By Jeremie Capron, Head of Research, ROBO Global
I recently had the opportunity to speak with Morten Paulsen, the Head of Research and a Managing Director at CLSA Japan. As a member of the ROBO Global Strategic Advisory Board, Morten’s insights into developments in robotics, automation, and AI in Asia help inform our research and alert us to the latest advancements coming to the forefront from both current industry leaders and emerging companies. Morten has covered the Japanese robotics and machinery industry as a Tokyo-based equity analyst since 1999. Following is an excerpt from our discussion:
Jeremie Capron:
Morten, you have followed the Asian factory automation markets for nearly two decades as a Tokyo-based equity research analyst. During that time, we have seen a dramatic surge in Asian robotics demand. Today, China has the largest industrial robot market in the world, accounting for a third of global demand. In addition, China’s operating stock of industrial robots is now larger than any other country. You spotted the trend early, documenting your findings in your 2008 trends report, “Automating Asia: Leading-edge firms driving growth.” What drove this historic shift in automation demand—and how much further can it go?
Morten Paulsen:
It’s common to hear that automation demand in Asia is driven by labor cost. That’s not incorrect, but the reality is more complicated and the drivers are constantly changing. When I started researching the “Automating Asia” report back in 2007, nobody in China talked about labor cost. Back then, labor was too cheap and automation equipment too expensive, so the economics did not support any labor substitution. The main reason why Asian manufacturers would purchase automation equipment ten years ago was because they wished to move up the quality curve. That started changing in 2010 and 2011 when labor cost reached a level where it became meaningful to calculate ‘payback time’ on investments. That drove another surge in automation demand. Increasingly, we hear less about labor cost and more about labor shortage.
Capron:
It sounds hard to believe that you can have labor shortage in a country with more than a billion people?
Paulsen:
True, but it’s important to keep in mind that there are several forces at play. The first force is demographics. The size of the Chinese workforce peaked in 2015. Due to China’s one-child policy, the size of the labor entry population (defined as people in the 18-23 year age group) is declining rapidly. Second is the service sector. This is the fastest growing sector in China. Young Chinese would rather work for a service business than on a factory floor. This trend is taking a lot of labor away from the manufacturing sector. Manufacturers in coastal regions who depend on migrant workers are also finding it harder to employ people due to the ongoing Hukou reform and faster wage growth in inland China.
Capron:
In China, the government is taking a very active role in the development of the manufacturing sector. From your perspective, how important are industrial policies, such as Made in China 2025?
Paulsen:
Government policy is partially responsible for the rapid increase in robot usage in China. Made in China 2025 is modeled after Germany’s Industrie 4.0 program, but it is vastly more ambitious in terms of reach, scale, and complexity. It is essentially the keystone of an audacious effort by China to avoid the ‘middle income trap’ and to become the global leader in a range of new and old industries and technologies.
Industrial policy affects both the supply side and the demand side, but quantifying the impact is time consuming since China has multiple layers of government policies in place to support its manufacturing industry, and the implementation is for the most part done regionally. As a result, the actual policies will vary depending on where in China we look.
From what I see, China recently has toned down the rhetoric around its government policies, and I believe we will see a renamed and revised version of the Made in China 2025 program introduced in the near future. However, I do not get a sense that the government in any way has moved away from its longer-term strategic targets.
It is important to stress that the Chinese automation and robotics markets are not dependent on government subsidies and support. It is the private sector, not the state-owned enterprises, that are leading the charge to automate.
Capron:
Where do you think we stand in the process of automating Asian factories?
Paulsen:
The market can go much further. China’s robot density just reached the world average in 2017, but the country will not stop there. The automotive sector is leading other industries in terms of robotics usage, but even here we see that China is behind in terms of robot usage. China is the world’s largest manufacturer of vehicles, with an annual production of around 26 million per year. The automotive industry in China has some 160,000 robots deployed. Put into context, the aggregate vehicle production of Japan and Germany would be about 16 million vehicles, but the two countries deploy more than 200,000 robots in automotive production.
Beyond China, other Asian markets are also emerging rapidly. India and Vietnam both are likely to become top-10 countries in terms of industrial robot purchases. The Indian automotive market has softened as a result of new insurance requirements, higher fuel prices, an uptick in interest rates, and other factors that are increasing the cost of ownership. However, automation capex is likely to accelerate due to stricter safety and energy efficiency regulations.
Vietnam is emerging as an important hub for final assembly. The country is close enough to China be linked to the Chinese supply chain, and labor cost is cheaper. The shift of assembly capacity to Vietnam is further accelerated by import tariffs on Chinese manufactured goods.
Capron:
To what extent are Japanese companies benefiting from this megatrend? What are the areas where Japan is truly dominant? Japan accounts for 23% of the ROBO Global Robotics & Automation Index, which includes best-of-breed companies that are powering the robotics and AI revolution.
Paulsen:
The high Japan weighting in the ROBO Index makes perfect sense to me. It’s been popular among global investors to ignore the Japanese equity market, but that is not something you can afford to do if you look at robotics and automation. Japan is the world’s largest producer of traditional industrial robots. Of the world’s big-6 manufacturers, 4 of them are Japanese.
Japan is even more dominant when you look at upstream component industries such as reduction gears, where Japanese suppliers command a 90% share of the global market. Reduction gears, also often called speed reducers, are mechanical parts that are critical for giving industrial robots the speed and precision they need. For a robot maker, the reduction gears are usually the most expensive components sourced externally. Japanese suppliers also command dominant positions in AC servomotors for industrial robots.
Although Japan has a strong position, I would highlight that there are areas in robotics where Japan is lagging behind. One area is robots for medical surgery, a market that has seen rapid growth. The other area is collaborative robots, a fast growing segment in the robotics industry where Universal Robots, a unit of ROBO index member Teradyne, continues to dominate.
Capron:
The advent of machine intelligence seems to be boosting the robotics market. How are Japanese and Asian companies in general positioned in this field?
Paulsen:
As elsewhere, Internet of Things, AI, and machine learning have been big buzzwords in the manufacturing sector in Asia. However, I do believe Asia is approaching smart manufacturing a little differently. In Europe, a lot of efforts are put into developing smart manufacturing as a tool for ‘mass customization,’ where online ordering is linked directly to the manufacturing process. In Asia, I see companies applying smart manufacturing primarily to control the supply chain and to improve efficiency though reduced cycle times. Japan, Korea, and Taiwan are the leading Asian countries in terms of adoption.
A lot of companies are rushing into this field, and there are lots of platforms and solutions competing with each other. I believe we have reached a turning point as manufacturers have become more critical. They want to see real results, not just nice graphics of machines being ‘on’ or ‘off’ on their smartphones. I believe we will see a separation of providers who are investing heavily in adding real value versus those that were just trying to free ride the IoT hype. In the longer run, I expect IoT to become a scale game, where profits will be skewed heavily to the suppliers with the largest installed base.
Capron:
After a phenomenal 2017, why was 2018 a more difficult year for Japanese robotics companies?
Paulsen:
I would highlight two important reasons for Japan’s difficulties in 2018. First, we saw significant multiple expansion in Japanese automation names throughout 2017, and in January 2018 the sector Price to Book ratio extended two standard deviations above the historical average. In most prior cycles, that is the point where we went from multiple expansion to sideways consolidation. It was the main reason why we downgraded the sector. The second reason was growth compression. YoY Growth in robot demand plummeted from +30% in 2017 to low single digits in 2018. The Chinese robotics market held up reasonably well in the first half of 2018 despite difficult comps in the electronics industry, but turned south in the second half—and the stocks de-rated rapidly.
Capron:
To what extent was the drop in Chinese robot demand caused by the trade dispute with the US?
Paulsen:
The USA-China trade dispute forced many manufacturers in China, especially those geared to exports, to rethink expansion plans. Speaking to people on the ground, I understand that among Chinese exporter manufacturers, as many as one out of three automation projects were postponed as a result. The Chinese manufacturers adopted a very negative view on the outlook for a resolution early on, and they are preparing for a long lasting conflict.
That said, the so-called trade war gets blamed for more than its fair share of the slowdown. I believe the Chinese Government has used the trade dispute to push through some painful but necessary policy changes such as the crackdown on shadow banking, a revision of subsidy programs, and a labor reform of the Hukou system. That’s easier to do when you can blame Donald Trump.
Capron:
What should we expect in 2019?
Paulsen:
A lot depends on the outcome of the trade negotiations of course. Putting that aside, industrial automation demand is primarily driven by capex trends in two verticals: automotive and electronics.
Capex for semiconductor production is likely to stay weak in 2019. Some major projects in Korea and Japan have been delayed by 6-9 months, and expansion plans in Taiwan may also be scaled back. Capex for smartphones is very subdued at the moment. I believe smartphone makers are reluctant to make any big investments before a clear roadmap for 5G is set and the network is installed. I do not see that happening before Q2CY2020 at the earliest.
When it comes to automotive capex, I am most concerned about the outlook for Europe. The uncertainty around Brexit and potential US tariffs is making it very difficult for European automakers to invest. Automotive capex in North America is stable at the moment if you include Mexico, but the market is saturated. The industry has already invested heavily in retooling production lines for SUV and trucks. Some of the slack in autos will be covered by growth in general industries, but growth rates are nevertheless set to decelerate.
Just like in the US, the Chinese automotive market appears to be saturated. Production volume peaked in 2017, and we should expect production volumes to decline further in 2019. Despite sluggish demand, automotive capex is likely to grow by double digits in 2019 because the CAFC carbon and NEV credit points system will come into play this year, forcing automakers to ramp up investments. China is also much more aggressive in promoting EVs, and that is positive for automation demand.
Capron:
Automation stocks have had a very strong start in 2019, although fundamentals do not seem to be getting better. How sustainable do you think the recovery is?
Paulsen:
In my view, the rally in automation stocks this year is a counter-reaction to how oversold the market was at the end of last year. Institutional investors are underweight in the sector, and they are very nervous about missing a rebound. It is clear that the demand situation in China could change quickly if a firm trade agreement is reached. From a momentum perspective, we are likely to see certain second derivative indicators, such as machinery order growth, to bottom out within the next 4 to 5 months as comps will get easier as we enter the second half of the year.
The rally we have seen so far in 2019 carries all the traits of a typical bear market bounce. As such, we should be prepared for more volatility assuming that a firm recovery may not materialize before 2020. That said, in my view, this is not the time to turn bearish. The focus should be on finding entry points and deciding when and what to buy. The consensus view is to ‘wait for confirmation,’ but I would recommend that forward-looking investors avoid the stampede and buy the sector before it is obvious to everyone that the cycle has bottomed.
Capron:
At the beginning of the year, you wrote that 2019 would bring the best buying opportunity in automation since 2015. Do you still hold that view?
Paulsen:
Yes, I do. 100%.