unlikely.
What I have of course described above is geopolitics – one the four megatrends that will have a major bearing on the growth in petrochemicals demand. The other three megatrends are the pandemic, sustainability and geopolitics.
Achieving better real-time visibility on demand in the supermarkets and consumer electronics stores will only get us so far, as the megatrends are also reshaping petrochemicals consumption.
The influence of one megatrend on another is a further grey area. For example, nobody knows the outcome of the clash between sustainability and demographics.
There is a lot of talk in the rich world about “less is more” which involves redesigning food packaging to reduce the consumption of virgin polymers. This is in parallel with the rise of recycling. Bans on single-use plastics are also gathering pace in the poor as well as the rich world.
But how will the “less is more” push fit with very youthful populations in Africa and India that want the same standard of living we take granted in the West? “Less is more” and recycling also only address single-use plastics; neither deal with what is sure to be booming demand in the developing world for the polymers that go into durable goods.
The same exceptional ambiguity applies to the impact on demand of government policies. The changing role of governments in shaping demand is a sub megatrend, a consequence of the headline pandemic megatrend.
It is one thing to measure all the details of government pandemic relief measures, but another thing entirely to assess the impact on consumer spending. Did people open their wallets last year or did they save what was handed out by governments? As the cash handouts continue, how much will be saved and spent?
Assuming that there is big surge in spending, might this lead to inflation that runs out of control, thereby leading to much-higher interest rates? Might we even end up with a return to the stagflation of the 1970s?
Nobody has a clue about any of the above. But one thing we do know is that GDP estimates have less value as tools for forecasting petrochemicals demand than before the pandemic. Even before the pandemic, GDP forecasts were a very blunt tool.
If GDP was a good guide to demand petrochemicals consumption would have cratered last year. But consumption boomed in 2020 up and down most of the petrochemicals value chains.
Something else we know is that people want to be told whether to move left, right or stand still because that is how businesses run. So, the temptation is to follow those with strong convictions and loud voices who say, “This is the only scenario, this is the only possible outcome”.
But because of the data holes and the variables related to the four megatrends, be wary of those who argue for only one possible outcome.
All the above makes it critically important that you prepare a range of scenarios. Your scenarios must include the right trigger points – the signs or events which will make one scenario more likely than the others.
In a further effort to support your scenario work, let me therefore consider three scenarios – best-case, medium-case and worst-case – for global linear-low density polyethylene (LLDPE) demand in 2021-2025. This follows my earlier scenarios for global polypropylene and high-density PE consumption.
A 49m tonnes LLDPE consumption gap: best to worst-case scenarios for 2021-2025
The five years up until 2025 is probably too short a time frame for the tectonic changes detailed above to have played.
The battle between the US and China over access to semiconductors and rare earths is for instance unlikely to have been resolved by 2025. The competing forces of sustainability and demographics will also, surely, take a lot longer to significantly reshape demand.
So, let us keep these LLDPE scenarios relatively simple by focusing only on the headline megatrend of the pandemic and the sub megatrend of government policies. Relatively is a key word as even this limited number of variables leads us into great complexity and a wide range of potential outcomes.
Take as an example the issue referred to above: the extent to which government stimulus money will be spent versus being saved. This useful article from The Economist details the framework for your analysis.
The Economist analysed the difference between post-tax income and consumer spending in 21 rich countries. Had the pandemic not happened, households would probably have saved $3tr in the first nine months of 2020, said the magazine. But they instead saved $6tr.
This implies excess savings of about $3tr—a tenth of annual consumer spending in the countries. But in some of the 21 countries, savings are higher than in in others. In the US, for instance, savings may soon exceed 10% of GDP, partly because of the new $1.9tr stimulus programme.
Normally, households do not save during recessions. But as I have been arguing since April last year, this recession is different because of the developed world’s income support and job-furlough schemes. The schemes explain booming luxury-food demand amongst low-income earners resulting in strong demand for polymers.
If consumers in the 21 countries surveyed by The Economist were to spend all the $3tr of excess savings, then global GDP growth would surge to 10% in 2021! This would be boom and bust, soaring consumer spending followed by high inflation and a rise in interest rates that promptly brought the party to an end.
But most of the rise in savings in the 21 countries is amongst rich people. The wealthy tend to save rather than spend their savings.
Against this trend must be balanced the support to petrochemicals demand that will be provided by “revenge retail therapy”, “revenge travel” and “revenge dining out”. After many months in lockdowns, we could see the wealthy spend a lot more in shops, on holidays and in restaurants as a reaction to the boredom of being stuck at home.
How, though, will revenge spending compare with declines in consumption linked to being at home? As lockdowns ease there will be less demand for laptops, computer monitors, tins of paints and rolls of wallpaper. Will the net impact of fewer lockdowns be positive, neutral or negative for petrochemicals demand? Again, nobody has a clue.
In some of the countries in The Economist survey, low-income earners are not expected to have any excess savings to spend. During the pandemic, the poorest 25% of European households were half as likely to increase their savings as the richest. In Britain, the poorest one-fifth of the population reported that they save less during the pandemic than before.
If this were not complicated enough, the picture is different in the US because of the exceptionally generous nature of its stimulus. In late December, bank balances amongst the poorest Americans were 40% higher than the same in month in 2019, according to JPMorgan Chase Institute. The poorest half of the US population have seen their liquid assets rise in value by 11%, nearly twice the increase for the richest 1%.
Had enough of complexity? Sorry, it gets worse. Studies have found that people are more likely to spend an increase in income (their pay) rather than an increase in their wealth (for example, the value of their home).
In Britain, savings have largely gone up because people have spent less. This will make them less likely to spend their higher saving, according to the research. But in the US and Japan higher savings are mainly the result of government stimulus. This makes an increase in spending more likely.
I believe it is necessary to build your upside, medium and downside scenarios for petrochemicals demand around these three broad frameworks, whilst also factoring in the kind of variables I’ve just described:
Scenario 1: Petrochemicals demand booms as high levels of personal savings are unlocked in the developed world and unprecedented levels of fiscal and monetary stimulus further buoy consumption. It has been suggested that the next few years could be like the “Roaring 2020s”. During the 1920s, the global economy boomed after the Spanish Flu pandemic.
Scenario 2: The outcome is neutral because consumption was already very good in 2020. As we emerge from the pandemic, there is less demand for some of the goods we needed during lockdowns but more demand for “Old World” services such as travel. The extra demand from travel is cancelled out by the decline in demand for the goods we no longer need.
Scenario 3: Demand collapses as economies overheat and interest rates return to the levels of the 1970s. Or demand falls because job losses caused by the pandemic are not fully compensated for by stimulus.
My three scenarios for 2021-2025 global LLDPE demand are built on 2020 consumption at around 40.2m tonnes, 5% higher than demand of 38.2m tonnes in 2019.
Under Scenario 1, global demand surges by 7% in 2021 and in 2022. Demand growth averages a very healthy 5% per annum in 2021-2025. Cumulative global demand during the whole forecast period reaches 238m tonnes with operating rates averaging 91%.
Scenario 2 sees growth at 5% this year and in 2021, and at an average of 2% in 2021-2025. Cumulative demand is at 221m tonnes, 18m tonnes lower when you look at the exact numbers than under Scenario 1. Operating rates average 84%.
Catastrophe! Scenario 3 involves growth at minus 2% in 2021 and minus 5% next year. Growth averages minus 1% in 2021-2025. Cumulative demand is at 189m tonnes, 49m tonnes lower than in 2021-2025. Operating rates average just 73%.
Demand workshops are the way forward
My latest attempt at estimating global demand is of course very much back-of-the-envelope. It is for demonstration purposes only in order to provide initial support for your planning process. The same applies to all the above commentary.
The data and the commentary are there to be challenged. The process of challenge is, in my view, critical.
This needs to take place in confidential workshops where internal and external political issues must be put aside. Everybody must have an equal voice without she or he who shouts loudest, and/or who has the highest job status, leading the debate.
Because of the data gaps that I outlined at the beginning of this post – and because of today’s exceptional ambiguity – demand workshops of this nature are essential if we are going to build the scenarios we need.