COVID 19 – How Will Global Manufacturing Output be Impacted?
This long-form article was written by our CEO and Senior Research Director Adrian Lloyd to give our clients, and any interested readers, the latest findings on the COVID-19 crisis from our analyst teams around the world | 11 min read
When clouds appear, wise men put on their cloaks;
When great leaves fall, the winter is at hand;
When the sun sets, who doth not look for night?
Untimely storms make men expect a dearth.
William Shakespeare; Richard III
I write from the comfort of my home, where I am in self-isolation. A family member recently said to me “what strange times we are living through”. We are indeed. My wife and I went for a walk the other day with our children, and an elderly couple politely acknowledged us while going to great lengths to keep their distance as we passed on the road…
Such strange times create fear and uncertainty – ultimately storms make men expect a dearth. With fear comes panic, overreaction and distrust. The 24-hour news cycle feeds on negative stories. They make the headlines and they dominate our reading, our viewing and our thinking. When a phenomenon of such magnitude as the Coronavirus epidemic emerges, our focus shifts almost exclusively to the daily news cycle. We become enthralled by it. We feed off it.
As technology analysts, we are tasked with interpreting this chaos and making some sense of it. We have the difficult function of making forecasts for the markets we cover, and in these times of uncertainty our job becomes particularly challenging. Our steady-state economic assumptions lose relevance as we stand on shifting sands, and we’re forced to concentrate on various short-term indicators and anecdotal insights to try to make sense of what we are hearing. One important role I take on during this period is to create a series of assumptions that we use internally to inform our thinking. It’s ineffective and inconsistent to rely on our analyst teams to do this in isolation, so I do my best to build up a sound position and we work from this. I rely on inputs from clients, our team, our partners (with a big thank-you to ITR Economics of course) and a whole host of secondary data to get a clear perspective. This insight is written with the intent of sharing these assumptions, which we hope helps explain some of our forecasting positions right now, but more importantly is useful intelligence for our clients and readership. I hope you find this content interesting and helpful.
Note that we tend to focus on the more advanced economies of the world, and these assumptions may be incorrect for developing regions that we are not studying as closely.
How will the Spread of Coronavirus play-out in the months ahead?
Our current view is that the number of infections we are seeing as a result of the Coronavirus will continue to accelerate over the next two months, and will start to be contained in regions that have been applying more stringent restrictions first, and then more widespread a further 2-3 months afterwards. By mid-August we predict that the fundamental problem of exponential infection growth will be resolved, and the number of new cases will be very low or near zero amongst developed economies.
This is not to say that by August, we can expect normality, as the process of preventing new infections and working this through the general population will be hugely disruptive for the vast majority of businesses, and will inevitably lead to a general recession in the global economy. I’ll discuss this impact later, but for now I’d like to visit the key points that drive our assumptions around timing.
It seems that China, Taiwan and Japan have all been effective at halting the spread of the virus. What has happened in these regions in recent weeks has helped navigate our thinking.
- China can be thought of as two separate regions: Hubei (the epicenter of the virus) and the rest of China. As confirmed by our analyst team in Shanghai, life in China for the most part is returning to normal. People are returning to work; indeed, President Xi has asked companies to reopen provided there is no evidence of danger from the Coronavirus. Ports are starting to return to a semblance of normality, and we estimate that about 60-70% of manufacturing capacity is operating again. The lockdown of the Hubei region, where most cases were centered, proved highly effective at limiting infections into the broader China population. The good news for the Hubei region is that, at the time of writing, more than four consecutive days of no new cases was achieved, suggesting the spread of infections is under control. Wuhan, where the virus is believed to have first occurred, is still being treated differently to the rest of the region. Last week, people were allowed to start moving around again, but the resumption of working life has not commenced yet and our projection is that it will take at least another month before the policies limiting working in the region are relaxed.
- Taiwan was incredibly proactive at responding to the news coming out China. As early as 31st December, measures were put in place to inspect flights coming in from Wuhan. On the 20th January, the government commenced a series of initiatives to protect against the virus, including provisioning of face masks, leading to the February closure of schools and mobilization of armed forces. Given the geographical proximity of Taiwan to the virus epicenter, and the limited information on hand at the time, the policy has been very effective at minimizing the number of infections. At the time of writing only 153 cases had been reported in the region; compare this with China having over 81 thousand. Taiwan is still reporting new cases, but the rate is very low compared with regions that have not effectively managed the risk from the start.
- Japan, much like Taiwan, was stricken with infected people early in the crisis due to the close business ties between Japan and Wuhan province, for example for Japanese automakers. Travel from Wuhan saw both Chinese and Japanese nationals seeding the virus in Japan. President Abe’s administration was slower to respond than Taiwan’s, which explains why the infection rate in Japan is materially higher than that of Taiwan. But the response, compared to many other regions around the world, has been robust. Many initiatives were enacted during February, and by the end of the month all schools had been closed. Given how early on Japan was affected by the crisis, its mitigation strategies have contributed to limiting infections to 1046 cases at the time of writing.
The critical point here is that once a series of rigorous containment strategies is implemented, the uptake of new cases can be slowed and reduced over time. China appears to have almost entirely halted new cases being derived from within its own population. It is thought that new recorded infections in the region are linked to repatriating Chinese nationals who caught the virus while traveling overseas. The fact that Wuhan is down to zero new cases for multiple consecutive days is a massive signal to the world that rigorous containment efforts work. The question is how quickly these efforts can be enacted elsewhere and to what extent. As the body of evidence continues to build that regions that are most proactive recover more quickly, there will be an inevitability that others will act accordingly.
When regions are proactive, we see that the timeline is about 8-10 weeks to halt the exponential growth of new cases. So, were a given country to implement aggressive action today, we could expect the spread of the virus to be contained by the end of May. The timing for Wuhan provides this line of thinking – the first action began at the end of December (suspension of passenger trains to and between Wuhan and Henan Province), although it wasn’t until 29th January for all Hubei cities to be placed in quarantine. China’s response to the crisis was criticized as slow initially, but despite this in less than two months the spread of the virus appears to have been controlled.
Our assumptions around timing also factor in that a quarantine period of 14 days ensures that a person or family no longer poses a risk of infection to the general population. Also, while noting that while there is an increasing body of evidence that certain anti-viral medications that are developed for other conditions, notably Hydroxychloroquine or a Hydroxychloroquine and Azithromycin combination, have shown to be effective at reducing the infection rates, we don’t believe that these will be sufficiently tested or made available in time to reduce infection rates within the period we are discussing here.
One final point on timing, and this is perhaps the largest risk to the forecast, is that we are assuming that once infection rates are reduced and controlled, this will remain the status quo. There is a significant concern that once containment strategies have been relaxed, and the virus has not been fully eliminated from a given population, it may start to resurface, and exponential infection rates could return. China is an important test case for this factor, as it is the first region to effectively drive new cases down to close to zero and is now relaxing these strategies. One could argue this is an optimistic position, but I like to think that following the severity of what we are going through people will remain vigilant.
What is the Status of Manufacturing Operations in China and to What Extent Will This Impact the Supply-side Equation?
China accounts for nearly 50% of global production, and with global supply chains so integrated, a major concern among manufacturing companies is that the supply of critical components required to create finished goods and equipment has been compromised. When one looks at the recently released industrial production numbers, with the region down 13.5% in January/February versus the same period last year, clearly there has been some disruption. So, the question is how long were manufacturing lines compromised? And how and when will this disruption manifest itself in the global supply lines?
I spoke to our Shanghai-based analyst Jan Zhang on this topic, and she shared some useful perspective on where the Chinese manufacturing economy is, and what we can expect in the weeks ahead. She made several key points that I wanted to highlight.
- At the time of writing, the vast majority of workers, Jan estimates as high as 80-90%, have returned to their place of work. There is no official data to point to that supports this, so this number is an estimate derived from what she is hearing and reading each day, as well as what she is experiencing through her own daily routine. With the exception of workers in the Hubei province, Jan expects by the end of the month all other provinces in China will have workers back to work as normal. There has been some reticence among workers to make this return, judging from social media content, with many concerned that there is a risk the virus could reemerge. But this seems to be a risk the Chinese government is willing to take, which now is clearly moving from containment mode to restoring the economy as best and quickly as possible.
- The wheels of production are moving slightly more slowly, as it takes a short but additional period of time for plants to reach output capacity, following a lengthy shut-down, once workers return. Again, there is nothing official one can reference right now, so we are relying on Jan’s estimate that by the end of last week (20th March 2020) machines were up and running by around 50-60% of usual capacity. She anticipates that by the end of this week it will be much higher. Hubei caveat aside, we are assuming that by the beginning of April, production capability will have been fully restored, and plants will be working particularly hard to recover lost time in meeting pent-up demand.
- Considering Hubei, our general assumption here is that the region will take an additional 4 weeks to return to normal. Slowly we are seeing officials relaxing various containment measures. For example, towards the end of last week, traffic checkpoints throughout Hubei began being removed, except for those limiting travel in and out of Wuhan. More recently, the lockdown appears to be loosening with people now starting to be allowed to move around within the region. We expect more caution within Hubei given its status as the epicenter of the virus, hence our expectation that it won’t be until the end of April before plants have got back to anything resembling normal operations. The good news here is that Hubei accounts for only 4% of manufacturing output, so its impact on the larger Chinese manufacturing economy is quite limited. Having said that, Hubei is one of four major manufacturing hubs for automotive in China and also has a strong electronics center, so these two industries will feel more impact than others.
- One of the most interesting points Jan shared, pertains to the timing of the Coronavirus outbreak. She argues that it was fortuitous that the broader shut-down of China to reduce infection spread coincided with the region’s Spring Festival and the Chinese New Year. This is a two-week holiday period akin to Christmas and New Year where businesses slow-down significantly, and workers are given time off to return home and spend time with the family. Given China’s major importance to the global supply chain, producers in the region are mindful that during this period they don’t want to cause disruptions. They therefore plan ahead. Inventories are built up in advance to the shut-down. Since this two-week period of planned inventory build-up took place in advance to the broader Coronavirus shutdown, the overall impact on supply chains will be reduced by 2 weeks. So, since it took eight weeks to return production output to normality (ignoring Hubei), we can assume only 6 weeks of supply chain disruption, but only from an inventory perspective (i.e. there are still logistical challenges that need to be resolved).
- As an interesting aside, assuming Jan’s assessment of 6 weeks of supply chain disruption is accurate, it would suggest we face the situation that certain overseas producers will, at some point in the next few months, feel the pain from this disruption and not be able to finish a given product due to parts shortages. All things being even, this could have a knock-on effect on lead times of up to 6 weeks for affected companies. However, and it’s a big “however”, this assumes that demand for these parts will continue to behave normally. This topic has come up during my own client conversations, although it occurred to me that this may have been over the last month or so, ever since we’ve heard about the shut-down. What doesn’t seem to have been considered up to this point, perhaps because people are still in denial about the extent to which the Coronavirus will impact their lives and businesses, is that in all likelihood their businesses will also experience periods of lost production that are synergistic with those experienced in China. For example, Italy announced at the beginning of this week that non-critical plants will have to close for a minimum of 2 weeks. Is it the case that the supply-side problem that seems to be a key focus for many people right now may be minimized as the disruption is universal on both sides of the supply chain?
- The final perspective from Jan relates to the logistical challenges still being faced within China. While many workers have returned, and plants will soon be fully up and running, there remain difficulties surrounding the movement of goods and these channels are operating below normal levels. The ports are a key supply-chain infrastructure and are rapidly being brought back on-line but have some way to go. This is a topic where our intelligence is still somewhat limited, and we are investing some additional resources to try and better understand this topic. We will report back through future written insights.
Our assumptions around both the timing of the Coronavirus’ impact and the regional severity of it are particularly important in providing a baseline for our manufacturing economic projections. What we are starting to see is that regions that are slow to cope with the spread of the virus are more significantly impacted, and may need to take drastic measures such as home quarantining people and closing down all businesses that are deemed non-essential. Those regions that are more successful at viral control are also feeling some economic hardship, but the impact will be much less as the general economy can continue to work.
Also, the length of the crises in each country will dictate the speed of recovery for its corresponding economy. The greater the disruption, the more difficult the recovery, and vice versa. Our position of an 8-10-week disruption per region (measured from the point when a de-facto country-wide lock down is brought into force) is sufficiently short that most developed economies should be able to carry this impact without a significant meltdown in core economic infrastructures, such as the financial markets. This ability to restore the economy to a semblance of normality is critical to minimizing the long-term impact of the Coronavirus epidemic and is another important assumption we are making in our own forecasts.
Leading into the start of 2020, one can reflect on a global economy that had slowed in many regions during the 2018/2019 period, although there was a sense that most regions had turned a corner and that the outlook for 2020 and 2021 was quite favorable. Our partners at ITR Economics had been predicting this behavior for a while, and this was reflected in our own manufacturing output forecasts. I was speaking to an Italian client recently who stated that prior to the virus taking hold in their region, their first quarter order book had been as strong as they’d seen in a long time. Italy’s manufacturing economy had contracted -3.2% and 2.5% in 2018 and 2019 respectively; so here was an economy that had been through hardship and, as expected, was seeing positive signs of recovery.
The fact that these last two years were challenging for many economies and we were turning a corner means that there were solid economic fundamentals in place at the start of the year. The US is arguably the single most important region as it is the largest in GDP terms and has the effect of driving the global economy, particularly in down periods. Right now, the fundamentals in the US economy remain sound. Most of the leading indicators ITR tracks were on the rise side going into the pandemic, and these point to a strong recovery. In addition, savings levels among corporations and consumers are sound and consumer spending was seeing a rising trend. All these factors point to a sound foundation for growth once the opportunity arises.
The graph below is an example of what ITR leading indicators look like; this is just one of many tracked by our partners and clearly shows a rising trend. As ITR updates this data, we will use it to help us assess how deep the current slowdown will be and adjust our models accordingly.
Looking back to 2008/2009, when we experienced an economic slowdown so deep it was named “The Great Recession”, we are of the opinion that the 2020 slowdown will not be anywhere near as severe. The situation back then was significantly complicated by being orchestrated or caused by the financial industry. This had the effect of amplifying the disruption, because the banks, which can help economies weather periods of hardship, were themselves in a state of crisis. Right now, the financial health of the banks, at least in the US, is much more favorable compared with 2008/2009. Factors that support this belief follow (source: Brian Beaulieu, CEO ITR Economics).
- US commercial banks are not reporting a tightening of standards for commercial and industrial loans to small businesses. This is somewhat balanced the other way by some tightening for large and medium firms but not to a level (yet anyway) that indicates overt concern.
- The Federal Reserve has dropped the reserve requirement to zero, effectively giving the banking system more latitude in handling their cash.
- The Fed lowered the discount window interest rate to a degree that strongly indicates it wants the banks to be aggressive when making loans. We think they would not have done this if the Fed was concerned about the health of the banks.
- Total assets for all commercial banks are up 4.8% year-on-year. This is elevated from the 2016-2019 period.
One more thing, before we start talking specific numbers for manufacturing. A quick question to you the reader. Put your hand in the air if you knew there was an oil crisis taking place right now…
This topic is worthy of an insight itself, so I won’t go into too much detail here, but with COVID 19 getting all of the headlines it may be easy to miss that the fact that oil prices during the month of March tanked to unsustainably low prices. This incident has parallels with what we saw in 2015/2016 period, and it had a material impact on the manufacturing economies for countries that are significantly involved in oil extraction. We have accounted for this in our updated forecast by considering the impact on these regions during the 2015/2016 period, while also factoring in our Coronavirus assumptions. Those regions most affected that we track within our manufacturing calculations are the USA, Brazil, Canada, Norway, and to a lesser extent, India.
The above chart shows that the US manufacturing economy is particularly sensitive to oil prices, as oil extraction and processing is a significant industry for the region, and there are many second and third-tier suppliers who rely on revenues from the sector. So, when oil prices collapse, investment in equipment and building of new wells are put on hold disrupting companies up and down the entire supply chain. We are not anticipating prices to stay at current levels for as long as the 2015/16 period, so our forecast for those affected regions will not be adjusted to quite the same extent as we saw in the past. We may need to adjust our forecast downwards for those affected regions, if we start seeing evidence that these low oil prices are here to stay.
Our Updated View of Global Manufacturing Output
The following figure shows our current assessment of what growth will look like for manufacturing output in 2020 and 2021, versus the forecast we published back in January. It is our intention to update this monthly, as we anticipate revisions will be required as we go along. As a reminder to our readers, we measure the aggregate value of all manufacturing activities, which does lead to double or triple counting within the same supply chains. Also, that we track each country’s output in local currency and aggregate using 11-year average currencies to eliminate the impact of currency movement over time, which can have a deleterious effect on understanding the real underlying growth rate.
The key points to observe, with additional caveats follow.
- Clearly, we are not predicting a contraction of the level seen in 2009, at least in growth terms. The manufacturing economies around the world have nearly doubled in value since then, so a smaller contraction in growth terms still amounts to a substantial reduction in overall value. Our current estimates suggest over $950 billion in production value will be lost in 2020, although we are anticipating 2021 will recover some of the lost value, but not all of it. When one looks at the speed of recoveries, if one thinks of the disruption as a “V”, it typically takes 2-3 years for the right side of the V to be recovered versus what took place on the left side of the V.
- Normally we present five-year forecasts, but at this stage we are only focusing on the critical 2020-2021 period. Long range forecasts are hard to do accurately when the economic environment is going through a period of significant instability. We will revisit the 2022-2024 period once we feel more confident about when the fallout from the Coronavirus is mostly contained.
- Our forecasting method is more instinctive than directly modeled. This is mainly because the data sources for each region are different, and we must interpret each country uniquely. We are building individual forecasts for 38 regions, which is especially challenging at this time. Our approach is to focus on the top 10 regions or so and use these as proxies for shaping growth rates for other regions.
- Taking the US as an example, we’ve revised down manufacturing output from 0.4% to -2.1%. Although in percentage terms this doesn’t seem an overly aggressive forecast, and some might say it is too optimistic, it still represents an output value loss of $114.8 billion. But as outlined in this insight, we are reasonably optimistic that manufacturing output will recover fairly quickly; other sectors of the economy may not fare so well, so we could see a situation where GDP falls much more than manufacturing output. It’s important for readers to realize that, during this period, some manufacturing will operate as normal or even be ramped up slightly, notably for food and medical suppliers. Food is one of the largest sectors for all regions, so strength in this sector helps buoy the overall manufacturing economy somewhat. Also, right now we don’t have many short-term signals to rely on, and the Coronavirus event is quite unique making it difficult to predict from previous events and corresponding market performances.
The table below provides a summary of our forecast changes for the top 10 manufacturing economies, as well as key contributing factors for developing our forecast.
As can be seen by our downward revision for Italy, regions significantly impacted are projected to have a much more difficult 2020, as one would expect. It is entirely possible other key regions, notably the US, UK and Brazil, could also see further downward revisions.
The other region worth highlighting is China, as it accounts for nearly half of global manufacturing output. Although we have revised down our forecast, we are still firmly of the mindset that the region will recover and achieve growth by the end of 2020. Feedback from our team in Shanghai is that there are positive early signs that the region will recover and that they are already seeing some rebound and recovery in key industries.
Over the coming weeks we will be looking more closely at the industries that contribute to manufacturing output and attempting to measure the impact within these sectors.