Not since the Industrial Revolution of the late eighteenth century have we stood on the precipice of such a monumental displacement of labour. Automation, the process by which human intervention in tasks is reduced or removed through greater utilisation of technology, is the next revolution. On the one hand, productivity gains that are being realised from automation are enormous. So much so in fact, that manufacturing is being re-shored to developed countries like the US, where rising automation has significantly reduced the need for, and hence cost of labour. A 2016 study by Deloitte looked at why re-shoring was occurring in developed economies and the results were clear. Though their cost competitiveness per capita was poor, they had automated to a point where a smaller, more highly skilled labour force could be used, making the cost more palatable.

Key factors that have contributed to reshoring in select countries
Figure 1. Key factors that have contributed to reshoring in select countries
Source: Deloitte


On the other hand, this is also indicative of the insidious job loss to automation. Yes, in this instance jobs are being transferred to developed nations but there is still net job loss in the global economy. This is the flip-side to the economic benefits of automation – unemployment. Generally speaking, there are three types of unemployment: cyclical, frictional and structural. Unemployment is cyclical if it is the result of poor economic growth, such as in a recession or similar downturn. It is said to be frictional if it is the result of normal turnover in the labour market as people transition between seasonal jobs, into new businesses, or into new industries. Both cyclical and frictional factors only result in temporary, intermittent job loss. Structural unemployment is considered more crippling because it occurs when there is an absence of demand for a certain type of worker.

Herein lies the greatest difference between the Industrial Revolution and what we are seeing with automation today. In the nineteenth century, workers transitioned from farms to factories – like for like manual labour. Unemployment did not rise dramatically because for each job made redundant, another was created in a new but comparably skilled industry. As robots essentially replace people however, they are not being offered comparable work in other industries.

Take road freight (trucking) for example. This is an industry particularly vulnerable to automation, with driverless transportation looming on the horizon. Not only could driverless trucks complete routes faster (robots don’t need to sleep), but they would undoubtedly offer safer and more reliable delivery. A best-case scenario might see this industry lose three quarters of its workforce as some workers transition to more of an oversight role (say one person per convoy on the roads and several at a central hub to monitor the entire fleet). In the US, this industry accounts for 8.7 million jobs directly (3.5m truckers, 5.2m support staff), many of which will be lost when driverless trucking inevitably becomes viable. The real problem is, when a truck driver loses their job, most will not re-qualify within the robotics industry. Many jobs will be taken away and only a few will be replaced.

Map of the most common job in each US state, 2014
Figure 2. Map of the most common job in each US state, 2014
Source: NPR, US Census Bureau

This is, of course, largely hypothetical at this stage but it illustrates the potential for an extraordinary transfer of wealth from labour to capital (employees to owners). Keeping with the trucking example, the average salary for a truck driver is around US$40,000. If we assume only the drivers lose their jobs, this would remove US$140 billion in salaries from the US market. Much of this wealth will be transferred to the robotics industry but perhaps just as much will be transferred to the owners of trucking companies. In the US, 9 out of the top 10 trucking companies are publicly listed, meaning the primary beneficiaries will be shareholders.

Investors tend to look at potential economic shocks in terms of both risks and opportunities. Automation need be no different. The risks are clear – if not managed carefully, automation could lead to mass unemployment, and accelerate the widening wealth equality gap. Inevitably, there will be considerable political interference designed to limit the extent to which this occurs but even so, a significant net loss of jobs appears unavoidable. For investors however, this could become a once in a generation opportunity. Simple analysis reveals which industries currently spend the largest proportion of their revenues on staff. Combine that with a view as to which jobs are most vulnerable to robots, and it paints a picture of which businesses are most likely to gain (and gain the most) from the automation revolution.

Wages make up 26% of fast-food sales in the US on average
Figure 3. Wages make up 26% of fast-food sales in the US on average
Source: Heritage Foundation


As illustrated above, another good example of where automation is likely to benefit shareholders is the fast food industry. For decades, global leaders like McDonalds have found ways to maximise the speed, consistency and price competitiveness of their offering. Though production of ingredients has been nearly fully automated, in-store the process is still surprisingly labour-intensive. Menu items are cooked, assembled and packaged by hand, ordering is done face to face and waste disposal remains completely manual. Companies like McDonald’s have built their brand on a formulaic, largely uniform offering and store layout, ideal circumstances for automation. Using touch screens for ordering, robots for food prep and conveyor belts for delivery, all the technology needed to replace the labour already exists. In Japan, the Uobei Sushi chain has fully automated its stores, allowing it to offer a faster, cheaper system with greater capacity and better asset utilisation.

An Uobei Sushi restaurant, Shibuya
Figure 4. An Uobei Sushi restaurant, Shibuya
Source: Cultural Xplorer

Some of the most striking examples of automation are physical manifestations, as illustrated above, but in reality, robots are also replacing white collar work. Law, an industry comprised of highly educated, well paid professionals, is just as vulnerable. Much of legal work is procedural in nature and can be learned and replicated by software and systems, with limited supervision. One example from Australia is conveyancing, the process of transferring ownership of property from one person to another. LawLab is a high-volume, paperless, online conveyancing service that manages $10 billion worth of property transactions nationwide. It is a platform as a service (PaaS) offering which eliminates the need for much of the more routine, standardised filing and administrative work.

Add to the list medical diagnostics, accounting, investment reporting, engineering and many others and it becomes obvious that all work that does not require innovation, creativity or opinion can be automated. Routine jobs – even complex ones – can be filled by robots if there are a defined set of rules, procedures and systems to follow. Even our own roles as investors are vulnerable, with most of the daily trading volume around the globe being done through fully automated algorithms with little human supervision. Global giants such as BlackRock and State Street are launching index funds and ETF’s run completely using algorithms and automated trading programs at low costs, with few staff.

US employment by type of work
Figure 5. US employment by type of work
Source: US Population Survey, Federal Reserve Bank of St Louis


All these examples show a transfer of wealth from the labour market to the capital market. Many industries will be able to cut their head counts and lower their labour costs, leading to gradual margin expansion. As investors, the best way to exploit this shift is to deploy capital over the long term into those industries standing to gain the most. The worst thing one can do on the precipice of an economic revolution is nothing. To quote LawLab co-founder Richard Bootle:

‘“Disruption to me seems to be something that everyone thinks happens to someone else” – we’ve all sort of said that in the smug, snug complacency of the back seat of an Uber watching the beginning of the end for the taxi industry’.

There are always opportunities to exploit seismic economic shifts. We can either watch from a distance as whole industries are transformed, including our own, or we can actively participate, and benefit from the inevitable transfer of wealth.