The ongoing selloff of technology takes on a different flavor in current market thinking. Unlike products on the goods side of the aisle, the service side of international trade is not about anti-dumping legislation, voluntary restraint agreements or countervailing duty laws. It is most certainly not about proposed tariffs regimes or possible trade wars. The issues at stake here are more nuanced, esoteric discussions around big data and privacy and futuristic technologies like autonomous vehicles (AVs). Rather than tariffs, the overriding fear is heightened government regulation—specifically targeting big data, privacy and futuristic technologies like AVs. Where companies such as Facebook (FB), Alphabet (GOOG) and Nvidia (NVDA) operated within a regulatory environment of laissez-faire oversight that sent each of these companies to the pinnacle of market valuations, consumers the world over are slowly coming to recognize the power of the personal information they provide through their written thoughts, profiles, search habits and even physical GPS position at any given moment of the day is being harvested by social media networks and sold to marketers the world over—the disposition over which they have precious little control to date. And that lack of control is unlikely to change any time soon.
The US maintains a literal patchwork of regulations regarding AV technology and real-world testing. The lightly-regulated streets of Arizona were home to hundreds of AVs in various stages of real-world testing by a plethora of companies racing to bring the technology to market. Regulators struggle to keep apace.
Figure 1: Facebook against the S&P 500
FB’s worldwide reach in social media space should not be underestimated. Daily active users of the FB platform came to 1.4 billion worldwide through December 2017, an increase of 14% YOY and cumulative increase of 57% since December 2014. In the US and Canada, daily active users number 184 million, down slightly from 185 million YOY over the period. The cumulative change since December 2014 came to 17%. In Europe, the numbers come to 277 million through the end of December 2017 with a cumulative change of 28%. The Asia/Pacific market is both FB’s biggest market and its biggest growth engine with 499 million daily users with a cumulative growth rate of 97%. The same statistic for Africa, the Middle East and Latin America comes to 441 million for the period with a cumulative change of 68%. Monthly world users come to 2.13 billion for the period for a cumulative growth factor of 53% since 2014.
By a company defined metric referred to as the average revenue per user (ARPU), the US and Canadian markets provide 49.1% of the company’s $12.779 billion revenue stream through the end of 2017, a stream that has jumped 140% since 2015. Europe provided 25% or $6.271 billion of the company’s total revenue stream for the period, followed by the Asia/Pacific region with just over 16% or $2.048 billion. Africa, the Middle East and Latin America generated just under 10% or $1.264 billion of total company revenue for the period.
Total revenue came to $40.65 billion, up 47% YOY, with ad revenue comprising 98.3% of total revenue for the period, or $39.94 billion. Harvesting the many facets of FB users’ platform footprint for sale to marketers worldwide who design highly targeted ads for FB users and friend networks has become a turnkey operation and is clearly the sine-qua-non of the company’s revenue stream. Net income came to $15.93 billion, churning out an EPS of $5.39 over the period. The effective tax rate for the period was 23%, on the high side for a technology company with broad international reach. The company’s cash hoard came to $41.71 billion. The company’s fundamentals leave little room for investor complaint.
FB (green line, main frame) market slide was in close proximity with the S&P benchmark (orange area, main frame) in response to the presumed sighting of inflation lurking deep within the January jobs report issued by the Bureau of Labor Statistics on the first Friday in February. FB peaked for the year to date occurred on the 1st of February at a price of $193.09. A short week later, the stock had slipped quickly into correction territory at $171.58 with a loss of just over 11%, plunging through its 50-day trading average in the process (blue line, main frame). Several hiccups attempt at recovery mirrored that of the S&P benchmark through much of February, peaking above its 50-day measure, before hitting a sustained downward plunge as the data breach by Cambridge Analytica, a British-based political profile consultancy co-founded by the conservative hedge fund manager Robert Mercer, splashed across the newswires. FB’s share price went from $186.09 on the 16th of March down to a trough of $152.22 in less than a week, losing about $80 billion in market capitalization in the process. The drop this time was less than two percentage points below bear territory and well below the stock’s 200-day trading average (red line, main frame). Market momentum (bottom frame) was close to perpendicular to the downside as investors stampeded for the exits. FB CEO Mark Zuckerberg’s stiff, even robotic performance first before a joint session of the Senate Judiciary and Commerce committees on Tuesday and then before the House Energy and Commerce committee on Wednesday was nonetheless warmly received by markets as momentum turned to the upside for the duration of April, poking positive for the first time in more than two weeks (see Figure 1, above). FB shares were up almost 6% and Zuckerberg’s net worth regained about $3.2 billion on the uptick. Oversold?
Figure 2: Alphabet against the S&P 500
Google has even more experience in collecting data on the consumers that use its platform—from tracking daily search history to tracking daily physical locations via the sheer ubiquity of smartphones using its Android operating system to harvesting media preferences through its YouTube subsidiary. The information is then sold to advertisers enabling the design of highly targeted ads based on synthesized browsing behavior. That said, total revenue increased 23% through the end of 2017 to $110.86 billion YOY, or two and a quarter times that of FB. Total revenue has almost doubled from $55.519 billion in 2013. Google segment comprised 98.9% of total revenues for the year at $109.652 billion, up 23% YOY. Revenue from the company’s other segment, euphemistically categorized “other bets” pulled in $1.2 billion for the year, up 49% YOY. The category combines revenues from other businesses under the Alphabet umbrella.
Geographically, the US market comprises 47% of GOOG revenue through the end of 2017 at $52.449 billion, up 23% YOY. Europe, the Middle East and Africa contributed 33% of total revenue or $36.05 billion for the period, up 19% YOY. Asia and the Pacific contributed 15% of total revenue at $16.24 billion, growing 29% YOY. Other Americas contributed 6% or $6.13 billion to total revenues, posting the highest geographic growth rate at 32% YOY. Overall, geographic growth remains flat with the US contributing 47% of revenue for the past two consecutive years. Europe, the Middle East and Africa growth declined slightly in 2017 to 33%, down from 34% in 2016 and 35% in 2015. Asia and the Pacific region rose slightly to 15% of total revenue in 2017, up from 14% in both 2015 and 2016. The other Americas remained flat at 5% for the three year period.
Interestingly, traffic acquisition costs (TAC), or the total cost of revenues as a percentage of total revenue generated continue to rise. TAC costs increased 29% YOY to $21.67 billion or 41% of total revenue through the end of 2017. The increase in TAC costs were mainly driven by the proliferation of, and searches conducted on, mobile devices that carry higher costs due to the increase number of toll gates, from browser providers to mobile carriers to original equipment manufacturers to software developers, that advertising passes through before ending up on a unique mobile screen device. TAC costs have increased 51% since 2015 and increased 29% YOY—outpacing the growth of revenue for the year by six percentage points through the end of 2017. If the upward trend of TAC continues, forward growth becomes more problematic.
After a 38% market performance in 2017, GOOG (green line, main frame) appeared hitched to the ebb and flow of the S&P 500 benchmark (orange area, main frame) since February’s market correction with a sharp plunge through the stock’s 50-day average trading measure (blue line, main frame). GOOG dropped directly into correction territory with a 14% nosedive. As investors staged hasty retreats from their premonitions of inflation, GOOG followed the market with two unsustainable forays in late February and early March above the 50-day trading measure only to lose the necessary market momentum (bottom frame), only to fall back for a third time on the coattails of the market now transfixed by the possibility of a trade war breaking out between the world’s two biggest economies. GOOG has noodled about its 200-day trading average from the beginning of April through Friday’s market close (13 Apr). At Friday’s market close, market momentum has also turned to the upside for April and squeaked positive ever so slightly positive (see Figure 2, above). Oversold?
Figure 3: Nvidia against the S&P 500
NVDA invented the graphic processing unit (GPU) back in 1999. Driven by the exponential demand for 3-dimensional graphic components of video gaming, the GPU format has evolved into a driving force itself in the cutting-edge fields of virtual reality (VR), high-performance (HP) computing as well as artificial intelligence (AR). The GPU format is used to enhance and accelerate a variety of scientific applications—from the prediction of weather to mapping of genomes to wind tunnel simulations. The company’s Drive AI platform allows the car to pilot in either full or partial autonomous mode through the use of an array of cameras and sensors that will futuristically sweep current staples of automotive equipment like side and rear-view mirrors, steering wheels and even pedals into the dustbins of history. NVDA currently works with over 300 auto manufacturers and suppliers, automotive research institutes, high definition mapping firms and myriad start-up companies working in the field of artificial intelligence as it applies to AV research and development.
The applications the GPU format in our technology-based world are broadly based moving forward, affording NVDA a rich array of paths for future growth. Total revenue through the company’s fiscal year ending the 28th of January 2018 was up 135% to $9.71 billion from $4.13 billion in FY2014. Net income for the period has soared 593% to $3.1 billion from $440 million. Net income per share has followed a similar path, up over 551% for period. Total revenue increased 41% YOY while net income was up 83%. This drove net income per share up 88% for the period to $4.82/share from $2.57/share through FY2017. Cash equivalents rose 8.42% while retained earnings increased 44%. Research spending came to $1.8 billion YOY, up 23% for FY2018. Research spending rose 10% YOY in FY2017. Long-term debt remained flat at $1.99 billion in FY2018 YOY. Dividends increased to $0.57/share or 18% for the period YOY, up from $0.45/share in FY2018.
By segment, the company’s GPU works, which includes the company’s vast gaming platform, grew by 40% YOY through FY2018 with $8.14 billion in total revenues. The sector grew by 39% YOY from FY2016. The GPU segment comprised 83.7% of total company revenue through FY2018. NVDA’s Tegra processor segment, which includes the company’s AV division, grew at an 86% clip YOY with $1.53 billion in revenue for the period. The sector produced 16% of the company’s total revenue, up from 12% in FY2016, with about 50% of the growth coming from AV platform development. Research costs increased 23% during FY2018 YOY and 10% in FY2017 YOY.
NVDA (green line, main frame) fell into correction territory over the course of a three-day market plunge that preceded the January jobs report release on the first Friday of February, dropping 13%. The fall stopped just short of falling though the stock’s 50-day trading average (blue line, main frame) before recovering about four percentage points to the upside from investors’ profit taking in preceding weeks. By the 20th of February, NVDA had hit a YTD high of $249.08. The stock would hit another YTD high of $250.08 on the 16th of March before falling steadily to a trough of $214.25 on the 6th of April. This time, NVDA’s market slide was in response to two fatal car crashes: One of a 38-year old San Mateo man slamming into a center divider on a Mountain View freeway driving a Tesla Model X in autopilot, while the other involving an Uber AV taxi killing a pedestrian in the early morning hours on a Tempe street. Both accidents happened three days apart (19 March for Uber and 23 March for Tesla). The news sent NVDA crashing through its 50-day trading average. Through Friday’s market close, NVDA has again clawed back eight percentage points on the news with market momentum (bottom frame) to the upside and eking positive for the first time since the closing days of March (see Figure 3, above). Oversold?
GOOG and FB are both chronic data aggregators. Advertising comprises almost the totality of their respective revenue streams. While one uses a web of seemingly innocuous personal connections, the other mines the sheer digital footprint of users on its many media outlets. The end result is spectacularly similar—the aggregation of mountains of data for the use of myriad digital proofs targeting consumers of every walk of life, irrespective of state, national or international borders. The single requirement of completing the circle is access to a digital device. Growth depends not on the mining of more data per se, but the placement of more digital devices in the pockets of an ever-greater number of people across the face of the planet.
Given FB’s sine-qua-non is the ability to harvest personal information from its user base to generate the data for marketers is the reason why government regulation becomes such a fundamental threat to the company’s existing business model. The whole notion of balance between user privacy and the ability of the platform to provide ever-detailed information for marketers to target audiences by age, product preferences, physical GPS positioning at any given moment in the day, through 100s of millions of "like" buttons and “cookies” becomes highly problematic. These are issues that underscore the commanding position technology plays in our social, political and economic lives to which we now have become unwitting products. No one knows just how much raw data is actually collected and what is is subsequent disposition. This is a running testament to technology’s omnipotent reach that has outpaced, perhaps hopelessly so, that of oversight at all levels of government. The inferences that can be drawn from the litany of presumably innocuous pictures, biographical data, textual postings and so-called shadow profiles that vacuum up contacts in messaging apps says much about a person and his/her proclivities toward a particular product, a vacation spot—or a particular political cause. While on-line targeted advertising has morphed into a literal art form in recent years to the chagrin of many, until the Cambridge Analytica breach that literally exploded into the news in mid-March, there was little for would-be regulators to point to in the way of public harm being done by uber-invasive technology. The alleged connection with the 2016 US presidential electoral cycle or the Vote Leave campaign in Britain in June of 2016 added a clear political dimension to the incident. Accordingly, where that balance is forged between profit, privacy and transparency will invariably be a political, rather than an economic, call. But before the political process renders a response, the nature of the public harm being done needs to be clarified. Most political decision makers understand the potential for guns to render public harm. Few have the same awareness in the technology space. What can be said at this juncture is a beginning recognition among some political decision makers that industry self-regulation appears not to be the same answer as it has readily been in the past. Looking beyond for the moment the basic tenets of the current political gridlock, the next step is identifying the problem which runs the gamut from privacy protection to misinformation to sharing of data with third parties to political bias, to foreign interference in the political process to governmental interference in the innovation process to the monopoly status of big-data firms to terror prevention. The list is by no means exhaustive.
At the Zuckerberg hearing before a joint session of the Senate Judiciary and Commerce committee with an additional 42 senators attending the Tuesday session, lawmakers clearly struggled to understand the many nuances in FB’s business model. The logistics of gathering and disseminating information gleaned from a user’s FB platform footprint was equally baffling as was the ability of marketers to direct their targeted pitches back into the platform toward a pre-determined cross-section of FB users. Democrats pressed hard with questions about the Russian meddling in the 2016 election. Republicans concentrated their ire on FB’s handling of conservative media feeds on its platform. Politics aside, the takeaway here after a ten-hour barrage of withering questioning was pretty clear: No one dominating theme was readily discernible. It was a clear signal that any determination on regulating social media in any of its myriad forms is absolutely dead for the remainder of the current Congress as the November mid-term elections loom large on the horizon. Beyond the election, a regulatory determination will be years—years—in the making. Stuffing the genie back into the bottle has always been a difficult task.
Across the Pond, the European Commission’s General Data Protection Regulation (GDPR) which updates a 1995 version of the bill, is the first real government attempt at regulating the collection of data on consumers in the EU. The law comes into force on 25 May. This is legislation that attempts to employ for the first-time a cross-border reach to protect personal data packed with the firepower of fines up to 4% of the company’s global revenue for lapses and breaches. GDPR essentially brings every firm--past, present and future--that has collected, collects or will collect personal date on EU citizens under its purview. The implementation of the law will certainly be telling and could form the basis of other copy-cat national versions across the developed world and beyond.
At this juncture, however, the cost of providing a single targeted ad to a user's mobile device is likely the biggest check on future growth of big data gatherers like FB and GOOG--not government regulation.
Self-driving technology appears to be good enough to lull the designated driver, even experienced test drivers, into a false sense of security. Falling victim of complacency is the same dilemma that operators of any technology with the potential of causing immediate loss of life of any scale face due to a momentary lapse invariably brought on from tedium and routineness. Currently, the US has no regulatory standards for AVs. The technology still lacks the reliability to deploy blanketly on US roadways, to which the Mountain View and Tempe accidents duly attest.
At the same time, last week’s joint announcement that Toyota and VW will join forces in the development of AV trucks and buses is not only an attempt at overcoming the shortage of long-haul truck drivers and the expanding need for their service by transport companies across the developed world. The joint venture between Toyota’s Hino Motors and Volkswagen Truck & Bus GmbH underscores the recognition of both companies of the potential demand for AV driving technology for long-haul routes as well as for the outsized capital outlay needed to bring such technologies to market. NVDA’s total revenue in FY2018 for AVs remains small at $1.5 billion. But the sector’s growth potential is white-hot at 83% YOY. The demand added four percentage points YOY to the company’s total revenue during FY2018. The pace will likely only quicken in the years to come. Current technology flaws will be fixed and the pace of growth will move forward.
And yes, each of these stocks was oversold.
Disclosure: I am/we are long FB, GOOG, NVDA. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.