Telco - Overall - Morgans

Telecommunications│Australia│Equity research│April 4, 2018

Telco - Overall What you need to know about 5G Neutral (no change)

■

5G has many purported benefits but the highlights are: 1) a ~10x increase in average speeds/ throughput; 2) a lower cost to serve than NBN; 3) the ability to handle many more devices; 4) the ability to more dynamically manage networks; and 5) lower latency (much faster response times).

■

On face value 5G looks like it could replace fixed line for the last mile. However fixed networks handle ~60x the traffic of mobile and there are currently not enough towers or spectrum for 5G to be an outright substitute. The two are complementary.

■

We explore the pros and cons of 5G overleaf but conclude that Telstra is the best placed given its substantial ownership of 3/4 key ingredients (fibre, towers and customer scale). Telstra lacks 5G spectrum but its balance sheet strength places it well to purchase spectrum as it comes to market later this year.

Highlighted companies TLS; Add recommendation; A$4.12 PT TPM; Hold recommendation; A$5.74 PT Vodafone and Optus (not rated) Evolution of internet technologies & speeds

Year Service

Average speed (Mbps)

1970s First email sent 1980 1G - Analogue mobile voice only Dial up (commenced at 14.4 Kbps / 1990 0.01 Mbps)

0.0

1993 2G - Digital mobile voice & SMS 2000 ADSL

0.0

2003 3G - Digital voice, SMS & Data 2006 ADSL 2

4.0

2010 ADSL 2+ 4G - higher speeds allowing video 2010 streaming

9.5

0.1 0.3 5.0

18.1

2017 Average Au fixed line internet speed 2017 Average Au 4G speed

11.1

2017 Average headline NBN speed 5G - higher speeds, lower latency, higher throughput, lower cost to 2020 serve and much more

31.2

34.0

200.0

What does 5G mean for speeds and how does it compare There are numerous claims that 5G will be ~100x faster than 4G with speeds of up to 2Gbps (2,000Mbps). This may be the case for theoretical or peak speeds but average or real world speed is what matters as this is what consumers actually get. Globally 4G speeds average around 20Mbps (~7x faster than 3G) and it seems likely that 5G speeds will average around 200Mbps (~10x faster than 4G). For context Australia’s average fixed line speed is 11Mbps and headline NBN speeds average 32Mbps. Our mobile network speeds are top 10 globally but our fixed line speeds sit low by global standards. Assuming regulators are not heavy handed then 5G will incentivise ongoing mobile network investment as it makes a difference. However our mobile networks are already faster than our fixed and speed isn’t the only consideration. Consumers and Telcos alike will continue to trade-off speed, reliability and price.

What about reliability and price The ACCC recently pointed out that reliability is the top priority for choosing an NBN plan whereas price is the top priority for choosing an ADSL (legacy broadband) plan. Evidently reliability on NBN matters more than on ADSL. We think this is a function of poor consumer perceptions of the NBN and therefore a temporary willingness to pay more in order to be less frustrated. We expect that over time, consumers will gain comfort with the NBN, and price will rise again to the top consideration. From a mobile perspective all offerings are constrained by available spectrum per subscriber (MHz as it relates to Mbps per subscriber per tower). 5G will be no different and even though 5G throughput will be higher, the available MHz/Mbps per subscriber per tower will likely be the bottleneck, assuming adequate backhaul fibre is provisioned to the towers. In the context of 5G reliability will likely be lower than fixed as it’s harder to reliably provision a moving target and there are unlikely to be sufficient spectrum and towers per person to supply the required volume. According to the ABS fixed subscribers download ~60x more Gb per month than mobile (156Gb on fixed versus 2.6Gb on mobile). It currently costs ~15x more to move a Gb of data over a mobile network than a legacy fixed network. NBN increases the price by ~3x (dropping the gap from 15x to 5x). All things being equal, 5G should drop the costs by ~10x, so 5G should cost about half the price per Gb of NBN. However handling the ~60x increase in volume (in today's terms and even more by 2020) is not currently possible and would require considerably more towers, fibre, and spectrum per subscriber. Consequently we view 5G not as an outright substitute for fixed line but as complementary. Software Defined Networks (traffic controllers for data networks) will allows telcos to dynamically reroute traffic onto the most useful network (automatically prioritising what matters most - speed, reliability or price). Fixed and mobile will converge over time and 5G will accelerate this.

Telstra is best placed assuming it acquires more 5G spectrum Nick HARRIS T +61 7 3334 4557 E [email protected] James BARKER T (61) 7 3334 4893 E [email protected]

The key ingredients for 5G are fibre (for routing last mile wireless traffic onto large central backhaul cables and directing traffic towards its destination), towers (for coverage and volume), customer scale (to better share the fixed cost base) and spectrum (for the last mile). Telstra has the most fibre (~200% more than average), the most towers (~60% more than average) and the most customers (35% more than average). Optus owns the most “5G spectrum” but Australian standards are still being devised. Once Australian 5G standards are set the spectrum will auctioned off starting mid-2018 and Telstra appears best funded to bid. It is around two years before 5G will become materially commercial and we think Telstra (Add recommendation; A$4.12 price target) is best placed for this step change.

IMPORTANT DISCLOSURES REGARDING COMPANIES THAT ARE THE SUBJECT OF THIS REPORT AND AN EXPLANATION OF RECOMMENDATIONS CAN BE FOUND AT THE END OF THIS DOCUMENT. MORGANS FINANCIAL LIMITED (ABN 49 010 669 726) AFSL 235410 - A PARTICIPANT OF ASX GROUP

Powered by EFA


Evolution of mobile speeds It was only in 2003 that mobile data became a commercial reality and since then speeds of mobile networks have moved from strength to strength.



2G was voice only (no data was moved over the mobile network);



In the real world average 3G speeds were closer to 3Mbps. 3G peak speeds stared at 10 Mbps and over time edged towards 22Mbps;



In the real world average 4G speeds are around 20Mbps (~7x faster than 3G). 4G peak speeds started at 50Mbps and over time edged towards 100Mbps;



In the real world 5G speeds are unknown as 5G is mostly theoretical at this stage (limited real world tests aside). We think that average 5G speeds are likely to be around 200Mbps or ~10x faster than 4G. 5G peak speeds are claimed at 10-20Gbps (10,000 to 20,000Mbps) but this assumes zero contention (not having to share the capacity with others) and sometimes the line between 5G and mmWave gets confused (mmWave being more commonly used for stationary or point-to-point last mile wireless rather than moving wireless); and



Put simply if 5G can move 10x the volume of data over a mobile network than 4G then the cost to move a Gb of data drops by a factor of 10 (ignoring capex).

Volume growth With 5G the obstacle becomes not the speed or cost per Gb but volume or the inability of mobile networks to handle the same amount of volume as a fixed line. In totality this gap seems unlikely to close anytime soon as without substantial increases in mobile towers and available spectrum per subscriber mobile in isolation will not be able to handle the increased volume. Currently the average household downloads ~60x more data on the fixed line network than the mobile network (156Gb per month downloaded on fixed vs 2.6Gb downloaded per month on mobile) according to the ABS’s December 2017 figures. Figure 1: Fixed line downloads per user per month in AU

Figure 2: Mobile downloads per user per month in AU

(43% CAGR of data downloaded from Dec 2010 to Dec 2017)

(57% CAGR of data downloaded from Dec 2010 to Dec 2017)

SOURCES: ABS, MORGANS, COMPANY REPORTS

SOURCES: ABS, MORGANS, COMPANY REPORTS

2


Figure 3: Extrapolating data downloaded (Gb) per fixed line subscriber per month

SOURCES: ABS; MORGANS; COMPANY REPORTS

If current volume growth trends remain intact then it seems unlikely that mobile will ever be able to handle the required volume of data. However there is the potential for low end users to be serviced on 5G instead of fixed. There is also the potential to shave NBN CVC costs (Connected Virtual Circuit or volume costs) by offloading some fixed line traffic off the NBN and onto 5G when the 5G spectrum is not fully used (at which point it has a near zero marginal cost to the Telco). For example a Telco may be able to buy 3Mbps rather than 4Mbps of CVC on the NBN and route a quarter of their fixed line traffic onto 5G in peak hour. The ability to dynamically reroute traffic is a benefit of 5G and Software Defined Networks (SDNs). At NBN Co’s February 2018 update it highlighted that an average of 200Gb was downloaded per customer and download volume was growing at 24% pa. Earlier NBN said that 50% of NBN plans downloaded less than 95Gb per month. Details of user download composition (low, median, high download volumes) are not provided (just the average is) so it’s not obvious what the distribution of subscriber downloads looks like. It is therefore difficult to predict what percentage of the user base could be serviced on mobile only (excluding stationary mmWave 5G). Our best guess is that downloads are likely to follow a normal distribution curve with some subscribers downloading lots and others little. We might be able to assume that ~16% of the base is very low end. This 16% number is not coincidently the “laggard” component of the technology adoption curve and mirrors the shape of a normal distribution curve. Laggards are the last to adopt and therefore likely to have the lowest downloads per subscriber. They are more likely to be using an internet connection for internet banking and email rather than streaming lots of high definition videos on Netflix. Should this be the case and 16% of subscribers can be served on 5G mobile only then this would mean that rather than having 8.1m active subscribers that the NBN currently forecasts it would have 6.8m (16% less) and EBITDA would drop from a forecast A$2.2m to ~A$1.4bn. All of this commentary is before taking into account any other variations of wireless and their likely impact on the NBN. Applying the same 16% offload could see Telstra save A$500m pa in NBN connection costs (of a total A$3bn NBN connection costs that it has flagged). There are other competing technologies sometimes included in the 5G discussion such as stationary 5G aka mmWave aka point-to-point wireless as well as more advanced versions of WiFi and even pre mmWave point-to-point wireless that has been used in commercial applications for decades. Assuming line of sight is available then these sort of technologies can offer, in many instances, faster installation, faster speeds and cheaper connectivity. However their reliability isn’t as high as fibre as environmental factors can block or distort line of sight and negatively influence performance.

3


Cost differential fixed vs 4G vs 5G Having ascertained that mobile can physically handle some of the fixed line volume, the next part of the equation comes down to cost. As the figure below shows the cost per Gb on mobile continues to trend lower due to competition (driving consumer price points lower) and technology advancements (driving the cost to serve lower). The cost per Gb of mobile has broadly dropped by 50% pa over the last six years. This rate should, in theory accelerate, given increased competition, the introduction of 5G and the NBN incentivising further mobile investment. Figure 4: Mobile price trends (cost per Gb/Mb) over time

SOURCES: ACCC, Morgans, Company data

The 5G opportunity for Telcos relates to using 5G to meaningfully lower the cost to serve. As the figures below show, there is a substantial difference in the current cost to serve a gigabyte of data on fixed line (pre and post NBN) versus through mobile. As Figure 5 shows, it currently costs 15 times as much to deliver one gigabyte of data over a 4G mobile network vs legacy fixed. However the NBN causes a three-fold increase in cost to serve (last mile access costs increase from ~$15 to ~$48 per month today – and even higher long term on the NBN’s corporate plan). Fixed costs increase by 3.2x so the cost differential drops from 15x to 4.7x more expensive on 4G vs NBN (Figure 6). 5G is expected to drop the cost by a factor of 10 so as Figure 7 shows should mean it becomes cheaper (about half the price) to move 1Gb of data on 5G than on the NBN. Figure 5: Legacy fixed vs 4G mobile today

Figure 6: NBN fixed vs 4G mobile today

NBN

Now

Fixed Mobile (4G) Mobile (5G) Differential (mobile vs fixed)

Figure 7: NBN fixed vs 5G mobile in 2020

1 15 na 15

SOURCES: MORGANS, COMPANY REPORTS


5G

3.2 15 na 4.7



3.2 15 1.5 0.47


Telstra discloses Average Revenue Per User (ARPU) and EBITDA for its fixed and mobile divisions which allows us to calculate costs (COGS plus OPEX) per subscriber. Fixed EBITDA is currently distorted by NBN one-offs but was historically (pre NBN) running at ~40%. Our assumption is this will be closer to 20% in an NBN reseller world.

4


As Figure 8 below shows Belong mobile offers consumers 15Gb per month for $40 ARPU (RRP of $2.67 per Gb). Belong NBN offers consumers 100 Gb per month for $50 (RRP of $0.50 per Gb), so consumers are paying 5.3x more to move a Gb of data over the mobile network than the fixed line NBN. If we assume both operate on a 20% EBITDA margin then the cost to Telstra is 5.3x more to move a Gb of data over the mobile network than the fixed line NBN. Figure 8: Cost differentials mobile versus fixed Note: Telstra does not disclose Belong or NBN margins (for comparison purposes we assume both operate at a 20% EBITDA margin). Belong NBN Gb included

Belong mobile

100

Belong NBN vs Belong mobile cost diff

15

ARPU

$50.00

$40.00

EBITDA

$10.00

$8.00

EBITDA margin

20%

20%

OPEX & COGS

$40.00

$32.00

Consumer cost per Gb (RRP)

$0.50

$2.67

5.3 x

Telco cost per Gb (COGS+OPEX)

$0.40

$2.13

5.3 x

SOURCES: Morgans, Company data

As Figure 9 below shows Telstra mobile offers consumers 15Gb per month for $59 ARPU (RRP of $3.93 per Gb). Telstra NBN offers consumers 1,000Gb per month for $79 (RRP of $0.08 per Gb), so consumers are paying 49.8x more to move a Gb of data over the mobile network than the fixed line NBN. If our assumed NBN margin of 20% is correct then it costs Telstra 37.3x more to move a Gb of data over the mobile network than the fixed line NBN. Figure 9: Cost differentials mobile versus fixed Note: Telstra discloses mobile EBITDA margins but not or NBN (for comparison purposes we assume a 20% EBITDA margin on TLS NBN). TLS NBN Gb included

TLS mobile

1000

TLS NBN vs TLS mobile cost diff

15

ARPU

$79.00

$59.00

EBITDA

$15.80

$23.60

EBITDA margin

20%

40%

OPEX & COGS

$63.20

$35.40

Consumer cost per Gb (RRP)

$0.08

$3.93

49.8 x

Telco cost per Gb (COGS+OPEX)

$0.06

$2.36

37.3 x

SOURCES: Morgans, Company data

The above figures (Figure 9 in particular) are illustrative only and attempt to prove the gap between fixed and mobile is already narrowing. We expect this gap will continue to narrow with the commercialisation of 5G and are already seeing this happening overseas (for example see Figure 25 where Verizon and ATT released unlimited 4G plans about 12 months ago). We also note that the above figures use included downloads and most consumers do not use their included allowances so Telstra’s cost to move a Gb is likely lower than illustrated above.

Generational change in mobile networks As the figures above show the cost to move data continues to decline as technologies get more efficient (higher throughput and lower hardware costs). The NBN actually increases this cost (which is unusual and, in our view, not sustainable) while progressive generations of mobile networks will encourage investment in mobile technologies and lower the cost of mobile. The value for Australian Telcos has been created by vertical integration and selling on-net products as on-net products is typically 3x more profitable than off-net or reseller products. 5


The NBN reverses this and in our view creates a clear incentive for Telcos to further build out their mobile networks at the expense of the NBN. 5G makes the allure of mobile over fixed even more compelling, in our view. Figure 10: Evolution of mobile network infrastructure

Figure 11: Evolution of mobile/internet speeds

2G

• A network protocol that allowed the most basic smartphone functionality: simple text messaging and very small amounts of data.

1980 1G - Analogue mobile voice only 1990 Dial up (commenced at 14.4 Kbps / 0.01 Mbps)

0.0

3G

• A network protocol supporting services that provide an information transfer rate of at least 2 Mbps, allowing larger data formats, including music and video.

1993 2G - Digital mobile voice & SMS 2000 ADSL

0.0

• The fourth generation of mobile communications technology (established in March 2008. 4G LTE currently delivers the fastest wireless connection for a mobile internet experience—up to 10 times faster than 3G.

2003 3G - Digital voice, SMS & Data 2006 ADSL 2

4.0

• A new protocol which is expected to be up to 100 times faster than existing networks and will have an additional advantage of low latency.

2017 Average Au fixed line internet speed 2017 Average Au 4G speed

11.1

2017 Average headline NBN speed

31.2

Average speed (Mbps)

Year Service 1970s First email sent

4G LTE

5G

2010 ADSL 2+ 2010 4G - higher speeds allowing video streaming

5G - higher speeds, lower latency, higher throughput, lower cost to 2020 serve and more SOURCES: VERIZON

0.1 0.3 5.0 9.5 18.1 34.0

200.0

SOURCES: MORGANS, COMPANY REPORTS, INDUSTRY REPORTS.

An overview of 5G The fifth generation wireless system (‘5G’) is the next generation of mobile internet connectivity. 5G is touted to provide improved connections through increased speed (volume) capacity and reliability (reduced latency). What are the practical benefits of 5G? 

Speed capacity – The International Telecommunication Union (ITU) has recently announced that the minimum requirements for peak data rates is downlink of 20Gbps and uplink of 10Gbps. There have been many different speed expectations touted in the media with respect to what speeds we should expect in a 5G world. The most consistent figure appears to be that peak speeds are likely to be c12x faster than base 4G speeds (i.e. 12x faster than the average 4G speed of ~18Mbps, which equates to 216Mbps).



Low latency: Latency or response time refers to the delay before a transfer of data begins following an instruction for its transfer. Under ideal circumstances, 5G networks should offer users a maximum latency of 4ms (vs c20ms on 4G LTE cells). The practical benefits of lower latency are (almost) real-time control of remote products (i.e. a surgeon may be able to use a device from a remote location to operate on a patient over 5G or the ability for a drone to accurately make a delivery onto a moving object i.e. hand delivered without crashing).



Can connect many more devices: The minimum requirement for 2 connection density is 1million devices per km . This provides the capacity for many devices (i.e. traffic lights, cars, health monitoring devices) to all be simultaneously connected to the network as the Internet of Things (IoT) thematic plays out over time.



Better integration with other networks – 5G networks are expected to more seamlessly handoff traffic onto other frequencies (e.g. flip between low and high bands as may be required due to interference like rain) or to other networks (e.g. fixed or 4G).

Better integration is a significant point as it allows for optimisation of network routes which is one of the key points of developing a Software Defined Network (SDN). SDNs are being rolled out globally and the key point of them is the ability to reroute internet traffic to the optimal network. The optimal network for the job could be the least congested (if you want speed), the cheapest (if price matters) or the fastest (if reliability matters).

6


In addition to this 5G should be able to run on multiple disparate networks (e.g. five specific fit-for-purchase networks rather than one general purpose network). In the case of 4G all connected devices run over the same frequency so low priority data (e.g. not time crucial file backups) can crowd out time critical data (e.g. remotely controlling a drone). However this seems less likely to be the case in a 5G world as 5G is expected to cover a range of spectrum (3.6 GHz; 24.527.5GHz) so non-time-critical data transfers can run on a totally different frequency and not impact time-critical data transfers. As part of this disparate network equation we need to point out that 5G is a two part process consisting of: 1) fully mobile (e.g. 3G, 4G, 5G) where the consumer is literally moving and is able to get theoretical maximum speeds of ~1Gbps on the fly; and 2) stationary mobile (e.g. mmWave which is expected by many to be a competitor to fixed line) and isn’t connecting moving objects but is point-to-point or line of sight wireless. In this instance the theoretical maximum speed is ~10-20Gbps. Obviously theoretical and actual speeds will vary greatly which is why we talk about average speeds. Stationary mobile is expected to deliver practical speeds of 1Gbs according to Optus and Verizon. Articles relating to 5G have claimed theoretical peak speeds of ~10Gbps and actual achieved peak speeds (in a controlled environment) of ~2Gbps. Achieved peak speeds are c40x faster than 4G (2,000Mbps vs 50Mbps) and 6x faster than 4G-LTE tri-band (2,000Mbps vs 335Mbps). Figure 12: Maximum theoretical downlink speed by technology (Mbps) 3.5G/DC-HSPA+

42.2 Mbps

4G/LTE

100 Mbps

4G/LTE Cat. 4

150 Mbps

4G/LTE Advanced

1000 Mbps

5G*

10000 Mbps

SOURCES: GSMA INTELLIGENCE * 10Gbps is the minimum theoretical upper limit specified for 5G but 20Gbps has also been achieved.

We flag that throughout the mobile network evolution, actual achieved speeds in a real life context have always been significantly lower than the quoted speeds before the new technology was released. We use c200Mbps as our estimated average speed and base this on recent commentary from industry players, the ratio of average to peak and likely contention ratios.

7


Figure 13: Technology shifts - Timing, speeds, generational uplifts and where relevant associated spectrum ranges

Frequencies used

900 Mhz

850, 900 & 2100 Mhz 700, 850, 900, 1800, 2100, 2300 & 2700 Mhz

39 GHz and 28 GHZ 3.5 GHz 3.6 GHz & 24 GHz , (26 GHz for mm Wave)

Average download speed (Mbps)

Speed uplift on prior generation technology

Average as % of Peak peak speed speeds

Year

Service

1970s 1980

First email sent 1G - Analogue mobile voice only

0.0

1990 1993

Dial up (commercially started at 14.5 Kbps / 0.01 Mbps)

0.1

2G - Digital mobile voice & SMS

0.0

0

2000

ADSL 3G - Digital voice, SMS & Data (initially deployed at 1Mbps in early 2000s but upgraded to 21Mbps around 2008)

0.3

1

60.0%

11

38.1%

2006 2010

ADSL 2

5.0

16.7

8

62.5%

ADSL 2+

9.5

1.9

20

47.5%

2010 2017

4G - higher speeds allowing video streaming

18.1

4.5

50

36.1%

Average Au fixed line internet speed

11.1

25

44.4%

2017 2017

Average Au 4G speed

34.0

Average headline NBN speed

31.2

2019 2019

Verizon stationary wireless (mmWave for last 60 metres) 1,000

10,000

10.0%

Optus stationary wireless (5G mmWave for last mile) 5G - higher speeds, lower latency, higher throughput, lower cost to serve and more

10,000

10.0%

1,000

20.0%

2003

2020

0

4.0

2.8

1,000 200

11.1

SOURCES: Morgans estimates; OpenSignal, ABS, ACCC, NBN Co, Company data, industry sources and Wikipedia

New revenue (including benefits of latency) The second aspect of 5G is the latency benefits it provides (i.e. the delay before a transfer of data begins) and the consequential new revenue opportunities for Telcos that this presents. 5G’s low latency, throughput and efficiencies present a number of opportunities in the development of the ‘Internet of Things’ (IoT) and as a result, new revenue cases for Telcos. Ericsson forecasts that by 2023 there will be over 30 billion connected devices, of which 20 billion will be related to the IoT. These devices may include connected cars, machines, meters, sensors, point-of-sale terminals, consumer electronics and wearables. These IoT devices are all collecting lots of data points and sending these data points back to a central location for interpretation and/or action. Of these IoT devices, there are currently ~0.5bn devices with cellular connectivity. This is forecast to increase to ~1.8bn by 2023. The majority of these connections is expected to be provided by LTE and 5G. 5G will also provide mechanisms for rapid – and cost-effective – introduction of new IoT services.

8


Figure 14: Examples of use case evolution and supporting network technologies Enhanced mobile broadband Automotive Manufacturing Energy and utilities Healthcare

Technologies

Current Browsing, social media, music, video Wi-Fi hotspots, on-demand GPS map data, over-the-air software updates Connected goods, intra/inter enterprise communication Smart metering, dynamic and bidirectional grids Remote patient monitoring, connected ambulances, electronic health records > > > >

On the road to 5G Fixed Wireless Access, interactive live concerts and sports events Predictive vehicle maintenance; capturing of sensor data for real-time traffic, weather, parking and mapping services Collaborative robots, distributed control system, remote quality inspection Distributed energy resource management, distribution automation

5G Experiences 4K/8K videos, mobile AR/VR gaming, immersive media Autonomous vehicle control, cooperative collision avoidance, vulnerable road user discovery Remote control of robots, AR in training, maintenance, construction and repair Control of edge-of-grid generation, virtual power plants, real-time load balancing

Tele surgery, AR aiding medical treatment

Precision medicine, remote robotic surgery, ambulance drones

> Multi-standard network > Cat-M1/NB-IoT > Cloud-optimized network functions > VNF orchestration >

Gigabit LTE (TDD, FDD, LAA) Massive MIMO Network slicing Dynamic service orchestration Predictive analytics

> > > > >

5G New Radio Virtualized RAN Federated network slicing Distributed cloud Real-time machine learning/AI SOURCES: Ericsson

Figure 15: Examples of 5G use case

SOURCES: Ericsson

International 5G trends and timing Historically emerging mobile technology (e.g. 4G) has been has been piloted in higher density urban areas, and then subsequently deployed into suburban and regional areas. We understand that 5G is expected to be trialled/rolled out in a similar manner. The first commercial networks based on 5G New Radio (NR) are expected to go live in 2019, with major commercial deployments from 2020. Early 5G deployments are foreseen in several markets, including the US, South Korea, Japan, and China. In Australia, Optus will use the 3.5MHz spectrum it has already purchased in Sydney and Melbourne to deploy city fixed wireless from early 2019.

9


Recent trials/news flow relating to 5G  Optus, AT&T and Verizon have already launched 4G mobile plans in competition with fixed line. Given the higher cost to serve these are, in our view, loss leaders designed for trials ahead of commercial stationary 5G deployments.  AT&T has recently announced that it expects to be the first US company to offer mobile 5G services in a dozen markets by late 2018. In addition, AT&T plans to trial 5G technology with businesses to help create ‘more engaging experiences for their customers’. In previous 5G trials, AT&T noted that reliable gig speeds were achieved, which provided AT&T with confidence that it could utilise 5G to provide broadband in the home.  Verizon has recently completed the acquisition of Straight Path Communications for ~US$3.1bn. Straight Path has 735 licenses in the 39 GHz band and 133 licenses in the 28 GHz band. Verizon CEO Lowell McAdam reiterated the company’s promise that “we’re very comfortable with being able to deliver a Gigabit of service to everyone that we’re providing service to,” when the company launches its 5G fixed wireless service in three to five U.S. cities sometime this year.  Huawei (Chinese Telco vendor) and NTT DoCoMo (Japanese Telco) have recently completed a joint field trial of 5G over a long distance. The trial reported downlink data transmissions of a maximum speed of more than 2Gbps on a testing vehicle. The two companies also recently completed a high speed data transmission over a distance of 1.2km in downtown Tokyo, achieving a downlink of 4.5Gbps and an uplink of 1.55Gbps.  SK Telecom in conjunction with Ericsson tested 5G technology to demonstrate at the 2018 Winter Olympics in Pyeongchang. Australia  Telstra has launched a 5G innovation centre on the Gold Coast from which it will trial new networking technologies in partnership with Ericsson. The centre's location was chosen due to Telstra's intention to run a trial 5G network during the Commonwealth Games on the Gold Coast in April 2018.  Optus is using mmWave/5G fixed wireless as part of its Commonwealth Games sponsorship. It intends to launch a 5G fixed wireless product in early CY2019. This is expected to operate in metro Sydney and metro Melbourne and the 3.5Ghz spectrum that Optus already owns. Optus said it believes it has sufficient spectrum to offer Gb services but did not comment on how many customers it could offer Gb services. We expect it's initially a fairly limited launch.  The Australian government has recently announced that 125 MHz of spectrum in the 3.6 GHz band will be sold at auction, with the process to commence in October 2018.

Towers and capex requirements Telstra has stated that there is fairly limited hardware capex for 5G and that it is mostly software upgrades. This would be the case should it upgrade existing 3G/4G sites to 5G. However we think Telcos are likely to build out more mobile phone towers to greater densify their mobile network as more towers per person create greater mobile throughput. This aside we estimate the biggest 5G capex spend is likely to relate to: 

Spectrum costs: This is likely to be the one-off largest capex spend in relation to 5G deployment. Pricing expectations vary greatly from hundreds of millions to billions. The key swing factor, in our view, is likely the trade-off between the government’s desire to maximise cash received (highest price wins) and the regulator's desire to enforce broader mobile spectrum sharing arrangements (why pay up if you’re forced to share);



Backhaul fibre network upgrades: To handle massive volume increases. Vodafone (through a TPG Telecom contract) has finished upgrading the 10


dark fibre backhaul to their towers. Vodafone’s lack of fibre was one of the main reasons for its mobile issues earlier in the decade. Telstra has recently announced that it intends to increase its mobile network capacity by five times; 

Micro-cell deployment: We believe that many more sites are required; however, these are likely to be micro-cells (i.e. on top of power poles). Optus has stated micro-cells cost ~A$100k per deployment. We expect a lot of these to be attached to existing fibre networks and therefore not require major fibre upgrade capex, just tower capex (see Figure 16 below);



The number of towers and required backhaul are a function of the desired average speed per subscriber. This dictates how many subscribers each tower can handle. Today in Australia there are on average about 1,500 subscribers served per macro tower and an average of 0.11 MHz per subscriber. Figure 16: Mobile architecture

Type

Usage

Typical range

Macro Cell

Suburb(s)

5-30km

Concurrent users 1,000-3,000

Lamp post / Venue / Small dense area

200-300m

100-2,000

100-200m

30-100

10-100m

1-10

Metro / Micro Cell Pico Cell

Building Home

Femto Cell




For more on potential spectrum sharing refer to our note “Regulatory ramblings cut short” 16 November 2017. An extract of this includes:



One of the key takeaways and its relevance to spectrum capex/value was a regulatory comment suggesting a need for “opportunities for licence design to incorporate or facilitate share arrangements, including some that could be managed by users or other third parties” and “greater use of sharing arrangements, including enabling technology innovations such as dynamic spectrum access applications to be supported by the new flexible licensing arrangements”. Overall commentary suggests more dynamic and shared use of mobile spectrum is likely going forward.



Rod Simm commented that “I do not want to see some players locked out of markets, or a new mobile entrant prevented from competing with incumbents, because they cannot get access to sufficient spectrum. This means it is vital to consider the impact of spectrum allocation on competition in spectrum-using markets”;



Progress is being made with respect to selection of 5G frequencies with formal standards likely in mid/late 2018, spectrum auctions from mid-2018 and commercialisation around 2020.

11


Spectrum ranges, usage and ownership in Australia Figure 17: Spectrum bands currently being considered for allocation Band Name

Spectrum parameters

Current use

Comments

Multi-band residual lots

Residual/unsold spectrum in 1800 MHz, 2 GHz, 2.3 GHz and 3.4 GHz bands

Unused

Allocation has commenced.

3.6 GHz band

3575-3700 MHz

850 MHz band

809-824 MHz and 855-870 MHz

900 Mhz band

890-915 MHz and 935-960 MHz

Point-to-point, FSS (satellite), Amateur, Regional WISPs Fixed links, point-to-point multipoint, land mobile 2G mobile (GSM), 3G, 4G

1.5 GHz band

1427-1518 MHz

Decision to re-farm has been made already, but a lengthy clearance process foreshadowed. 2G shutdown imminent, large portion of the band un-utilised.

Space research, FSS satellite (ESA, CSIRO, NASA), NBN Point-to-point, some multipoint, Defence

Millimetre wave band 24.25-27.5 GHz

Likely first 5G band.

Second of the 5G bands. Broader interest to be determined. Technology standarisation progressing. Lower level of nearterm domestic interest. SOURCES: ACMA, MORGANS, COMPANY REPORTS

Spectrum ownership in Australia currently focuses on older bands (pre 5G spectrum) with Optus being the only owner of “5G” spectrum. Optus purchased this 5G spectrum before 5G existed and intends to use it in Sydney and Melbourne in early CY2019 for point-to-point stationary 5G wireless trials (in competition with fixed line). Figure 18 (below) is a generic representation of spectrum ownership across Australia. It does not differentiate between the areas this spectrum is owned in (i.e. Optus ownS 3.6GHz spectrum on which it will release a fixed 5G product in metro Sydney and Melbourne) but does not own this 3.6 GHz in Brisbane or other locations so cannot offer this as a national product. NBN Co also owns some regional spectrum for its fixed wireless connect but this is not included in our forecasts. For more specific details refer to: https://web.acma.gov.au/rrl/browse_licences.cat_listing

Figure 18: Spectrum bands currently owned or being considered for allocation Usage

4G 3G 2G / GSM / CDMA 4G 4G 4G 4G 4G 5G 5G 5G 5G 5G mm Wave 5G mm Wave 5G mm Wave 5G mm Wave

MHz

TLS

700 850 900 1,800 (1.8 GHz) 2,100 (2.1 GHz) 2,300 (2.3 GHz) 2,600 (2.6 GHz) 3,400 (3.4 GHZ) 3,400 (3.4 GHz) 3,500 (3.5 GHz) 3,600 (3.6 GHz) 3,700 (3.7 GHz) 24,200 (24.2 Ghz) 25,000 (25 GHz) 26,000 (26 GHz) 27,500 (27.5 Ghz) Total MHZ owned

Optus

40 30 16.8 73.3 22.1 80 10.5 TBA TBA TBA TBA TBA TBA TBA TBA 272.7

TPG Telecom

Voda

20 16.6 45.8 17.5 5.8 40 5.4 TBA 100 TBA TBA TBA TBA TBA TBA 251.1

10 20 16.6 55 45

Expiry

20

10

20 TBA TBA TBA TBA TBA TBA TBA TBA 146.6

TBA TBA TBA TBA TBA TBA TBA TBA 50

2029 2028 Ongoing 2028 2032 2029 2029 TBA 2030 TBA TBA TBA TBA TBA TBA

SOURCES: ACMA, ACCC, MORGANS, COMPANY REPORTS

12


Key mobile assets in Australia The figure below looks at key mobile asset ownership in Australia as at 31 December 2017 and based on publically disclosed data. Figure 19: Key mobile assets in Au (and national average of 4G spectrum per tower) Telstra

Optus

Vodafone

Market share (by subscribers)

47%

33%

19%

Mobile subs (pre+post+mob BB) Dec 17

14.1

10.0

5.8

240,000

48,000

4,900

Towers

9,000

7,000

4,000

Average subscribers per tower

1,565

1,421

1,450

Total 4G MHz per tower

225.9

134.5

110

MHz per sub per tower

0.14

0.09

0.08

Average Mb per sub per second

28.9

18.9

15.2

Kilometers of fibre installed across their network

SOURCE: MORGANS RESEARCH, COMPANY

Our view is that as 5G is progressively phased in and additional mobile infrastructure is progressively built out, the line between fixed and mobile will converge as consumers will just have connectivity which automatically traverses the most appropriate path. Looking outside of Australia there are very few telcos that split fixed and mobile earnings. Instead most report and bundle them into consumer centric products as illustrated below by Telstra. Figure 20: Moving from network to end user facing (regardless of the delivery technology/technologies)

SOURCES: TELSTRA, MORGANS

13


Figure 21: Where we see TLS’ ability to capitalise on an increased shift to mobile data

SOURCES: Ericsson traffic measurements (Q3 2017); MORGANS RESEARCH, COMPANY

Taking a look at international trends in mobile Mobile data internationally continues to grow Increases in mobile data/usage globally is nothing new. The following chart demonstrates the significant growth in mobile data over the last 10 years. Total mobile data traffic is expected to rise to at a compound annual growth rate (CAGR) of 42%, increasing 8x between 2018 and 2023 (Source: Ericsson). Vodafone has stated that the typical industry growth rates in mobile broadband (MBB) are in the range of 40-70% per annum, driven by data performance capabilities in 4G, the increased use in smart phone technology and strong competition in the metropolitan mobile services market. Figure 22: Mobile traffic (Voice and Data) – data traffic grew 65% in 3Q17 on the pcp

SOURCES: Ericsson traffic measurements (Q3 2017) 1 Traffic does not include DVB-H, Wi-Fi, or Mobile WiMAX. VoIP is included in data traffic

Australia’s positioning in a global context According to Akaimai’s State of The Internet Connectivity Report, the United Kingdom had the fastest connection speeds on mobile networks. Notably, while th Australia was 11 on the list, it recorded the fastest mobile speeds in the Asia th Pacific region, just in front of Japan at 12 .

14


Figure 23: Connection speeds on mobile networks (1Q17 average Mbps) 30

26 25

24.2

24.1 22.4

20

21.6 17.4

17.3

16.6

16.2

15.9

15.7

15.6

15

15

14.8

14

10

5

0

SOURCES: AKAMAI

Current examples of 4G mobile only as a substitute for fixed There are already examples of Telcos in Australia and overseas offering mobile (and fixed wireless) plans that are directly competitive to fixed line. Despite the cost per Gb of 4G mobile being ~15x higher than legacy fixed and 5x higher than NBN, Optus offers intermittently, a mobile only plan with unlimited data for $90 and 200Gb for $70 per month. Currently these sort of offerings are able to serve a small number of consumers rather than the mass market. Figure 24: 4G mobile only comps are already in the market thanks to Optus

SOURCES: Company data

Optus is already experimenting with 5G as part of its Commonwealth Games communications sponsorship package. Optus CEO, Allen Lew, commented on its February 2018 result call, that Optus will in early CY19, roll out a 5G product that competes directly with fixed line in certain locations. Optus will launch this is Sydney and Melbourne where it already owns 200MHz in the 3.5GHZ range (a 100MHz block in Sydney and a 100MHz block in Melbourne). 15


Looking globally, in the United States AT&T and Verizon both launched unlimited mobile Broadband offerings about 12 months ago. These clearly compete with fixed line offerings. Spectrum available was initially unchanged and the higher consumption per user slowed average network speeds by ~ 3Mbps. Average speeds have gradually recovered, albeit still below speeds available pre unlimited plans, presumably as Telcos upgraded their networks. Figure 25: The impact of unlimited 4G mobile plans on speeds in the USA (AT&T and Verizon speeds dipped as they jammed substantially more demand on-net)

SOURCES: OPENSIGNAL.

Telstra’s positioning in a global context Telstra, on a number of key metrics, sits largely at the top of its key global peers. Figure 26: Global telcos – EBITDA margin (3-yr historical avg)

Figure 27: Global telcos – FCF/Sales margin (3-yr historical avg)

45.0%

20.0% 40.0%

41.3% 40.2%

35.0% 30.0% 25.0%

18.0% 39.3%

17.5%

37.5%

16.0%

36.3% 35.9% 35.7%

32.6% 32.2% 31.8%

14.0% 30.4% 30.3%

29.7% 29.5% 29.3% 29.2%

12.0% 25.9%

10.0%

15.2% 14.6% 14.5% 13.4% 13.3% 12.6% 12.5%

12.2% 11.0% 10.8%

20.0%

8.0% 8.1%

15.0%

6.0%

7.3% 6.1%

10.0%

4.0% 5.0%

5.6%

5.6% 4.4%

2.0%

0.0%

0.0%

SOURCES: MORGANS, FACTSET, COMPANY REPORTS

16



Figure 28: Global telcos – Capex (% of sales)

Figure 29: Global telcos – ROIC 14.0%

30.0%

25.0%

12.0%

26.9%

13.0% 12.4% 11.6% 10.9%

10.0% 20.0%

10.6% 9.6%

20.7%

9.1%

8.0%

19.7%

8.6%

18.8% 17.9%

15.0%

17.2% 17.2% 17.2% 17.2% 16.8%

6.0%

16.1%

5.4%

13.6% 13.6% 13.5% 13.5% 13.5%

4.0%

12.4%

4.1%

10.0%

3.8%

2.0%

3.4% 3.4% 3.1% 2.4%

5.0%

0.8%

0.0% -2.0%

0.0%

-2.0%



Telstra peer benchmark Telstra has a superior ROIC (Return on Invested Capital) to peers so mathematically should trade at a higher multiple. However, Telstra trades on lower multiples so Australian investors are clearly expecting this ROIC to decline. Figure 30: Comparative company analysis – key metrics vs global peers Company

Market Cap US$b

PE

EV/EBITDA

EV/EBIT

EPS Growth EBITDA Margin

NTM FY+1 FY+2 NTM FY+1 FY+2 NTM FY+1 FY+2 FY+1

Net Debt ROIC P/BV /EBITDA FY+1 FY+2 FY+1 FY-1 AVG FY+1 Div Yield

FY+2

FY+1

FY+2

AT&T Inc.

217.2

10.2x 10.2x 10.0x 6.1x

6.2x

6.0x 11.1x 11.2x 10.6x 13.3% 1.9%

32.6%

32.9% 5.7% 5.8%

2.0x

11.7%

1.5x

BCE Inc.

38.3

15.5x 15.7x 14.9x 7.6x

7.7x

7.4x 13.2x 13.4x 12.7x 2.1%

5.7%

40.5%

41.0% 5.5% 5.8%

2.5x

8.2%

2.8x

BT Group plc

31.4

4.3x

4.4x

4.3x

2.6%

31.5%

31.7% 7.0% 7.0%

1.4x

9.6%

2.7x

Deutsche Telekom AG

76.7

13.1x 13.6x 12.0x 4.9x

5.0x

4.7x 11.0x 11.3x 10.3x -24.5% 13.5% 30.3%

30.8% 5.4% 5.9%

2.3x

4.4%

1.8x

NTT DoCoMo, Inc.

98.5

14.2x 14.0x 14.2x 6.4x

6.7x

6.4x

31.3% 3.6% 3.9%

-0.3x

11.6%

1.7x

Orange SA

44.8

12.2x 12.4x 11.5x 4.7x

4.8x

4.6x 10.1x 10.3x 9.6x 19.2% 8.0%

31.5%

32.0% 5.1% 5.4%

2.0x

3.2%

1.1x

Rogers Communications Inc. Class B17.8

13.9x 14.2x 13.3x 7.5x

7.6x

7.2x 12.4x 12.7x 11.7x 8.8%

6.7%

38.5%

38.9% 3.5% 3.6%

2.5x

8.6%

3.8x

Singapore Telecommunications Limited 42.0

13.7x 14.8x 13.8x 12.5x 12.6x 12.5x 23.7x 23.7x 23.7x -5.6%

7.2%

29.3%

29.2% 5.9% 5.4%

1.9x

10.9%

1.9x

Swisscom AG

25.1

16.0x 16.0x 15.9x 7.5x

0.8%

36.3%

36.4% 4.7% 4.7%

1.8x

11.4%

2.9x

Telecom Italia S.p.A.

14.2

9.8x 10.0x 9.3x 49.5% 5.6%

Telefonica SA

8.0x

8.0x

8.2x

8.2x

8.2x

-5.5%

9.7x 10.3x 9.7x 16.9% -1.3% 31.4%

7.5x

7.4x 15.3x 15.4x 15.2x -4.5%

9.9x

9.4x

4.7x

4.8x

4.5x

43.9%

44.5% 0.5% 0.7%

3.0x

2.3%

0.7x

50.9

10.6x 10.9x

9.9x

5.3x

5.4x

5.1x 12.0x 12.3x 11.2x -2.0% 10.4% 32.1%

32.6% 5.0% 5.0%

2.8x

4.7%

2.0x

Telenor ASA

34.6

16.2x 16.5x 15.4x 6.4x

6.5x

6.2x 10.9x 11.1x 10.3x 37.1% 6.8%

39.4%

40.0% 4.6% 4.8%

1.0x

11.1%

4.6x

Telia Company AB

20.3

14.3x 14.4x 14.0x 8.1x

8.2x

8.0x 14.2x 14.4x 13.8x 2.4%

32.7%

33.0% 5.9% 6.0%

1.7x

4.4%

1.6x

T-Mobile US, Inc.

50.9

16.8x 17.8x 14.4x 6.2x

6.4x

5.8x 13.2x 13.8x 11.9x 23.8% 23.2% 27.4%

28.2% 0.0% 0.0%

2.1x

9.1%

2.1x

196.1

10.4x 10.4x 10.2x 6.4x

6.4x

6.2x 10.0x 10.1x 9.7x 21.6% 2.1%

37.0%

37.1% 5.0% 5.2%

2.3x

21.1%

3.8x

72.6

20.8x 22.5x 20.9x 6.2x

6.3x

6.2x 18.1x 19.3x 18.1x 24.7% 7.8%

31.2%

31.4% 6.6% 6.6%

2.2x

-2.0%

0.9x

Average

13.5x 13.9x 13.0x 6.6x

6.6x

6.4x 12.7x 13.0x 12.3x 11.1% 6.5%

34.1%

34.4% 4.6% 4.7%

2.0x

8.1%

2.2x

Median

13.8x 14.1x 13.5x 6.3x

6.4x

6.2x 11.5x 11.8x 10.9x 11.0% 6.2%

32.4%

32.7% 5.1% 5.3%

2.1x

8.8%

2.0x

5.2x

4.8x

37.7%

39.0% 7.0% 7.0%

1.6x

13.0%

2.5x

Verizon Communications Inc. Vodafone Group Plc

Telstra Corporation Limited

28.6

9.8x

8.2x

9.8x 10.5x

9.6x

4.9x

8.7x

9.3x

8.5x

-7.9%

2.9%

9.0%

SOURCE: MORGANS RESEARCH, FACTSET, AS AT 3 April 2018

17


Queensland Brisbane

New South Wales +61 7 3334 4888

Stockbroking, Corporate Advice, Wealth Management

Sydney

Victoria +61 2 9043 7900


Melbourne

Western Australia +61 3 9947 4111


West Perth

+61 8 6160 8700


Brisbane: Edward St

+61 7 3121 5677

Armidale

+61 2 6770 3300

Brighton

+61 3 9519 3555

Brisbane: Tynan Partners

+61 7 3152 0600

Ballina

+61 2 6686 4144

Camberwell

+61 3 9813 2945

Perth

+61 8 6462 1999

Brisbane: North Quay

+61 7 3245 5466

Balmain

+61 2 8755 3333

Domain

+61 3 9066 3200

Adelaide

+61 8 8464 5000

Bundaberg

+61 7 4153 1050

Bowral

+61 2 4851 5555

Geelong

+61 3 5222 5128

Norwood

+61 8 8461 2800

Cairns

+61 7 4222 0555

Chatswood

+61 2 8116 1700

Richmond

+61 3 9916 4000

Caloundra

+61 7 5491 5422

Coffs Harbour

+61 2 6651 5700

South Yarra

+61 3 8762 1400

Gladstone

+61 7 4972 8000

Gosford

+61 2 4325 0884

Southbank

+61 3 9037 9444

Gold Coast

+61 7 5581 5777

Hurstville

+61 2 8215 5079

Traralgon

+61 3 5176 6055

Ipswich/Springfield

+61 7 3202 3995

Merimbula

+61 2 6495 2869

Warrnambool

+61 3 5559 1500

Kedron

+61 7 3350 9000

Mona Vale

+61 2 9998 4200

Mackay

+61 7 4957 3033

Neutral Bay

+61 2 8969 7500

Milton

+61 7 3114 8600

Newcastle

+61 2 4926 4044

Australian Capital Territory

Noosa

+61 7 5449 9511

Orange

+61 2 6361 9166

Canberra

Redcliffe

+61 7 3897 3999

Port Macquarie

+61 2 6583 1735

Rockhampton

+61 7 4922 5855

Scone

+61 2 6544 3144

Northern Territory

Spring Hill

+61 7 3833 9333

Sydney: Level 7 Currency House

+61 2 8216 5111

Darwin

Sunshine Coast

+61 7 5479 2757

Sydney: Grosvenor Place

+61 2 8215 5000

Tasmania

Toowoomba

+61 7 4639 1277

Sydney Reynolds Securities

+61 2 9373 4452

Townsville

+61 7 4725 5787

Wollongong

+61 2 4227 3022

South Australia

Hobart

+61 2 6232 4999

+61 8 8981 9555

+61 3 6236 9000

Disclaimer The information contained in this report is provided to you by Morgans Financial Limited as general advice only, and is made without consideration of an individual’s relevant personal circumstances. Morgans Financial Limited ABN 49 010 669 726, its related bodies corporate, directors and officers, employees, authorised representatives and agents (“Morgans”) do not accept any liability for any loss or damage arising from or in connection with any action taken or not taken on the basis of information contained in this report, or for any errors or omissions contained within. It is recommended that any persons who wish to act upon this report consult with their Morgans investment adviser before doing so. Those acting upon such information without advice do so entirely at their own risk. This report was prepared as private communication to clients of Morgans and is not intended for public circulation, publication or for use by any third party. The contents of this report may not be reproduced in whole or in part without the prior written consent of Morgans. While this report is based on information from sources which Morgans believes are reliable, its accuracy and completeness cannot be guaranteed. Any opinions expressed reflect Morgans judgement at this date and are subject to change. Morgans is under no obligation to provide revised assessments in the event of changed circumstances. This report does not constitute an offer or invitation to purchase any securities and should not be relied upon in connection with any contract or commitment whatsoever.

Disclosure of interest Morgans may from time to time hold an interest in any security referred to in this report and may, as principal or agent, sell such interests. Morgans may previously have acted as manager or co-manager of a public offering of any such securities. Morgans affiliates may provide or have provided banking services or corporate finance to the companies referred to in the report. The knowledge of affiliates concerning such services may not be reflected in this report. Morgans advises that it may earn brokerage, commissions, fees or other benefits and advantages, direct or indirect, in connection with the making of a recommendation or a dealing by a client in these securities. Some or all of Morgans Authorised Representatives may be remunerated wholly or partly by way of commission.

Regulatory disclosures Analyst owns shares in the following mentioned company(ies): -

Recommendation structure For a full explanation of the recommendation structure, refer to our website at http://www.morgans.com.au/research_disclaimer

Research team For analyst qualifications and experience, refer to our website at http://www.morgans.com.au/research-and-markets/our-research-team

Stocks under coverage For a full list of stocks under coverage, refer to our website at http://www.morgans.com.au/research-and-markets/company-analysis/ASX100-Companies-under-coverage and http://www.morgans.com.au/research-and-markets/company-analysis/EX-100-Companies-under-coverage

Stock selection process For an overview on the stock selection process, refer to our website at http://www.morgans.com.au/research-and-markets/company-analysis

www.morgans.com.au If you no longer wish to receive Morgans publications please contact your local Morgans branch or write to GPO Box 202 Brisbane QLD 4001 and include your account details. 12.12.17

18