三季度业绩发布之际对无线设备投资者关注热点的更新 -...

2017 年 10 月 19 日

全球：通讯技术：设备证券研究报告

三季度业绩发布之际对无线设备投资者关注热点的更新 (摘要)

投资者关于无线设备和 5G 的热点话题本报告中，我们重点讨论了最近数周的会议会谈中投资者关于无线设备行业的五

个热点话题。我们看到 5G 基站的部署有加速的可能，新设施将为在固网和无线

市场均有布局的设备供应商提供最佳机遇。我们对诺基亚、中兴通讯(H)和爱立信

的买入评级以及高于市场的预测反映了我们的观点，即参与各方都可受益于 5G所带来的无线开支的周期性反弹。我们认为如果无线开支部分转向固网，则诺基

亚/中兴通讯的布局较好，而爱立信的利润率也将因自身举措而强劲扩张。

5G 商用服务的推出是否会早于预期？我们目前预计 5G 基站将于 2019 年开始商用，较我们此前的预期早一年。诺基亚

预计美国/中国将在这一时间框架迎来大规模商用部署，而我们认为中国的发展也

将带动其它地区的部署。全球电信运营商的 5G 试点进展也可能会促进相关标准

的清晰化。

5G 相关的设备开支周期规模将有多大? 我们认为 5G 资本开支水平将与 4G 类似，新频段的使用将带来大基站的部署，从

而我们对诺基亚/爱立信/中兴通讯的预测高于预期。我们的情景分析显示，如果无

线支出转向固网，则这些公司的增长也仍将迎来复苏。

5G 网络的结构特点？ 5G 的特点是网络将更像“云”。非实时功能（如用户管理）或许会在数据中心运

行，从而带来额外的前程/回程固网开支。

未来几年中国市场的开支趋势？我们认为非传统企业的投资可推动 5G 技术加速采用。我们的敏感性分析表明中

兴通讯可得益于中国市场更快速的普及/市场份额的变化，而诺基亚/爱立信受到的

影响较小。

行业定价如何？新的市场结构是否对此有影响？尽管定价形势仍然严峻，但随着中兴通讯市场份额扩大推动全球市场结构改善，

我们预计定价会相对稳定。

*全文翻译将随后提供

相关研究

5G: How 100x faster wireless can shape the future, 2016 年 4 月18 日

What we learned at our 3rd Innovation Symposium, 2017 年 6 月

28 日

Nokia: Compelling value post earnings reset; synergy upside, Up to Buy, 2016 年 6 月 29 日

Ericsson: Restructuring potential with market improvement, reiterate Buy, 2017 年 7 月 10 日

中兴通讯: 在正确的方向上采取三项举措；上调评级至买入, 2017年 8 月 7 日

各代蜂窝技术的峰值数据速度

资料来源：华为、高盛全球投资研究

侯雪婷执业证书编号: S1420515060001 +86(21)2401-8694 [email protected] 北京高华证券有限责任公司

北京高华证券有限责任公司及其关联机构与其研究报告所分析的企业存在业务关系，并且继续寻求发展这些关系。因此，投资者应当考虑到本公司可能存在可能影响本报告客观性的利益冲突，不应视本报告为作出投资决策的唯一因素。有关分析师的申明和其他重要信息，见信息披露附录，或请与您的投资代表联系。

北京高华证券有限责任公司投资研究

3G (HSDPA) 4G (LTE cat 4) 5G

14Mbps

150Mbps

10Gbps

2017 年 10 月 19 日全球：通讯技术：设备

全球投资研究 2

Key investor debates and feedback pre 3Q results season

Could 5G commercial rollouts begin earlier than expected? Wireless equipment spending is in its third year of decline, following the initial benefit to

CommTech equipment industry revenues from the first 4G rollouts (2012-13). As such, many investors are trying to gauge whether the industry is at a cyclical trough and pinpoint the timing of the next-generation network technology upgrade – 5G. We now believe 5G networks will be commercially rolled out in meaningful scale in 2019 (vs our prior view of 2020), suggesting Ericsson, Nokia and ZTE will all benefit from the next upcycle in the coming years.

In our report, ‘5G: How 100x faster wireless can shape the future’, April 18, 2016, we initially

argued 5G networks (as per Exhibit 1) will bring 100x the speed (delivering better consumer experiences e.g. for ultra fast video), ultra low latency (key for e.g. autonomous driving, as well as VR), extreme reliability (key for mission critical IOT e.g. connected healthcare) and a high degree of scalability (facilitating a step change in number of sensors/devices connected - e.g. Factory 4.0 applications). As such, given the potential to deliver value to telcos (and a broadening pool of customers), we concluded that it was realistic to expect the first commercial rollouts to occur in

2020. However, as per takeaways from the 5G panel at our GS Innovation conference, attended by Nokia, Verizon and Qualcomm (see ‘What we learned at our 3rd Innovation Symposium’ June

28, 2017) we believe initial 5G commercial rollouts will begin earlier - in 2019.

Exhibit 1: 5G enables critical use cases such as IoT based on high reliability/low latency Wireless network generation roadmap

Souce: Goldman Sachs Global Investment Research.

Key reasons for the faster rollout timeline are a) Nokia (one of the top 3 global equipment vendors) recently stated it expects meaningful commercial 5G deployments in the US and China in 2019; b) progress on 5G trial activity by telcos globally can help standards crystallize; and c) 5G enabled

devices (smartphones, routers, etc.) will be available in 2019.

1980s 1990s 2000s 2010s 2020s

1G

2G

3G

4G

5G

Mobile Analog Voice

Mobile Digital Voice

Mobile Messaging

Mobile Internet Access

Mobile Broadband

All-IP (Integration of Voice and Data)

Enhanced MobileBroadband

Support for MassiveIoT Ecosystem

Ultra-Reliable, Low-Latency

5G will bring fiber-like speeds and reliability to wireless networks while enabling the deployment of

a massive ecosystem of IoT devices

5G networks offer 100x the speed vs 4G, with 50x lower latency and 2,500x the throughput, enabling ultra-fast video, mission critical IoT & Factory 4.0 use cases.



Nokia expects meaningful 5G deployments in the US and China in 2019: Nokia

at its 2Q17 results highlighted that while it had previously expected initial 5G commercial rollouts in 2020, with scale rollouts in 2021, the timeline has been pulled in by a year,

with telco discussions suggesting meaningful initial builds in 2019 (see summary of vendor and telco statements on timelines in Exhibit 2).

Moreover, Rajeev Suri, CEO of Nokia, subsequently highlighted China could be one of the first movers on 5G and other regions could potentially follow too

(per our recent London roadshow, where he highlighted China as a potential early mover joining US, Korea and Japan). The impetus towards 5G in China appears to be the

recent equity investment by webscale players Baidu, Alibaba and Tencent in the China telco space, where more aggressive moves towards 5G rollout could catalyse other carriers to make progress in this area.

Furthermore, while key European telcos such as Vodafone have so far stated commercial rollouts are targeted for 2020/21, industry discussions across the region suggest a greater desire to move towards 5G than hitherto expected, with

India another region where there could be a positive surprise (the government is pushing towards nationwide 5G coverage by 2020).

Meanwhile, Ericsson has not specifically stated its expectations for the timeline of 5G commercialization (we look for further color at the upcoming Capital Markets day on November 7-8, 2017), but has also argued China will likely be at the forefront. Nokia recently indicated its developers were experiencing a backlog in terms of workflows,

related, among other things, to 4.9G equipment (which can be rolled out near term with the option to upgrade later to 5G using software).

Exhibit 2: Industry players suggest accelerating timeline for 5G Summary of company commentary on 5G

Souce: Company data, compiled by Goldman Sachs Global Investment Research.

Company Recent commentary on 5G

Nokia(CEO, 2Q17 conf call)

• “Our original expectation was that 5G would only really take off in the 2020 to 2021 time period. That is now

changing…United States and China preparing to move earlier and that should drive others to respond”

• “I'm increasingly convinced that 5G trials will accelerate in 2018 and that in 2019 we can expect to start to see

meaningful deployments in the U.S. and China and potentially several other markets like Japan”.

Ericsson(CEO, 2Q17 conf call)

• “We have started to increase investments in R&D in Radio. And here, it is, again, to prepare ourselves both for a stronger

product offering in LTE as well as getting ready for the 5G launch, which will happen in the next few years.”

Huawei(President Wireless Network;

news article)

• “Huawei is confident that the partnership with Deutsche Telekom can fully prepare the commercial launch of 5GNR

services in Europe by 2020 thanks to 3GPP standardization efforts.”

Qualcomm (CEO, Citi Global Tech conf)

• “You'll start to see the first commercial devices in 2019.... already seeing people doing trials and early development in

the marketplace now. But the real standard compliant, new radio 5G will happen in the 2019 time frame.”

AT&T(CEO, GS Communacopia conf)

• “…as we move into a world of 5G…it's going to be operating at millimeter wave spectrum and we've secured a nice

footprint of this millimeter high frequency spectrum with high bandwidth…You'll see us begin building that out in the end

of 2018 time frame. And so this is how over the next few years, you get 1-gig speeds ubiquitously.”

Sprint (CEO, GS Communacopia conf)

• “…we are greatly positioned for 5G….We expect -- this is not 2019 or 2020. We expect in 2018 to deliver 1 gigabyte

LTE speeds to our customers in the places where we've been able to deploy massive MIMO.”

The timeline for 5G

commercial rollouts is

accelerating, with

meaningful rollouts in

US/China now planned

for 2019.



Progress on 5G trial activity by telcos globally can help standards crystallise: We believe ongoing solid progress on 5G trial activity in key regions can support the 3GPP body in setting up 5G standards in a shorter than originally expected timescale. For example, in the US, AT&T is working with Ericsson, Nokia, Intel and Samsung to roll out fixed wireless 5G (i.e. delivering speeds similar to those available using fiber access to homes via wireless networks instead of fixed line) in the US by late 2018 (fixed wireless is directed specifically at home usage rather than smartphone usage outside the home, as per Exhibit 3). Verizon also aims to roll out fixed wireless 5G to customers by next year. T-Mobile, with Nokia and Ericsson, aims to deploy its nationwide 5G network in the US by 2020 and plans to help in driving 3GPP certification for 5G in the 600 MHz spectrum (auctioned by the US government last year). Further, Korean telcos are moving forward on trial implementations: Korea Telecom (KT) plans to launch its pre-standard 5G trial network by 2018 Winter Olympics (with commercial rollouts beginning in 2019) and indicated this could help to solidify standards.

5G-enabled devices (i.e. smartphones, routers, etc.) could be available in 2019: The development of the ecosystem supportive of 5G (i.e. 5G capable devices and chipsets) is a key factor determining the timing of commercial rollouts. We believe the industry is progressing here, further supporting the case for initial ramp up of base station infrastructure in 2019. Qualcomm stated it expects commercial rollout of 5G-enabled devices, chipsets, modems and routers in 2019. Meanwhile, Samsung, at the Mobile World Congress (MWC) in Barcelona in 2017, announced 5G chipsets for smartphones and routers, based on pre-5G standards set by Korea Telecom and Verizon, and it expects these to be rolled out next year.

Qualcomm has stated

that 5G-enabled devices,

chipsets, modems and

routers will be

commercially available

in 2019.



Exhibit 3: 5G fixed wireless service expected to available in 2018 in the US 5G fixed wireless structure vs fixed-line broadband


How big will the 5G-related equipment spending cycle be? Although it may be too early for telcos to quantify precisely the magnitude of 5G investment ahead, recent investor debate has included discussion as to the risk that this could be of a lower size than was seen for 4G. By contrast, we believe recent vendor and telco commentary suggests

the cumulative capex envelope will be broadly similar on 5G to that seen on 4G, but with the mix of spend potentially shifting slightly towards fixed-line networking equipment. As such, our view of a meaningful cyclical spending rebound drives our above consensus estimates for Nokia, Ericsson and ZTE. While we see some risk that wireless spend could migrate to fixed-line equipment, in which case Nokia/ZTE would be better positioned than Ericsson (given their higher fixed exposure per our scenario analysis below – assumptions on p. 10), our positive view on the

latter reflects the fact it can still benefit greatly from self-help.

We believe a significant macro cell buildout will be needed for 5G. Some

investors have asked whether the fact that small/micro cell deployments could be prevalent in 5G networks (given the higher density configuration capabilities of using small cells – i.e. helping to penetrate buildings at such frequencies) means there will be fewer large-scale macro network/base station deployments. However, while 5G

networks will build upon the existing macro cell networks (i.e. large base station networks) seen in a 4G world, the need for broad macro cell coverage will be necessitated by the spectrum bands that will be used. Thus, in our view, there will still be a significant further 5G macro base station buildout (as small cells are likely to be insufficient). In addition, initial white papers from vendors and supply chain participants stress the 5G network will move from a model of population coverage (focusing most on

densely populated areas) to geographic coverage (covering the whole country) in order to deliver new use cases that move beyond the smartphone towards e.g. connected cars

5G Networks will include significant small cell buildouts but macro cells will remain important, given the spectrum bands that will be used, and geographic coverage needed to facilitate e.g. autonomous driving.



for autonomous driving, connected machines for advanced IoT, and connected healthcare (see Exhibit 4).

Exhibit 4: Latency and reliability requirements for various use cases 5G Use cases requiring low latency and/or high reliability

Souce: Nokia, Goldman Sachs Global Investment Research.

Specific vendor commentary from Nokia and ZTE suggests aggregate spending levels on 5G will be at least as large but mix of spend could shift –Nokia has stated (as per Exhibit 7) it sees 5G driving a ‘super-cycle’ in the industry with spend (on wireless plus fixed-line) at least as large as that seen on 4G, while ZTE stated the scale of spend on 5G will be greater than 4G. That said, we believe the move towards certain functions being centralized in the datacenter as part of 5G could have implications in terms of spending mix. In particular, Cloud RAN is envisaged by industry

players whereby certain functions of the base station, related to provisioning the user with services, could be located in a centralized datacenter location (see p. 10-12). As such, value could be transferred at the margin from the base station towards centralized hardware/software and there could be a need for incremental fronthaul/backhaul (and fixed-line equipment e.g. routers) to connect the cell site nodes to the datacentre. Thus vendors' ability to position themselves in one of these profit pools will be key (especially

if value is lost from the base station). Vendors with fixed plus wireless offerings (e.g. Nokia, Huawei) may benefit from this, with potential top-line risks for vendors such as Ericsson (however, we believe cost-cutting efforts can still lead to significant EPS growth even in such a scenario, as per Exhibit 16).

While the precise quantum and speed of telco 5G rollouts for these companies is not currently clear, our estimates for Nokia and Ericsson assume a broadly similar aggregate amount of spending on wireless and wireline (i.e. fixed) equipment for the 5G cycle vs what was seen during the prior (4G) cycle. While

we expect wireless spending to be cumulatively slightly lower (5% lower) for the 5G cycle than the equivalent 4G cycle, we expect this to be mainly compensated for by spending on fixed line (given the need for incremental fronthaul/backhaul – see above). As seen in Exhibit 5, we assume that while wireless spend in the 5G upgrade cycle will be 5% lower

(vs 4G), overall capex spend (i.e. wireless + fixed) for these companies will be broadly

Autonomous Vehicle

• Latency < 5-10ms

• Reliability up to BLER

Augmented Reality / Virtual Reality

• Latency < 5ms

• Reliability requirements less stringent

(but need to detect failures reliably)

Industry Control / Automation

• Latency partially < 0.5ms


Remote robotics / surgery

• Latency < 1ms (due to need for

haptic feedback in certain

implementations)


less stringent

ve

ry s

trin

ge

nt

less s

trin

ge

nt

very stringentLatency requirements

Reli

abil

ity R

equir

emen

ts



similar to that seen in 4G (i.e. 1% lower on our base case). Even in a scenario where wireless overall spend is 10% lower (and partially migrates to fixed), aggregate capex related to 5G would only fall 5% for these companies.

Exhibit 5: Aggregate 5G spend to remain similar for EU Tech equipment vendors to that seen on 4G… Aggregate wireless and wireline spend in 5G vs 4G; Black box (Base Case; GSe)

Exhibit 6: … with wireless capex to grow at a c.2% CAGR in 2018-21E (assuming capex peaking in 2024) Wireless spend CAGR 2018-21E; Black box (Base Case; GSe)



Our implicit assumptions as to the potential trajectory of wireless equipment spending are based on the peak capex year being one year later for 5G than was seen on 4G (capex peaked in 2014 in the 4G cycle i.e. four years after the initial rollouts began). We note that 4.5G/4.9G solutions have already started to

proliferate and these are upgradable to 5G, which could mean the take-up of standalone

5G equipment is somewhat more gradual/elongated (given potential to upgrade 4.5G/4.9G boxes). Thus, in our base case scenario, we expect wireless capex to peak in 2024 (five years after the expected initial rollouts in 2019). We expect the wireless equipment market to grow at a c.2% CAGR over 2018-21 on this basis (see Exhibit 6). For illustrative purposes, however, we show in Exhibit 6, that if capex peaks one year earlier (i.e. in 2023), this would accelerate capex growth in the earlier years of the cycle,

thereby raising the expected wireless market CAGR (2018-21E) to c.3%.

Ultimately, we also believe that the scale of spending will hinge on telcos' perceived ability to garner additional value from shifting to 5G. Key factors will

include telcos’ ability (1) to monetize new incremental TAMs in the form of subscription models related to connected machines/IoT, autonomous cars, etc., alongside (2) to boost existing telco sub ARPUs via provision of significantly enhanced service quality

(e.g. for ultra-fast video). We also think investors should monitor (3) the path to wider-scale VR/AR adoption, which has the potential to significantly drive data usage and thus large-scale capacity investment.

9% -10% -7% -5% 0% 3%

Low -5% -4% -3% 3% 5%

Moderate -3% -2% -1% 4% 6%

High 0% 0% 0% 5% 8%

Aggregate (Wireless and Wireline) spend related to 5G vs 4G

Cumulative Wireless spend in 5G cycle vs 4G cycle

Fix

edli

ne

in 5

G

-10% -7% -5% 0% 3%

2025 -1% 0% 1% 3% 3%

2024 0% 1% 2% 4% 4%

2023 1% 2% 3% 5% 6%

Wireless capex CAGR (2018-21E)

Pea

k

cap

ex




Exhibit 7: Recent commentary from Nokia, Ericsson and ZTE suggests 5G will drive solid aggregate spend Summary of recent commentary from major equipment vendors

Souce: Company data, Goldman Sachs Global Investment Research.

We reflect our view of broadly similar aggregate spend in the 5G upgrade cycle vs 4G, and a longer time to peak capex in our above-consensus 2018-21E estimates for Nokia, Ericsson and ZTE as follows:

Nokia: We estimate a 2.4% group revenue CAGR (2018E-21E), higher than Bloomberg

consensus at c.1.8%, factoring in our relatively positive base case for wireless market growth in the upcoming 5G rollout cycle. While we estimate only a c.2% CAGR in wireless revenues (including services) over 2018-21E, we see a higher CAGR in wireline equipment, which has higher gross margins than wireless (i.e. c.4%/c.3%/c.2% CAGR in fixed access/routing/optical revenues). This is congruent with our view that fixed-line network elements could be of greater relative importance in a 5G world (see p. 10).

o As such, our estimates are 9%/6% ahead of SME consensus for 2018/19 Networks adjusted EBIT, and we note Nokia’s fixed-line sales are growing faster than its wireless revenues.

o We estimate Nokia’s group sales can return to growth (following our estimate of

4.2% yoy declines in FY17) even in our most bearish wireless scenario (where spend would shift towards fixed line from wireless base stations), given c.35% exposure to wireline. Moreover, in such a scenario, EBIT would only fall c.3% in 2021 vs our base case (per Exhibit 12).

o That said, were 5G rollout momentum to be faster than we expect, which is a potential outcome given increasingly bullish commentary from equipment vendors

on 5G rollout timelines for China/US etc., this could materially benefit Nokia’s wireless revenues (>10% uplift to our 2021 wireless revenue estimate in our most

Company Recent commentary from major equipment vendors

Nokia

• Nokia stated at 2Q17 results that it believes 5G will drive the next growth cycle in the industry

• Nokia expects 5G networks to have a larger proportion of macro cells than small cells vs. earlier expectations, which we believe is supportive of

demand. Hence, management does not see 5G as a “hotspot” technology (i.e. it will not represent a smaller generation of technology than 4G from

a US$ spend point of view)

• One reason Nokia sees 5G as a significant driver is the potential scope for its customers (telcos) to open up various new revenue sources (i.e. IoT,

autonomous driving, etc.)

Ericsson

• In 5G Networks, Ericsson is confident that coverage will remain important so there will still be a need for radio units (depending on the success of

various 5G use cases e.g. IoT)

• We note that 5G may offer scope for some hardware reuse (of 4G/4.5G boxes) by telcos, which could reduce the capex required for coverage

rollouts (which are generally low margin) but with an associated need for software upgrades (which are much higher margin)

ZTE

• ZTE expects overall telco capex on 5G to be greater than that on 4G, although demand could be gradual to come on stream (we expect a

significant portion of telco spending to come from higher spectrum requirements for full network coverage e.g. for use cases such as IoT)

• ZTE management expects the number of 5G base stations doubling vs that seen in the 4G cycle, but it is expected that a significant portion of these

investments will comprise small cells (as opposed to macro cells)

• That said, it may be difficult to quantify the number of 5G base stations that will be deployed e.g. when 4G buildouts occurred in China, original

plans to deploy these eventually ended up in some cases being far larger than the originally announced number



bullish scenario). In our most bullish wireless scenario, group EPS would increase by c.5% in 2021 (vs our base case – per Exhibit 13).

Exhibit 8: Our base case for Nokia is a 2% wireless CAGR, as wireless spend will be only slightly lower on 5G vs 4G Nokia Wireless (incl. Services) revenues CAGR 2018-21E; Black box (Base Case; GSe)

Exhibit 9: … but see a 3% CAGR for Nokia fixed-line, which could be even higher if wireless spend migrates to fixed Nokia Wireline (incl. Routing, Optical and Fixed Access) revenues CAGR 2018-21E; Black box (Base Case; GSe)



Exhibit 10: We believe Nokia group sales can return to growth even in our most bearish wireless decline cases… Nokia Group revenues CAGR 2018-21E; Black box (Base Case; GSe)

Exhibit 11: … with limited sensitivity on top line to lower wireless given wireline (35% of sales) can compensate Nokia Group revenues 2021E uplift; Black box (Base Case; GSe)



Exhibit 12: Nokia group EBIT/EPS show limited sensitivity to wireless revenues migrating into fixed… Nokia Group EBIT 2021 uplift; Black box (Base Case; GSe)

Exhibit 13: … e.g. if centralizing functions in datacenters drives a capex shift from wireless to fronthaul/fixed-line Nokia Group EPS 2021E uplift; Black box (Base Case; GSe)



Ericsson: We model c.2% CAGR in 2018-21E in Ericsson’s Networks segment (excluding

Patents), which is congruent with our base case for growth of the wireless market during the initial years of the 5G rollout cycle (at 2% p.a. in 2018-21E). We believe this is a more positive assumption than is embedded in consensus estimates.

o As such, our estimates are 3%/5% ahead of SME consensus for Ericsson’s Networks segment revenues (2018/19E). We see Ericsson as a pure play on wireless (i.e. no fixed-line exposure) and it would therefore be relatively more sensitive (vs Nokia) to any scenario with lower wireless spend (e.g. due to capex migrating towards fixed-line parts of the network). However, as detailed in our report Restructuring potential with market improvement; reiterate Buy, July 10, 2017, we argue company-specific self-

help measures can significantly help margins. As such, we are >15% ahead of SME consensus on 2018/19E group adj. EBIT, and still see >20% EPS CAGR in 2018-21E in more bearish scenarios.

-10% -7% -5% 0% 3%

2025 -1% 0% 1% 3% 3%

2024 0% 1% 2% 4% 4%

2023 1% 2% 3% 5% 6%

Nokia Wireless revenues 2018-21E CAGR


Pe

ak

cap

ex

3% -10% -7% -5% 0% 3%

Low 4% 3% 2% 2% 2%

Moderate 5% 4% 3% 3% 3%

High 6% 5% 4% 4% 4%

Nokia Wireline revenues 2018-21E CAGR


Fix

edlin

e in

5G

0% -10% -7% -5% 0% 3%

Low 2% 2% 2% 3% 4%

Moderate 2% 2% 2% 3% 4%

High 2% 2% 3% 4% 4%

Nokia Group revenues 2018-21E CAGR


Fix

edlin

e in

5G

0% -10% -7% -5% 0% 3%

Low -2% -1% -1% 2% 3%

Moderate -1% -1% 0% 3% 4%

High 0% 0% 0% 3% 5%

Nokia Group revenues 2021E uplift


Fix

edlin

e in

5G

0% -10% -7% -5% 0% 3%

Low -3% -2% -1% 2% 3%

Moderate -1% -1% 0% 3% 4%

High 0% 0% 0% 3% 4%

Nokia group EBIT 2021E uplift


Fix

edli

ne

in 5

G

0% -10% -7% -5% 0% 3%

Low -3% -2% -1% 2% 4%

Moderate -2% -1% 0% 3% 4%

High 0% 0% 0% 3% 5%

Nokia group EPS 2021E uplift


Fix

edli

ne

in 5

G



o Were 5G rollouts to be faster than expected, this could lead to EPS CAGR (2018-21E) increasing to c.27% (most bullish scenario) vs our base case of c.23%.

o That said, as highlighted in the scenario tables below, Ericsson would see lower

top-line growth (vs Nokia) in a weaker wireless spend environment (vs our base case) given it has no fixed-line exposure to compensate for wireless declines. However, we note Ericsson’s 2018-21E EPS CAGR would still be >20% in all our scenarios given self-help actions.

Exhibit 14: Ericsson would see lower topline growth (vs Nokia) in a bearish wireless spend scenario… Ericsson Group revenues CAGR 2018-21E; Black box (Base Case; GSe)

Exhibit 15: …given it has no fixed-line exposure to compensate for declines in the wireless market Ericsson Group revenues 2021E uplift; Black box (Base Case; GSe)



Exhibit 16: …but the EPS CAGR would still be >20% given self-help actions Ericsson Group EPS CAGR 2018-21E; Black box (Base Case; GSe)


ZTE: We are 2%/2% ahead of Reuters consensus on 2018/18 group revenues given upside

skews due to scope for market share gains in China, as well as the potential for domestic

market spending (70% of revenues) to be more aggressive in light of investments by tech players into the telco space.

o In our sensitivity analysis, we focus on the impact of ZTE’s wireless and group topline in 2020E under different market scenarios. On a group level, we believe there is further upside to our 2020 group revenue estimates under scenarios where the cumulative wireless spend in China in the 5G cycle is equal to or more than that

in the 4G cycle.

o In scenarios where the cumulative wireless spend in the 5G cycle is lower than that seen during the 4G cycle, we see ZTE’s wireline exposure offsetting the majority of these declines. As in the scenario tables below, we see scope for a 2% uplift to our ZTE group revenue estimates in our most bullish scenario.

Exhibit 17: ZTE could see material upside to revenues if the 5G cycle progresses faster than expected (vs 4G)… ZTE Wireless revenues uplift; Black box (Base Case; GSe)

Exhibit 18: …with the company’s Wireline exposure (c.25%) helping offset wireless if spend migrates to fixedZTE Group revenues uplift; Black box (Base Case; GSe)



-10% -7% -5% 0% 3%

2025 0% 1% 1% 3% 3%

2024 1% 1% 2% 3% 4%

2023 2% 2% 3% 4% 5%

Ericsson Group revenues 2018-21 CAGR


Pe

ak

cap

ex

-10% -7% -5% 0% 3%

2025 -6% -4% -2% 1% 3%

2024 -4% -2% 0% 4% 5%

2023 -1% 1% 3% 6% 8%

Ericsson Group revenues 2021 uplift


Pe

ak

cap

ex

-10% -7% -5% 0% 3%

2025 21% 22% 22% 24% 25%

2024 22% 23% 23% 25% 26%

2023 23% 24% 24% 26% 27%

Ericsson Group EPS 2018-21 CAGR


Pea

k

cap

ex

-10% -7% -5% 0% 3%

2025 -7% -4% -2% 3% 6%

2024 -5% -2% 0% 5% 8%

2023 -2% 1% 3% 8% 11%

ZTE Wireless revenues 2020E uplift


Pe

ak

cap

ex

0% -10% -7% -5% 0% 3%

Low -2% -1% 0% 2% 3%

Moderate -1% 0% 0% 2% 3%

High 0% 0% 0% 2% 3%

ZTE Group revenues 2020E uplift


Fix

edli

ne

in 5

G



Assumptions: Our key assumptions for the scenario analysis (for all companies) include:

o Our assumed range of -10% to +3% for cumulative wireless spend in the 5G

cycle vs levels seen on 4G captures: (1) Bearish scenario (i.e. -7% and -10%)

wherein we expect a higher mix of software in 5G spend (vs base stations and

macro cells) and more value migrating from wireless base stations to the

datacenter (and associated fixed-line equipment); (2) Base case scenario (i.e. -

5%) wherein we believe that the cumulative wireless spend related to 5G will

be 5% lower vs than in 4G but overall spend (i.e. wireless and wireline) will

remain similar and (3) Bullish scenario (i.e. flat to +3% cumulative wireless

spend) wherein we expect 5G wireless spend to be equal/greater to that seen

on 4G (as expected by ZTE).

o As per our analysis, in a scenario where cumulative wireless spend in the 5G

cycle (i.e. 2019-27E) will be lower than was seen on 4G (i.e. below 0% delta),

we add c.50%/75%/100% (i.e. low/moderate/high fixed-line usage in 5G) of the

delta in the form of incremental wireline spend during the same period (given

our view that part of the wireless capex decline in 5G will be offset by extra

fixed-line spend for fronthaul/backhaul, as explained earlier).

What will the structure of the 5G network look like? Initially identified in our report ‘5G: How 100x faster wireless can shape the future’, April 18, 2016,

the move to 5G will see networks become more “cloud like” vs. the 4G world. In particular, there is scope to separate the radio and the baseband, and also for non-real-time functionality (e.g. subscriber management) to run in cloud RAN (radio access network), i.e. a virtualized data center. GS believes that although wireless base stations could at the margin lose some value there can

be opportunities in fixed and cloud for those vendors that offer a broad/converged offering. These developments will likely bring certain cost and efficiency advantages, as we discuss in more detail below.

Splitting the radio and the baseband: 5G will offer scope to separate the radio (for

transmitting/receiving) and the baseband (for signal processing), and connect the two with fiber, whereas hitherto these two tended to be within the same radio unit. By

separating these two elements, it is possible to centralize the signal processing. The benefit of this is that one is able to have one location where all the intelligence on the network is located, and processing can be directed to the parts of the network where it is most needed, thus allowing more efficient utilization of equipment. A further advantage is this allows for better spectral efficiency as different cells will be able to talk to each other (such a plug and play self-organizing network is sometimes referred to as “tight

coordination”). One derivative effect of this will be an increased requirement for dark fiber due to the fact that latency between the radio and the baseband will need to be very low. Indeed, Verizon’s pending deal to acquire XO Communications and plans to expand its FiOS footprint in Boston are likely precursors of this architectural shift.

Cloud RAN: Another key architectural change will be the advent of Cloud RAN (C-

RAN), a data center aggregating traffic from the radios, where non-real-time functionality

will be virtualized and run on commercial off the shelf servers. Clearly, there are certain functions that cannot be migrated to the cloud easily in this way. For example, radio scheduling (which involves knowing exactly when to burst energy in different directions in a radio network) should take place close to the radio and needs to be purpose built, to avoid energy efficiency issues. In contrast, certain functions do not need to take place in real time and can work on a slower cadence. Examples of functions that can be

virtualized and centralized in this way include mobility management, subscriber

5G could see higher spend on fixed line equipment as some processing is centralized in datacenters which must be connected to base stations.



management, and other layer 4-7 network services. As such, some of the radio functionality will be software defined and migrate to a server. By contrast, those functions that are real-time critical will tend to take place on purpose built high

performance hardware.

BBU to be divided into DU + CU: 5G C-RAN BBU (base band unit) will be divided

into the functional entities of DU (distributed unit) and CU (centralized unit). Accordingly, the fronthaul domain will include two levels:

1. Domain I between RRU and DU

2. Domain II between DU and CU

Exhibit 19: Two-level fronthaul architecture for 5G Fronthaul network architecture


o Distributed Unit for real-time radio functions

Real-time functions would require a large amount of specialized computing. According to Heavy Reading, these functions are more efficiently processed on custom silicon than

on general purpose processors today. Since radio units are relatively high volume and performance oriented, power efficiency is critical. As a result, specialized hardware will typically be deployed at the distributed site.

o Centralized Unit for non-real-time functions

For functions that do not need ultra-low latency and can tolerate dozens of milliseconds of delay, they are ideal for centralized deployment on general-purpose processors.

o DU and CU will be connected by optical fiber

As the BBU is sliced into two parts and deployed separately, they will need to be connected by optical fiber at very high speed to ensure that they talk to each other and data is transmitted back and forth in an ultra-fast manner to facilitate the high speed and low latency of 5G network.



Exhibit 20: 5G networks will drive upgrades in both mobile fronthaul and backhaul wired networks 5G Network architecture


How will China spending levels trend in coming years? The total amount of capex to be spent in China on 5G and the intensity of spending (i.e. time period) is a strong focus for investors, as it will not only have implications for telcos’ cash flow and potential ARPU growth, but also on the rate of revenue growth and gross margin trends for telecom equipment vendors, optical fiber/cable suppliers, and optical component suppliers. We expect the spending timeframe for the next wireless upcycle in China to be more spread out than

was seen on 4G, but with investment by non-traditional players in the space suggesting potential to drive an earlier start to adoption of more advanced technology. We analyse a range of scenarios around growth rates of China wireless equipment spending, but also market share assumptions (given ZTE’s recently expressed aspirations to increase its domestic share). We conclude that the impact on Ericsson/Nokia is relatively small, but with potential upside skews for ZTE.

For the last round of 4G investment, China telcos in aggregate grew wireless capex for three consecutive years: For 5G, we expect the total capex spend to be

similar to that seen on 4G, and potentially higher due to the fact that 5G

deployment will be on a higher spectrum band than 4G, translating to more base

stations needed for the same coverage; and full coverage is needed to enable

mission critical machine-to-machine communications. Although the separation of

BBU into DU and CU may lower the equipment cost to some extent for the general

purpose part, extra optical fiber and cable is needed to connect DU and CU, which

may offset the cost saving from Cloud RAN.

Following the introduction of new sources of capital in the China telco space there could be an evolution in the approach of telcos to technology deployment: We believe a more rapid move towards 5G than hitherto could be

possible given such new sources of capital. As noted above we argue this could

help catalyse the broader global 5G ecosystem development. While some players

aim for 5G initial trial and commercial deployment in 2020/21, others target these

in 2019/20 (we believe that given competitive dynamics may be different now,

those that have invested relatively lower amounts in their networks may seek a

faster ramp in 5G spending whereas those who are further ahead may adopt a

more measured ramp cycle on 5G than was seen on 4G).

The investment by

Tencent, Alibaba and

Baidu into the Chinese

telco space could offer

scope to catalyse faster

5G take-up and boost

capex.



Exhibit 21: Wireless spend in China troughed in 2016… (in RMB billions)

Souce: Company data.

Exhibit 22: …and we see the declines in China wireless capex improving in 2017/18E GS wireless capex forecasts by regions (in USD mn)


Our China sensitivity analysis suggests a relatively small impact for Nokia/Ericsson’s 2018 EPS but potential upside skews for ZTE: We model 8%

declines in China Wireless spending in 2018. However, we present a scenario analysis below to understand the impact on our covered companies of various market spending/market share outcomes. Our conclusion is that there is a relatively small impact for Nokia/Ericsson’s 2018 EPS (both of which we estimate have a 10% share in

the Chinese market, constituting c.10% of revenues vs ZTE with 30% market share and 70% revenue exposure to China). That said, outsized market share gains/a better wireless spending environment could offer significant potential upside skews for ZTE. We reiterate our Buy rating on ZTE (H shares), given our view that there could be scope for better-than-expected China market evolution in the coming years, driven by tech players investing in the space (and targeted market share gains, as articulated by the

company).

5%

‐6%

4%

‐13%‐12%

‐4%

4%

‐2%

2%

‐15%

‐8%

‐3%

‐20%

‐15%

‐10%

‐5%

0%

5%

10%

US Europe Japan India China Total

2017E 2018E



Exhibit 23: ZTE could see a 5% uplift to our 2018 revenue estimates in the most bullish scenario… Scenario table for ZTE 2018E sales, sensitized by wireless market share & China wireless capex


We analyze in parallel different aggregate wireless spending scenarios in China, pegged against various market share assumptions for

ZTE/Nokia/Ericsson (noting ZTE has stated that it targets growing its domestic China

market share from 30% to 40% over the next three years).

Market share and wireless spending in China inputs: In our base case, we model

current market share for ZTE/Nokia/Ericsson in China to be c.30%/c.10%/c.10%. We assume a range of scenarios for market share which, alongside the size of the China wireless market (2018 wireless spending), drive the topline estimates for our scenario analyses for each of the three relevant companies. Our base case assumption for 2018E

China wireless capex is for 8% yoy declines, i.e. c.RMB131 bn (i.e. c.€17/SEK131) of capex. We keep gross margin, opex/sales, EBIT margin and tax rate (etc.) constant in each scenario for ZTE, Nokia and Ericsson (corresponding to that in our published estimates).

ZTE could see upside potential if there is better-than-expected wireless capex spending and/or incrementally higher China market share by 2018. We note

that c.23% of ZTE’s total revenues are from wireless equipment in China, where it has strong gross margins of 44% (relative to Nokia/Ericsson). Our sensitivity analysis (Exhibit 23-25) suggests scope for a 5%/20% uplift to 2018E revenues/EBIT in the most bullish scenario (i.e. -4% Chinese wireless spend, 35% market share).

Exhibit 24: We see potential for c.20% uplift to our 2018E EBIT estimate… Scenario table for ZTE 2018E Group EBIT

Exhibit 25: …and a c.19% uplift to 2018E EPS forecast for ZTE in our most bullish scenario Scenario table for ZTE 2018E Group EPS



Nokia and Ericsson are relatively less impacted by China wireless market dynamics and market share fluctuations. We assume both Nokia and Ericsson to

each have c.10% market share in China (in our base case model), with lower gross margins than their respective group averages given competitive dynamics and telco spending behavior. Further, only c.10% of their revenues is from China Wireless, as both

players are more exposed to the European and US Wireless markets.

2018E Sales -13% -11% -8% -6% -4%

25% -5% -4% -4% -3% -3%28% -3% -2% -2% -1% -1%30% -1% -1% 0% 1% 1%33% 1% 1% 2% 3% 3%35% 2% 3% 4% 4% 5%

ZTE China Wireless Market

share

China telco wireless capex yoy growth (2018E)

2018E EBIT-13% -11% -8% -6% -4%

25% -18% -16% -14% -13% -11%28% -11% -9% -7% -5% -3%30% -4% -2% 0% 2% 4%33% 3% 5% 7% 10% 12%35% 9% 12% 14% 17% 20%



share

2018E EPS-13% -11% -8% -6% -4%

25% -17% -16% -14% -12% -10%28% -11% -9% -7% -5% -3%30% -4% -2% 0% 2% 4%33% 2% 5% 7% 9% 11%35% 9% 11% 14% 16% 19%


share




Exhibit 26: Our analysis suggests minimal impact from …Scenario table for Nokia 2018E Group Sales

Exhibit 27: …China for both Nokia and Ericsson Scenario table for Ericson 2018E Group Sales



What is the state of industry pricing and will the new market structure have an impact? A final key topic is the extent to which the consolidating wireless equipment market structure can lead to pricing dynamics improving. Specifically, the top three players now comprise c.90% of market share (Exhibit 28). Our view is that while pricing remains tough, as it has done for the last 1-2 years, consolidation can mean pricing pressure may not accelerate. That said, we do not assume that industry pricing will markedly improve, given ZTE now appears to be winning market

share in certain markets. As such, we believe gross margin improvement for infrastructure vendors will likely result from company specific factors.

The top three players in wireless equipment, Ericsson, Nokia and Huawei now share c.90% of the market: As per our view, while pricing remains tough, as it has

done for the last 1-2 years, consolidation can lead to the continuation of the dynamic whereby pricing pressure may not accelerate. Moreover, we note that Huawei, which

was historically very price aggressive, does not appear to be driving down prices currently due, in our view, to the fact that its installed base is by now so large that discounting would tend to disrupt its installed base and be dilutive to profits.

We do not assume that industry pricing will markedly improve, however, noting that ZTE has been working on a turnaround (some of its management

have recently changed and restrictions on its purchasing of US semis are now removed).

Thus, we argue ZTE has made itself available as a credible fourth supplier to telcos (and in some cases it has been winning market share in certain markets). As such we believe gross margin improvement (or in the case of Nokia, broad stability) for infrastructure vendors will likely result from company specific factors. In the case of Nokia, this is more likely to be the result of continued progress on cost synergies and highly profitable patents revenue growth, whereas Ericsson's profitability can improve if it exits low

margin contracts and activities (albeit any efforts to centralize/drive more discipline around its pricing could potentially improve the broader market profitability environment, this is not our base case).

2018E Sales-13% -11% -8% -6% -4%

8% -1% -1% -1% -1% -1%9% 0% -1% -1% 0% 0%

10% 0% 0% 0% 0% 0%11% 1% 1% 1% 1% 2%12% 2% 1% 1% 2% 2%


Nokia China Wireless Market

share

2018E Sales-13% -11% -8% -6% -4%

8% -2% -1% -1% -1% -1%9% -1% -1% -1% -1% 0%10% 0% 0% 0% 0% 0%11% 0% 0% 1% 1% 1%12% 1% 1% 1% 2% 2%


ERIC China Wireless Market

share

10 years ago, the top 3 players shared c.70% of market value, but now this group accounts for c90%.



Exhibit 28: Wireless market structure improvement given top 3 players have >90% market share Global wireless market share (1996-2015)


ZTE’s share gain momentum in both local and in overseas markets may lead to some pockets of pricing pressure but we do not see this getting significantly worse overall: We note that ZTE has been gaining some market share

(per recent management commentary) in its local Chinese and overseas end markets

and hence we believe it could use price to some degree to do this, supporting our view that while pricing may not get significantly worse this could prevent a significant improvement in pricing near term. According to IHS, ZTE expanded its market share to 15% (vs 12% prior) in 1H17 in the Wireless equipment market. In China, ZTE has stated that it targets to grow its wireless market share from 30% to 40% over the next three years. Overseas, ZTE aims to double its shares from 6% to 12% in the next 2-3 years.

For example, it won a contract in December 2016 worth €1bn for CK Hutchison / Vimpelcom in Italy (original vendor was ERIC). That said given ZTE’s margin structure some contracts may be accretive to it which are less attractive to Ericsson given higher profitability aspirations.

Industry consolidation may lead to telcos wanting to keep a diversified pool of suppliers, with ZTE a potential beneficiary pre 5G ramp: Overall, we believe

that as the industry becomes more consolidated (noting the acquisition of Alcatel-Lucent by Nokia in 2015), telcos may want to keep their supplier pool diversified and pricing under control. Hence, we see ZTE as a natural addition to the pool of equipment vendors before telcos begin large scale 5G investments. In that context, GS believes ZTE could have some success breaking into Tier 1/2 telcos and hence keep pricing under control from the point of view of carriers).

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

Ericsson Nokia-Siemens Huawei Alcatel-Lucent ZTE Other

Top 3 players have >90% market share following Nokia/Alcatel deal



Nokia OYJ (NOKIA.HE): Well positioned for converged 5G Networks; routing recovery; Capital return potential; Buy

What’s changed We reiterate our Buy rating on Nokia. We note its potentially advantageous position midterm due to a converged product portfolio (including routing, optical and fixed access) that could gain increasing importance in 5G, where

some spend may shift to fixed line (from wireless). Our positive view also derives from our expectation of further progress on cost execution/synergies (from ALU acquisition) and margin expansion. Despite challenging conditions in the Wireless networks market, we believe Patents revenues can continue to expand (noting recent deals with Apple, Xiaomi and LG), driven by further progress at Chinese handset makers. A recovery in Routing/Webscale also

helps drive our above-consensus estimates. Our latest Nokia estimates reflect our view that revenues in 3Q17 will be broadly in line with normal QQ seasonality in Networks (+1% qoq growth vs. normal 1Q growth rate of 1%-2%). In particular, we assume -2% QQ decline in wireless revenues (includes Services), implying c10% decline yoy, noting that the company had previously flagged certain project delays in 2H (negatively impacting topline).

However we model c2% yoy growth in IP routing, noting easy comps vs last year (-11% yoy in 3Q16). As such, our 3Q ultra-broadband revenues estimate (wireless hardware and fixed) of c€2.2bn is broadly in line with company-compiled consensus at c€2.3bn.

Implications We keep our estimates unchanged. Our 2017/18E estimates are c1%/9% ahead of SME consensus on Adj. Group EBIT. Overall, while our FY17

Networks growth est. of -5% yoy growth is towards the lower-end of NOK’s guidance at -3% to -5%, we continue to see Nokia’s EBIT benefitting from ongoing cost savings. We believe investors will look for further commentary on potential for the wireless infrastructure market to stabilize into 2018, with the 5G ramp ahead, and see incremental cash returns as a potential catalyst (given €1.7bn cash inflow from the positive Apple patents negotiation result).

We model c€1.6bn cumulative share buybacks in 2018-19E. We look for further colour on capital returns and the LG patents deal at Nokia’s 3Q results.

Valuation We remain Buy with 12m PTs of €5.9/$7.1 ADR based on a SOTP with Networks at 7.5x (unchanged) 2018E EV/EBITDA and a DCF for Patents.

Key risks Key downside risks to our view and price target include worse-than-expected

wireless capex, integration risk and pricing pressure.

Growth

Returns *

Multiple

Volatility Volatility

Multiple

Returns *

Growth

Investment Profile

Low High

Percentile 20th 40th 60th 80th 100th

* Returns = Return on Capital For a complete description of the investment

profile measures please refer to the

disclosure section of this document.

Nokia (NOKIA.HE)

Europe Technology Peer Group Average

Key data Current

Price (€) 5.00

12 month price target (€) 5.90

Upside/(downside) (%) 18

Market cap (€ mn) 18,544.9

Enterprise value (€ mn) 24,990.1

12/16 12/17E 12/18E 12/19E

Revenue (€ mn) New 23,944.0 22,944.3 23,237.0 23,736.4

Revenue revision (%) 0.0 0.0 0.0 0.0

EBIT (€ mn) New 2,171.7 2,388.0 3,098.1 3,271.0

EBIT revision (%) 0.0 0.0 0.0 0.0

EPS (€) New 0.22 0.27 0.38 0.41

EPS (€) Old 0.22 0.27 0.38 0.41

EV/EBITDA (X) 7.5 8.5 7.0 6.3

P/E (X) 22.7 18.3 13.1 12.1

Dividend yield (%) 5.2 3.0 4.2 4.5

FCF yield (%) (6.6) 0.9 4.8 12.5

CROCI (%) (5.3) 5.3 7.0 12.9

400

420

440

460

480

500

520

3.0

3.5

4.0

4.5

5.0

5.5

6.0

Oct-16 Jan-17 Apr-17 Aug-17

Price performance chart

Nokia (L) FTSE World Europe (EUR) (R)

Share price performance (%) 3 month 6 month 12 month

Absolute (9.3) 1.0 10.9

Rel. to FTSE World Europe (EUR) (10.1) (1.5) (4.2)

Source: Company data, Goldman Sachs Research estimates, FactSet. Price as of 10/17/2017 close.



Ericsson (ERICb.ST): Pure play on 5G cycle; cost-cutting plan remains in focus; Buy

What’s changed We remain Buy on Ericsson noting its status as a pure play on the upcoming 5G cycle that could start faster than we had previously expected. As such, we see its incremental investments in LTE and 5G as beneficial but see

broader strategic moves to streamline the business as key to unlocking value. We remain positive that Ericsson's operational efforts can drive margin improvement and look for further colour at its 3Q results (20 October) and the November CMD. A recent news article from Reuters highlighted that Ericsson is currently looking to merge its Spanish network rollout and maintenance arm, Abentel (which it bought in 2016). We view this news flow

as positive given that Network rollout and Managed Services (i.e. related to Network maintenance) are both lower margin businesses for Ericsson (GSe single digit gross margin segments), implying scope for margin improvement at a group level. In addition, the article also states that Ericsson intends to sell/merge its loss making (or/and low margin) businesses to improve its group profitability level, which we have argued will be a key margin driver for

the company mid-term.

Implications We keep our estimates unchanged. Our FY18/19 Group Adj. EBIT estimates are >10%/6% ahead of SME consensus. As per our Buy thesis, we argue Ericsson can exit part of its Media business (SKr c2.4bn EBIT losses ex patents in 2016), which could improve group 2018 EBIT. Furthermore, we expect Ericsson to streamline its IT/Cloud segment (SKr c4.6bn EBIT losses

ex patents in 2016) by cutting/reducing exposure to loss making businesses like IT Cloud Hardware. In Networks, while ERIC plans to temporarily ramp-up R&D, this is to improve (or/and maintain) its product competitiveness, a positive in the light of an accelerating 5G deployment timeline. We believe investors will look for further colour at its 3Q results on initiatives/efforts to improve gross margins e.g. via pricing discipline. Our FY17 Networks yoy

growth estimate of -8% is line with the company’s guidance (high single digit declines). We model c5% qoq group top-line declines in 3Q vs. normal seasonality of 0%-1% declines primarily due to FX headwind (weaker SEK vs $). We note that SME consensus EBIT estimates for 2017/18 have been revised down 40%/23% since last results.

Valuation We remain Buy with 12m price targets of SKr55/ADR $6.90 based on an

8.5x 2018E EV/EBITDA multiple (unchanged).

Key risks Key downside risks to our view and price target include worse cost control, wireless market dynamics and market share loss.

Note Alexander Duval, Hameed Awan and Piyush Maheshwari are responsible for the content on

Nokia and Ericsson in this note.

Growth

Returns *

Multiple


Multiple

Returns *

Growth

Investment Profile

Low High





Ericsson (ERICb.ST)

Europe Technology Peer Group Average

Key data Current

Price (Skr) 48.42

12 month price target (Skr) 55.00

Upside/(downside) (%) 14

Market cap (Skr mn) 159,253.4

Enterprise value (Skr mn) 168,538.6

12/16 12/17E 12/18E 12/19E

Revenue (Skr mn) New 222,608.0 203,610.3 196,512.9 199,909.1


EBIT (Skr mn) New 13,866.0 4,470.0 14,732.5 19,210.0

EBIT revision (%) 0.0 0.0 0.0 0.0

EPS (Skr) New 1.27 (3.51) 2.27 3.63

EPS (Skr) Old 1.27 (3.51) 2.27 3.63

EV/EBITDA (X) 8.8 9.9 6.3 5.5

P/E (X) 50.6 NM 21.3 13.3

Dividend yield (%) 5.9 2.1 3.1 4.1

FCF yield (%) 1.7 3.9 6.3 6.3

CROCI (%) 3.9 (1.2) 5.5 6.3

440

470

500

530

560

590

40

45

50

55

60

65



Ericsson (L) FTSE World Europe (GBP) (R)


Absolute (20.6) (16.7) 1.0

Rel. to FTSE World Europe (GBP) (22.6) (23.0) (11.6)




ZTE Corp (0763.HK): Mk share gain, cost control, 5G all in favor; Buy

What’s changed We remain Buy on ZTE Corp (H shares) for its turnaround not only on its top line, but also on its bottom line. ZTE has taken a series of actions that we believe are steering it in the right direction: (1) Going on the offensive in overseas markets to gain share (e.g. EUR1bn contract win at Hutchison & Vimpelcom in Italy). As the communications technology equipment market

gets more and more consolidated, telcos are looking for diversity in its vendor structure. ZTE, as the fourth largest vendor, is in a good position to step up and take the benefits through increased market share. In fact, ZTE’s 1H17 results have proven that it is indeed gaining market share both on wireless equipment and on optical equipment. In 1H17, ZTE’s carriers’ network revenue was +13% yoy, vs. Huawei +1%, Nokia -2%, and Ericsson -

10%. According to IHS, ZTE grew its optical revenue by 35% yoy in 1H17, the fastest among the top 10 network equipment vendors, and hence expanded its market share to 15% from 12% in 1H16. We note that given differing margin aspirations of ZTE vs. other vendors it may be able to accept deals that e.g. Ericsson would now not compete for in light of new measures to drive towards new EBIT margin targets. (2) effective cost control

measures on SG&A with new strategy for its consumer segment. In 1Q 2017, ZTE cut its labor force by c.3,000, which it noted would translate to RMB700-800mn savings on an annual basis. In addition, we expect ZTE to record lower advertising expenses, mainly in China, as it shifts its smartphone strategy away from open channels and back to telco channels; (3) spinning off Nubia and refocusing on telco channels to improve

profitability of its smartphone business. After the successful spin-off of Nubia (which is currently pending shareholder approval), the company noted there will be another c.3,000 headcount taken off of ZTE’s payroll. We have not incorporated the spin-off of Nubia in our estimates given the outstanding contingencies.

Implications We keep our estimates unchanged. Our 2017E/18E/19E earnings estimates

are 3%/8%/13% above Bloomberg consensus.

Valuation We remain Buy on H share and Neutral on A share with an unchanged 12m price target of HK$23/RMB23 based on an unchanged 15x/17x 2018E P/E multiple.

Key risks Key downside risks to our view and price target include continued telco capex cut, further margin pressure, higher SG&A spending.

Note Tina Hou and Bowen Bao are responsible for the content on ZTE Corp in this note.

Gao Hua Securities acknowledges the role of Alexander Duval, Hameed Awan, Piyush

Maheshwari and Bowen Bao of Goldman Sachs in the preparation of this product.

Growth

Returns *

Multiple


Multiple

Returns *

Growth

Investment Profile

Low High





ZTE Corp. (0763.HK)

Greater China Telecoms and Technology Peer Group Average

Key data Current

Price (HK$) 29.45

12 month price target (HK$) 23.00

Upside/(downside) (%) (22)

Market cap (HK$ mn) 122,240.8

Enterprise value (Rmb mn) 105,336.6

12/16 12/17E 12/18E 12/19E

Revenue (Rmb mn) New 101,233.2 113,688.0 121,114.1 133,098.6


EBIT (Rmb mn) New 2,556.2 5,545.0 6,231.1 7,396.5

EBIT revision (%) 0.0 0.0 0.0 0.0

EPS (Rmb) New (0.57) 1.12 1.34 1.61

EPS (Rmb) Old (0.57) 1.12 1.34 1.61

EV/EBITDA (X) 9.2 13.0 11.2 9.4

P/E (X) NM 22.3 18.7 15.5

Dividend yield (%) 2.5 1.4 1.7 2.1

FCF yield (%) 2.2 2.8 3.5 4.0

CROCI (%) (5.7) 16.9 18.0 19.2

9,000

10,000

11,000

12,000

13,000

14,000

10

15

20

25

30

35



ZTE Corp. (L) Hang Seng China Ent. (R)


Absolute 57.8 113.4 176.3

Rel. to Hang Seng China Ent. 47.1 88.2 127.9




信息披露附录申明本人，侯雪婷，在此申明，本报告所表述的所有观点准确反映了本人对上述公司或其证券的个人看法。此外，本人薪金的任何部分不曾与，不与，也将不会与

本报告中的具体推荐意见或观点直接或间接相关。

投资摘要投资摘要部分通过将一只股票的主要指标与其行业和市场相比较来评价该股的投资环境。所描述的四个主要指标包括增长、回报、估值倍数和波动性。增长、

回报和估值倍数都是运用数种方法综合计算而成，以确定该股在地区研究行业内所处的百分位排名。

每项指标的准确计算方式可能随着财务年度、行业和所属地区的不同而有所变化，但标准方法如下：

增长是下一年预测与当前年度预测的综合比较，如每股盈利、EBITDA 和收入等。回报是各项资本回报指标一年预测的加总，如 CROCI、平均运用资本回报率

和净资产回报率。估值倍数根据一年预期估值比率综合计算，如市盈率、股息收益率、EV/FCF、EV/EBITDA、EV/DACF、市净率。波动性根据 12个月的历史

波动性计算并经股息调整。

并购评分在我们的全球覆盖范围中，我们使用并购框架来分析股票，综合考虑定性和定量因素（各行业和地区可能会有所不同）以计入某些公司被收购的可能性。然后

我们按照从 1到 3对公司进行并购评分，其中 1分代表公司成为并购标的的概率较高(30%-50%)，2分代表概率为中等(15%-30%)，3分代表概率较低(0%-15%)。对于评分为 1或 2的公司，我们按照研究部统一标准将并购因素体现在我们的目标价格当中。并购评分为 3被认为意义不大，因此不予体现在我们的目

标价格当中，分析师在研究报告中可以予以讨论或不予讨论。

Quantum Quantum 是提供具体财务报表数据历史、预测和比率的高盛专有数据库，它可以用于对单一公司的深入分析，或在不同行业和市场的公司之间进行比较。

GS SUSTAIN GS SUSTAIN是一项侧重于通过发现优质行业领先企业而实现长期超额收益的全球投资策略。GS SUSTAIN 50关注名单列出了我们认为凭借出色的资本回

报、具有可持续性的竞争优势和对 ESG（环境、社会和企业治理）风险的有效管理而有望在长期内相对于全球同业表现出色的行业领军企业。候选企业主要基

于对企业在这三方面表现的综合量化分析筛选而出。

相关的股票研究范围

侯雪婷：A股电信与科技、大中华电信与科技。

A股电信与科技：光迅科技、中联通(A)、大华股份、烽火通信、歌尔股份、海康威视、立讯精密、欧菲光、网宿科技、中兴通讯(A)。

大中华电信与科技：中国通信服务、中国移动、中国移动(ADR)、中电信、中电信(ADR)、中联通(H)、中联通(ADS)、华虹半导体、联发科、谱瑞科技、矽力

杰、中芯国际、中芯国际(ADR)、舜宇光学、台积电、台积电 (ADR)、联电、联电 (ADR)、中兴通讯(H)。

信息披露

与公司有关的法定披露

以下信息披露了高盛高华证券有限责任公司（“高盛高华”）与北京高华证券有限责任公司（“高华证券”）投资研究部所研究的并在本研究报告中提及的公司之间

的关系。

高盛高华在今后 3个月中预计将从下述公司获得或寻求获得投资银行服务报酬：中兴通讯(A) (Rmb30.63)、中兴通讯(H) (HK$29.50)

高盛高华在过去 12个月中与下述公司存在投资银行客户关系：中兴通讯(A) (Rmb30.63)、中兴通讯(H) (HK$29.50)

没有对下述公司的具体信息披露： Ericsson ($5.98)、Nokia ($5.86)

公司评级、研究行业及评级和相关定义

买入、中性、卖出：分析师建议将评为买入或卖出的股票纳入地区投资名单。一只股票在投资名单中评为买入或卖出由其相对于所属研究行业的总体潜在回报决

定。任何未获得买入或卖出评级且拥有活跃评级（即不属于暂停评级、暂无评级、暂停研究或没有研究的股票）的股票均被视为中性评级。每个地区投资评估

委员会根据 25-35%的股票评级为买入、10-15%的股票评级为卖出的全球指导原则来管理该地区的投资名单；但是，在某一特定分析师所覆盖行业中买入和卖

出评级的分布可能根据地区投资评估委员会的决定而有所不同。此外，每个地区投资评估委员会管理着地区强力买入或卖出名单，该名单以总体潜在回报规模

和/或实现回报的可能性为主要依据确立各自研究范围内的投资建议。将股票加入或移出此类强力买入或卖出名单并不意味着分析师对这些股票的投资评级发生

了改变。

总体潜在回报：代表当前股价低于或高于一定时间范围内预测目标价格的幅度，包括所有已付或预期股息。分析师被要求对研究范围内的所有股票给出目标价

格。总体潜在回报、目标价格及相关时间范围在每份加入投资名单或重申维持在投资名单的研究报告中都有注明。

研究行业及评级：每个行业研究的所有股票名单可登陆 http://www.gs.com/research/hedge.html通过主要分析师、股票和行业进行查询。分析师给出下列评级中

的其中一项代表其根据行业历史基本面及／或估值对研究对象的投资前景的看法。具吸引力(A)：未来 12个月内投资前景优于研究范围的历史基本面及／或估

值。中性(N)：未来 12个月内投资前景相对研究范围的历史基本面及／或估值持平。谨慎(C)：未来 12个月内投资前景劣于研究范围的历史基本面及／或估

值。

暂无评级(NR)：在高盛于涉及该公司的一项合并交易或战略性交易中担任咨询顾问时并在某些其他情况下，投资评级和目标价格已经根据高盛的政策予以除去。暂停评级(RS)：由于缺乏足够的基础去确定投资评级或价格目标，或在发表报告方面存在法律、监管或政策的限制，我们已经暂停对这种股票给予投资评级和价

格目标。此前对这种股票作出的投资评级和价格目标(如有的话)将不再有效，因此投资者不应依赖该等资料。暂停研究(CS)：我们已经暂停对该公司的研究。没有研究(NC)：我们没有对该公司进行研究。不存在或不适用(NA)：此资料不存在或不适用。无意义(NM)：此资料无意义，因此不包括在报告内。



一般披露

本报告在中国由高华证券分发。高华证券具备证券投资咨询业务资格。

本研究报告仅供我们的客户使用。除了与高盛相关的披露，本研究报告是基于我们认为可靠的目前已公开的信息，但我们不保证该信息的准确性和完整性，客

户也不应该依赖该信息是准确和完整的。报告中的信息、观点、估算和预测均截至报告的发表日，且可能在不事先通知的情况下进行调整。我们会适时地更新

我们的研究，但各种规定可能会阻止我们这样做。除了一些定期出版的行业报告之外，绝大多数报告是在分析师认为适当的时候不定期地出版。

高盛高华为高华证券的关联机构，从事投资银行业务。高华证券、高盛高华及它们的关联机构与本报告中涉及的大部分公司保持着投资银行业务和其它业务关

系。

我们的销售人员、交易员和其它专业人员可能会向我们的客户及自营交易部提供与本研究报告中的观点截然相反的口头或书面市场评论或交易策略。我们的资

产管理部门、自营交易部和投资业务部可能会做出与本报告的建议或表达的意见不一致的投资决策。

本报告中署名的分析师可能已经与包括高华证券销售人员和交易员在内的我们的客户讨论，或在本报告中讨论交易策略，其中提及可能会对本报告讨论的证券

市场价格产生短期影响的推动因素或事件，该影响在方向上可能与分析师发布的股票目标价格相反。任何此类交易策略都区别于且不影响分析师对于该股的基

本评级，此类评级反映了某只股票相对于报告中描述的研究范围内股票的回报潜力。

高华证券及其关联机构、高级职员、董事和雇员，不包括股票分析师和信贷分析师，将不时地根据适用的法律和法规对本研究报告所涉及的证券或衍生工具持

有多头或空头头寸，担任上述证券或衍生工具的交易对手，或买卖上述证券或衍生工具。

在高盛组织的会议上的第三方演讲嘉宾（包括高华证券或高盛其它部门人员）的观点不一定反映全球投资研究部的观点，也并非高华证券或高盛的正式观点。

在任何要约出售股票或征求购买股票要约的行为为非法的地区，本报告不构成该等出售要约或征求购买要约。本报告不构成个人投资建议，也没有考虑到个别

客户特殊的投资目标、财务状况或需求。客户应考虑本报告中的任何意见或建议是否符合其特定状况，以及(若有必要)寻求专家的意见，包括税务意见。本报告

中提及的投资价格和价值以及这些投资带来的收入可能会波动。过去的表现并不代表未来的表现，未来的回报也无法保证，投资者可能会损失本金。

某些交易，包括牵涉期货、期权和其它衍生工具的交易，有很大的风险，因此并不适合所有投资者。外汇汇率波动有可能对某些投资的价值或价格或来自这一

投资的收入产生不良影响。

投资者可以向高华销售代表取得或通过 http://www.theocc.com/about/publications/character-risks.jsp取得当前的期权披露文件。对于包含多重期权买卖的期权

策略结构产品，例如，期权差价结构产品，其交易成本可能较高。与交易相关的文件将根据要求提供。

全球投资研究部提供的不同服务层级：根据您对接收沟通信息的频率和方式的个人偏好、您的风险承受能力、投资重心和视角（例如整体市场、具体行业、长

线、短线）、您与高华证券的整体客户关系的规模和范围、以及法律法规限制等各种因素，高华证券全球投资研究部向您提供的服务层级和类型可能与高华证

券提供给内部和其他外部客户的服务层级和类型有所不同。例如，某些客户可能要求在关于某个证券的研究报告发表之时收到通知，某些客户可能要求我们将

内部客户网上提供的分析师基本面分析背后的某些具体数据通过数据流或其它途径以电子方式发送给他们。分析师基本面研究观点（如股票评级、目标价格或

盈利预测大幅调整）的改变，在被纳入研究报告、并通过电子形式发表在内部客户网上或通过其它必要方式向有权接收此类研究报告的所有客户大范围发布之

前，不得向任何客户透露。

所有研究报告均以电子出版物的形式刊登在高华客户网上并向所有客户同步提供。高华未授权任何第三方整合者转发其研究报告。有关一个或多个证券、市场

或资产类别（包括相关服务）的研究报告、模型或其它数据，请联络您的高华销售代表。

北京高华证券有限责任公司版权所有 © 2017 年

未经北京高华证券有限责任公司事先书面同意，本材料的任何部分均不得(i)以任何方式制作任何形式的拷贝、复印件或复制品，或(ii)再次分发。

三季度业绩发布之际对无线设备投资者关注热点的更新 -...

Documents