Portfolio Update: 2019 Research Review – So Far, So Good
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Tejas Raut Dessai
FOR IMPORTANT DISCLOSURES 203.901.1633/.901.1634
psagawa@ / firstname.lastname@example.org
December 9, 2019
Portfolio Update: 2019 Research Review – So Far, So Good
Our model portfolio is built to exploit the paradigmatic shifts that are remaking the TMT landscape. In 2019, we wrote at length on five of these shifts – 1) The enterprise cloud – hosting platforms, hybrid cloud management, and SaaS applications; 2) 5G – strong early buildout spending, device uptake in 2021, gaming & broadband replacement as long term drivers; 3) Fintech – revolution at POS, the imminent threat of digital payments, the future risk to consumer banking; 4) Streaming – NFLX is VERY well positioned, streaming gaming is coming; 5) Smartphones have plateaued – services and peripherals have more opportunity, AAPL’s walled garden approach is risky. Performance of stocks tied to these themes has been generally good. Our 15-stock model portfolio is ahead of the S&P 500 but behind the tech elements of the index both for the past 3 months and YTD – we suffered for not playing the chip value chain and payments processors. We are removing CIEN, as shifting carrier investment weakens the optical market demand. We are adding back NVDA, which we had removed at the start of the year in the face of channel inventory problems but which we believe faces easier compares and resurgent data center demand for 2020.
- 4Q model portfolio performance was okay. Since our last update, the portfolio is up 608 bp, 40 bp better than the S&P 500 and 265 bp behind the tech components of that index. As noted in Sept., our performance vs. the tech benchmark has suffered in 2019 from our low exposure to semiconductors and payments processors. Still, the names associated with our main 2019 research themes, both in and out of the portfolio, have been strong performers – with the notable exception of NFLX. Top-line surprises in the portfolio have picked up since 1Q and EPS surprises have remained strong.
- Enterprises are accelerating their move to the cloud. The best-in-class SaaS players have delivered extraordinary growth, overdelivering against the expectations of analysts who annually predict a negative inflection point. We believe that the TAM is much larger than consensus models, with considerable further runway for outperformance (SaaS: Handicapping the Unicorn Races). Likewise, the big three cloud platforms (AMZN, MSFT and GOOGL) are delivering faster sales growth than many have expected, with a strong likelihood that the CAPEX rebound that began in 2H19 will accelerate in 2020 (Cloud Software: A Long Journey with Three Layers). Finally, enterprise demand for tools and services to manage their complex hybrid cloud IT assets are gaining great traction. The performance of focused players like NTNX and PVTL has been obscured by messy business model transitions, while old line IT suppliers with skin in the game, like IBM, CSCO and especially VMW, have their hybrid cloud-based growth tempered by declines in legacy businesses. (Hybrid Cloud Platforms: Herding Cats and Dogs Together – Efficiently and Securely, SaaS: Handicapping the Unicorn Races)
- 5G investment is still in its first phase. We believe carriers and the analysts that follow them have undersold the capex that will be necessary to remain competitive in the coming shift to 5G. (5G: Why TMUS will Win) Our early phase picks – KEYS, CIEN and XLNX – have posted strong growth, although investors appear to be concerned for 2020. At the same time, we believe many are jumping the gun on 2020 5G smartphone forecasts. We believe that device demand (and thus, component demand) will remain moribund until the very end of next year, as early network coverage and device pricing will dampen interest. We believe demand will pick up in 2021, and that gaming and broadband replacement will be big drivers.
- Digital players threaten the financial services industry. New solutions that help retailers integrate their point-of-sale systems with e-commerce and their business back office, while improving the checkout experience for both customers and employees have enormous traction with merchants. (Digital Payments: Revolution at the Register). This is great for disruptors like SQ (who have posted strong growth) and a serious threat to merchant acquirers, who are also squeezed by merchant banks on the other side. Card issuing banks will have their own problems, as digital franchise players like AMZN, GOOGL, and AAPL expand from simple credit cards to offer much broader financial product portfolios in the future (Digital Consumer Banking: What Will GAFA Do?). News flow continues to reveal the ambitions of the internet giants.
- The streaming era has begun. As expected, DIS, T, CMCSA, AAPL and others have announced streaming video services. (The End of TV) With 160M+ subs worldwide, an original programming budget of $15B (more than all competitors combined), enormous momentum in international markets (more than half its sales come from countries where its US rivals barely play), NFLX is and will be the dominant streaming video platform, with plenty of runway to continue its growth. Expect the competition to jockey for position behind NFLX (and likely DIS) as PayTV cord cutting accelerates. Gaming is going streaming too, but the transition will take much longer as tech obstacles (e.g. complex I/O, low latency tolerance, bandwidth/processing requirements, etc.) are resolved. (Game Streaming: Not Everyone Can Win) We expect 5G, which offers 75%+ reduction in network latency vs. even wired broadband, to catalyze cloud gaming adoption both for mobile and residential use cases.
- Mature smartphone market looks to services and peripherals for growth. The global installed base of smartphones has plateaued at just over 4B users. Any growth in global sales will come from sub-$100 models replacing the ~1B feature phones and from fluctuations in the rate of replacement, which has been declining steadily over the past 3 years. (Smartphones: The Market is Mature – Now What?) We believe that the inflection point won’t come until 2021, when 5G is broadly enough available to spur new upgrades on a global basis. Meanwhile, the massive base of smartphones is driving engagement for web-based services, thus far benefitting AAPL and GOOGL, whose app stores (with their 30% fees) dominate access to mobile platforms. We believe this model will face pressure. Auxiliary access devices (e.g. watches, smart earbuds, home speakers, etc.) should also see major growth, further strengthening the case for AI assistants.
- Replacing CIEN with NVDA. CIEN shares have stalled, with Chinese trade issues a considerable overhang. With 5G buildouts likely to move into phase 2 deployments by 2021, we believe optical investment could decelerate. In its place, we are returning NVDA to the portfolio. We had removed it at the beginning of the year in response to the inventory glut of graphics cards in the channel. With much easier compares for 2020 and a reacceleration in datacenter investment on tap, we see upside vs. tempered expectations.
Our Major 2019 Research Themes are Playing Out … So Far
Our model portfolio performance hit a rocky patch in mid-2019, with the implosion of then constituent PVTL in March spoiling an otherwise strong first half and the whipsaw effect of the China trade war driving investors to names perceived to be safer over the summer. Moreover, as we noted in our September update, our skepticism on the widely expected semiconductor cycle turn for 2020 and for the prospects of the newly consolidated merchant payments processors has been a counter-consensus drag vs. the broader tech benchmark. Still, the portfolio is up 29.1% YTD and 6.1% since our last update (Exhibit 1, 2, 3), well ahead of the buoyant S&P 500 (+375 bp YTD and +40 bp since September while trailing the tech components of that index (-1204 bp YTD and 265 bp since September).
Over the course of 2019, we published research on a variety of TMT topics, with five key themes predominant. 1) The Enterprise Move to the Cloud – extraordinary growth in 3 areas: SaaS applications, cloud hosting and hybrid management solutions; 2) The Inevitable Move to 5G – carriers will spend more than they want to, device sales will disappoint before they surprise, and gaming/residential broadband replacement will be the killer apps; 3) Digital Players are Coming for Financial Services – internet platforms will build from credit cards to grab the primary consumer relationships, superior digital point of sale solutions (integrated with online and back office) will squeeze merchant acquirers from one side while commercial banks push from the other; 4) All Media Will Migrate to Streaming Platforms – NFLX is in great position, not all streamers will win, and cloud gaming will be a big thing … in time; and 5) Alternative devices and Services are More Interesting Than Smartphones – The smartphone installed base has plateaued and 5G won’t accelerate replacement until 2021, and 3rd party services/peripherals will test walled gardens.
Companies well positioned against the opportunities created by these vectors of change delivered strong growth in 2019, broadly exceeding market expectations on both top and bottom lines with a few exceptions. However, investors often responded cautiously to this business performance, in many cases driving sales multiples lower despite considerable upside surprise. We believe this caution is misplaced and expect our favored stocks in each category to continue to surprise to the upside in 2020. At the beginning of this year, we published our take on valuation (Model Portfolio Update: What is Value in Tech?), asserting that the widely used DCF is a flawed methodology for valuing high growth companies resulting in a significant downward bias as applied by sell-side analysts. This is a substantial opportunity for investors.
We are making one change to the portfolio with this update. CIEN has been a great stock for us, up 61.1% since we added it two years ago. However, we are concerned that sales acceleration with the ramp in 5G spending amidst an upgrade cycle has peaked and that the exposure to China adds risk. We will remove it. In its place, we are adding back NVDA. At the beginning of 2019, NVDA management revealed a significant glut of graphics card inventory in the distribution channel, the consequence of a dramatic pullback in demand by cryptocurrency miners. We removed it at that time, taking the majority of market’s negative reaction before we could make the change. Since then, NVDA has recovered and then some. We see upside to consensus sales expectations, as we believe analysts are too conservative on datacenter capital investment. Better late than never.
Exh 1: SSR TMT 15 Stock Model Portfolio Performance Since Last Update
The Exodus to the Cloud
We started the year with major pieces on SaaS applications and Hybrid Cloud Integration (Hybrid Cloud Platforms: Herding Cats and Dogs Together – Efficiently and Securely, SaaS: Handicapping the Unicorn Races) and addressed important aspects of the changing enterprise landscape within our work on cybersecurity (Cybersecurity: Dumb Users and State Sponsored Cyberweapons) and semiconductor demand (Semi Forecasts: Too Bullish on Smartphones, Too Bearish on Cloud). Essentially, we see the enterprise IT market in the early days of what we expect to be a nearly comprehensive shift to the cloud. In this, the addressable market for cloud-focused solutions in not just enterprise spending on on-premises hardware and software, but also all of the related spending toward supporting IT operations – e.g. people, facilities, communications services, power, etc. (Exhibit 4) In sum, this suggests a global addressable market of more than $3.5T annually, a heady opportunity for products and services that today generate $460B in revenues.
In this context, we believe consensus growth expectations for many cloud-focused companies – best-of-breed SaaS players, leading edge hybrid cloud management solutions, and hyperscale datacenter operators chief amongst them – are far too conservative (Exhibit 5). This is particularly evident in the SaaS arena, where the top names have delivered consistent, and, in some cases, even accelerating growth in the face of consensus projections for a sharp inflection into deceleration. This has yielded a consistent run of upside surprises yet estimates for 2020 still embed a sharp market deceleration. A third of our portfolio – MSFT,
Exh 3: TMT portfolio cumulative performance relative to benchmark – YTD
NOW, ZEN, AYX and OKTA – derive most of their sales and profits from enterprise applications and are leaders in the transition to the SaaS model. We could easily list another 15 that deserve consideration for inclusion, including former constituents CRM, WDAY and ADBE. Similarly, broad semiconductor forecasts from industry analysts anticipate very modest single digit growth in hyperscale datacenter investment over the next few years. Given 2019 revenue growth of better than 40% for a hosting market that is on a very gentle glide path, this seems almost absurdly pessimistic. This suggests substantial upside for component vendors with exposure to this market – we remain bullish for XLNX, NVDA and others. Moreover, this dim view of necessary capex may also flag pessimism for the cloud platforms themselves. We believe AMZN, MSFT and GOOGL could each grow to better than $100 billion in long term annual revenue (Exhibit 6).
Finally, we are bullish for the companies that provide tools and services to enterprises struggling with the transition to hybrid architecture. At the core, a handful of companies are offering “plumbing” to help distribute processing and storage between internal datacenters and cloud hosts. We have NTNX in the model portfolio – it struggled in early 2019 as investors were befuddled by the near-term implications of its shift from bundled hardware and licensed software to a purely subscription software model. It seems to have finally gotten its feet beneath it. VMW is the other big name in hybrid cloud integration – it is beginning to show its promise as its moribund private datacenter-oriented virtualization franchise gives way to its cloud business. Other classic IT names also play in this space – IBM, CSCO, and HPE come to mind, perhaps in that order – all waiting for growth businesses like hybrid cloud outweigh their declining core units. Finally, the big three hosts, MSFT, GOOGL and AMZN all see hybrid cloud integration as both an opportunity and a vulnerability and will be aggressive in the space.
Exh 4: Global Public Cloud Spending in all 3 Categories Forecast, 2017 – 2021E
Exh 5: Consensus Forecasts for SaaS Sales Growth are overly Conservative
Exh 6: SSR Forecast of Total Addressable Market for Cloud Services by Category
Here Comes 5G
For carriers, 5G is an inconvenient necessity. Each generational upgrade – from 1G to 2G to 3G to 4G – has been about addressing network congestion and this one is no exception. 5G offers more than a tenfold increase in per cell capacity, enabling broad swaths of previously ineligible spectrum to be included and facilitating much finer cell splitting in the process. With mobile data traffic increasing at nearly 50% per year, operators will have no choice but to commit to 5G to cope. Verizon and AT&T may downplay the capex that will likely be required but should the T-Mobile/Sprint merger finally be consummated; they will have little choice but to follow aggressively or risk hemorrhaging market share (Exhibit 7). We wrote at length about this prisoner’s dilemma earlier this year (5G: Why TMUS will Win).
While we believe broad 5G investment is inevitable (Exhibit 8, 9), the timing will follow in phases. (The 3 Phases of 5G: Coverage, Density and Applications) Phase one is well underway, as component suppliers introduce necessary parts, and carriers bid out initial coverage buildouts to equipment vendors, sign tower leases and secure backhaul capacity for their planned networks. Our picks for this phase – testing leader Keysight, optical backhaul supplier Ciena and base station component player Xilinx – have been strong contributors to our portfolio performance since their addition, although Xilinx hit a rough patch earlier this year and Ciena’s momentum may be petering out, as evidenced by a recent high-profile downgrade (Exhibit 10). While the next phase will see substantial macro cell deployments, we are too gun shy to recommend any of the suppliers – Huawei, Ericsson or Nokia – fearful that overzealous bidding for new contracts will result in margin disappointments, a phenomenon that has punished investors in every previous building cycle.
Ex 7: SSR Analysis of Options and Outcomes of Wireless Prisoners’ Dilemma
Exh 8: SSR Thesis of 5G Opportunity, broken down in deployment phases
Exh 9: Snapshot of US Carrier Capex Spending during 3G and 4G buildout
Exh 10: CIEN and XLNX benefitted from early 5G buildout, but growth is slowing
Analysts and investors seem to expect 5G smartphones to drive a major upgrade cycle beginning in 2020. We are skeptical of this timing. Service availability will be very limited for most of the year, both by the number of markets served and the completeness of coverage. Consumer use cases have not been well articulated and only a few flagship devices are in the market. We expect carriers to have to return to subsidies to induce users to switch to the new standard, but it will be a while before they are anxious enough to open that expensive Pandora’s box. Apple won’t even have a 5g iPhone model until 4Q. This means Qualcomm’s hopeful forecast of more than 200M 5G phones sold in 2020, echoed by both Gartner and IDC, is quite optimistic indeed.
Longer term, we are 5G believers. The new wireless standard is designed to offer a dramatic improvement in latency, down to 4ms vs. the 40ms typical on 4G LTE and the 15-30ms experienced on fixed residential broadband networks. This is a big deal for the $150B gaming market, and we believe, likely to be the biggest driver of accelerating phone upgrades and residential broadband cord cutting. On the latter point, we have been writing of 5G’s potential as a fixed network replacement for years. We have reached a point where network connection speeds (easily 200Mbps+ for 5G) are much more than necessary for all known applications (8K video will be well below 100Mbps as new compression standards are adopted). A combined T-Mobile/Sprint will be a powerful competitor in US 5G with superior performance, coverage and costs vs. the currently prevailing leadership duopoly. We patiently hold TMUS in the model portfolio, awaiting merger approval.
The Bank of Amazon (and Google and Apple)
Over the summer we wrote on fintech (Digital Consumer Banking: What Will GAFA Do?) highlighting the threat that digital big digital consumer platforms Apple, Google and Amazon posed to traditional banks. Credit cards are the start, but deposit accounts and loans are obvious opportunities for companies that would see enormous synergies to their core businesses. Indeed, these giants need not make meaningful profits from consumer financial services – increased customer loyalty, reduced friction for transactions and valuable user data would more than enough incentive. Moreover, digital players have several major advantages over banks – massive consumer reach, technology leadership, no costly branch banking infrastructure, etc. Surveys of younger demographics reveal an enthusiasm for banking with these omnipresent online brands that should be very troubling for banking executives (Exhibit 11, 12). We have both Google and Amazon in our model portfolio, both of which have been modest performers this year.
We also wrote on a revolution at the retail point of sale (Digital Payments: Revolution at the Register), as merchants look to offer better checkout experiences while tying in-store systems to their e-commerce solutions and integrating the whole back to their back-office software. Portfolio constituent Square, added 3 months ago, is at the forefront of this change and is delivering 43% YoY growth as it moves from small merchants into bigger chains. The traditional merchant acquirors have responded to this threat by consolidating – 7 global players have combined into 3, spurring a furious rally in the shares of the buffed-up leaders. Still, we are concerned that they will be squeezed between the tech players fomenting change at POS and merchant banks flexing from the other direction. For example, Chase Paymentech just launched a payments processing service that promises to settle credit transactions and deposit payments to a merchant account on the same day.
Exh 12: Average cost per digital customer acquired for banks is significantly high
The narrative on Netflix turned south in 2019 after a negative surprise on net subscriber additions in the seasonally weak 2Q and in the face of anxiety over the pending entry of media giants Disney, AT&T (Time Warner/HBO), Comcast/NBCUniversal, and others (including Apple) into the US streaming market. Netflix numbers were back on track for 3Q19 but the damage was done and with the looming launch of Disney+ and AppleTV, investors have treated the stock with great caution (Exhibit 13). We wrote about the streaming market in the wake of all this (The End of TV: Media Strains to Adapt to Streaming ERa), drawing the conclusions that a) the market was never going to be winner take all, b) that consumers would pay for new streaming by paring back PayTV rather than cancelling existing streaming services, and c) that NFLX is by FAR the best positioned platform, not just in the US but worldwide. It remains in our model portfolio and we believe it is poised for a run as the anxiety around the Disney+ launch subsides.
We also published a report on streaming gaming (Game Streaming: Not Everyone Can Win). While significant technical obstacles have kept gaming platforms years behind streaming video, we see the rise of cloud-based gaming as inevitable. 5G is likely to be a significant enabler for streaming games, as it offers a dramatic reduction in network induced latency, a big deal for gamers used to nearly instantaneous performance from their consoles. To date, the obvious allure of online play across broad geographies has been tempered by frustration at the lag between pressing a button and seeing the results on screen. 5G
Exh 13: NFLX Cumulative Historical Quarterly Subscriber Base until 3Q19
Exh 14: Key Elements Driving a Paradigm Shift in Gaming
Exh 15: SSR Estimate of Global Gaming Sales, 2019 – 2025E
Exh 16: Leading Cloud Gaming Platform Players Scored on Critical Criteria
should open online play for much more sophisticated games (Exhibit 14, 15). While it is very early days, we see Microsoft as the obvious winner. That archrival Sony has chosen Microsoft Azure as its partner for its own game streaming services is a clear sign of its strength (Exhibit 16).
For the rest of the players, our crystal ball is murkier. Nintendo has extraordinary momentum behind its hit Switch platform and an enviable slate of tentpole game franchises but it has zero experience in cloud computing. Sony’s PlayStation has the biggest installed base but a less impressive roster of exclusive content. The big publishers (which includes Sony, Microsoft and Nintendo) will see top developers gain leverage in negotiations, just as proven video content producers, like JJ Abrams, have been able to cash in as Netflix, Disney and others ramp up. Look for a margin squeeze to emerge with time. The non-console publishers may struggle to build independent streaming platforms and could be squeezed from that direction as well.
Exh 17: Global Smartphone Unit Shipments growth has declined, 2010 – 2019
Smartphones and Services and Alternative Access Devices
The smartphone market is largely saturated and any significant further growth in the installed base will come from very cheap smartphones replacing very cheap old school feature phones (Exhibit 17). Sales of new smartphones will ebb and flow with replacement rates, which will be driven by carrier incentives and by compelling innovation. We have written that we do NOT expect an inflection point in replacement (which has been in decline for 3 years) in 2020 (Smartphones: The Market is Mature – Now What?). We believe that 5G will induce an upcycle starting in 2021, but it will eventually run its course as any cycle does (Exhibit 18).
Exh 18: Analysts expect 2020 total mobile phone unit sales to rebound, which we believe is too early
In this context, we would not buy stocks tied to a presumption of near-term smartphone growth – e.g. AAPL, QCOM, much of the semiconductor industry.
We do see real growth in services delivered via the huge installed base of smartphones. This has been a saving grace for Apple this year in the face of weak iPhone volumes and the absolute driver of revenues for all of the big internet franchises from Amazon to Facebook. Still, we have a degree of caution around the fees being generated from Apple’s AppStore and Google’s Play. Apple faces an anti-trust lawsuit over its monopoly over software distribution for iOS. A negative outcome could open the door to much lower markups for apps and, thus, lower service revenues for Apple (and by extension, Google).
We also wrote about peripherals, which we broadly expanded to include all wearables and alternative access devices like smart home speakers. This market is booming. The Apple watch has been a success, with health as a main driver, and Google’s pending deal for Fitbit should only stimulate the wellness/wearable market even more. The reception for Apple’s AirPods has been even more impressive. We expect big things from a new generation of smart wireless earbuds that can tap into the increasing usefulness of AI assistants. This is also powering smart speakers, which are becoming nearly ubiquitous. We think these markets have real legs. We note that we are NOT enthusiastic for AR glasses for the consumer market (The Long Road to AR Glasses). Display technology is still years away from enabling a lightweight and inobtrusive form factor.
The Model Portfolio
Our portfolio has had a wild ride since our last update in September. As of a week ago, we were ahead of both the S&P 500 and the tech components of that index, which we use as our benchmark. However, a sharp downturn in SaaS stocks and a CIEN downgrade pulled us below the benchmark by 265 bp for the period, although we were up 6.1% in absolute terms and beat the S&P by 40 bp. For the year, we are up by 29.1%, comfortably ahead of the S&P broader index by 375 bp but, again, behind the tech benchmark by 1203 bp. As we noted with the September update, this underperformance is the result of our skepticism on the semiconductor cycle turn and on merchant acquirors, which are part of the index (Exhibit 20).
With this update, we have decided to remove Ciena, which is up 61.1% since we added it two years ago. Unfortunately, a downgrade by a major brokerage hit the stock 6% last week, before we were able to publish this update making it our worst performer for the period. Our rationale for removing Ciena is that optical spending to support 5G roll outs may have peaked and that an upgrade cycle to 800Gb could be less vigorous than expected. Moreover, its narrow customer base has led to significant volatility in results quarter to quarter. While we still believe Ciena to be well positioned in a market that will be attractive long term, we see better near-term opportunities for the portfolio.
Exh 19: NVDA is returning to meaningful growth and conservative consensus estimates leave enough room to surprise
Exh 20: A brief Summary of Major TMT Themes on the Horizon
One of those opportunities is Nvidia (Exhibit 19, 21). We had included the graphics processing chip leader in the portfolio for more than two years before removing it in March after the company revealed a significant glut of graphics card inventory in its distribution channels in the wake of a significant decline in demand from cryptocurrency miners. Excess channel inventory is often more difficult to resolve than management or investors presume, with the double whammy of reducing near-term shipments below actual demand after having inflated the previous quarters used as comparison. In this case, our caution was over played in the context of chip industry euphoria and Nvidia shares recovered even as quarterly revenue declined year on year. We are adding it back now, with the inventory glut behind us and as we believe investors underestimate the likely 2020 demand from datacenter operators
Exh 21: Summary of Updates to Model Portfolio – out with CIEN, in with NVDA
Exh 22: SSR TMT 15 Stock Model Portfolio Reconstituted