All LSEG websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.
The Financial & Risk business of Thomson Reuters is now Refinitiv
All names and marks owned by Thomson Reuters, including "Thomson", "Reuters" and the Kinesis logo are used under license from Thomson Reuters and its affiliated companies.
Ryan Shea has been employed in the financial sector for more than two decades. During this time he has held senior positions in several major sell-side and buy-side financial institutions in a variety of roles encompassing macroeconomic research, FX portfolio management and investment strategy. In addition to being partner and Head of Research in Amareos Ltd, Ryan is Managing Director of Black Swan Economics Consultants Ltd (BSEC), a global macro advisory company established in 2013. Prior to this he was Head of Macroeconomic Research for the Fixed Income Department of Abu Dhabi Investment Authority(ADIA) in the UAE. Ryan completed his BSc and MSc in Economics at Queen Mary College, University of London.
Often one hears the phrases “dumb” money and “smart” money when commentators are discussing financial market participants. Such classifications are not only arbitrary they are also meaningless. More than twenty years toiling away both on the sell-side and the buy-side side has taught us that asset markets are great levellers.
There are countless instances where “smart” money suffers traumatic P&L underperformance (one prominent, and unquestionably smart, UK hedge fund manager has had a rather torrid time of it over the past 18 months) while “dumb” money has shined [1] [2]. Additionally, even those investors with proven strong investment track records are far from infallible. Take Warren Buffett, someone we mentioned in last week’s Market Insight [3].
Buffett’s investment career is without doubt impressive. He, and his Berkshire Hathaway colleagues have generated consistent market-beating returns such that one wonders if some academic economists, intellectually beholden to the efficient market hypothesis, consider him nothing more than a ukulele-playing, cherry-cola-drinking figment of their imagination. Despite that, he has one notable blind spot – technology – one he acknowledged in his recent shareholder meeting as per his “We missed it” comment in relation to Amazon and “I blew it” in relation to Google.
Technology and humankind have a long and chequered history. Without it our lives would be infinitely poorer – we would be inhabiting a pre-stone age world. Aside from a few extreme Ecowarriors who dream of return to a simpler (simplest?) life, who would argue against the merits of fire, the wheel, the internal combustion engine, electricity or the internet [4]. And yet, at the same time, we seem to fear every step down the technology path.
Present public concern centres on the impact of technology on employment, especially from robotics and artificial intelligence, amid worries it will lead to mass job losses – employment extinction if you like [5].
The historical record suggests such fears are largely unfounded. Two centuries ago in England, similar concerns gave rise to the Luddite movement [6], but what happened then – and indeed has happened in other industries since – is that employees displaced by technological innovation were simply reabsorbed into the labour market and engaged in different activities to earn a living. This transition is neatly mapped by the loss of agricultural jobs, first to manufacturing and then, in turn, to services, where the vast majority of people – in developed economies at least – are now employed.
Despite such experience, the foundation for arriving at this pessimistic assessment is that unlike previous episodes of technological innovation it is not just blue collar workers involved in manual activities who are at risk. It is any job that has a standardized pattern or routine and hence is amendable to automation. A well-known study published in 2013 by Carl Benedikt Frey and Michael Osborne concluded that almost half of US workers had jobs that were at risk [7].
We are by no means experts in the field of robotics, but just take a look at the following recently uploaded footage showing a very capable [8] robot developed by Boston Dynamics:
Worse, there is now a robot capable of flipping hamburgers – the ultimate downside employment hedge [9]!
Even more concerning, as Bill Gates noted in a recent interview where he proposed taxing robots [10], is the likely speed of the substitution of technology for labour.
“You cross the threshold of job-replacement of certain activities all sort of at once.”
A fast transition of the sort Gates seems to be anticipating will make it much harder to absorb displaced workers, so it is easy to see why there is concern about the likely profound socioeconomic impacts upon the global economy. Nevertheless, technology is, and will continue to, play an ever-increasing role in our lives.
In light of such trends, and unlike in the late 1990s when Buffett’s aversion to technology stocks served him well – it helped him deftly sidestep the dot.combubble in the late 1990s – it has cost him and his shareholders by missing opportunities such as Alphabet. This is something he appears to be trying to make up for by increasing his stake in Apple.
Yet, while Buffett appears to be latterly – and narrowly – learning from his past mistakes (the most effective way to become smarter) and overcoming his aversion to leading technology stocks there is still much, supposedly “smart”, money firmly in the bearish camp when it comes to this sector.
For example, short interest in FANG stocks, the bell-weather tech stocks comprising Facebook, Amazon, Netflix and Google (subsequently renamed the mnemonic-destroying Alphabet) has risen sharply over recent months to more than USD17bn (up just over 30% year-to-date). Without doubt, this has been a painful trade for supposedly “smart” hedge funds as these stocks have done remarkably well, handsomely beating the overall market having risen by an average of 25% year-to-date.
What’s more, despite the well-recognized dangers of doubling down on a losing trade that is exactly what the shorts have been doing, with aggregated positions having increased every month since the start of the year.
Clearly, hedge fund managers shorting FANG stocks are not motivated by Ludditeian (we just made that word up) thinking [11]. Rather, their bearish view reflects their belief that investors are overpaying for future earnings, even though those earnings streams will almost certainly be rising as technology companies take an increasing slice of the expanding global economic pie.
In last week’s Market Insight we discussed the historically elevated valuation of the S&P500, which on Shiller’s CAPE has reached 29.5. Given such levels have only been exceeded twice before in over 130 years, it sounds stretched, but compare that with the PE ratios for the FANG stocks. Only Apple, with a PE ratio of 18 is below the market average. Google’s PE stands at 31, Facebook at 38, and Amazon and Netflix’s having eye-watering ratios of 180 and 205 respectively. To even remotely justify such valuations, the future earnings pie of Amazon and Netflix in particular are going to have to grow…. a lot!
In short, hedge fund managers are betting on the fact that these stocks are in a bubble where rationality has been replaced by hype. It is not hard to see why they might think that given such strong price momentum and in the case of Netflix and Amazon very lofty valuations.
As regular readers will appreciate, our crowd-sourced sentiment indicators are uniquely placed to pick-up “bubble behaviour” because asset price bubbles are typically characterized by extreme sentiment skews (very high indicator readings). By invalidating one of the prerequisites for the predictive power of “the many over the few”, this sets-up the condition we denote “crowd fail”, or as Benjamin Franklin more eloquently put it,
“If everyone is thinking alike, then no one is thinking”
So, if the “smart money” hedge funds are correct and the FANG stocks are in a bubble, one that they will successfully exploit when it bursts, as all bubbles must do at some stage, then we should expect to see strongly positive sentiment readings for the four constituent stocks.
What do we observe? Take a look at the following exhibits.
Exhibit 1: Crowd-sourced Sentiment By Media Type – Facebook
We have distinguished between media types, reflecting our judgement that the tone of the mainstream media is more reflective of a professional investor audience – the “smart” money using the language introduced at the start of this note, whereas social media is more reflective of a retail investor audience – the “dumb” money.
The common thread in our sentiment indicators is that for each of the four tech companies social media sentiment is higher than mainstream media. This strongly suggest that that these stocks are more favourably viewed by joe public compared with investment professionals.
However, when we look at the overall level of sentiment, that is to say the skew in the crowd’s thinking only Netflix’s, and to a lesser extent Amazon, can be considered elevated. Indeed, for Google and especially Facebook, social media sentiment is close to zero, indicating no strongly held views, while mainstream media sentiment is strongly negative.
Drilling down into tone of the online posts in relation to the four companies, we also observe notable differences in the intensity of crowd feelings towards the four companies at the individual emotion level. Again, to capture potential differences between the two media types we plot both separately rather than in the usual aggregated form – see exhibits below.
While the crowd has high levels of anticipation regards all four companies, as one would naturally expect for leading technology companies, posts about Facebook in both media types and in Google on social media have notable readings for Anger and Disgust – two negative emotions that actually are positive price signals because they are usually observed towards the end of a stylized market psychology cycle – see exhibit below.
Such strong negative emotional readings, which are absent for Amazon and Netflix, are not the sort one expects to see during a bubble. Combined with the relatively low overall sentiment reads, this suggests that, if anything, the risk is that in these two companies’ share price remains on the upside given the potential for position capitulation ie a short squeeze is tangible.
That said, if one simply cannot resist the temptation to join the bearish side of the trade in these technology leaders, with overall sentiment positive but losing momentum and given the near absence of negative emotional reads of the four companies it is Netflix that appears to most vulnerable.
[1] Think: the repeated surveys showing more than 50% of drivers consider themselves to be above-average despite its logical impossibility.
[2] Were it not for some well-entrenched cognitive biases that serve to override logic when assessing past performance, investors would be amongst the most humble people on the planet.
[4] This is obviously an abbreviated list. Although we are fans of Apple products we would not go as far as to add the iPhone, or their other sleekly designed products to that list, although we are sure many would.
[5] Or, even more extreme – outright extinction.
[6] The Luddites smashed machinery in textile factories in the belief that this would protect their jobs.
[8] Capable of what is clearly the question. We doubt that the video will do anything to assuage concerns held by some as to our longevity on this planet.
[10] Gates’s view is that robots should be taxed in much the same way as labour presently because otherwise the government would potentially lose a massive amount of revenue (through lost income tax receipts not to mention increases in unemployment benefits to be paid). While accepting that taxing robots may, by introducing an inefficiency, stifle innovation which is detrimental, we see merit in Bill Gate’s proposal, but not just because of the impact on government finances. At hand is a much more fundamental problem.
Just as light displays duality – it behaves as both a wave or a particle – so too do humans. We are both wage earners and consumers. Absent such a redistribution mechanism the value-add generated by robots doing the work formerly done by humans will accrue to their owners. This has serious social and economic implications. It will further exacerbate economic inequality which, in our opinion, has been a significant contributing factor to the nationalist-populist backlash observed in many developed countries. Moreover, if even close to 50% of jobs are lost to technology – estimates vary depending on the employment characteristics of individual countries, but the lower limit for estimates is in the mid-30% range – over a relatively short period of time, then the impact upon aggregate demand will be catastrophic. Indeed, it would be almost the perfect Marxian scenario which implies the global economic system collapses due to the impoverishment of labour to capital. In comparison with such a possibility, the idea of a tax being applied to robots is perhaps not as negative as some would have us believe.
