15 Stock TMT Model Portfolio: Winning Enterprise Cloud and Betting on Self-Service Analytics
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Tejas Raut Dessai
FOR IMPORTANT DISCLOSURES 203.901.1633 /.901.1634
psagawa@ / email@example.com
May 30, 2018
Through 5 months of 2018, the reported financials and stock performance of companies associated with the paradigmatic themes underlying our model portfolio continues to be outstanding. 1. Enterprise Cloud: Sustained, blisteringly fast revenue growth for top hosting platforms and widespread upside surprises from leading SaaS names are evidence of a generational paradigm change. 2. AI: Autonomous robocabs, AI assistants everywhere, uncanny image recognition, next-level business analytics – AI is real now and ready to drive real revenue. 3. 5G: The TMUS-S deal may hold up US investment a bit, but 5G spending will be big in 2019. 4. Blockchain: It’s early, but coin-mining is still a thing, IT consultants are busy, and hosts are getting ready for commercial applications. 5. Digital Media: There is still a major chapter yet to play out, as ad dollars follow viewers from linear TV to streaming platforms. 6. E-commerce: Reported revenues underestimate AMZN’s dominance – by GMV it will pass WMT by year end. Our 15-stock model portfolio appreciated by 6.3% since the end of February, beating the S&P500 by 841bp and the tech components of that index by 424bp over that time. All 15 stocks exceeded consensus expectations for both sales and EPS during their most recently reported quarters. Midway through the quarter, we changed out FB in favor of QCOM on fear of an advertiser backlash, a move that proved premature. Today, we are adding AYX – growing more than 50% 6-months after its IPO, already cash flow positive and trading at a P/S in line with more mature companies. To make room, we are removing WDAY. While we expect it to continue to deliver strong results, we see less upside and less likelihood of acquisition than AYX at its current market cap.
- Enterprise cloud migration – full steam ahead. Industry analyst Synergy estimates that public cloud computing sales accelerated to 51% annual growth in 1Q18, up from mid-40’s growth a year ago. Market leader AMZN’s AWS grew 49% YoY. Number 2 MSFT’s Azure grew 93%. Third place IBM’s cloud “as-a-service” revenues, which includes slower growing applications, was up 25%. Number 4 GOOGL does not report revenues for its GCP cloud service, but 3rd parties estimate its sales more than doubled. At the same time, nearly every major SaaS application vendor exceeded consensus expectations, including CRM, ADBE, WDAY, NOW, ADSK, DATA, SHOP, VEEV and ZEN. In support of this, cloud datacenter CAPEX was up big (see GOOGL, MSFT, FB, AMZN and others) driving big sales in the segment for chipmakers like INTC, NVDA, MLNX, and others. We believe that this transition will have legs to last for years.
- AI is coming of age. Big developer conferences from GOOGL, MSFT and FB in the quarter highlighted big advances in AI applications. AMZN and GOOGL announced major enhancements to their AI virtual assistants, with AAPL believed poised to do the same for Siri this week. GOOGL’s Waymo started offering fully autonomous rides to consumers in Phoenix, in anticipation of a full commercial launch later this year. All the top cloud hosts now tout their major investments in AI related infrastructure, APIs and development tools, as CEOs across disparate industries call out their interest in AI infused applications. GS and JPM each hired high profile AI scientists to head their fledgling development programs.
- TMUS-S will delay US 5G, but big investment is coming. TMUS is making the pending 5G transition a key plank in its case for regulatory approval of its deal for S. In the meantime, the uncertainty will limit its progress on the technology, giving both VZ and T leeway to be less aggressive themselves. Still, China and Korea are moving quickly, so expect the DoJ to expedite its review, given the importance the administration has placed on 5G leadership. The political climate is also favorable to QCOM – approval of its deal for NXP appears to be a quid pro quo for easing sanctions on ZTE, perhaps also presaging a new licensing deal with Huawei.
- It’s still early for Blockchain. While crypto-coin investors cope with skeptical news flow and volatile trading, hundreds of commercial enterprises are exploring high potential applications for the underlying blockchain tech. By its nature, blockchain involves cooperation across multiple organizations creating an obstacle as leaders work to recruit critical masses to their proposed transaction networks. So far, the beneficiaries are the IT consultants – IBM and ACN – which are playing a key role in development.
- Digital media is killing TV. TV audiences have been shrinking since about 2010, the declines obscured by NSLN’s skewed ratings methodology. TV ad sales have been surprisingly robust, saved by rising CPMs and higher ad loads, but the inflection point seems upon us. Buyer commitments from the recently completed Upfronts are not yet available, but continued audience deterioration and the clear enthusiasm for digital – GOOGL and FB killed it in 1Q18 ad sales – portends weakness. NFLX, which doesn’t sell ads, continues to syphon eyeballs from linear TV at an impressive rate.
- AMZN is even more dominant. 3rd party sales grew 43.9% YoY in 1Q18, making gross merchandize volumes through AMZN ~$93B for the quarter, just 24% shy of WMT. At this pace, AMZN will pass the leader by year end, with the rising % of 3rd party sales fueling higher margins to boot. Prime continues strong growth, as AMZN keeps adding perks, and should sustain momentum despite a price hike. At this point, government intervention may be the only real check on the company’s potential size. For would-be retail rivals, partnering with GOOGL may be the only viable option.
- Our model portfolio beat the tech benchmark by 350bp. The 15 stocks in our model portfolio were up 632bp since our 2/23/18 update, beating the S&P500 by 841bp and the tech elements of that index by 424bp. This includes the effect of swapping in QCOM for FB on 4/10/18, which hurt performance by -253bp. ZEN, NFLX, KEYS and CRM were the top performers, while FB, IBM, GOOGL, ACN, and XLNX were all down for the quarter.
- Adding AYX, dropping WDAY to make room. SaaS data analytics application developer AYX is up 130% in the 14 months since its IPO, on better than 50% sales growth and surprisingly positive cash flow. We believe that the strong enterprise SaaS market and AYX’s unique product that allows enterprises to adapt their structured data to the needs of modern analytics will allow it to beat consensus expectations for decelerating growth. Furthermore, we believe AYX is a strong candidate for acquisition by any number of potential buyers. To make room for AYX, we are removing WDAY. While we are still fans of the company, its products and management, its sales are now large enough to make sustaining its 30%+ sales growth difficult and its $28B market cap makes it a less likely acquisition.
Exh 1: SSR TMT Model Portfolio Performance – Before Reconstitution
Six Themes Driving TMT Model Portfolio Performance
We believe that we are in the third inning of a new generation in TMT. The previous era – sparked by the introduction of PCs and cell phones, the break-up of AT&T, and the deregulation of Cable TV – saw the democratization of computing and the rise of networking give birth to the Internet. The Cloud/AI Era had its seeds in the invention of a datacenter architecture that could break through scale limitations and the smartphone revolution. Essentially limitless and super cheap computing resources matched to oceans of data made AI viable, supplying a revolutionary software architecture to match the tectonic shift in hardware. This is defining a new dominant paradigm for information use – nearly ubiquitous access to powerful, intuitive, convenient and labor-saving applications running in the cloud. Now about a decade into this Cloud/AI Era, several major themes for the next decade are apparent.
- Enterprise Cloud: Cloud computing is now normal. Traditional IT shops are planning their transition from their legacy apps and systems to SaaS apps and public cloud hosting. There are switching costs and other obstacles that will stretch out the migration for many, many years, but the exodus has begun. (http://www.ssrllc.com/publication/37440/, http://www.ssrllc.com/publication/the-cloudai-era-a-perspective-on-the-next-decade-of-tmt-investing-2/)
- AI: The Deep Learning systems at the leading edge of AI demand huge datasets, copious computing resources and still scarce scientific talent. These are concentrated in a handful of platform players that are leading the expansion of AI into exciting new applications – autonomous robocabs, virtual assistants, AR, predictive analytics, etc. GOOGL, MSFT, IBM, FB, and AMZN are leading. (http://www.ssrllc.com/publication/a-deep-learning-primer-the-reality-may-exceed-the-hype/, http://www.ssrllc.com/publication/ai-where-the-brains-are/, http://www.ssrllc.com/publication/ai-as-a-service-deep-learning-is-fundamental/, http://www.ssrllc.com/publication/ai-data-food-for-artificial-thought/)
- 5G: Wireless network technology continues to progress network generation to network generation. 5G is upon and is an unusually big step, bringing with it huge increases in network capacity and connection speed, along with similarly large decreases in per unit network costs and in latency. Deployment is proceeding more quickly than many presupposed, with major implications for wireless carriers and technology suppliers. (http://www.ssrllc.com/publication/5g-hyperscale-ai-5g-sea-change/, http://www.ssrllc.com/publication/5g-rising-global-carrier-competition-to-drive-capex/, http://www.ssrllc.com/publication/5g-test-equipment-adding-keys-to-our-model-portfolio/)
- Blockchain: Blockchain technology was born of necessity during the financial crisis, in many ways orthogonal to the Cloud/AI paradigm, but its usefulness is amplified by the rise of cloud platforms. Future transaction networks, spanning many organizations, will very likely be hosted in the cloud, and while we think it will take a few years for most commercial implementations to gain critical mass, the opportunity for the major cloud platforms (IBM, MSFT, AMZN, GOOGL) will be substantial. Meanwhile, IT consultants (IBM, ACN) are busy laying groundwork. (http://www.ssrllc.com/publication/commercial-blockchain-objects-may-be-further-away-than-they-appear/, http://www.ssrllc.com/publication/nvda-ai-upside-greatly-outweighs-crypto-risks/, http://www.ssrllc.com/publication/ai-blockchain-and-the-cloud-banks-in-the-new-technology-era/)
- Digital Media: Reach to billions of connected consumers from powerful cloud platforms has already spurred a revolution in advertising and media. Global digital ad spending passed TV spending about a year ago, and is now poised to begin eroding linear television revenues directly, as streaming video continues to syphon viewers and engagement away from the traditional medium. (http://www.ssrllc.com/publication/advertising-the-golden-age-of-tv-enters-its-golden-years/, http://www.ssrllc.com/publication/war-on-tv-part-iv-the-biggest-pipe-will-not-win/, http://www.ssrllc.com/publication/war-on-tv-part-ii-streaming-is-coming/, http://www.ssrllc.com/publication/the-war-on-tv-the-attack-of-the-boxes/, )
- E-Commerce: In the US, this means AMZN. Considering that 20% of its revenue are 3rd party sales commissions (about 17% of the sale price), its gross merchandise volume for 2018 should top $450B, striking distance from WMT’s expected $513B in sales. AMZN’s 30%+ growth rate will see its GMV blow past WMT in 2019. For struggling retailers, partnering with GOOGL may be the last, best option to the AMZN marketplace. (http://www.ssrllc.com/publication/tmt-worrying-about-the-government/, http://www.ssrllc.com/publication/the-cloudai-era-a-perspective-on-the-next-decade-of-tmt-investing-2/, http://www.ssrllc.com/publication/ten-investible-things-that-we-think-will-happen-in-2018/, http://www.ssrllc.com/publication/amzn-winning-christmas/)
Exh 2: SSR’s Key Investment Themes for TMT in 2018
Progress in 1Q18 – By Theme
Enterprise Cloud – Synergy Research Group estimates that 1Q18 global cloud hosting revenues were up 51% YoY to $15B+, an acceleration from 43-45% growth in Q1-3 in 2017 (Exhibit 3, 4). Market leader Amazon Web Services was just under that at 49%, but still holds a 33% share of the market. Microsoft’s Azure, growing at a blistering 93%, gained ground to more than 13% share. IBM, growing about 30%, is down a tad YoY to 8% share. Google Cloud Platform went from 5% to 7% of the market YoY, implying a more than doubling of sales. On that trajectory, cloud hosting should be a better than $75B market for the full year. We believe that the long-term opportunity is more than 10 times that.
The top 20 SaaS application players almost universally topped expectations in 1Q18, delivering average sales growth of 29% YoY. Just yesterday, Salesforce delivered 25.4% YoY growth for a top and bottom line beat, on the back of its announced deal for Mulesoft (Exhibit 5). Fold in the success of Microsoft’s cloud applications – commercial Office 365 sales were up 42% YoY – and the enterprise migration to the cloud is even clearer. We believe that there is considerable further runway for SaaS market growth.
AI – There was a wash of AI news during 1Q18. Google’s I/O conference showed off big steps forward for its AI-powered Assistant – it can now make phone calls to schedule appointments, amongst other things – and integrated new AI capabilities to Photos, Maps (AR walking directions), and News. It also announced a new iteration of its TPU ASIC for AI processing and confirmed that Waymo will begin offering commercial robocab service across Phoenix by year end. Earlier in the quarter, Waymo pulled human monitors from its autonomous test fleet in Arizona and ordered 20,000 new vehicles from Jaguar. Indeed, Alphabet is rebranding its entire Google Research organization as Google AI.
Exh 3: Cloud Infrastructure Services Quarterly Market Share Trend, 4Q15 – 1Q18
Exh 4: Public Cloud Services Market Size Forecast, 2017 – 2021
Exh 5: Latest Reported Quarterly Surprise by Top 20 SaaS Companies
Microsoft’s Build conference also focused on AI. It announced a dedicated Azure hardware platform for AI, a range of AI-powered tools and APIs for Azure customers, and the integration of its Cortana AI assistant with Amazon’s Alexa. Facebook’s F8 confab was another AI fest – natural language processing, language translation, AR, image recognition, AI developer tools, etc. were all day one keynote discussion topics. Amazon didn’t host a conference but continues the relentless improvement of its popular Alexa AI assistant and its AI support on AWS. Apple begins its annual WWDC event next week and has hinted at major improvements to its Siri AI assistant platform.
5G – T-Mobile’s planned merger with Sprint will slow network deployment while the deal works its way through judicial review and while the two companies define plans for combining their network assets. In turn, this will mute pressure on Verizon and T-Mobile to move more quickly on their own 5G plans (Exhibit 6). Still, we expect DoJ approval and believe that a combined T-Mobile/Sprint will be a stronger rival, likely with aggressive 5G build-out commitments tied to the government’s sign off on the deal. This should mean a sharp acceleration in spending in 2019 for all the US carriers. Meanwhile, Chinese, Korean and Japanese carriers view 5G as significant imperative, with government support to push country technology champions – like Huawei, ZTE, Samsung and NTT-DoCoMo – forward. We see early cycle investment opportunities, such as test equipment (Keysight, National Instrument, Anritsu) or optical equipment (Ciena, Infinera), as the most promising.
Blockchain – Cryptocurrency gets most of the headlines – coin prices were volatile during the quarter, with skeptical comments from Warren Buffet and others serving to dampen a recovery from the severe downturn that started the year. Still, we believe mining demand for GPUs will see minimal impact, as softening retail prices for graphics cards are cutting into distributor mark-ups (as much as 95% at one time) and not chipmaker revenues. We believe ASIC-based mining rigs had already taken the volume end of the coin-mining market anyway, and that the GPU-based miners likely use their computers for other purposes
Exh 6: Capex as Percentage of Sales Forecast for US Carriers
as well, making their graphics card purchases less dependent on the strict economics of the crypto market. Moreover, Fortnite is quite popular.
As for the blockchain technology that underlies cryptocurrencies, commercial applications are taking their time to gain critical mass. IT consultants, i.e. IBM and Accenture, are very busy working on prototype implementations, from which their customers hope to recruit a necessary critical mass of competitors/partners to form viable transaction networks. Eventually, we expect cloud hosts – i.e. IBM and Microsoft – to provide platforms for high-performance, secure blockchain transaction networks. This may be a meaningful business in 3-5 years, but today it is nascent.
Digital Media – Netflix is a juggernaut. In 1Q18, it grew subscribers 6% QoQ, blowing past consensus expectations. Its 125M subscribers are streaming more than a billion hours of programming per week, growing its average audience at a better than 30% YoY pace. Along with Alphabet’s YouTube, which streams more than a billion hours of video per DAY around the world, Netflix is eroding the audience for traditional linear TV at a furious pace (Exhibit 7). Poor ratings for TV network shows are an old story – the once thought impregnable NFL has been in decline for years, as has nearly every important viewing event. US linear TV viewership has been in decline since at least 2011, the trajectory semi-hidden by Nielsen’s methodology which we believe systematically overcounts passive viewers. Even the audience that is still watching is less attentive, particularly to the ad spots that pay the bills.
Meanwhile, subscriber numbers for Pay TV are also down, with consumers also opting for skinnier bundles without expensive channels that they seldom watch. This is a double whammy, further cutting the audience for advertising purposes, while paring back subscription fees. The saving grace for media companies has been their ability to squeeze more minutes of advertising per hour of programming and higher prices from advertisers. This is starting to show in media numbers – NBC (with Olympics and Superbowl revenues) and CBS have bucked the trend, but soft audience numbers suggest that the dark clouds will head that way next.
Exh 7: US Time Spent per Day on Major Media, US Q4, 2017
Exh 8: Global Digital Ad Spending Forecast, 2016 – 2022
At the same time, digital advertising – led by the dominant Google and Facebook platforms – is going gangbusters (Exhibit 8). Both companies easily topped ad sales expectations for the quarter, up 26% and 49% respectively. Together, the two behemoths represent more than 70% of US digital ad spending, which recently passed TV to become the biggest category of media advertising at 41% share. While the runway is not endless, we believe TV, still 35% of ad spending, will provide ample nourishment to fuel digital ad growth for the intermediate term.
E-Commerce – Amazon posted extraordinary results in 1Q18, with total North American on-line sales up 46% YoY to more than $30B. While this is obviously impressive, observers continue to understate the company’s retail presence. Nearly 20% of Amazon’s retail revenues were from 3rd party sales where it recognizes only the fees that it earns for booking and fulfilling sales (Exhibit 9). With these fees estimated at about 15% of the total price, the gross merchandise volume for products sold through Amazon was nearly $62B. Given the blistering sales growth, we expect Amazon to challenge Walmart for its spot at the top of the developed world’s retailers by year end. This is a big deal, and with B2B wholesale, drugs, and other major markets available to it, we expect Amazon’s e-commerce business to keep growing for a long time, absent government intervention.
This is the obvious big risk. The Trump administration continues to harrumph about Amazon’s use of the US Postal Service – although there doesn’t seem to be much evidence of preferential treatment. European regulators who have come down hard on Alphabet and turned their focus to Facebook could take on Amazon as well. A new administration in the US could prove even more hostile than the current one. Still, we are skeptical that restricting competitive analysis to the on-line segment dominated by Amazon would be appropriate under US antitrust law.
Exh 9: AMZN 3rd Party Seller Services Quarterly Sales, 1Q16 – 1Q18