Citi Research Equities 1 December 2015 76 pages Consumer Electronics Global Virtual and Augmented Reality Are you sure it isn't real? What will be the next new device likely to develop a market comparable to smartphones, a market said to be worth US$400bn annually? We believe virtual reality and augmented reality (VR/AR) will create major new markets that replace the smartphone, eventually developing into a combined terminal market of US$19bn (CY2035). VR and AR overlap in part with the concept of the internet of things (IoT). These are new spheres of technology likely to have a major impact on the future of e-commerce and mobile commerce, believed to be worth US$3trn annually. In this report, we discuss the long-term outlook for VR/AR technologies at companies likely to be of interest when related markets begin to take off in 2016, as well as business trends at those companies. Our main recommendation related to the VR/AR theme is Sony. It intends to launch a headset with a 120Hz refresh rate, with the business based on a model of recovering the initial investment primarily via game content. We believe Sony will also play a key role in platform and content supply. Other names that could benefit are Facebook and Samsung Electronics in headsets; LG Display, Japan Display, TDK, Murata, Alps Electric, and Analog Devices in hardware; and Live Nation, Walt Disney Co, Time Warner, Twenty-First Century Fox, SAP, and Salesforce in software and services. We believe hardware will drive the market in the near term so most attention is likely to be on this segment. In this report we also introduce our global VR/AR theme basket. Details about the basket and performance data can be found under the Bloomberg ticker CGRBVRAR. Please see Figure 69 on page 66 for more. Figure 1. Immersive experience Kota Ezawa +81-3-6270-4804 kota.ezawa@citi.com Christopher Danely +1-415-951-1885 christopher.danely@citi.com Jason B Bazinet +1-212-816-6395 jason.bazinet@citi.com Ashwin Shirvaikar, CFA +1-212-816-0822 ashwin.shirvaikar@citi.com Walter H Pritchard, CFA +1-415-951-1770 walter.h.pritchard@citi.com Dennis Chan +886-2-8726-9086 dennis.chan@citi.com Mark May mark.may@citi.com Takao Kanai takao.kanai@citi.com Jim Suva, CPA jim.suva@citi.com Takero Fujiwara takero.fujiwara@citi.com Source: 2015 Sony Computer Entertainment, Inc. Click to Play Video See Appendix A-1 for Analyst Certification, Important Disclosures and non-us research analyst disclosures. Citi Research is a division of Citigroup Global Markets Inc. (the "Firm"), which does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the Firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. Certain products (not inconsistent with the author's published research) are available only on Citi's portals.
Contents 1. Technological advances are spurring interest in VR/AR 4 1-1 Technological trends preceding VR/AR 4 1-2 VR: At first, only visual; interaction came later 6 1-3 Augmented reality to replace smartphones and have commercial applications 6 2. What is VR/AR? 8 2-1 VR uses high-performance displays and sensors 9 2-2 AR uses transparent lenses that do not block real-world vision 12 2-3 CPU, display, battery the common technological issues 14 2-4 Trends in major VR/AR headsets 18 3. Potential size of VR/AR market is $600bn 22 3-1 Headsets to drive VR/AR market growth in the near term 23 3-2 VR to take off first, followed by AR and commerce 25 3-3 AR to eat into smartphone market? 29 3-4 VR/AR outlook 32 4. The mechanism of VR/AR services 35 4-1 Conversion of media content to VR 36 4-2 AR commerce a big part of market 41 4-3 Voice phone, video phone, chat, SNS 43 4-4 Software 44 4-5 Services: What can a services company do in AR/VR? 45 5. Potential beyond headsets 46 5-1 Motion capture critical to future of VR/AR 47 5-2 Enhancing HMI: HUDs and Augmented Reality for Automotive 51 6. Lots of participants, spanning hardware, software, content, retail, and services 56 Sony(6758.T, Buy, TP Y3,900) 57 HTC(2498.TW, Sell, TP NT$28.00) 59 Advanced Micro Devices (AMD.US, Sell, TP$1.5) 60 Microsoft (MSFT.US, Sell, TP$40.00) 62 Appendix A-1 70 2
Figure 2. Virtual reality and augmented reality Source: Citi Research. 3
1. Technological advances are spurring interest in VR/AR 1-1 Technological trends preceding VR/AR We anticipate a $600bn market by 2025 Before discussing AR/VR, we look back at the evolution of the virtual experience and technology. AR/VR experiences feel familiar We have already recently seen Virtual Reality/Augmented Reality (VR/AR) escape its previous confines as a toy for hard-core gamers. We now expect VR/AR to forge an intimate presence in people s lives in a variety of work situations and also in the field of such entertainment as movies and games, with users obtaining assistance with things such as vehicle navigation and daily lifestyle information. We see the market for hardware, networks, software & content and services reaching $200 billion by 2020. This suggests that VR/AR developers are not aiming for the technology to be confined to devices such as goggles and eyewear. The related markets are broad in scope: in addition to infrastructure and devices that display virtual information and allow users to hear, touch, and communicate with them, there is a market for content and software for sound and graphics for virtual spaces and information to be built out on these devices and infrastructure and one for service industries related to acts of consumption that are mediated by these experiences in virtual reality. We believe destructive creation, in which merchandise and experiences in existing and inconvenient reality are swiftly replaced, is occurring. People have long been familiar with virtual reality-like experiences. They say things like "it felt just like flying in the sky," or "the baby is happy, as if being held in its mother s arms." Media such as sound recordings, photography, and video recording have long provided pseudo-live experiences. Until about 30 years ago, monaural recordings, analog photos, 8mm film footage, and analog TV were the main providers of these experiences. With the aid of the user s imagination, these media provided a strong enough sense of reality to engage people. Game consoles had already appeared at about that time, creating a new market. Recording and playback equipment evolved with the introduction of CDs, digital cameras, digital camcorders, and digital television. Games and movies moved into the digital realm. Higher image quality allowed a great leap in the sense of realism created. The typical household device for enjoying movies, dramas, sporting events, and concerts shifted from a 21-inch CRT TV to a 52-inch LCD TV, and 4K TV appeared. Music moved in the direction of downloads and streaming. Highresolution audio appeared at the same time. In addition, consumer-generated media began to dominate personal computer and smartphone screens. 4
Figure 3. Valuations for VR/AR-related names Company Ticker Analyst Rating Currency Target price Price (Nov 30) Market Cap (mn USD) PER EV/ EBITDA PBR ROE Sony 6758.JP Kota Ezawa 1 JPY 3,900.0 3,177.0 32,617 17.6 2.78 1.36 7.7% Facebook FB.US Mark May 1 USD 134.00 104.24 294,790 35.5 24.81 5.96 16.8% Samsung Electronics 005930.KR Henry Kim 1 KRW 1,900,000 1,284,000 163,492 8.6 2.81 0.99 11.5% LG Display 034220.KR Henry Kim 1 KRW 35,000 25,400 7,856 12.2 2.08 0.68 5.6% Japan Display 6740.JP Kota Ezawa 1H JPY 610 375 1,836 5.1 0.86 0.50 9.8% TDK 6762.JP Kota Ezawa 1 JPY 12,000 8,850 9,080 13.7 5.12 1.30 9.5% Murata 6981.JP Kota Ezawa 1 JPY 21,400 19,100 32,916 17.8 8.38 2.79 15.6% Alps Electric 6770.JP Kota Ezawa 1 JPY 4,700 3,870 6,055 13.5 5.16 2.51 18.5% Analog Devices ADI.US Christopher Danely 1 USD 70.00 61.63 19,232 19.1 11.59 3.25 17.0% Live Nation LYV.US Jason Bazinet 1 USD 30.00 25.39 5,138 95.0 11.02 3.23 3.4% Walt Disney DIS.US Jason Bazinet 1 USD 125.00 113.47 187,586 20.1 10.95 3.89 19.4% Time Warner TWX.US Jason Bazinet 1 USD 95.00 69.98 55,948 12.5 9.13 2.43 19.4% 21st Century Fox FOXA.US Jason Bazinet 1 USD 36.00 29.51 34,549 17.0 7.37 4.38 25.7% SAP SAPG.DE Walter Pritchard 1 EUR 83.00 74.59 97,068 19.5 14.62 3.57 18.3% Salesforce CRM.US Walter Pritchard 1 USD 99.00 79.69 52,914 105.6 78.64 10.43 9.9% Note: Based on forecasts for next fiscal year. Prices as of November 30 close. Source: Citi Research Figure 4. VR/AR an extension of previous electronics technologies Source: Citi Research. 5
1-2 VR: At first, only visual; interaction came later Early VR was just a small dip into the world of computer graphics; the future should bring simultaneity Virtual reality has its roots in the use and experience of this content and these media. In the early stages of the VR market, we believe games and movies will be central. VR technologies are based on the assumption that the user will view moving images. Adding sound to the visual element establishes the minimum conditions for a VR experience. In some contexts, technology will add simultaneity, a feature not available through existing media and content. Experiencing video images and sound in real time with other people and being able to converse with others and see them within a VR environment allows a visceral experience of simultaneity. 1-3 Augmented reality to replace smartphones and have commercial applications Glasses make anything possible Augmented reality technologies, like virtual reality technologies, often use eyeglass or goggle-type terminals, and AR is like VR in that the basic configuration is a video and sound package that creates simulated experience. AR systems allow the user to move while wearing simulation equipment, and such systems can be used without interfering with work or everyday life, suggesting a potential to replace smartphones. Prospects are good also for commercial AR applications. We believe augmented reality technologies are likely to penetrate the e-commerce and mobile commerce markets. AR has potential uses facilitating work in factories and manufacturing plants, in operation of machinery and electronic equipment, and in such areas as material handling in warehouses and at distribution hubs. The potential for commercial applications is far greater than the entertainment potential, we believe. Importantly, we expect AR technologies to replace the smartphone handset in the long term. It is no exaggeration to say that AR technologies will allow the evolution from smartphones to wearable devices. We therefore believe the AR market will ultimately be larger than the VR market. We expect the economic effects of VR/AR technologies to extend into two separate markets: a hardware market for body-mounted devices on the one hand, and a content and services market on the other. We explore both potential markets in this report. We hope readers will understand that the creative destruction associated with VR/AR technologies (potential for disruptive innovation) is, indeed, real. 6
Figure 5. AR glasses help find the book you are searching for Source: Citi Research. 7
2. What is VR/AR? Virtual reality and augmented reality refer to the experience of a highresolution graphics environment through a cutting-edge device (such as a headset or eyeglass-type terminal) that draws on advanced displays, graphics semiconductors, motion sensors, and headphone technologies. VR and AR allow users to take in new information via recognition of artificially created objects. Virtual reality involves a goggle-type display that completely covers the field of view. Wearing the display, users experience the illusion of inhabiting an artificial space. Users see computer graphics that move in tandem with movements of their head, neck, and body, and hear sounds in the virtual world through headphones. If the quality of the graphics is high enough, users come to accept the idea that they are actually within an artificially created world. Augmented reality, while similar, does not necessarily immerse the user in a virtual space. Instead, users experience AR through computer-generated images superimposed over their view of the real world. AR is characterized by the use of computer graphics while recognizing the real world at the same time. AR allows computer graphics to appear like real images within real spaces. Perhaps the most important feature of AR is that it allows mobility. Figure 6. Example of VR headset device (HTC Vive) Figure 7. Example of AR headset device (Microsoft Hololens) Source: HTC. Source: Microsoft. 8
2-1 VR uses high-performance displays and sensors What is virtual reality? VR visualizes using two displays so that it appears computer graphics are being viewed with both left and right eyes. Users don goggle-type displays that completely cover their fields of vision, so they feel, illusorily, as if they have placed themselves in an artificial space. As users turn their heads to the right, in tandem with this the images move smoothly rightward into the field of vision. If users look upward, the skies open up, studded with stars. If users don headphones they can hear the sounds of a virtual world (sometimes birds chirping and sometimes machine guns and combat aircraft), and they completely lose confidence in the knowledge of where they are. Figure 8. VR: Playstation VR Wearing Goggle Figure 9. VR: Getting lost in an imaginary world Source: Sony Computer Entertainment, Inc. Source: 123RF. Structure of VR goggles The necessary hardware for a VR environment includes a terminal in the form of goggles, a computer for data processing, and network equipment. Looking ten years or more into the future, we expect data processing computers to be incorporated into the viewing goggles. We also expect to see standalone devices that require no network connection. Goggle-type headsets use two small displays, one in place of each of the goggle lenses, and an array of sensors, including acceleration sensors, angular velocity sensors, electronic compasses, image sensors, and others. 9
Figure 10. VR headset teardown Source: Citi Research. Displays require high refresh rates Better sensors than those used in smartphones Preventing VR sickness Users look into VR goggle displays at close range, as if looking into a kaleidoscope tube. Displays used in the left and right goggle sockets are either organic EL displays or LCDs. They must have very high resolution and high refresh rates. It is also important for the two displays to be perfectly synchronized. Displays in use today have a resolution of 386ppi or more and use refresh rates of 60~120fps. The main sensors mounted on VR goggles include an acceleration sensor, an angular velocity sensor (gyro), and a geomagnetic sensor (or electronic compass). Others include proximity sensors, ambient light sensors, image sensors, and inertial sensors. The acceleration sensor supplies information about orientation and movement. The angular velocity sensor detects inclination and rotation. The geomagnetic sensor determines compass direction and detects changes in the direction the user faces. Acceleration, angular velocity, and geomagnetic sensors are common in smartphones, along with GPS units, but virtual reality headsets are generally used indoors in a fixed position, so they do not require GPS input. VR experiences operate more directly on the senses (vision, hearing) than using the screen of a smartphone, and, for that reason, the sensors used in VR equipment must be of higher quality than the sensors used in most smartphones. Gear VR, marketed by Samsung Electronics, relies on the sensors and display built into a smartphone to create a VR experience. Products developed and announced by Sony, Oculus, and NVidia create VR experiences using the 120fps PS and other equipment several times more sophisticated than the average smartphone. VR systems can cause unpleasant sensations in some people, known as VR sickness. High-resolution displays, fast refresh rates, precise synchronization, high- 10
precision sensors, and fast reaction times are all important tools to help ameliorate VR sickness. The success of virtual reality depends to a large degree on the technological level of the electronic components and devices used. What can we do with VR? The key characteristic of VR is that users truly feel as if they have placed themselves in a virtual space. Viewing high-resolution computer graphics in 3D, users are enveloped in a sense of reality that differs from the feeling of really being there or of a powerful performance. This is generally known as immersion. By drawing users into virtual spaces and information, the experience is changed into one in which games or movie-watching become highly immersive. This has lots of applications in the commercial sphere alone, and we expect ramp-ups to be relatively rapid in areas such as remote medicine, military drills and pilot training, simulators for auto driving instruction, etc. Leading candidates for future VR adoption includes as participation in meetings from remote locations and business discussions. 11
2-2 AR uses transparent lenses that do not block real-world vision What is augmented reality? Augmented Reality (AR) resembles VR but it is not necessary for users to immerse themselves in virtual spaces. Indeed, it can be used by overlaying the requisite information on the scenery (field of vision) in the current world, with computer graphics being used while users are aware of the current world. In AR, it is possible for users to view objects, people and information such as words and arrows created by computer graphics as real-life images that appear to have risen up in real space. Unlike in VR, users view AR images and information such as computer graphics while remaining aware of the real-world space, so they are expected to use AR while walking in real spaces, riding in vehicles, and engaging in communication. Figure 11. AR: CG overlaid onto real world scenes Figure 12. AR: 3D holograms you can reach out and touch Source: By Edward (via Wikimedia Commons). Source: 123RF. Structure of AR glasses Technological hurdles to AR are high; we expect market growth in the long term AR glasses use transparent lenses resembling eyeglass lenses equipped with a power source and backlights to brightly overlay computer-generated edge information onto the transparent lenses. Because these glasses usually allow the user to see and react to the external environment (unlike virtual reality goggles), they also include image sensors. Because mobility is a requirement, AR glasses are equipped with GPS sensors. Connecting to a personal computer or a PS4 limits mobility. It therefore becomes necessary to embed central functions such as image processing within AR glasses. Supporting high-level image processing over extended periods of time requires very high-capacity batteries and batteries of a high energy density given that they have to be part of the worn apparatus. The ultra-high-performance electronic components and semiconductors required are not yet available with current technology. Therefore, AR is properly considered a promising technology over a longer time horizon than VR. Early versions with basic functions but certain limitations are expected to appear in 2016 2017. Full-fledged AR systems are likely to appear between 2019 and 2022 (based on model numbers and assuming consistency of development). Mobility will be a strong source of appeal. A significant number of applications people use smartphones for today may be replaced by augmented reality technologies, making a long-term growth scenario plausible. 12
AR glasses use transparent displays AR systems are characterized by computer graphic displays superimposed on the user s view of the real world which involves more specialized computer graphics projection equipment than required by VR technologies. Various approaches have been explored using transparent polymers or glass to create prism-shaped elements that insert images into a portion of the user s field of view. Devices that mount a prism in front of the viewer s eyes can be considered early types. The next generation is expected to include information tools (called scouters and by other names) that cover one eye or that the user views with both eyes to see projected computer graphics over a broader field of view devices sometimes depicted in movies and comics today. Ultimately, AR glasses are expected to be eyeglass-type devices that cover the entire field of view with transparent lenses that allow projection of images of non-transparent objects or people, creating the illusion of their actual existence within real space. Sensors include GPS sensors and image sensors Outlook for VR/AR Sensors used in AR headsets are similar to those used in VR equipment. They include acceleration sensors, angular velocity sensors, and geomagnetic sensors. Because of the need for mobility in the case of AR, GPS sensors will be important. AR systems are also very likely to include image sensors in order to accurately analyze the user s surroundings. AR equipment will need to continuously capture a scene very closely replicating the user s view in real time and use the data to determine the appropriate position of computer graphic overlays, suggesting that pressure to improve the precision of AR systems will continue to grow. We believe advanced AR systems will also require proximity sensors, ambient light sensors, and inertial sensors, although these may not be core components. The leading market for VR is games and images that look to pursue the most intimate experiential development, using the sense of immersion. Both the gaming and imaging industries have been taking the lead for a while in the generation of visuals that create this immersive sense via the technological evolution of VR. We think VR using 3D goggles could engender significant innovation in the way content is created, watched, and played. Games and movies can only be experienced in one place, but they match well with VR. We also think there is considerable potential for the use of VR in amusement parks and theme parks. However, we think that AR has greater commercial potential and envisage the market growing larger than that for VR. Moreover, AR is delineated on the skein of a real space, so placement precision and the quality of the sensor response, for instance, need to be at an even higher level than in VR. As a consequence, we think the hardware devices such as smart eyeglass frames that realize AR will form larger markets than VR ones. After hardware, we are most optimistic about consumer markets and think AR here has the potential to be a major constituent element of contemporary mobile commerce. Moreover, we would expect the information itself that is overlaid and displayed as AR on real spaces to form a major constituent element of the market value. That is, we envisage a structure whereby the user puts on an eyeglass-style device and spends money because he or she obtains information in front of their eyes about the degree of popularity of a product and how well it has sold to date. We also think advertisements might be displayed. It is not easy to make projections of the size of the markets for these constituent elements but the main constituents are in our view likely to be hardware, commercial services, and entertainment services and we note the potential for rapid market expansion. 13
2-3 CPU, display, battery the common technological issues VR will require $1,000 PCs Whether market penetration of VR/AR increases greatly will depend on technological progress. It is particularly important for the initial cost to come down. For VR, the cost of the headset is likely to be some $450, and it will be necessary to connect it to a high-end PC costing about $1,000. This is because a highperformance graphics card will be necessary to process the CG that is projected onto the VR headset display. There are few problems associated with connecting the VR headset with a cable to a PC, but reducing the total cost will be critical to increase in market uptake, and the cost of buying the PC cannot be ignored. We would not expect significant uptake of VR by the consumer market if the combined cost of the high-end PC and the headset was in excess of $2,000. In the case of AR, mobility is the key to market uptake. For this, the performance of the semiconductors and other electronic components in today s smartphones would be ample for the key components in AR headsets, but reducing power consumption is the critical determinant of mobility. PS4 already provides hint to solution of cost problem We are optimistic regarding the cost problem. In the technology world, innovation has repeatedly reduced the cost of realizing the required functionality and performance. In addition, the PlayStation4 will be used as the CG processing computer for the PlayStation VR headset to be launched in 2016, and the price cut in 2015 means that a machine with 1,830GFLOPS (single precision) of processing power can bought for $349. Figure 13. Sony PlayStation 4 Improvement in GPU performance key to market uptake of VR/AR Source: Sony Computer Entertainment Inc. We think that in the future all necessary functions for VR and AR will be built into the headset. But during the transition period over the next 10 years, the speed of CG processing by the GPU and other components and the extent to which power 14
consumption can be reduced will be important. Our analysis concludes that technological development will be insufficient in the next 10 years for the GPU processing capacity necessary for VR to be powered by a battery. Nevertheless, the cost of a PC to connect with using a cable will fall sufficiently. For the graphics card that contains the GPU to have the recommended specifications, a product costing around $500 is currently necessary, but this could fall to $50 over the next 10 years. Reduction in AP, MCU power consumption key to market uptake of AR For the power consumption issue critical to AR, the applications processor (AP), which also functions as a graphics processor, and the microcontroller unit (MCU) are important. Right now power consumption for the digital signal processor is about 50mW, but we note that power consumption fell by 99% between the 2000s and 2010s. If there is another 99% drop over the next ten years, these components could be built into the headset and powered by a battery for a sufficiently long time. This would create the mobility that would provide a powerful driver of AR market uptake. Power consumption for PC GPUs fell by 92% over the last ten years, and could fall by another 50% by around 2018. Figure 14. GPU performance evolution Figure 15. GPU: Power consumption relative to performance GFLOPS 16,000 14,000 12,000 GeForce Radeon Roadmap by GeForce W/GFLOPS 0.400 0.350 0.300 x 1/12 10,000 0.250 8,000 6,000 4,000 2,000 0 00/1 04/1 08/1 12/1 16/1 Source: NVIDIA, AMD, Citi Research 0.200 0.150 0.100 x 1/2 0.050 0.000 05/1 07/1 09/1 11/1 13/1 15/1 17/1 19/1 Source: NVIDIA, Citi Research VR/AR the next hope for display industry? As we discuss below, if the VR/AR headset market grew to 300mn units, the market for the displays used would be 600mn units. If high-end smartphones currently account for 390mn of the overall smartphone market of 1.4bn units, then the VR/AR headset display market would be larger than the current high-end smartphone display market. 15
Figure 16. LCD panel market by major application (VR/AR figure is for 2025, rest are 2015 estimates) (units, mil.) 1,600 1,400 1,200 1,000 800 600 400 200 0 Total Smartphone VR/AR (2025) Featurephone High-end Smartphopne PC TV Tablet DSC Source: Citi Research. Ample room for development of VR/AR displays The displays currently used in VR/AR do not fully satisfy requirements. For VR, LCDs or OLED displays are used. Only a few products use a refresh rate sufficiently high to ensure users do not experience motion sickness. In AR, Google Glass used a prism, but this is a transitionary technology and a transparent display will ultimately be adopted so that AR CG can be displayed on a wider field of view. We think that if it becomes possible to manufacture high-performance transparent displays at low cost, the VR/AR headset market will take off rapidly. We would see a figure of around $20 as one benchmark, enabling cheaper headsets and extending battery life. Energy density of batteries an issue Figure 17. Size and price of small displays and VR/AR displays ($ per inch^2) 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 4.3" 4.5" 5.0" Q1Q2Q3Q4 CY11 CY12 CY13 CY14 CY15 CY16 CY17 CY18 CY19 CY20 Source: DisplaySearch Citi Research. Future trend (our estimate) VR/AR using high-end display price trend (our estimate) A rechargeable battery will be necessary for headsets. A power supply from a battery that can ensure mobility will be particularly important for AR, as it is intended for use when the users is out and about and moving around. A battery that is sufficiently compact and high capacity to do away with a connecting cable and extend usage times will drive more rapid market growth. We think it would be 16
sufficient for the battery to last long enough to watch a two-hour rugby match using VR and spend three hours of instrument panel operations via AR. An AR headset would not necessarily use more power than a current smartphone. The display size would not be as large as 5, and the amount of power used for backlighting (or self-lighting in the case of OELD) would be smaller. However, the batteries would need to be smaller and lighter than those used in handheld smartphone in order for it to be integrated into a device worn on the head. Ultimately, we assume that power consumption would be about the same as for a smartphone, and that market uptake of high-performance AR headsets would be greatly aided by a reduction in the size of batteries to around one-quarter thanks to advances in energy density. This may not seem that big a change relative to performance improvements for other electronic components. However, this is because battery performance is determined by the underlying principles of chemical reaction. Chemical reactions are not impacted by efforts to increase efficiency or spending more money. Therefore, we think improvement in battery performance will be relatively slow. One potential workaround is using wireless charging, or promoting the uptake of replacement batteries. Figure 18. Lithium-ion battery cost reductions slower than those for other components? (US$) 1.2 $90c -23% (CAGR-2.5%) 1.0 $70c 0.8 0.6 0.4 0.2 0.0 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E Others PTC Copper foil Electrolyte Separator Anode Cathode Note: We use the expected cost trend for an 18650 cylindrical 2.2Ah battery. Source: TSR, Citi Research. 17
2-4 Trends in major VR/AR headsets VR headsets have appeared in the past three years The Oculus Rift VR system and the Google Glass AR system both appeared in 2012, three years ago. We believe VR systems will start to become generally available at retailers around the end of 2015 and into early 2016, which will mark a turning point for the industry. Different products have different characteristics and strategies will differ. PS VR, Oculus Rift, and HTC Vive mainly have gaming applications, while Samsung Gear VR is used through a smartphone and will be positioned as a kind of VR lite, mostly upgrading access to content not available on an "unaugmented" smartphone. Even among the three devices mainly aimed at the game market, strategies will differ. PS VR assumes use with the PS4 system. Content will have to be developed within a closed environment, but that content is guaranteed to work on the associated hardware. Oculus Rift development is an open market and a range of content is already functioning on model systems provided to developers. However, because processing capacity depends on the computer Oculus Rift runs on, there has been some concern about possible frame dropping and other problems. HTC Vive is designed to allow use of Steam VR, Valve s VR game platform. AR is being tested ahead of diffusion On the other hand, AR equipment is still at the trial and error stage of development. Although Google Glass was released to the general public, Google halted sales in early 2015. Epson's MOVERIO AR terminal and a number of other eyeglass-type terminals have appeared, but none of these products appear ready for widespread consumer or business use at present. Until now, AR has been based on the idea of taking in information from the real world and displaying an interpreted version of it. Microsoft s Hololens, announced in January 2015, relies on a more advanced concept of projecting virtual objects onto real objects, integrating the two. Playing Minecraft on the Hololens platform, users accumulate structures on an actual desktop. Players see the virtual and real elements merged. Magic Leap, backed by Google, is under development using a similar concept. This integration of the virtual and the real has become a core AR concept. 18
Figure 19. Major VR/AR headsets Company Sony Facebook Samsung Valve / HTC Google Seiko Epson Product PlayStation VR Oculus Rift Gear VR HTC Vive Google Glass Moverio VR or AR VR VR VR VR AR AR 2011 BT-100 Released 2012 Prototype announcement Prototype announcement 2013 Released for Developers Released for Developers 2014 New Proto Announcement New Model (Project Bought by FB Morpheus) Released Gear VR announcement Developers 2015 Discontinued New Prototype Alliance with Released for HTC Vive announcement MSFT Developers announcement Announce consumer model PlayStation VR announcement 2016 Released for Consumers To be released Gear VR announcement Released for Released for Consumers Released for Developers Released for Consumers Released for consumer BT-200 Released BT-2000 Released (for business) Source: Citi Research. Figure 20. Sony PlayStation VR Source: Sony Computer Entertainment, Inc. PlayStation VR Sony announced Project Morpheus, a prototype PlayStation VR model in March 2014 at GDC 2014. In the following year, at GDC 2015, the company unveiled a new prototype. In September this year Sony introduced the PlayStation VR name and announced plans for its full release in the first half of 2016. The system will use a 5.7-inch organic EL display with a 1,920 x 1,080 resolution (960 x 1,080 in each eye), a 120Hz refresh rate, and a viewing angle of about 100 degrees. Because the system will be hard-wired (with an HDMI connection) and must connect to PS4 equipment, users will have to own a PS4 system, but we believe problems caused by inadequate graphics processing capability (such as dropped frames) are unlikely to occur. Oculus Rift Oculus VR, Inc. announced its first prototype at E3 in June 2012. In March 2013, it launched the Oculus Rift DK1 as a developer model. Priced at $300, resolution was 1,280 x 800 pixels on a 7-inch LCD display. Facebook acquired Oculus VR in March 2014 for US$2bn. The company launched the Oculus Rift DK2 in the following July, priced at US$350 with a resolution of 1,920 x 1,080 pixels on a 5.7-inch organic EL display. The device connects to a personal computer with an HDMI cable and is treated as an auxiliary display with a 75Hz refresh rate. Oculus recommends the controlling PC use an NVIDIA GeForce GTX970 or AMD Radeon R9 290 graphics processor or better. The company also has demanding CPU and memory recommendations. In June 2015, the company announced an Oculus Rift model for general release, scheduled to become available between January and March 2016. Oculus VR has announced a partnership with Microsoft as well. The final release version will include an Xbox One controller. Edge of Nowhere (Insomniac Games), 19
Chronos (Gunfire Games), and other titles are under development. We believe other game studios are developing titles. Available content appears set to expand. Figure 21. HTC Vive Source: HTC. HTC Vive HTC announced the HTC Vive at MWC in March 2015. HTC Vive was developed in collaboration with Valve, which runs the Steam game distribution service, part of the SteamVR project. Gaming will be the main application. HTC Vive is scheduled for release in the second half of 2015, but no release date has yet been announced. HTC Vive uses a 2,160 x 1,080 pixel organic EL display with a 90Hz refresh rate. It will require a wired connection to a personal computer. The Lighthouse position detection system will allow the user to move more while wearing the equipment than a PS VR or Oculus user and the system is capable of recognizing multiple devices, allowing it to reflect the movements of other players within the VR environment. Figure 22. Samsung Gear VR Samsung Gear VR Samsung Electronics announced its Samsung Gear VR in September 2014 and began selling models for developers in December last year. Samsung introduced new models in March 2015, planning a release for the general public in November 2015. Gear VR has been developed in collaboration with Oculus. It is used in conjunction with some models in Samsung's Galaxy line of smartphones. The headset can be used wirelessly. The device will offer Netflix, Vimeo, Hulu, and other content. Affordably priced at US$99, it is likely to be positioned as introductory VR equipment. Source: Samsung Electronics. Google Glass Google announced a prototype under the name Project Glass in April 2012. Google launched a developer model in April 2013 and released a model for general sale in May 2014, both priced at US$1,500. However, the company halted Google Glass sales in January this year. The terminal is in the form of a pair of eyeglasses with a prism mounted in front of one eye for information display. The prism is semitransparent, allowing the user to see through the information display even when active. Development has continued since the suspension of general sales. Google is now working on a model for enterprise use. Figure 23. Microsoft HoloLens Source: Microsoft. Microsoft Hololens Microsoft announced Microsoft Hololens in January 2015 during a Windows10 press event. The company is expected to release a developer model in January March 2016, priced at US$3,000. Hololens is an independent goggle-type computer equipped with multiple sensors to detect outside world information and equipped with a lens for displaying holograms (rather like a projector). Demo versions allow playing Minecraft, Halo, and other games with real world objects treated as virtual objects. In addition to games, the equipment has been shown in demo sessions to be compatible with various simulations and graphics-rich applications. Other VR/AR-related companies are ramping up activity with each passing day. We have seen large capital procurements with the idea of starting up a business within the next several years. 20
Figure 24. VR/AR timeline Category Year Mon Event Hardware 2012 Apr Google announces Project Glass May Leap Motion procures $12.75mn, launches Leap Motion controller Jun Oculus launches test model 2013 Mar Oculus Rift DK1 launched Apr Google Glass developer model launched 2014 Feb Magic Leap procures $50mn from Google Mar Facebook acquires Oculus for $2bn Sony announces Project Morpheus May Consumer version of Google Glass launched Jul Oculus Rift DK2 launched Sep Samsung Gear VR announced Samsung Gear VR developer model launched Oct Magic Leap procures $542mn from Google and Qualcomm 2015 Jan Google Glass sales suspended Microsoft announces Microsoft Hololens Mar Sony announces new format for Project Morpheus HTC announces the HTC Vive Jun Consumer version of Oculus Rift announced FOVE procures capital from Samsung Ventures (amount undisclosed) Starbreeze acquires InfinitEye, announces Star VR Sep Sony announces the launch of PlayStation VR as part of Project Morpheus Oct Facebook CEO Mark Zuckerberg notes the firm is working on AR (at Vanity Fair New Establishment Summit) AuraVisor begins procuring funds via Kickstarter Nov Next VR procures $30.5mn from Comcast Venture and Time Warner Investments, among others ~Dec HTC Vive launch expected 2016 Q1 Consumer version Oculus Rift expected to be launched Q1 Microsoft Hololens developer model expected to be launched (price of $3,000) H1 PlayStation VR launch expected Content 2015 Apr WorldViz procures several million dollars from Intel Aug Resolution Games (Sweden) procures $6mn from Google Ventures and others Sep Jaunt procures $65mn from Walt Disney Co. and others Source: Company data, Citi Research. 21
3. Potential size of VR/AR market is $600bn We feel the VR/AR market is one with a long tail and that it is not easy to estimate market scale. Taking a big picture view, we think AR has the properties required to substitute for the current smartphone market and that VR has the properties required to substitute for the current game and movie industries. VR game applications are the field where we think the obstacles to diffusion are lowest and here we expect 2016, when makers of major game machines such as Sony and Microsoft are slated to announce goggle-like game terminals, to be the year VR takes off in earnest. Taking this as the starting point and envisioning the spread of VR/AR over the next decade, we see the end-market (excluding AR commerce) reaching $400 billion and $6.3bn in the first five years. Even excluding the retail shopping market, where there is little dependence on VR/AR technology, we put the overall VR/AR market at about $600bn by 2025 or so. Figure 25. Estimated market for VR/AR (US$, tn.) 1.6 1.4 US$514bn market excluding commerce 1.2 1.0 Commerce 0.8 0.6 0.4 0.2 US$73bn market excluding commerce US$600bn 0.0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. 22
3-1 Headsets to drive VR/AR market growth in the near term Focus on headset market expansion over the next five years Headsets is the key Attractive hardware as well as contents and software are necessary for the VR/AR market to expand. Widely distributed personnel and elemental technologies (including services, retail, tourism, amusement parks, and sporting events) are a feature of the VR/AR industry. Over the next five years, we forecast hardware, mainly headsets, will by the driver of industry growth. We expect VR headsets to become established first and we believe improving VR headset performance and lowering prices will be key to getting the market off the ground. Now we turn to why we think hardware (in particular headsets) will be vital for the overall VR/AR market in the short term. The e-commerce market did not take off until PC uptake was sufficient. The market has now hit $1trn and has been growing at 10%-20% annually. Within e-commerce is the growing field of mobile commerce using mobile phones. This market has been expanding in line with the uptake of smartphones, with little difference in timing between increases in device usage and expansion for commerce. The mobile commerce market is currently around $300bn and has been expanding at a rate of 30%-50% annually. This suggests that increasing popularity of VR/AR headsets is likely to mean rapid expansion for AR commerce at roughly the same time. In our view, the key to AR commerce expansion is headset expansion. In addition, in terms of investment efforts in VR/AR are not exactly new in commerce. Additional investment is necessary, but we think it will be smaller than that needed to launch the first online shop, and visibility on the future of the business is better. The psychological hurdle to spending on VR/AR content and services seems low, and we think it will expand as headset uptake does. Headset expansion to look similar to smartphone expansion We think market expansion for VR/AR headsets is likely to look like that for smartphones in terms of speed and scale. We also anticipate similarities in terms of application and costs for electronic components and semiconductors. 23
Figure 26. Prices and volumes for major consumer electronics products Smartphone Featurephone Tablet Notebook PC LCD TV VR/AR headset Smartwatch Game Console 2015 2030 2025 2035 2010 Market units (mn units) 10,000 1,000 2010 2015 Tablet 100 Game console 2025 Featurephone Notebook PC LCD TV 10,000 1,000 VR/AR headset 100 10 10 ASP ($) Source: Citi Research. 24
3-2 VR to take off first, followed by AR and commerce Massive markets for both hardware, software We project long term growth in the VR/AR market to $674bn in 2025, and further growth thereafter. Market is set to ramp up from 2016. For the 10 years through 2025, we forecast a CAGR of 122%. The market will begin the ramp up with the simultaneous launch of both VR and AR headsets, but we think that VR will be the main driver over around five years. We forecast a market for hardware (headsets) of $85bn in 2025, assuming sales of 375mn units in 2025 and that the average unit price declines from $342 in 2016 to $227 in 2025. We think the primary VR content will initially be game software, but that the market will then expand to include movies, TV programs, and music before broadening further to encompass concerts, zoos, and theme parks. We think the market will remain small at just $3.8bn in 2016, but we forecast growth to $29.3bn in 2020 and $107.0bn in 2025. AR is likely to be trialed in areas with special needs, such as construction sites and warehouses (material handling) and medicine from around 2020 and gradually spread from there. Thereafter, we expect miniaturization and reductions in the power consumption for AR devices (glasses, goggles and headsets) worn on the eyes and head to promote explosive market uptake by consumers. Figure 27. Breakdown of VR/AR market into hardware, software and services, and commerce (US$, bn.) 800 700 600 m-commerce, e-commerce 500 400 300 VR headset market size 200 100 0 2010 2011 Source: Citi Research. 2012 2013 2014 2015 2016 2017 2018 The earliest VR headsets appeared on the market around 2013 in a venture-like trial, but we expect the genuine ramp up of the market to start in 2016. This is when products by Oculus, which was acquired by Facebook, and Sony are to be launched. Given the cost of the display, sensors and other materials and the scale of volume production (small initially), we put the average unit price at $316 in 2016. Other than the Samsung Gear VR, which uses the display, sensors and other key devices from 2019 2020 2021 2022 2023 2024 2025 Software, Service Hardware 25
smartphones and is priced at $200, we expect other products to be in the $300 $500 range. We put the market size at 10mn units in 2017, 20mn around 2020 and 30mn around 2025. We think long-term equilibrium will be 40mn 50mn annually, and that the unit price will ultimately fall to $150 $150. VR games We think the game industry will be able to develop the market first and most reliably. We expect this market to grow to $14.2bn in 2020 and $57.9bn in 2025. In 2020, we expect game title disc purchases, downloads, and streaming to account for 43% of the overall VR market. The nature of the graphics design for the first-person shooter games that have enjoyed sustained popularity in the video game market means that the gamer moves the character on the screen in the same way as moving his or her own body. As this design is a good match for VR, we thing first-person shooter games will be very easy to convert to VR. Accordingly, we think that around half of the titles for game consoles will be converted to VR by 2025. Over the long term, the vast bulk of games are likely to be converted to VR, which is likely to become the standard for home games. VR movies, TV, music videos While 3D flat panel TVs for the home did not prove popular, watching 3D movies at the cinema has now become an established form of entertainment worldwide. As it is essentially an extension of 3D, we think that the movie watching experience is the next easiest area for the incorporation of VR after games. We also expect the production of some TV programs and music videos using VR. In the movie, TV, and music markets we expect a partial fusion with what are currently seen as secondary retail markets in the form of DVDs, VOD, and streaming. We think a business model could emerge under which movies are distributed to consumer PCs over the Internet on the day of the premier so that they can be watched with the same realistic, immersive experience as a cinema using a VR device. The shift in the product distribution structure from cinemas to the consumer channel could well cause a major change in the industry itself, and as many uncertainties remain, we factor in only the shift in the market in this analysis. Entertainment, commercial applications, tele-existence We think the VR market related to movies, TV, music videos, DVDs VOD, and streaming will start to grow strongly around 2017. We put market size at $2.3bn in 2017, $11.3bn in 2020 and $39.2bn in 2025. We think VR will spread to theme parks, sporting events, and music concerts. For activities that need to be in real time, such as watching sports and live music, technological advances and reductions in telecom costs will be necessary. We therefore think that modest growth in this market will center on theme parks and the like, so we do not assume a big market in this area. Overall, we forecast a market of $2bn in 2020 and $4bn in the longer term. However, we think there are many potential industrial applications and if technological progress accelerates this market could be many times larger than currently thought. We think there is considerable potential for growth in commercial applications if technological advances improve quality and reduce initial costs. When we are booking a restaurant, want to know how a hotel room would feel before we book, or want to check if an airline seat will have sufficient leg room, we are currently forced to make a passive decision based on limited information on websites. If the cost were feasible, the company providing the service could replace the virtual tour on 26
its website with genuine VR. In the long term, we think that VR is a leading candidate for conferencing systems in commercial uses. Tele-existence is the ultimate incarnation of VR. Tele-existence involves identifying (seeing) objects and people in a VR worlds as well as being able to touch them, speak to them and mutually interact with them. It is a VR world in which one can talk and walk with people and touch objects to confirm their texture. It would require the user to wear not only a headset, but a range of sensors on their body. This is a field for the long term, but even now there is much active development underway, so it is by no means a dream. We put the combined market for entertainment, commercial applications and teleexistence at some $10bn around 2020. However, this is based on conservative assumptions for uncertain technological elements and business models, so the potential market is much larger. AR device market size We expect the market for AR headsets to be comparable to that for smartphones. Ultimately, some consumers will do away with smartphones and use AR headsets for calls and data transmission. We currently envisage AR headsets replacing smartphones in 2025 2030. Through 2025, however, we position AR headsets as a consumer device that will generate net growth in the same way as VR headsets. In the long term, we expect the extent of smartphone market substitution by AR to rise from 0% in 2025 to 25% in 2035. AR headsets require specialized displays so we expect unit prices to be relatively high. However, as with smartphones, they will serve multiple purposes, so we think volume production will bring prices down in the long term. Unlike VR, the effectiveness of AR will decline markedly once hardware quality falls below a certain standard, so we see less potential for growth in low-end products than for VR. In around 20 years time, we would expect low end users to continue to use low-end smartphones and midrange and top end users to shift to AR headsets. AR commerce (shopping, advertising) AR voice calls, video calls, email, SNS AR as work tool (writing, data handling, graphs/spread sheets, electronic conferencing, data base use) We forecast the AR headset market at $16.9bn in 2020 and $79.1bn in 2025. Over this period, we expect the smartphone market to grow from 2bn units to 2.25bn. Under our current model the biggest market segment is commerce. We forecast that the AR commerce market will grow to a massive $49.3bn by 2020. But this would still be not more than 2.7% of the overall e-commerce market, which we put at $1.8trn in 2020. By 2025, we project growth in AR commerce to more than $400bn, which would make it a major channel in the e-commerce market with a weighting of around 15%. Our forecast for the size of the AR estimate is based on the assumption that AR headsets are used for five years (there is a five-year replacement cycle) and on an assumption for an average annual amount spent per unit of the installed base. AR headsets display CG without completely blocking the user s field of view, so they could be used to make telephone calls and post entries on SNSs. This does not necessarily involve using functions unique to AR, but we think it is likely to be more convenient than using a smartphone, so we expect AR headsets to replace smartphones in some cases. We estimate this part of the market for conventional communication services at $5bn $20bn. We think the use of AR could spread among companies as a work tool. The author is currently tethered to an LCD monitor while writing this report, but in the future may well write into a space displayed with CG. It would most likely be possible to display 20 years of corporate results horizontally in an Excel earnings model. In 27
meetings and presentations, BYOD may well mean that each participant can connect their own AR headset to the internal network to view notes on a laptop PC that would currently be displayed using a projector and screen. We conservatively estimate the market at $10bn in 10 years time, but we see ample potential for growth. Figure 28. VR/AR market forecasts (US$bn) CY CY CY CY CY CY CY CY CY CY CY CY CY CAGR CAGR 13 14 15 16 17 18 19 20 21 22 23 24 25 15->20 20->25 VR Hardware 0 0 1 2 3 4 5 5 5 5 6 6 6 43% 4% VR Game industry 0 0 0 2 4 7 11 14 19 27 39 50 58 72% 33% VR Motion video 0 0 0 1 2 4 7 11 17 22 28 34 39 108% 28% VR Entertainment 1 1 1 2 2 2 3 4 5 7 8 9 10 28% 20% AR Hardware 0 0 0 1 3 5 10 17 25 36 49 63 79 111% 36% AR Commerce 0 0 0 1 4 10 23 49 90 152 240 356 401 165% 52% AR Communication 0 0 0 0 0 1 2 6 12 21 33 48 66 200% 61% AR Commercial/Enterprise 0 0 0 1 1 2 2 4 5 7 10 13 15 67% 33% Total VR/AR 1 1 2 8 20 35 63 110 179 277 412 578 674 118% 44% Hardware 0 0 1 3 6 9 15 22 30 41 54 69 85 93% 31% Software, Service, Contents 1 1 1 5 10 16 26 39 58 84 118 154 188 93% 37% Commerce 0 0 0 1 4 10 23 49 90 152 240 356 401 165% 52% Source: Citi Research. 28
3-3 AR to eat into smartphone market? Net gains or cannibalization? When wearing a VR device, the user is typically restricted to using it in a single location. This represents a new way for people to use a device. As such, VR devices are a new type of product that do not overlap with established consumer electronics products in terms of hardware. That is, VR devices can be considered as a genuinely new industry capable of driving net growth in the consumer electronic equipment market. Conversely, VR content will frequently be drawn from established industries like games and movies. If established movie theaters and game software add VR compatibility, then conventional movies and games would no longer sell. In other words, VR will erode established content markets. Meanwhile, AR differs from VR in another respect. Consumers could use AR in areas like mobile commerce, email, phone/texting, and in mobile meeting tools at work. AR could also be used for mobile advertising, as on smartphones. In other words, the pattern of use of AR will resemble that of smartphones. For this reason, we think AR will start to erode the smartphone market on both the hardware and software sides. In our current market model, in working out the size of the hardware market for AR headsets we assume that the majority of growth in the AR headset market from 2025 onward will be achieved at the expense of the smartphone market. The 2025 timing is based on anticipated progress in AR headset technology; however, the shift could happen earlier if improvement in the quality of transparent displays or the development of miniature computers gathers pace. Figure 29. AR headsets to replace smartphones? Figure 30. Smartphone market trends should AR headsets eat into demand (units, mn.) 1,200 1,000 800 600 400 200 0 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. But VR/AR can still be a new growth driver for the consumer electronic equipment market new market creation replacing smartphone (units, bn.) 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. current smartphone market view AR replacement scenario Even though we think AR headsets will erode some of the smartphone market, we still think VR and AR are the keys to the resumption of growth in the consumer electronic equipment market. In our simulation we expect VR/AR headsets to play a driving role in the digital appliance industry between 2020 and 2025. We do not expect much positive news for the consumer appliance market in 2015. We see a struggling smartphone market coupled with a headwind from PC market contraction. There is also substantial concern about negative growth for products like conventional satellite navigation systems, DVD players, and TVs. Wristwatchstyle wearable devices and wristband-type wearables are starting to gain traction, but with volumes of around 20mn units at the largest player the market is still some way short of the hundreds-of-million-unit scale of the smartphone, PC, and TV 29
markets. VR/AR will not offer direct support to the industry as a whole either. We therefore expect the stagnation in the hardware industry to deepen. Meanwhile, economic activity using digital appliances is on the rise. This is because the amount of content consumed via digital appliances is growing while e- commerce (including AR commerce) and use in the enterprise domain is still expanding. This shifting of added value from hardware to software is likely to continue. In these conditions, hardware manufacturers are being called upon to move beyond the so-called selling boxes model of simply producing and selling hardware to a model that includes the selling of content and software, thus allowing revenue to be generated whenever the hardware is used (sometimes called a platform business). The range of economic activity that can be accomplished using VR/AR hardware is extremely large. This makes it difficult for hardware manufacturers to be able to cover all the necessary content, services, and applications. Similarly, on the commerce side it is unlikely that shopping tied to specific hardware will become popular. These conditions will make it difficult for hardware firms to generate synergies. At the same time, we think companies capable of running diverse non-hardware businesses will be able to find a competitive edge through their business models. It could be argued that one reason that Microsoft, Google, and Facebook are pursuing VR/AR so aggressively is because they have plenty of management resources and expect earnings opportunities to grow beyond the hardware itself. Sony s portfolio of content including games, movies, and music will allow the company to build a platform business to compete against Google. We think these companies will be in the vanguard when these kind of businesses are farther along in terms of development. 30
Figure 31. Trends for major consumer electronics products (USD, bil.) 1,000 900 800 700 Game, Car Elec, DVD, Camera, TV Tablet PC 600 VR/AR headset 500 400 300 Smartphone 200 100 Wearable Consumer IOT 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E Note: We use a simplified formula for estimating the markets for game devices, auto electronics, DVDs, cameras, and feature phones in 2018 and beyond. Source: Citi Research. 31
3-4 VR/AR outlook VR/AR mapping Current market forecast models do not necessarily factor in changes to VR/AR device applications and demand for them as their performance improves. This is understandable as there are limits to how far out a meaningful roadmap can be drawn. Over the next 10 years or so, however, we believe both VR and AR devices will steadily evolve. VR starts with users viewing a virtual scene in isolation and ultimately evolves into an immersive experience they feel part of a virtual world. AR starts by displaying part of a users actual field of vision in a text box. Semi-transparent landmarks, figures, and other computer graphics are then introduced into the real-world scene in such a way as to appear natural. In the final stage, objects and people that have a convincing life-like quality are introduced and users are able to interact with them (move objects, talk to people, etc.). Figure 32. VR and AR are evolving VR immersive ---- virtual world tele-existence (being at the place and interact) ambient ---- immersive AR (solid objects/human CG in the real fileld) isolated ---- video call FPS game automotive HUD real world text box in the air transparent objects in the real field solid objects in teh real world AR Source: Citi Research The three stages of VR Skype and Facetime video call services share concepts found in VR. Halo, Metal Gear Solid, and other first-person shooter (FPS) video games are also close to virtual reality. But video calls/fps video games are a far cry from immersion in a virtual world and are more accurately described as isolated experiences. The next stage is to create a virtual setting (ambience) around the user. This is done by looking at a VR headset display rather than a laptop PC or TV screen. In the case of a roller coaster computer graphic, the user would be engulfed by the screams of other people on the ride. The final stage of VR evolution is the immersive experience. Users enter a virtual world in which they can roam over hills, hold business discussions with customers, or dance with their grandchildren. This is not just the sensation of watching a virtual world but of actively engaging with it. We expect this will involve users wearing a headset as well as clothing devices (gloves, shoes, pants) that enable the virtual world to respond to their movements. 32
The three stages of AR We expect AR to evolve in three stages. The first stage is digital display functionality on vehicle windshields (speed, etc.). This allows drivers to know if they are travelling at a safe speed without having to divert their eyes to the dashboard. However, this type of computer graphic does not follow the driver s line of sight. The next stage is using semi-transparent images to display directions, signals, road signs, etc., in the driver s field of vision. A feature of the second stage of AR is that the computer graphic works in unison with the movement of vehicle. The final stage is the introduction of life-like computer graphics. In ten years, we think a number of items could be orbiting the kind of globe one buys for children. Figure 33. Semi-transparent information overlaid on actual scenery Figure 34. Immersive AR concept (augmented reality globe) Source: Pioneer Corporation. Source: By Wilson Teresi (via Wikimedia Commons). (https://commons.wikimedia.org/wiki/file:augmented_reality_world_globe.png) Potential applications diverse Changing our world We believe VR/AR is likely to have many applications in the leisure and entertainment field, including concerts, sports, horse racing, and amusement parks, as well as commerce and business (conferences, training, exhibitions, travel, etc.). For example, people could watch sporting events in VR. This would not replace the thrill of watching an event live and should thus increase the size of the sports market in net terms. Also, we see potential for pop concerts to be streamed using VR services. We believe the use of VR for sports and arts that are primarily meant to be consumed visually is not far away. By projecting virtual objects or people in real space through the use of AR, it has become possible to intuitively convey images of objects or people that previously had to be imagined. By superimposing 3D computer-generated imagery over realtime images, AR allows users to view and touch objects from different angles in an almost realistic fashion and to use a gyroscope to adjust views and aspects. This not only offers greater convenience, it may also affect human sensibility. Put another way, humanity s desire to own material things and infrastructure could wane and the importance of objects existing physically before one s eyes could also diminish. In the world of AR, it is also possible to superimpose text, statistics, or other information forms over real-life images or background scenes. We believe this will bring life changes similar to those triggered by the birth of the Internet. As this technology not only allows explanations of immediate events/phenomena even where there is no prior memory or knowledge, it also makes it possible to obtain 33
information to augment existing knowledge such as up-to-date information and statistics (support ratings, sales, traffic news, etc.). Customizing such information to an individual s needs also appears feasible, especially when linked to profiles, data sets and productivity applications relevant to that user. The use of VR can simulate environments that give the impression almost of a novel scene. Using a headset or similar device, the VR visitor normally stands in a certain space, or in some cases sits in a chair or lies on a bed. VR not only includes sight and sound, but if sensors corresponding to each of the five senses are applied to the body, the overall virtual experience can deliver a level of awareness that transcends the physical body. For the user, VR heightens the perception of being physically present in a non-physical world, a perception that is created as the user s awareness of physical self is transformed by being immersed in a virtual space. We believe the use of VR in entertainment applications such as games and movies will deliver a relatively more realistic experience. VR/AR to be used in films and animation in the near future The world of VR/AR has been inspired by science fiction films and animation, and we believe future VR/AR technologies are likely to be first used in these media. Figure 35. Examples of VR/AR in film and animation Contents Summary Year Battleship Spaceship chose targets using an AR display 2012 Inception Entering into a dream is a kind of VR 2010 Avatar Not VR, but humans able to control living creatures in another world remotely 2009 Ironman Helmet visor AR style 2008 irobot Set in the near future, and AR shows up often 2004 Minority Report AR shows up in various places 2002 A.I. Artificial Intelligence Second half unfolds as if entering into VR 2001 Matrix Quintessential VR depiction 1999 Predator Predator helmet boasts sophisticated AR functions (mainly temperature display) 1987 Terminator Terminator eyes can see various AR displays 1984 Star Wars Holographic conference (AR) 1977 NHK Special "Next World" Discusses technological developments expected in 2035, introduces penetration of AR and VR into everyday life. 2015 Dragonball (cartoon) Determines enemy capabilities using a scouter (AR) 1984~ Source: Citi Research. 34
4. The mechanism of VR/AR services The VR/AR market is poised to grow to encompass various content and services. We expect these markets to be constituted mainly from certain parts of established markets that have been converted to a VR/AR format. In this way, VR/AR has an element of creative destruction. However, the potential to create new markets should not be ignored. The question is how much of these content and services markets are capable of transforming into a VR/ARrelated version of themselves. The basic view seems to be that content and service industries will see VR/AR as an opportunity to raise average sales prices by increasing the added value of their offerings. Moreover, if new markets are created, then the strategic and psychological barriers to entry are likely to be low. Figure 36. VR music concert to create new markets Source: 123RF, Citi Research. 35
4-1 Conversion of media content to VR VR games Figure 37. VR ratio for major game types 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. Console game PC game Mobile game The game market can be divided into PC games, console games (as typified by the Xbox and PlayStation), and smartphone and other mobile games. We think the segments that will see the soonest growth in VR content (game titles) will be PC and console games. Many mobile games do not require VR, and given the processing power of smartphones, VR would not be feasible immediately, but there are possibilities over a longer time frame more than 10 years into the future. We think that the percentage of VR games will rise more quickly for console games than PC games. The same number of VR titles may emerge for PC games as for console games, but the PC game market is larger, so in percentage terms conventional games are likely to remain the mainstream for longer. Many PC are multiplayer and massively multiplayer online games, and it will not be easy to use VR headsets given their performance limitations in the near term, so we expect the ramp up of VR to be more rapid for console games. We estimate that game software market at $83.6bn in 2014, with the breakdown $31.5bn for PC games, $24.5bn for console games, $24.5bn for mobile games and $3.1bn for other. We forecast CAGRs of 5.6% for this market in 2015-2025 and 1.2% in 2025 2035. This would put the 2025 market at $157.9bn. We think the market will center on console games through 2020 and that PC games will begin to contribute in 2020 2025, with mobile VR games emerging after that. So we project growth in VR games from $14.2bn in 2020 to $57.9bn in 2025 and $93.3bn in 2035. This would represent 11%, 37% and 52%, respectively, of the overall game market. We expect VR titles to replace 16% of all console games by 2020 and 40% of all PC games by 2025. In reverse, we forecast that games will account for 43% of the overall VR market by 2020 (disc purchases, downloads and streaming). Figure 38. Game content market (VR, non-vr) (US$, bn.) 200 180 160 non-vr game 140 120 100 80 60 40 20 0 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Source: Company data, IDC, Gartner, VZ Charts, Newzoo, Citi Research. 2031 2032 2033 2034 2035 VR Game replacing conventional game VR Game as new makt creation 36
VR movies, DVD/VOD/streaming We expect VR to drive big changes in cinema. Not only will the form in which movies are watched likely change, but cinema content itself is likely to change to suit VR headsets. While the impact that the latter change will have on the movie market is unknown, in this analysis simply assume that movies become VR compatible at the same rate as market uptake of VR headsets increases. We also expect movie theaters to provide customers with VR headsets in the same way to do 3D glasses. We estimate the movie screening market at $35bn $40bn, and expect the number of VR movies to begin to grow strongly around 2019. Rather than market uptake of VR headsets, we think the key factor will be a reduction in the price of headsets to the point that it is commercial viable for movie theaters to provide them to customers. We assume that the percentage of movies that are VR compatible is 8% in 2020 and 35% in 2025. Figure 39. Scenario for shift in movies to VR (US$, bn.) 50 45 40 35 30 non-vr box office 25 20 15 10 VR Box Office 5 0 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. We also expect uptake of VR for secondary uses of movies after they have been released, such as DVDs. Other than movies screened in cinemas, we think other VR content will be produced for distribution via VOD channels, and that VR will be adopted more quickly in the secondary use market than for movie screenings. We project the movie-related VR market at $11.3bn in 2020 and $24.1bn in 2025. While we expect TV content and music videos etc. to become VR compatible in the same was a movie content, our market projections factor in only DVDs and other secondary uses (we do not factor in TV broadcasts and public screenings etc.). 37
Figure 40. VR movie and DVD market (US$, bn.) 70 60 50 40 DVD, VOD, Streaming 30 20 10 Box office 0 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. Concerts and festivals one major market We think that VR is highly compatible with music concerts. Once the installed base of VR headsets reaches to critical mass, we expect the market to begin to ramp up around 2019, and ultimately grow into a VR market in excess of $10bn. Our US media analyst Jason Bazinet estimates the world concert market at $13.6bn in 2014 and the festival market at $3.47bn, and forecasts a growth rate of around 5% in the concert market in both developed and developing nations (please see his October 7, 2015 report Live Nation Entertainment, Inc. (LYV) - Two Drivers of Equity Upside with a Safety Net. Figure 41. Global concert and festival market (US$, bn.) 40 35 30 25 Market for Concert 20 15 10 5 0 2012 2013 2014 Source: Citi Research 2015 2016 2017 2018 2019 2020 2021 2022 2023 We forecast growth in the world concert and festival VR market to $3.5bn in 2020 and $7.7bn in 2025. While much of the VR concert and festival market will come from the substitution of existing markets, we also expect the creation of a reasonably large new market. This essentially means the production of events such as concerts specifically for VR. We project this market at $2bn in 2020 and $4bn in 2025, as part of a total market of $3.5bn and $7.7bn. 2024 2025 2026 2027 2028 2029 2030 Market for Festival 2031 2032 2033 2034 2035 38
Figure 42. VR concert and festival market trends (US$, bn.) 14 12 10 8 6 4 2 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research Other entertainment, commercial applications, tele-existence VR markets for which we did not do a detailed analysis but simply factored in tentative assumptions are commercial applications such as watching sports and conferences, and service areas in which VR looks like being useful, such as restaurant and hotel bookings. We assume that these markets total $2.2bn in 2025. Jason Bazinet Media, Cable and Satellite sector jason.bazinet@citi.com +1-212-816-6395 Tele-existence is the ultimate incarnation of VR. Tele-existence involves identifying (seeing) objects and people in a VR worlds as well as being able to touch them, speak to them and mutually interact with them. It is a VR world in which one can talk and walk with people and touch objects to confirm their texture. It would require the user to wear not only a headset, but a range of sensors on their body. As this is an area for the very long term, we did not include it in our forecasts. However, there is much active development already underway, so it is not simply at dream. Perspectives from Hollywood Jaunt, founded in 2013, is a virtual reality start-up based in the US. So far, the firm has raised $100 million in capital, the largest capital raise so far for a VR startup. Jaunt has robust ties to Hollywood including Disney (the lead investor), Evolution Media Partners, and Participant Media. European media firms, including ProSienben and Axel Springer, are also investors. However, Disney s relationship with Jaunt goes beyond financing. Indeed, ABC News (owned by Disney) has used the startup s production technology to produce content. Jaunt isn t the only startup getting funding from Hollywood. NextVR has raised about $35 million from Formation 8, Comcast, Time Warner, Dick Clark Productions, and MSG. NextVR has streamed a few live events including a US presidential debate and an NBA game. So, why are Hollywood firms keen to explore VR? Well, we don t expect every piece of content that comes out of Hollywood to be ripe for the VR experience. That said, we do think two factors will determine what type of content is most suitable for VR investments: First, does the narrative unfold in real time? Second, will the consumer find value in virtual presence? 39
Figure 43. Classification by value of virtual presence and need for real time narrative Value of Virtual Presence No Replay of extreme sports unavailable to all (skydiving, scuba diving) 2 1 Yes Replay of concerts Relive life moments (wedding, birth of baby) Video games Theme parks Traditional 4TV content Traditional film content Professional sporting event Olympics College sporting event Award shows (Oscar, Emmy s) Local newscast 3 Broadway show No Yes Real Time Narrative Source: Citi Research. If the answer to both questions is yes, then we expect VR to emerge as a popular option for consumers. This suggests sporting events professional, collegiate or Olympic will all emerge as hotbeds for VR investment. However, there are also types of content that are not live where virtual presence is apt to prove popular as well. Video games, replays of popular concerts, or the ability to experience extreme sports like skydiving or scuba diving without having to take on risks or purchase the requisite equipment could also prove popular. In the future, VR applications that allow consumers to relive life experiences, like a wedding or birth of a child, may also emerge as popular applications. This doesn t mean, however, that all forms of content are suitable for VR. For example, we don t expect live newscasts, Broadway shows, or traditional TV and film content (that takes the perspective from a single camera) to be conducive to VR. 40
4-2 AR commerce a big part of market AR commerce poised for sharp growth Mobile commerce is steadily increasing its share of the overall e-commerce market, and this is clearly a driver of smartphone market uptake. We think that AR headsets are a promising platform that will replace some smartphones, so there is likely to be a shift in the market from mobile commerce to AR commerce. In the case of the Japanese market, for which the METI provides a detailed analysis, the relevant commerce market in FY3/15 was 215trn*, within which the e-commerce market was 9trn/$73bn (market penetration of 4%). At present the global mobile commerce market is about $298bn. We forecast ongoing rapid growth in the mobile commerce market to $1.2trn in 2025. Within this, we forecast AR commerce using an AR headset at $40bn. We also project further subsequent growth to $975bn in 2035. *This is not nationwide transactions, but covers retail and service. Figure 44. Global trends in AR commerce, mobile commerce, e-commerce (US$, tn.) 4.0 3.5 e-commerce 3.0 2.5 2.0 1.5 m-commerce 1.0 0.5 AR-commerce 0.0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Euromonitor, Citi Research. AR commerce to grow at same pace as AR headset market penetration Forecasting AR commerce market based on e-commerce, mobile commerce We forecast CAGRs for the AR commerce market of 52% in 2020 2035 and 9% in 2025 2035. We think growth in the market will be in line with the increase in the market penetration of AR headsets. We forecast the market units of AR headsets at 99mn units in 2020 (estimated number of headsets in use based on cumulative sales assuming a lifespan of from four to five years), 1bn units in 2025, and 4bn units in 2035 and think the AR commerce market will begin to grow as soon as headsets become available with very little time lag. It took seven years from the time PCs became readily available before the e-commerce market started to grow, and around 2 years after smartphones became available for the mobile commerce market to ramp up. As the performance of devices has advanced, the time lag before the relevant commerce market starts to grow has decreased. AR headsets will have native broadband Internet connectivity, so we think that the AR commerce market will grow simultaneously with the ramp-up for AR headsets. With think AR commerce consumer behavior will be similar to current mobile commerce. Our market forecast assume that the AR market expands at around the same rate as the e-commerce and mobile commerce markets did. AR commerce will enable a range of consumption behavior using graphics displayed on the headset. We expect users to use AR headsets in the same way that they current use smartphones to make purchases, enjoy content and make bookings for a 41
variety of services. This AR commerce is not intended to be used when sitting on a chair at home, but rather when out and moving about and taking a break. In this sense usage looks like that of smartphones. 42
4-3 Voice phone, video phone, chat, SNS Using AR headset to make phone calls 140mn new lines to be created In the future, we think things currently done on a smartphone (voice and video phone, email/messaging/chat, SNS) will be done using AR headsets. Smartphone users pay monthly fixed rates to telecom operators, and we envision similar contracts for AR headsets. Our forecast for the AR headset market is 52.5mn units in 2020, and we believe that half of this will involve contracts with telecom operators that are similar to existing mobile phone contracts. By 2025 we think the market will increase to 343mn units, with the telecom contract rate rising to 44%. By 2035 we think AR headsets with telecom contracts will reach 14.4bn units. Figure 45. Installed base for AR headsets and volume of voice/data contracts 10,000 1,000 100 AR headset install base AR with telecom operator contract 4bn 1.4bn 10 1 10mn+ subscribers 100mn+ subscribers 0 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. Significant net growth even with some smartphone cancellations We assume that AR headsets will to some extent eat into the smartphone market. Therefore, the market expansion for AR headsets will not necessarily mean a similar level of expansion for telecom operator subscribership. However, we do think that AR headsets will be more accretive for the telecom and telephone market than concepts like IoT or wearables. What will the business model look like for using AR headsets for functions like telephony or email? We assume consumers will enter into fixed-rate contracts with telecom operators to pay for usage of both voice and data services. Therefore, the revenue from content and commerce discussed earlier will be separate and will not overlap. 43
Figure 46. Annual market for AR headsets, expansion for voice/data market 250 AR headset line contract market (US$, bn.) AR headset value market (US$, bn.) 200 150 $66bn subscription revenue 100 50 $6bn subscription revenue 0 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source: Citi Research. 4-4 Software Walter Pritchard US, Software walter.h.pritchard@citi.com +1-415-951-1770 The theoretical software applications for AR/VR are numerous as users find themselves in a more immersive user interface. Initially, high value applications industries such as field service (a mechanic receiving real-time instructions on how to fix a car with a hologram projected to help guide the repair), healthcare (patient assessment/diagnoses), e-commerce (trying on clothes virtually) and architecture/construction (virtually walking through the proposed structure as its designed / built) are a few examples of the many that intuitively seem viable. Although, as with any new technology, initial integrations will most likely come in the form of niche applications and over time, as the AR/VR interface becomes more like cameras on smartphones (with standardized APIs), software applications would see deeper integration with existing applications. Ultimately, it is the applications software companies that will benefit the most as the new interface increases the value of the application, much the same way that mobility increased the value of some of the same applications. We note the companies making the most investments here are in areas where there are early, viable AR/VR applications noted above. MSFT is making significant investments here, with consumer implications first (see below) while CRM is making significant investments in order to integrate AR/VR into customer service and other related business processes. Finally, ADSK is making early investments to bring the technology to the architecture/construction space in a similar fashion to the example provided previously. 44
4-5 Services: What can a services company do in AR/VR? Ashwin Shirvaikar, CFA US, Computer Services & IT Consulting ashwin.shirvaikar@citi.com +1-212-816-0822 There are multiple different roles available for IT services companies that are doing AR/VR work. At a broad level, this work involves a systems integrator stepping in to integrate disparate systems and components such as sensors (gyro, accelerometer, etc.); applications (by vertical, e.g., manufacturing, healthcare, military); and components (heads-up display, special gloves, camera, etc.). This is still a cuttingedge area that not every services company has invested in or is focused on. However, those that have invested (Globant and Accenture come to mind) now have capabilities in specific use-case areas like i) gaming, ii) digital marketing, iii) social/mobile platforms, iv) corporate training, and v) vertical-specific applications in healthcare, retail and manufacturing. In terms of technical skillsets required, a general understanding of human-machine interfaces (HMI), photo-rendering, 3D rendering, 3D graphics programming (including specific game engines like Unity 3D and Unreal and experience with Oculus / Samsung Gear / familiarity with HoloLens), etc. is needed. Technical skills are only one aspect of the job though: the more important skill is around user interface and user experience (UI/UX) design with both the design and technical skills mattering. The opportunity set is difficult to measure currently because this is still cutting-edge work that is necessarily early stage and smaller in size. However, it is not uncommon to see industry analyst estimates that run into multiple billions of dollars of services-related work by the end of this decade. Services companies have done work with AR/VR in a number of ways so far. Some good examples include: 1. Globant has created a commercial-use AR app to help car manufacturers where customers go to a retailer, point their phone at a car, and receive additional information on the car. 2. Globant has used AR to create videos that make the job of a healthcare provider somewhat easier. For example, this app makes it easier for a child to take medicine through a mobile application where a phone is pointed at medicine and an animated cartoon is played on the phone to make the experience more enjoyable for the child. 3. Accenture is working on VR for manufacturing technicians where they are aided with VR headsets that decrease the effort required and eliminate errors on high-end aerospace assembly. 4. Globant has used VR to aid planetariums through a virtual space simulation system that processes data in real time. 45
5. Potential beyond headsets In this analysis we have limited the VR/AR market universe to headsets to ensure the rationality of our market forecast model. Our analysis covers headsets and peripheral devices as well as related content, communications, and e-commerce. However, VR/AR is not limited to headsets. There are other promising technologies and applications over the long term. These are not included in our market forecast model, although we believe there could be opportunities to use them as the headset market expands. Below we introduce several non-headset VR/AR technologies that are worthy of attention. Figure 47. HUD-AR example: Pioneer's AR navigation system Source: Pioneer Corporation. 46
5-1 Motion capture critical to future of VR/AR Wearable sensors that relay body movement VR/AR will enable users to look at a virtual world and CG objects using the display, but it may also be possible in the future to touch, manipulate and move objects. In addition to the headset, this will require sensors on the body and hands in order to communicate the body s movement to the computer. To do this, users will wear gloves with sensors and devices with actuators (haptic devices) that can relay the sense of touch. These are so-called wearable devices. If all that is required is to recognize movement in a room or the movement of hands in front of the user s body, sensors that use cameras (image sensors) and infrared sensors could be placed in the corner of a room or in front of the user s body to measure movement. We do not specifically factor volumes or values for these peripheral devices into our industry forecast model. However, once VR/AR headsets that enable a more immersive experience are development and computer process capacity increases, there will be demand to be able to touch objects and be touched in the virtual world. Here we outline a range of such devices. Leap Motion: VR-related The US venture company Leap Motion has launched is compact sensor device also called Leap Motion. This enables a user to move both hands in the space covered by the sensor, which will recognize the movement and perform a range of tasks on the PC monitor. If a user starts up Google Earth and moves his hands with Leap Motion, he can move on the screen as though he is flying around a town in the sky. If this sensor were mounted on a VR headset (Oculus Rift) and connected to VR there would be many future possibilities. By establishing the sensor recognition region in the area in front of the user s eyes, at the same time as the sensor recognizes the movement of hands held in front of the body, the virtual world created by VR in line with the movement of the head shows the user s hands. Thus the hands in the VR display move just as if they are the user s own hands. Figure 48. Leap Motion sensor recognizes space Source: Leap Motion. 47
Figure 49. Using your own hands in VR with Leap Motion Source: Leap Motion. Xbox Kinect: VR-related While the recognition accuracy of Kinect, which was launched in 2010 and drove sales of the Xbox 360, is lower than Leap Motion, it was the first to introduce a mechanism to track human joints. It detects human movement using image sensors and infrared sensors. The information collected is displayed as CG on a screen. It was announced under the name Project Natal in June 2009 and attracted much attention. Launched in November 2010, sales were very strong. Cumulative sales are more than 10mn units, and the next generation Kinect for the next generation Xbox One is now on sale. Kinect is a device that is highly compatible with VR, and we think Microsoft will use it as a technological base point when expanding its business in the VR area. VR will be used mainly within only a limited range of portability or no portability at all, so detecting a user s movement using a fixed sensor device such as Kinect will be effective. We do not specifically factor Kinect into our market model, but we think this mechanism of sensing movement within an overall space and displaying it in a virtual space is a promising component of VR. Figure 50. Kinect Figure 51. Kinect detects motion by scanning joints Source: Microsoft. Source: Citi Research 48
NEC s ARmKeypad: AR-related The Japanese IT services company NEC has developed AR software that is useful when using measuring equipment. By combining an AR headset with a smartwatch, it is designed so the user can see CG through AR glasses. A keyboard and switches are displayed on a worker s arm, and these are used to control the actual measuring device. It is not a semi-transparent hologram, but rather shows a clear keyboard attached to the arm via CG, and the camera sensor on the AR headset and accelerometer in the smart watch are used to recognize the movement of the fingers when using the keyboard. The system will be particularly useful for workers who need their hands to remain free, such as facility inspectors, home delivery drivers and healthcare professionals. There is no need to stop work in order to pull a device out of one s bag or pocket, and the task can be completed using the virtual keyboard. This will help with the accuracy of the job, and would probably also have hygienic benefits in the medical industry. NEC has developed the software, but says the hardware such as AR headsets and smartphones will be from other companies. We have not specifically included this type of software in our industry model. However, we think that the software used by AR headsets will in many cases function in this way to assist workers to do their jobs. Figure 52. NEC: ARmKeypad Source: NEC. Fujitsu s ring shaped controller and globe shaped sensor: AR related Fujitsu, which is involved in IT services, PCs and semiconductor fabrication, has developed a globe shaped sensor and a ring shaped controller. The globe shaped sensor links to an AR headset, and displays the necessary information on a display according to the movement of the user s hands. The sensor on the wrist detects hand gestures, and will display information in conjunction with specific movements as well as displaying information according to what is touched using the sensor worn on a finger. The sensor on the wrist can detect the movement of the arm so the user can, for example, turn the pages of a manual displayed on the AR headset, and has functions such as enabling a user to enter results such as test passed via CG using arm movement. The sensor on the finger has an NFC tag reader, so if NFC tags are attached to specific objects, there is no need to supplement cameras and the like with hand and finger movement, and the information in the NFC tags that are touched can be displayed in the headset. 49
We expect this device to be used in manufacturing plants and for healthcare, education and retailing. It will be most useful when it is inconvenient to take out a terminal or controller to complete a task. The ring shaped terminal can be used to write words in the air without holding anything in the hand to enter letters and numbers and select information shown on the display. The fundamental mechanism is to use the incorporated accelerometer, gyroscope, and magnetometer to detect movement, but its miniaturization into a ring shaped device enables information entry using hand and finger movement. Figure 53. Fujitsu's glove-type sensor device Source: Fujitsu Figure 54. Fujitsu's ring-type sensor device Source: Fujitsu. 50
5-2 Enhancing HMI: HUDs and Augmented Reality for Automotive Itay Michaeli US, Automotive itay.michaeli@citi.com +1-212-816-4557 The heads-up display (HUD) concept has been around for quite a while, in fact the first known automobile HUD went into production with GM in 1988. From that point forward HUDs have been a very niche product mainly due to size, cost and benefit constraints. With recent trends of infotainment, a more focused effort to improve safety by helping to keep eyes on the road, the declining cost of components, and the reduced size of components/end system, this niche product is on track to become a more mainstream product which is widely accepted by many vehicles. Factor in the evolution of the HUD to encompass augmented reality and you have an explosive growth product within the global automobile industry. HUDs can either be in one of two fittings: 1) windshield HUD (currently the only one with real production volume) a HUD that projects information directly on the windscreen of the vehicles; and 2) a combiner HUD, which projects the information to a separate piece of glass or plastic, known as the combiner. Windshield HUDs, as shown below, project directly to the windshield, with no combiner piece. The ASPs on windshield HUDs (as per an HIS survey) carry a premium of c2.3x combiner HUDs, even as the volume mix shifts from windshielddominated to a more combiner-dominated. The Cadillac CT6, unveiled at the NYC Auto Show has a windscreen HUD Figure 55. Evolution of windshield HUDs Source: Continental, Citi Research. 51
Combiner HUDs shown to the right this feature can be added to a MINI Cooper for $500 Figure 56. Combiner HUD display Source: MINI, Citi Research. Sizing up the markets: The current and the future The overwhelming majority of HUDs currently produced are windshield HUDs, which typically carry an ASP of c2.3x more than combiner HUDs. In sizing up the market we believe that units (1 unit per vehicle) are expected to grow at a 2014-2018E CAGR of ~39% (1.6 million units to ~6 million units). At the same time, the revenue TAM is expected to grow at a CAGR of c27% (from c$650 million to nearly $1.7 billion). It is important to note that the growth trends vary by HUD technology; for example, the 2014-2018E revenue TAM of the windscreen HUD market in expected to grow by a CAGR of c21% to $1.3 billion; the combiner HUD market CAGR is estimated at c98%, growing to $375 million. 52
As expected, North America, China and Western Europe remain the most important regions for both unit TAM and revenue TAM, accounting for 75% of the total TAM in 2018E. Figure 57. HUD market overview: Revenue TAM, volume, ASP $1,800.0 5,809 7,000 6,000 $1,200.0 5,000 $600.0 1,573 $1,681.9 4,000 3,000 2,000 $0.0 $643.6 2014 2015E 2016E 2017E 2018E Revenue TAM ASP Trend Unit Volume (in #K) 1,000 0 Source: IHS, Citi Research. North America, China and Western Europe remain key regions for adoption of HUD technology. By 2018 it is estimated that overall HUD penetration will be c6%. The major regions adopting this technology are no surprise: Western Europe, Japan/ South Korea, and North America remain the highest regions by percentage of sales adoption; however, it remains imperative to consider adoption at the absolute unit level, as this is how suppliers generate incremental sales revenue and because without a unit volume base one would miss the explosive growth in China. So while the adoption rate is important to see which countries have a more favorable view on advanced automobile technology, one must just as equally focus on actual units shipped. As such, North America, China, and Western Europe are estimated to have the highest unit volume sales by 2018, accounting for c75% of total shipments (c1.4 to 1.5mn units per country). A further step to the future augmenting reality Figure 58. Global sales penetration of HUDs (windshield + combiner) 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% Indian Subcontinent Central / Eastern Europe Source: IHS, Citi Research. South America Greater China 2014E Sales Penetration Rest of the World The beautiful thing about HUDs is that the very nature of the product allows for OEMs to leverage it as an ideal platform for multi-modal technological convergence. What we mean by this is that as the infotainment and active safety megatrends continue to develop, they will eventually converge where all information will be available and accessible in one place, allowing for faster adoption and increased penetration. For HUDs, this can be achieved by leveraging the prime real estate Other Asia Pacific North America Japan / Korea 2018E Sales Penetration Western Europe Global 53
within a vehicle to not only display driver information, but also provide active safety cues to help the driver gauge potential troublesome situations, or provide navigation and other directions in one easy to view location. We already know the trajectory for both windshield and combiner basic HUDs, but the next step in the evolution of the HUD is to augment the reality of the driver s viewpoint, allowing for 3D effects to cover the depth of the scene ahead. The 3D aspects of the augmented reality will not only help the driver minimize distractions and help keep them focused on the road, but it will also help to reduce the uncertainty of the road ahead. As seen in the images below, the augmented 3D reality is able to help enhance both infotainment (real time navigation becoming part of the road) and active safety technology (3D distance between cars to help prime AEB or ACC) into one easy-to-view GUI/HMI. Figure 59. Augmented reality HUD: ACC Figure 60. Augmented reality HUD: NAV Source: Continental, Citi Research. Source: Continental, Citi Research. A small, but hyper-growth market Augmented reality HUDs (AR-HUDs) will all start off as windshield HUDs, and will be at a premium relative to the typical windshield HUDs; the exact ASP of these AR- HUDs are currently unknown as only very limited shipments are expect to occur from 2014E to 2019E. From 2019E through 2021E unit shipments should ramp significantly, driven by increased penetration relative to standard windshield HUD shipments, but will still remain a niche market. As the windshield technology does carry a much higher ASP versus combiners, we believe unit shipments will likely remain low as these costs could be a headwind to more mass market adoption, thus limiting this technology primarily to luxury vehicles at first. By 2018 we estimate that 3.5 million windshield HUDs will be shipped, of which a mere c5,800 units will have some type of augmented reality. This number of units could reach 455,000 by 2021 or c10% of total windshield HUD unit shipments (or c0.4% of global sales). 54
Figure 61. Windshield HUD sales (LHS) and windshield AR-HUD Sales (RHS) / Windshield AR- HUD sales are included in the total windshield HUD sales 6,000 600 4,500 450 3,000 300 1,500 150 0 0 2014 2015E 2016E 2017E 2018E 2019E 2020E 2021E Windscreen HUD Sales Windscreen AR-HUD Sales Source: IHS, Citi Research. As HUDs are very software-driven product platforms, we tend to favor this technology in our "Car of the Future" investment framework. As such, we think the main beneficiaries could be companies with reflection-based technology (windshield or combiner vs. light bars) such as Nippon Seiki, Continental, and Visteon. Additionally, this software-heavy solution should allow for higher margins and higher ASPs on future iterations of these products for example, any increased penetration of AR- windshield HUDs in favor of typical windshield HUDs would help expand the revenue TAM for total HUDs. 55
6. Lots of participants, spanning hardware, software, content, retail, and services As VR/AR has the potential to replace parts of many existing industries, the range of related technologies, businesses, and companies is wide. Hardware not only encompasses displays for visual content and headphones for sound, but also includes key electronic components sensors such as accelerometers and gyro sensors which ensure that images adjust as the user s head position (line of sight) changes. As with goggle- and eyeglasstype headsets, we believe the importance of sensors is likely to come to the fore when VR/AR advances to include hand and feet wearables. We also note the importance of software and content. Sophisticated software is vital for the operation of wearable machines and infrastructure. In the creation of highly immersive games or movie content, we feel technical quality will play a central role overall. Finally, an industry to manage such platforms is indispensable. We surmise that such a role could be performed by the retail industry, which handles e-commerce and that a variety of service industries will broaden business breadth by bringing on board VR/AR. Figure 62. VR/AR supply chain Hardware device, Components Network infrastructure Software, platform CG, Game & other info contents Consumer IT service Source: Citi Research. Commerce New market creation Data (distribution, transaction) Commercial IT service 56
Sony (6758.T, Buy, TP 3,900) Kota Ezawa Japan, Industrial & Consumer Electronics kota.ezawa@citi.com +81-3-6270-4804 Figure 63. Financial Summary Year to 31 Mar 2014A 2015A 2016E 2017E 2018E Sales ( mn) 7,767,266 8,215,880 8,140,000 8,104,000 8,006,000 Oparating Profit ( mn) 26,495 68,548 389,700 468,100 509,800 Net Income ( mn) -128,369-125,980 202,000 232,300 258,900 Diluted EPS ( ) -124.99-113.04 164.67 180.82 201.53 PE (x) NM NM 19.3 17.6 15.8 EV/EBITDA (x) 9.1 8.2 4.3 3.4 2.9 DPS (Y) 25.00 0.00 33.00 36.00 40.00 Net Div Yield (%) 0.8% 0.0% 1.0% 1.1% 1.3% Note: Based on the November 30 closing price. Source: Company data, Citi Research. Overview of Sony s VR/ARrelated ops VR/AR SWOT Sony s overall earnings Sony says it will launch the PlayStation VR (previously called Project Morpheus) business by June 2016. VR headsets might be launched on the market at the same time. This will represent the commercial launch of a major platform business that has been in preparation for two years since being announced. Competition will be a key focus given that other VR devices are poised to launch at around the same time, including products by Facebook (Oculus Rift) and HTC (HTC One). We see Sony as a prime candidate to spearhead VR industry growth. We think the company has the competitive edge to go toe-to-toe with any other global player. The PlayStation VR is handled by Sony Computer Entertainment (SCE), which runs the game business. Unlike the firm s electronics divisions, which excel in short-term hardware sales strategies, SCE is adept at running a business model built around software/services as well as hardware. This is because SCE s experience with the PlayStation business has equipped it with the strategic mindset and patience that are required to establish a platform business. Sony also has a substantial track record in electronic component and semiconductor technologies that can help hone the competitive edge of its VR business. The company s standalone electronic component technologies include GPU technology developed through its sensor and games console offerings as well as compact display technologies developed through smartphones, portable gaming machines, goggle-type home theater devices other than the PlayStation VR, and digital camera viewfinders. The firm also has a rich technical foundation in mechatronic devices (like sensors), partly due to its experience in marketing robot products in the early 2000s. We think there will be marked variations in user opinions on different companies VR/AR headsets at first. We believe Sony has a distinct edge over other companies in this area. However, Sony lags the US Internet majors in terms of sales ability, customer reach in the Internet domain, and spending power. Given that in the market s early stages it will likely be a long time until benefits of scale kick in, we see some concern as to whether Sony can remain competitive while navigating the capital recovery cycle in the games business. Consolidated earnings are in a growth phase thanks to improving profit contributions from multiple businesses, including the games segment which is currently preparing for the PlayStation VR launch. We forecast sales of 8.14tn ($67.8bn) and OP of 389.7bn ($32.5bn) in FY3/16. In FY3/17 we expect sales to fall 0.4% and OP to grow 20.1%, putting the margin at 5.8%. We think OP will go on to head for the 500bn benchmark, driven by growth in image sensors, games, 57
music and other entertainment businesses and helped by the benefits of TV and smartphone business restructuring. Investment stance: Buy recommendation Our rating is Buy and our target price is 3,900. We use an EV/EBITDA-based SOTP valuation model, assigning high multiples to media, gaming, and internetrelated businesses. Our target price equates to an FY3/16E P/E of 23.7x, which values the company more highly than Japanese technology sector peers. 58
HTC (2498.TW, Sell, TP NT$28.00) Dennis Chan Greater China, Communication Equipment & Components dennis.chan@citi.com +886-2-8726-9086 Figure 64. Financial Summary Year to 31 Dec 2013A 2014A 2015E 2016E 2017E Sales (mn NT$) 203,403 187,911 117,422 88,193 85,197 Oparating Profit (mn NT$) -3,971 669-12,601-6,596-6,575 Net Income (mn NT$) -1,324 1,483-14,789-5,613-6,287 Diluted EPS (NT$) -1.59 1.81-17.86-6.78-7.59 PE (x) NM 44.8 NM NM NM EV/EBITDA (x) 23.6-5.6 0.3-3.7-4.5 DPS (NT$) 0.00 0.38 0.00 0.00 0.00 Net Div Yield (%) 0.0% 0.5% 0.0% 0.0% 0.0% Note: Based on the November 30 closing price. Source: Company data, Citi Research. Vive, HTC s virtual reality headset built in partnership with Valve, is expected to premiere on December 28, 2015, according to Digitimes. In the same report, Digitimes also reported expectations that a series of Vive-powered games will be launched in December along with the Vive device. We understand the first will be launched in the US, UK, and Germany. The devices will first find its place in gaming. However, HTC has partnered with Google, HBO, Lionsgate, and the National Palace Museum in Taiwan in order to bring VR into other mainstream applications such as traveling, education and shopping. HTC showcased its Vive product at MWC 2015 and dispatched developer kit to developers a few months later. The development kits are equipped with two OLED displays (1080x1200 pixels for each eye) supporting a 110-degree field view with 90Hz refresh rates to ensure smooth motion. Unlike Oculus and Sony s VR products, HTC Vive can be used while walking around, which means the system can track the user s position in a 15 square feet area using two lighthouses (base stations that monitor user s movement via laser position sensors) that let users walk around the virtual objects. The handheld motion controllers combined with gyrosensors and accelerometers enable the system to capture precise user motions (i.e., position, head movement and vision) and manipulate objects on the virtual plane with a realistic sense of touch. According to an ITpro report quoting an HTC-connected products marketing executive director, Jeff Gattis, the Vive will reportedly target the high end of the consumer market, delivering a premium VR experience. This will, he says, mean a higher initial entry price for early adopters. 59
Advanced Micro Devices (AMD.US, Sell, TP$1.5) Christopher Danely US, Semiconductor Christopher.danely@citi.com +1-415-951-1885 Figure 65. Financial Summary Year to 31 Dec 2013A 2014A 2015E 2016E 2017E Sales (mn USD) 5,299 5,506 3,983 3,600 3,600 Oparating Profit (mn USD) 103 163-332 -253-204 Net Income (mn USD) -83-403 -677-449 -400 Diluted EPS ($) -0.11-0.22-0.73-0.57-0.51 PE (x) NM NM NM NM NM EV/EBITDA (x) 8.5 8.0-20.2-20.6-31.5 DPS ($) 0.00 0.00 0.00 0.00 0.00 Net Div Yield (%) 0.0% 0.0% 0.0% 0.0% 0.0% Note: Based on the November 30 closing price. Source: Company data, Citi Research. AMD seeks to leverage its strengths in the core PC gaming segment into the virtual reality market. In March 2015, AMD announced its LiquidVR virtual reality technology initiative. AMD s LiquidVR software development kit helps developers effectively combine AMD s GPU with its virtual reality software to enhance user experience with features such as low latency and high refresh rates. We would note AMD is not building its own virtual reality hardware. Instead it is partnering with multiple virtual reality headset makers such as Oculus VR and Valve. Virtual reality headsets can run at up to 90 frames per second, which is roughly three times as fast as standard PC monitors (which generally run at 30 frames per second). As a result, AMD should benefit from the high graphics processing power requirements for virtual reality rendering. We believe one of the main reasons AMD created the Radeon Technologies Group in September was to focus on emerging opportunities in virtual reality. The Radeon Group will be led by chief architect, Raja Koduri, who has more than 20 years of total experience in graphics architecture at Apple and ATI/AMD. AMD seeks to tackle common virtual reality user issues such as motion sickness by offering low-latency head tracking, whereby scenes quickly update when a user turns their head. The company is also offering scalable rendering whereby developers can utilize multiple CPUs and GPUs to render images at faster refresh rates. Lastly, in the PC environment, AMD seeks to make an easy transition from the PC to the virtual reality headset whereby a user can boot Windows directly to their headset. According to AMD, the ideal virtual reality system offers a 120 degree horizontal field of view, 135 degree vertical field of view, 16k resolution, and less than 20ms of motion latency. AMD offers several high end products targeting virtual reality applications, including the R9 290 card, priced at $399, and R9 290X card, priced at $549. Oculus is collaborating with PC OEMs such as Dell, Alienware, and ASUS, and GPU makers AMD and NVIDIA, to bring Oculus Ready PCs market. An Oculus Ready PC is one that meets the required graphics and computing requirements to support stereoscopic 3D rendering. According to Oculus, these requirements include at least an Intel Core i5 processor and either an AMD R9 290 card or a NVIDIA GTX 970 card. We estimate AMD s GPU business should generate revenue of roughly $535mn in 2015 (13% of total sales), with low single-digit operating margins. Of that $535mn, we estimate virtual reality represents less than $5mn in annual revenue. We 60
believe it will be another two to three years before virtual reality represents a more meaningful portion of AMD s revenue as the industry is still nascent, and headsets are not yet being widely adopted by PC gamers due to high costs and lack of games that support it. 61
Microsoft (MSFT.US, Sell, TP$40.00) Walter H Pritchard, CFA US, Software walter.h.pritchard@citi.com +1-415-951-1770 Figure 66. Financial Summary Year to 31 Dec 2013A 2014A 2015E 2016E 2017E Sales (mn USD) 77,849 86,833 93,580 92,534 95,649 Oparating Profit (mn USD) 26,764 28,013 38,183 28,227 29,492 Net Income (mn USD) 21,863 22,328 31,685 21,516 22,231 Diluted EPS ($) 2.58 2.66 2.63 2.70 2.89 PE (x) 21.1 20.4 20.7 20.1 18.8 EV/EBITDA (x) 12.0 10.8 8.1 10.6 10.1 DPS ($) 0.88 1.06 1.20 1.39 1.55 Net Div Yield (%) 1.6% 1.9% 2.2% 2.6% 2.8% Note: Based on the November 30 closing price. Source: Company data, Citi Research As mentioned previously, Microsoft announced Microsoft Hololens in January 2015 during a Windows10 press event. The company is expected to release a developer model in January March 2016, priced at US$3,000. Hololens is an independent goggle-type computer equipped with multiple sensors to detect outside world information and equipped with a lens for displaying holograms (rather like a projector). Demo versions allow playing Minecraft, Halo, and other games with real world objects treated as virtual objects. In addition to games, the equipment has been shown in demo sessions to be compatible with various simulations and graphics-rich applications. While on the surface Hololens appears a competitive product, we hesitate to conclusively determine its relative competitive positioning in this nascent space. AR/VR is becoming a UI requirement for operating systems, and so we believe MSFT will continue to invest here through Windows 10 / Hololens. At a high level, this incremental investment is an attempt to arrest the decline in Windows relevance due to the broader shift away from PCs towards mobile devices like smartphones, tablets and now AR/VR headsets. We believe Hololens will see the majority of its early adoption in the consumer market (specifically in the gaming space) though as with the majority of MSFT s business, this initial consumer exposure is an attempt to create customer familiarity with the interface to eventually drive adoption and monetization in the enterprise. From a financial perspective, we believe the revenue impact is negligible in the short term given it is unclear the appetite consumers will have for a US$3,000 gaming product, and while the impact on expenses is tough to precisely pin down, Hololens fits in with the broader MSFT trend of funding unprofitable, consumer-oriented businesses in an attempt to keep its operating system relevant as it looks to transition further to the cloud space. 62
Figure 67. SWOT analysis Source: Citi Research Other related companies NVIDIA Christopher Danely US, Semiconductor Christopher.danely@citi.com +1-415-951-1885 NVIDIA (not covered) has created a software developer kit branded as GameWorks VR that enables headset makers and video game developers to optimize virtual reality performance on NVIDIA GPUs. Similar to AMD, NVIDIA s software developer kit helps developers tackle issues such as low-latency head tracking, scalable rending on multiple GPUs, multiple resolution shading, and ensuring seamless plug and play experience for headsets. NVIDIA enables a feature they call asynchronous warp, which reduces latency by adjusting images as users move their heads (just-in-time) rather than continuously rendering new frames. NVIDIA claims asynchronous warp reduces image latency by 50% from 50ms to 25ms. NVIDIA also announced its DesignWorks VR developer kit in August 2015, which is targeted at the professional virtual reality environment rather than gaming. DesignWorks VR is built off many of the same libraries and features of GameWorks VR, but targets professional use cases such as automotive designers and movie developers, and computer-aided design (CAD) designers. For example, Ford can use virtual reality to create digital 3D prototypes to analyze the interior and exterior of car without the need for a costly physical prototype. The movie industry is also working on a virtual reality enabled film experience, whereby moviegoers are fully immersed in the virtual reality experience. NVIDIA offers multiple virtual-reality enabled GPU products with the GeForce GTX 970 card priced at $329, GeForce GTX 980 priced at $499, GeForce GTX 980Ti priced at $649, and flagship GeForce GTX Titan X priced at $999. The latter cards 63
are specifically optimized by NVIDIA for virtual reality applications and currently available. We would note the Oculus VR headset requires a GTX 970 GPU or higher. Asahi Kasei Takao Kanai Japan, Chemical Asahi Kasei was the first in the world to bring an electronic compass to market, and it has become the de facto standard. This is because in addition to the Hall effect sensor and signal processing IC technology amassed over the years, Asahi Kasei has successfully developed an algorithm that indicates direction. We believe it continues to boast the top share in the electronic compass market. Initially this was adopted in some mobile phones, and volume has grown along with smartphone market expansion. However, competition with peers in Japan and abroad has gotten fierce, so the business is now struggling in terms of both prices and volumes. Asahi Kasei intends to maintain growth as an electronic component maker via horizontal deployment to other products like magnetic sensors. It is unclear what the costs of dealing with the piling business problems will be, but excluding construction materials we see little impact on the overall business, including housing. We look for earnings to grow steadily driven by chemicals, textiles, and critical care. While the acquisition of Polypore will weigh on earnings for now, overall earnings should be firm. We do not expect a major recovery for the shares until the impact of the piling business issue are clearer. However, fundamentals are sound. In addition to chemicals, fibers, and critical care, we think Polypore will contribute to growth over the medium term. Even excluding one-off losses related to the piling business we think valuations look appealing. Figure 68. Electric compass by Asahi Kasei Source: Asahi Kasei 64