By John Richardson
MONITORING demand has never been harder because of the pandemic. One of my colleagues on the ICIS Pricing team was, for example, last week told by one of his reliable contacts that polyethylene (PE) demand was good, bad or indifferent. It used to be that if you talked to the people you trusted you would get a consistent view of what was happening.
It is impossible to tell whether PE demand – or indeed demand for most other petrochemicals and polymers – is strong, weak or indifferent.
What is crystal clear to me, however, is that using multiples over GDP growth to estimate demand is even less reliable than before. This has long been a flawed way of assessing consumption. But the pandemic has blown what I believe is even a larger hole in the credibility of this approach. The hole could be irreparable. Let me explain why.
GDP growth collapsed to historic lows in Q2. In the US, for instance, GDP fell by no less than 32.9% on a year-on-year basis. The EU was nowhere near as bad as growth shrank by 11.9%, again year-on-year, but this was still a historic decline.
So, of course, Q3 will be better as lockdowns have eased and we should see some pretty astounding quarter-on-quarter recoveries in headline GDP, as well as all the composite numbers on retail sales, imports and exports that go into headline GDP.
Under the old approach, one might be tempted to assume major declines in petrochemicals demand in Q2, based on the historic collapses in GDP. You may then assume equally strong rebounds during the third quarter.
Or you could decide to write 2020 off as an awful year because of the knock-on damage to the whole year from the disastrous second quarter. You might then predict a very strong recovery in 2021, based a rebound in full-year 2021 GDP. You may instead assume a more gradual return to demand levels from before the crisis by spreading the full recovery in GDP – and therefore petrochemicals demand – over 2021 and 2022.
Any of these approaches would give you V-shaped recoveries similar to the patterns we saw after the Global Financial Crisis. But I don’t believe these approaches will give you a realistic picture of what’s going to happen.
To begin with, even if GDP does recover from the low points of the crisis – a highly likely outcome – it will still, in my view remain a lot smaller than before the pandemic began. I fear it will take several years to find, produce and globally distribute a vaccine. China rode the rescue of the global economy in 2009, but is unable to do so again.
End-use applications for polymers that go into single-use packaging haven’t declined in line with the collapses in GDP. As I’ve been discussing since March, no-durable applications have been doing well because of lockdowns that have led to stocking up on food and greater hygiene needs – e.g. more demand for detergent bottles made from high-density polyethylene (HDPE) or polypropylene (PP).
Problems caused by government money
And, of course, non-woven grades of PP have done extraordinarily well because of the demand for face masks. Demand for most grades of PE, most grades of polyethylene terephthalate (PET) and some grades of PP and polystyrene (PS) have benefited.
A further boost to packaging demand may have also come from people eating at home as restaurants shut down. Think of the surface areas of large tubs of ice cream used in a restaurant and then think of the additional surface areas of smaller tubs of ice cream we buy for the home.
Next think of the unprecedented amount of government money that has been spent in Western countries to support lower-wage workers. In the US, EU and Australia, lower-paid workers are better off than they were before the crisis.
This has led to a very different dynamic playing out in supermarkets than what was seen during the Global Financial Crisis.
Back then, an acute shortage of credit across the West and the surge in foreclosures in the US hit lower-wage earners very hard. They were forced to cut back on spending on basic groceries. As one supermarket executive told me, “We saw more purchases of ‘own brand’ or ‘white label’ tins of beans and less of bars of chocolate, ice cream and cakes.”
On this occasion, though, government money has led to an increase in spending on luxuries such as singles buying family bars of chocolate, according to some of the fast-moving consumer goods companies I’ve spoken to.
When we do a see a strong rebound in GDP growth, these complications mean that there may not be a corresponding strong rebound in demand for the polymers that go into packaging. We might instead see demand for polymers that go into packaging decline rather than increase as GDP recovers because of the reduction in government stimulus programmes.
This might already be happening in the US due to the failure of Congress to agree on a new stimulus package. President Trump’s executive orders on stimulus have also fallen well short of fully replacing all the cancelled government cheques.
Can Western governments continue to afford so much largesse as debt levels soar?
We must also consider the distorting effects of stimulus even where governments can afford to keep the tap fully open.
In Australia, which has relatively low levels of government debt, there is reported to be a shortage of bricklayers because bricklayers are receiving more money to stay at home, via government cheques, rather than work. Some people believe this is why another stimulus scheme designed to boost construction of new homes is struggling to get off the ground
Such distortions may force governments to adjust stimulus programmes which reduce spending by lower-paid workers.
This only scratches the surface of the complexity
All of this only scratches the surface of the complexity out there. It is not just the case of non-durable end-use applications for petrochemicals enjoying strong demand versus weak demand for durable applications, which, based on my discussions with the market, was my first theory back in March.
What we must also consider is staying at home versus moving around. Many of us have spent much more time at home because of the lockdowns. And even now that lockdowns are easing, working from home remains more prevalent than working from offices for health, cost and efficiency reasons. There is also the big collapse in travel for leisure and business.
This explains why it is not just polymers that go into packaging that have benefited from the lockdown. So, too, have sales of office furniture and personal computers as people turn their homes into offices. Household appliance sales have also done well.
This means that declines in demand for petrochemicals that go into office furniture, such as polyurethane (PU) foams, may have not been as bad as the falls in GDP. The opposite may, in fact have happened – demand could be stronger than before the crisis.
Drawing a V-shaped pattern of recovery for flexible PU foams, made from toluene diisocyanate (TDI), therefore doesn’t much sense to me.
What happens next for the sales of office furniture and personal computers? As GDP growth recovers, we may experience a fall in demand as we reach the point where most people have equipped their homes to function as offices.
And against every extra office chair that has been sold to turn homes into offices, we must balance out the loss of sales of chairs to new restaurants starting up or expanding. Think also of the loss of demand for office furniture in all the corporate head offices that had planned to expand, but are instead downsizing, quite probably permanently.
Permanent changes to demand are another layer of complexity. This is turning out to be like making a lasagne where you add layer upon layer of cheese (in my opinion, the more layers of cheese, the better the lasagne!).
Another layer of cheese includes the permanent collapse of demand for commercial real estate, which is bad news for all the construction petrochemicals that go into this sector.
Then consider all the losses of demand from the closure of restaurants, cafes and bars that used to serve busy city-centre offices. This will amount to less need for all the fixtures and fittings that go into food and beverage outlets, along with lower demand for packaging materials for food and drinks sold by these businesses.
But because many more people will permanently work from home, will the loss of demand at retail outlets in city centres be cancelled out by better sales in restaurants, cafes and bars in the suburbs? In my home city of Perth, Western Australia, suburban cafes etc are booming as home-bound workers pop out for lunch as a break from physical social isolation.
An unmitigated downside is the loss of demand for all the petrochemicals that go into transportation. I believe we are likely to never fly as much again because of health, cost and efficiency reasons. Travel to and from work will be greatly reduced, lowering the demand for automobiles and for public transport.
The developing world and China
What will the permanent changes in our lifestyles wrought by the pandemic mean for the clothing industry?
Take away all the demand for clothes we used to buy for the office and consumption of polyester fibres could be a lot lower for many years to come, if not permanently lower. On the subject of synthetic fibres, how will the fall in demand for office space affect demand for nylons and PP that go into making carpets?
Some economies will do better than others. What I believe will be a permanent decline in tourism will significantly damage tourism-dependent economies such as Spain, Greece and Thailand in every industrial and service sector in these countries – and so in every petrochemical.
Every extra PET-made mineral water bottle sold in the UK because people are not travelling could mean one or more bottles not sold in Spain and Thailand.
We must also not forget the split between the developed and the developing world. There will be a severe effect on the developing world because of the rise of extreme poverty, even in demand for the most basic of consumer non-durables.
Anybody who suddenly finds themselves earning less than $1.90 a day, the definition of extreme poverty, won’t even be able to afford a bottle of shampoo – even a supermarket own-brand label.
Using GDP growth to measure demand growth in the developing world has always had its limitations because of the role that income distribution plays in petrochemicals demand. We have seen situations where headline GDP growth has been weak, but there has been big increases in the number of people escaping extreme poverty and thus being able to afford modern-day consumer goods for the first time.
The effect of this crisis on the developing world must not be underestimated as it is the developing world that is the primary driver of global incremental demand growth. The IMF warns that poverty reduction could be set back a decade in low-income developing countries as a result of the collapse of remittances and containment measures which forced the closures of businesses and schools.
When you are living on the margin, with very little or no social safety net, you have to work every day to feed yourself and your family. More than 70% of people in an IMF survey across 20 African countries said they were at risk of running short of food.
And finally, last but far from least is China, the country that is a region by itself and is by far the most important petrochemicals market in the world.
We have no idea whether China has enough financial flexibility to close the gap between its lower-paid workers and its rich minority over the next 12-18 months, which has been widened by the pandemic. If it fails to close the gap, retail sales growth will remain depressed.
The long-term challenge relates to the success of the 14th Five-Year Plan. Can China make its economy much more consumption and less investment driven? Success is far from guaranteed, as I shall discuss in a later post. Failure would mean lower consumption growth than would be implied by China’s official GDP growth numbers, the accuracy of which has long been questioned.
No room for the lazy
Petrochemicals customers must, as a result, spend a lot more time talking to their customers and their customer’s customers, all the way down to the final consumer, a low-paid worker who can no longer be able to afford family-sized bars of chocolate.
For many years we have, quite frankly, got away with being lazy as petrochemicals demand has increased and fallen with GDP to a degree of accuracy that most people have found acceptable.
What has increased the cosiness of the comfort zones has been the reliability of economic cycles. Up until now, downturns have been very brief, the Global Financial Crisis being the most recent example. Demand lost in one year was usually made up in the next one year – with a tolerable degree of accuracy.
But the pandemic, along with the longer running changes resulting from ageing populations and sustainability, has left no space for the lazy. We instead need to understand intimately what is happening in the supermarkets, the car showrooms, the real-estate sales offices and the restaurants and cafes. We need to walk away from our spreadsheets and talk to people.
We then of course need to build new models. We need to create demand forecasting models that work better than just relying on multiples of growth over GDP.
And, as the chart at the beginning of this blog post illustrates, we need to then work out the implications for each end-use application (here I have just PE and PP as examples). Let’s think of this as a traffic light system:
- We know for certain that, for instance, impact copolymer grades of PP will be worse off than before because of the pandemic. This is why it is in red at the right-hand side as a definitive loser because of what I believe will be a permanent decline in auto sales.
- But the other grades will be amber (neutral, neither forward nor back), green (a go for more investment and growth) or red, indicating a decline. Think here of linear low-density films. Will post-pandemic demand be greater because of working from home, greater concerns over hygiene and less sensitivity over plastics pollution? Or will demand be weaker due to the problems in tourism-dependent economies, commercial real estate and developing economies?
This is a very long preamble to a free advert: We are working on a new bottom-up approach to demand, the details of which I will provide in a later post.
The first iteration of this model is certainly not going to win the gold medal and so we will work with our customers to continuously improve our approach, One of the great things about ICIS is that we conduct literally tens of thousands of market engagements every week which we will use to help build this new model.
Petrochemicals producers must, as I said, also themselves spend more time talking to people all the way downstream to the struggling working parent making the hard choice of whether to buy a new washing machine or leave it on the shelf. New demand teams will need to be set up.