Chapter 2.ppt of macroeconomics by mankiw 9th edition
Wall Street Mastermind Sector Spotlight - Technology (October 2023).pdf
1. CONFIDENTIAL
WALL STREET MASTERMIND
Sector Spotlight: October Recap
Sector Leads
Jagger Lambert | Media& Entertainment
James Concepcion | Media& Entertainment
Pan | Technology
Teddy Kesoglou | Technology
Avi Krishna | Healthcare
Michael Reed | Healthcare
JoeAmes | Healthcare
Project Founders
Jagger Lambert
James Concepcion
3. 3
I. Large Language Models 5
II. 5G 12
III. CyberSecurity 17
IV. Deal Coverage 23
TABLE OF CONTENTS Technology
4. 4
Large Language Models Overview Technology
I. Large Language Models 5
II. 5G 12
III. CyberSecurity 17
IV. Deal Coverage 23
5. 5
Large Language Models Overview Development & Scalability/Timeline
2017
• Google introduced a special type of deep learningarchitecture
called "transformer" through the publication "Attention isAll
You Need"
• Through Positional Encoding and Self-Attention, transformers
can analyze sequential data, sentences, words, and expressions
• After 2017, AI developers began implementingtransformersin
their designs
ELMo(94M)
GPT-1(117M)
Bert (345M)
GPT-2 (1.5B)
XLNet (340M)
RoBERTa
(345M)
Megatron-LM
(8.3B)
Turing NLG
(17B)
GPT-3 (175B)
Megatron-Turing NLG
(530B)
GPT-4 (Estimated
1T)
0.01
0.1
1
10
100
1000
2018 2019 2020 2021 2022 2023
Size
in
Billions
of
Parameters
Development & Scalability of LLMs Over Time
• ELMo is the last non-transformer-based learningarchitecture,
comprised of a collection of LSTMs and 94 million parameters
• OpenAI released GPT-1, which hasa size of 110 million
parameters,similar tothe original Transformer Model, and is
focused on Next Word Prediction
• BERTwas a breakthrough for LLMs, establishinga relationship
between itsinputsand outputs,thusbecoming bidirectional
2019
• Transformer-XL wasreleased, introducing relative positional
encoding and solving the original transformer'sissue of fixed
context length
• GPT-2 was released, similar to GPT-1, with 1.5B parameters
• Microsoft established a partnership with OpenAI, investing$1B
• NVIDIA released Megatron-LM, which mimics GPT's structure
and has a size of 8.3B parameters
• In 2020, Microsoft's TuringNLG, Google's T5, and OpenAI's GPT-
3 were released, bringingLLMsto significant scales with 175B
parameters
• Between 2021 and 2023, Megatron Turing NLG wasreleased;
ChatGPTbecame the first LLM chat-simulator open to non-
technical users.Cohere, Google, Meta, and DataBricks released
their simulators. GPT-4 is the largest LLMat 1T parameters
2018
2020 -
Present
*Sources: InsideBigData, Medium
6. 6
LLMs Trading Comps & Market Drivers
Increased
Competition
in AI
Initial attention with
OpenAI's ChatGPT
sparked market
competition
Other firms begin
competing: Google's
Bard, and Significant
Gravitas' AutoGPT
BARD <- Google’s
LLM Model in
response to the
potentialof
ChatGPT replacing
Google as a source
of information
New Competitive Forces Dominant Position Threatened
In 2019 and 2023, Microsoft
extended an OpenAI
partnership in a $10B+ deal
In return, OpenAI is running
on Microsoft's Azure
computing systems, and
shares initial profits
*Sources: CapIQ, SEC Filings
Companies
Day Close Price
($ USD)
52w High Price
($ USD)
% change of
52-Week
Shares O/S
(Millions)
Market Cap ($M
USD)
Enterprise Value
($M USD)
EV/ LTM
EBITDA
EV/EBITDA
(FY)
EV/Revenue
(LTM)
$ 125.30 $ 142.38 -12.00% 12,516 $ 1,565,127 $ 1,474,638 14.7x 12.1x 5.0x
$ 133.09 $ 145.86 -8.75% 10,334 $ 1,371,429 $ 1,473,325 17.1x 14.0x 2.7x
$ 338.11 $ 366.78 -7.82% 7,432 $ 2,506,976 $ 2,468,712 22.1x 24.2x 11.3x
$ 37.01 $ 52.88 -30.01% 9,398 $ 2,776,650 $ 2,875,44 17.9x 12.6x 4.6x
$ 170.77 $ 198.23 -13.85% 15,634 $ 2,662,353 $ 2,605,090 18.5x 20.8x 6.8x
$ 180.83 $ 241.86 -25.23% 183 $ 32,442 $ 44,968 16.1x 17.1x 3.1x
Median 17.5x 15.6x 4.8x
Average 17.7x 16.8x 5.6x
Apple Inc.
7. 7
LLMs Investment Outlook & FutureGrowth
Top Opportunities to Expand From LLMs
*Sources: Cision
$12,750
$40,810
$-
$5,000
$10,000
$15,000
$20,000
$25,000
$30,000
$35,000
$40,000
$45,000
2023 2024 2025 2026 2027 2028 2029
Market
Value
($M
USD)
How Large Language Models will Change Working in Finance
• LLMscan revolutionize patient care byanalyzing symptomsand cross-
referencingthem with vast medical databases
• LLMscan assist medical professionals to quicklyreview and summarize
the latest medical literatures
1
Healthcare
• LLMscan play a pivotal role in the finance sector by analyzingvast
amountsof market data to predict trends
• LLMscan assist in automatingtaskssuch as credit risk assessment, fraud
detection, and financial forecasting
2
Finance
• LLMscan serve aspersonalized tutors, offering studentstailored
assistance based on their learningstylesand needs
• For research, LLMscan sift through large amountsof academicliterature,
summarizingfindingsand highlightingkeyinsights, acceleratingthe pace
of discovery
3
Education
&
Research
Large Language Models (LLMs) Market Growth
Work Before
the use of
LLMs
Data Collection:
Manually gather financial
data, reports, and news
Analysis: Manually sift
through the collected
data to make predictions
Risk Assessment:
Evaluate credit risks or
investment
opportunities based on
the analyzed data and
human judgment
Research: Spend hours
investigating market
trends, historical data,
and economic
Work After
the use of
LLMs
Instant Data Collection:
LLMs quickly pull most
relevant financial data
and news for your needs
Enhanced Analysis:
LLMs provide real-time
insights from the
collected data
Data-Driven Risk
Assessment: AI-driven
evaluations for more
accurate decisions
Automated Research:
Instantly access
summarized market
trends and predictions
8. 8
Cost Comparison Analysis of Top LLMs & Specific Industry Benefits
*Sources: OpenAI, Google API, McKinsey
The division responsible for
promoting products/services
and securing customers
Advantages Practical Applications
Marketing &
Sales
Brief Description Cost Information
Uses data analytics and customer
insights to develop targeted
campaigns and increase sales
conversions
Operations
IT &
Engineering
Risk & Legal
Human
Resources
Employee
Optimization
The day-to-day management
of production and distribution
processes of a company
Streamlines supply chains, optimizes
production processes, and ensures
timely product/service delivery
The discipline concerned with
designing, developing, and
maintaining technological
systems
Implements and maintains software
and hardware systems to enhance
company operations and product
offerings
The sector managing
regulatory compliance and
potential threats to an
organization
Assesses and mitigates risks to
protect the company from potential
legal disputes and operational
disruptions
The department focused on
managing employee relations,
benefits, and organizational
development
Recruits, trains, and retains talent,
while fostering a positive company
culture and ensuring compliance with
labor laws
Strategies and tools used to
enhance employee
performance and satisfaction
Implements training programs,
feedback systems, and productivity
tools to maximize employee efficiency
and engagement
Using GPT-4
Open-
Sourced API
• Cost: $0.03 per 1K tokens for input
and $0.06 per 1K tokens for output
• Context: 8K tokens
• Note: 1,000 tokens are about 750
words
Creating
Proprietary
LLM Model
with Google
Collab Pro
• Cost: 1.) Developing a specialized model
(Falcon 7B) on Google Colab Pro: $9.99
2.) Deploying the model on-demand with a
single GPU machine: $1.006/hr
• Context: Model based on Falcon 7B, fine-
tuned on proprietarydata
• Note: Cost-efficient deployment ispossible
with a single GPU machine
Cost Comparison: GPT-4 API vs. Colab Pro (Per Tokens)
$7.2
$14.4
$21.6
$28.8
$36.0
$43.2
$50.4
$57.6
$64.8
$72.0
$1.0 $2.0 $3.0 $4.0 $5.0 $6.0 $7.0 $8.0 $9.1 $10.1
$-
$20.0
$40.0
$60.0
$80.0
80k 160k 240k 320k 400k 480k 560k 640k 720k 800k
Price
of
Tokens
Total Tokens Used
GPT-4 API Cost (USD)
Google Colab Pro LLM Cost (USD)
9. 9
Imaging
Feature
OpenAI Updates
New AI
Capabili
ties
OpenAI is currently considering in-
house chip manufacturing amid the
global shortage. Some options its
considering is an acquisition,
forging its ties to NVIDIA, or
diversifying its chip production
Expand
to Other
Areas
Race to Capture Market Share in Booming AI Market
March 21, 2023: Google launched Bard in order
to compete with OpenAI's ChatGPT generative AI
chatbot
September 19, 2023: Google released updates to
Bard (it can now talk in multiple languages, fact
check, and automate tasks in Google’s suite of
tools like YouTube and Google Drive)
ChatGPT is beginning to roll out new
capabilities in both voice and image
recognition and production. OpenAI
plans to offer a new type of interface
which will allow casual voice
conversations with ChatGPT, as well as
showing it visual cues
New imaging features are very similar
to what Google's Bard already has
rolled out, but it will enable users to
show ChatGPT images to help interpret
what is being shown and how to
interact with objects in the real world
Meta Updates
Race for Consumer Facing AI
May 4, 2023: Microsoft released Bing AI powered
by OpenAI and its own Prometheus model in
response to Google launch of Bard
September 21, 2023: Microsoft launched Microsoft
Copilot to integrate AI into Microsoft Office,
allowing for easier workflows
Oct 16, 2023: Baidu Inc’s founder, Robin Li,
declared that his company’s large language model
(Ernie 4) can rival OpenAI’s advanced GPT-4.
Ernie is infused into Baidu’s products (search,
maps) and Ernie chatbot had surpassed 45 million
users
Recent Developments in LLMs Artificial Intelligence
February 2023: Meta introduced LLaMA in
2023, a Large Language Model system. The
smallest model, 7B, is trained on 1 trillion
tokens
September 27, 2023: Meta released own
chatbot (Meta AI) and a cast of celebrity-
based AI characters
Meta also introduced AI Studio, platform for
developers to build third-party AI for their
specific brands and needs
September 27, 2023: Meta added additional
functionalities to Emu, its image generation
model. People can now edit photos and create
custom stickers in Instagram
*Sources: Bloomberg, Meta
10. 10
Popular Language Models Comparison
Model GPT-3 BLOOM LLaMA LLaMA-2 T5 PaLM
Developer OpenAI BigScience META GOOGLE
Model Size (# Parameters) 175B 175B 65B 70B 11B 540B
Training Data (# Tokens) 300B 350B 1.4T 2T 34B 795B
Training Compute (FLOPS)2
3.20E+23 3.70E+23 9.90E+23 1.50E+25 2.20E+2 2.60E+24
Processor
Manufacturer NVIDIA NVIDIA NVIDIA NVIDIA GOOGLE GOOGLE
Type GPU GPU GPU GPU GPU GPU
Model V100 A100 A100 A100 TPU v3 TPU v4
Processor Hours 3,552,000 1,082,990 1,770,394 3,311,616 245,760 8,404,992
Grid Carbon Intensity
(kgCO2e/KWh)
0.429 0.057 0.385 0.423 0.545 0.079
Data Center Efficiency (PUE) 1.1 1.2 1.1 1.1 1.12 1.08
Energy Consumption (MWh) 1,287 520 779 1,400 86 3,436
Carbon Emissions (tCO2e) 552 30 300 593 47 271
Popular Large Language Model Comparison1
Commentary
▪ Google’s PaLM is the most comprehensive
model, since it has the biggest modelsize &
training data
▪ PaLM, being the largest model, exhibits oneof
the highest energy usage & carbon emissions.
▪ NVIDIA is the only independentmanufacturer
of processorsin these advanced LLMs
▪ Meta & Open AI has the most data center
efficient models
Utilizing
closed-source
models (latest
GPT-4 model
from OpenAI
via API access)
Building model
based on a pre-
trained open-
source model &
hosting it within
own IT
infrastructure
Ways to
Build Large
Language
Models
Source: 1: ‘Reducing the Carbon Footprint of Generative AI’ by Boris Gamazaychikov
2: floating point operations persecond(Speed which computercan docomplexcalculations)
API Access On-Premise
Total Cost
Breakdown
Project Setup & Inference
1
2
3
Maintenance
Other Misc. Costs (Salary)
12. 12
The Timeline & Evolution of 5G explained
Speed – 5G at its best is
ultra-fast. Everyday tasks
will feel super-fast like
watching videos without
having to wait for it to load
Capacity - 5G can handle a
massive number of devices
connected simultaneously.
Which is important in
urban settings with lots of
users connected
Low Latency - very low
latency, which means there
is minimal delay in data
transmission. This means
calls, and video calls feel
more natural, without lag
Coverage - range is
somewhat limited
compared to older
technologies like 4G. To
provide widespread
coverage, 5G networks
require more infrastructure
Everyday uses – it isn't
just about faster
smartphones. It opens the
door to new applications
and technologies, including
augmented and virtual
reality, smart cities,
connected vehicles
Early
2010-2015
• Groundwork for 5G begins with research and development.
Various companies explored for fasterwireless networks
• International Telecommunication Union (ITU) set the official
standards for 5G technology, defining the requirements
2016-2018
• Field trials and testing of 5G begins around the world. This starts
implementations of practical applications
• Commercial 5G deployments began in certain countries, with very
limited coverage in major cities. The focus was initially on
providing faster mobile broadband connectivity
2019
• 5G coverage expanded to more regions, and more compatible
devices, like smartphones and hotspots, became available. The
technology began to be a bit more adopted
2020
• Challenges arose as the pandemic caused supply chain issues
specifically with computer chips required to make the antennas
operate. Some early use cases like fixed wireless access and
enhanced mobile services started to gain traction
2021-2023
• More advanced use cases and applications, such as Internet of Things
(IoT)and industrial applications
• 5G networks became more prevalent in many countries, with broader
coverage and more devices supporting the technology. Smart cities
and autonomous vehicles began to see real-world testing
13. 13
5G Trading Comps
1. Stand-alone 5G 2. High revenue
3. Artificial
intelligence
4. IOT 5. Cybersecurity
A 5G core and radio
allow the 5G stand-alone
network to fully utilize
the capabilities of the
most recent cellular
technology
5G service revenues
will reach $315 billion
by 2023. This growth’s
primary driver is
cellular subscription
upgrades to 5G
networks
Deep learning AI will
significantly cut energy
usage and enhance
performance when old
wireless methods are
replaced
A vital component of
these technologies is
massive multiple-input
multiple-output, which
increases cell towers’
capacity
5G cybersecurity
delivers increased
network safety through
international mobile
subscriber identity
Stand-alone
5G
High
revenue
Artificial
intelligence
IOT
Cyber
Security
Companies
Day Close Price
($ USD)1
52w High Price
($ USD)
% change of
52-Week
Shares O/S
(Millions)
Market Cap ($M
USD)
Enterprise Value
($M USD)
EV/ LTM
EBITDA
EV/EBITDA
(FY)
EV/Revenue
(LTM)
$ 4.46 $ 6.82 -34.60% 99 $ 7,351 $ 6,985 NM NM NM
$ 3.31 $ 3.34 -0.90% 5,533 $ 17,037 $ 15,992 4.7x 4.7x 0.7x
$ 108.99 $ 139.94 -22.12% 1,116 $ 120,338 $ 127,151 9.5x 9.7x 3.3x
$ 0.60 $ 3.97 -84.89% 97 $ 68 $ 234 N/M N/M .87x
$ 1.70 $ 2.38 -28.57% 84 $ 141 $ 171 8.4x 6.6x 0.5x
Median 8.4x 6.6x 0.7x
Average 7.5x 7.0x 1.3x
*Sources:CapIQ
1. StockPrice as of 10/31/2023
14. 14
5G Industry Overview
5G Industry Overview
•Elaborate
• 5G is expected to become the backbone of the digital
economy, enabling faster data speeds and low-latency
connectivity. This technology has the potential to disrupt
industries like IoT, autonomous vehicles, healthcare, and
more
As industries adopt enterprise solutions more and more, 5G enterprises need 5G
to stay competitive, improve operational efficiency, support remote work, and
leverage emerging technologies like IoT, AR/VR, and edge computing. The biggest
company currently in the market is China Mobile Ltd. At 185.72 Billion
Investment Factors
• 5G networks requires significant infrastructure
development. Companies involved in building out the 5G
infrastructure like cell towers and antennas, such as cell
tower operators and network equipment manufacturers
2
Infrastructure
Development
• 5G technology relies heavily on advanced semiconductor
components. With semiconductor manufacturers producing
chips for 5G devices, such as smartphones, IoT devices, and
base stations. The chip industry will play a crucial role for
5G
3
Semiconductor
Industry
*Sources: Precedence Research, Statista
$3.1
$34.6
$0.0
$5.0
$10.0
$15.0
$20.0
$25.0
$30.0
$35.0
$40.0
2021 2022 2023 2024 2025 2026 2027 2028 2029 2030
5G Enterprise Value Market Size ($B USD)
Top 5G Providers by MarketCap ($B USD)
$142.5
$109.0
$165.7
$1,361.0
$101.9
$177.6
Verizon AT&T T-Mobile
China Mobile Deutsche Telekom Comcast
1
Growth
Potential
15. 15
Top Opportunities of Growth From 5G
• Provides secure wireless network access across campuses and
homes
• Enables IoT sensors to track classroom attendance, study room
availability, and public transportation
1
Education
• Improves fleet tracking and efficiency, optimizing supply chains
and reducing costs
• Provides real-time data for dynamic insights and accurate
inventory management, enhancing customer satisfaction and
reducing waste
2
Transportation
• Provides reliable and high-speed network access across large
areas, enhancing the guest experience and operational efficiency
• Supports a large number of devices, even during peak times,
ensuring uninterrupted service for guests and staff alike
3
Entertainment
• Pioneering Companies:
• Amazon, Microsoft, and Google: These tech giants have been
consistently pushing the boundaries of what's possible in both cloud
computing and telecommunications
• At the Forefront: Their investments and research into 5G and cloud
integration set them apart from the competition
• Key Players: Their vast infrastructure, combined with their
technological prowess, positions them uniquely. They're not just
adopting 5G; they're shaping its future
Leaders in 5G and Cloud Synergy Cloud's Role in 5G Deployment& Key Benefits
• Foundational to 5G:
• Cloud infrastructure makes 5G networks adaptable and scalable
• Dynamic allocation of resources for efficiency
• Strategic Collaborations:
• Major telecom operators partnering with cloud providers
• Collaborations ensure a blend of telecom and cloud services
• Benefits of 5G via Cloud:
• Network Efficiency: Optimized resource utilization and reduced costs
• Peak Performance: Real-time data processing and faster delivery
• Low Latency: Instant communication and minimized delays
5G Areas for Growth & News The Evolution of 5G
Downlink Speeds by Technical Generation (mbps)
42 100 150
1,000
10,000
3.5G / DC-
HSPA+
4G/LTE 4G/LTE Cat. 4 4G/LTE
Advanced
5G
Downlink Speeds are increasing
Exponentially with every technical
generation
17. 17
21 20
25 23
30
25
32
39
66
104 105
0
20
40
60
80
100
120
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
#
of
Cyber
Secruity
Attacks
Cyber Security Industry Overview
Cyber Security Overview
1
Cyber Security
2
Cyber Attacks
Types
3
Preventable
Measures
• CyberSecurity:Safeguarding of computer systems, networks and information
from cyber-attacks(maliciousefforts to compromise sensitive information, often
to extort money from users or disrupt business operations)
• Malware:Softwarethat infectand damage devices or steal important info
• Phishing:Misleading attempts to trickpeople into revealing important
information through the means of emails and phony websites
• Firewalls:Protectivebarriersthat prevent access to unauthorized users
• Antivirussoftware:Softwaretodetect and remove malware
• PasswordProtocols:Strongcompany-widepassword policies
Timeline of Cyber Security Developments
Cyberattack Incidents with >$1M in Reported Losses
2011: Hackers
broke into
PlayStation
network to steal
personal info from
millions of users,
forcing Sony to
shutdown their
networks for weeks
2018: Facebook data compromised,
resulting in exposure of personal data
for 50 millions users
2019: Singapore’s Healthcare system
breached >80 times (the worst
cyberattack hit country in the world)
1960s
1980s
1990s
2000s
2010s
*Sources: InfoSec, Google
• The U.S. Department of Defense launched ARPANET, the
precursor to the modern internet. By the early '70s, the
first computer worm, Creeper, emerged, prompting the
creation of Reaper, an early antivirus solution
• Massive data breaches, sophisticated ransomware
attacks, and state-sponsored cyber threats have
dominated recent years. Concurrently, advancements in
technologies, such as AI, have been harnessed to
strengthen cybersecurity measures
• The scale and sophistication of cyberattacks increased
significantly. The world grappled with threats like the
ILOVEYOU virus and the rise of botnets. State-
sponsored cyberattacks marked for cyberwarfare
The advent of the World Wide Web brought new security
vulnerabilities. Incidents like the Morris Worm highlighted
these challenges, but the decade also saw the
establishment of essential security protocols, such as SSL
• The first computer viruses for personal computers
appeared, with "Brain" emerging in 1986. The decade
also witnessed the rise of hacker culture and the
realization of the need for stronger security measures
18. 18
Cyber Security Niche Industry Players 12-Month Stock Performance Overview
*Sources: CapIQ
-50.00%
-40.00%
-30.00%
-20.00%
-10.00%
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
Share
Price
Percentage
Change
(%)
CrowdStrike Holdings, Inc. (NasdaqGS:CRWD) - Share Pricing Rapid7, Inc. (NasdaqGM:RPD) - Share Pricing
Okta, Inc. (NasdaqGS:OKTA) - Share Pricing Palo Alto Networks, Inc. (NasdaqGS:PANW) - Share Pricing
Tenable Holdings, Inc. (NasdaqGS:TENB) - Share Pricing CyberArk Software Ltd. (NasdaqGS:CYBR) - Share Pricing
Zscaler, Inc. (NasdaqGS:ZS) - Share Pricing Fortinet, Inc. (NasdaqGS:FTNT) - Share Pricing
Oct-Nov 2022: Cybersecurity stocks
moved together due to overall market
volatility, concerns about a slowdown in
cybersecurity spending, increased
competition, and potential consolidation
Sep 2023: Cybersecurity stocks could trend together in
September 2023 due to the release of a new major
cybersecurity report that highlights the growing threat
of cyberattacks and the need for businesses to invest in
cybersecurity solutions
19. 19
The Global Cyber Security Market Size is projected to grow from
$172.32Billionin2023to$424.97Billion in 2030
Drivers
153.7
172.3
200.9
239.2
278.3
317.2
359.0
391.3
425.0
$0
$50
$100
$150
$200
$250
$300
$350
$400
$450
2022 2023 2024 2025 2026 2027 2028 2029 2030
Market
Size
in
USD
Billions
Cyber Security Forecasted Market Size Growth Drivers
1
Enterprise
Solutions
As industries adopt enterprise solutions, the need for Cyber Security in areas
such as manufacturing, banking, financial services and insurances will drive
the growth of this industry
Banking, Financial Services, Insurance (BFSI) holds the largest share and is
expected to rise due to the importance of security across financial,
banking and insurance institutions
2
BFSI
3
Healthcare
The healthcare sector is estimated to see considerable growth as well with
the adoption of digital customer healthcare records. Implementation of
Cyber Security protocols will be needed to ensure confidential data is not
leaked
Cyber Security Deep Dive Trading Views, Market Potentials, and Growth Engines
*Sources: Statista
1. StockPrice as of 10/31/2023
Companies
Day Close Price
($ USD)1
52w High Price
($ USD)
% change of
52-Week
Shares O/S
(Millions)
Market Cap ($M
USD)
Enterprise Value
($M USD)
EV/ LTM
EBITDA
EV/EBITDA
(FY)
EV/Revenue
(LTM)
$ 134.25 $ 138.64 -3.17% 117 $ 15,425 $ 12,436 13.4x 11.4x 5.2x
$ 52.13 $ 58.19 -10.41% 4,051 $ 208,886 $ 192,004 10.7x 11.2x 3.4x
$ 144.64 $ 153.21 -5.59% 911 $ 129,937 $ 177,559 12.4x 13.4x 2.9x
$ 176.77 $ 191.99 -7.93% 239 $ 41,695 $ 39,350 NM NM NM
$ 57.17 $ 81.24 -29.63% 785 $ 44,230 $ 41,981 33.7x 27.4x 8.4x
$ 20.26 NM NM 310.8 $ 74,145.1 $ 74,086.0 NM NM NM
Median 17.5x 15.9x 5.0x
Average 12.9x 12.4x 4.3x
20. 20
22.13 22.00
18.36
20.67
17.59
14.68
18.00 17.72
24.95
35.00
31.00
0
5
10
15
20
25
30
35
40
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Valuations
(Billions
USD)
Post Deal Valuations
Evolving Deal Landscape
Median Post-Valuation & Deal Count
Commentary
Strategic M&A and sponsor
activityare driving overall
deal flow and valuations
M&A and sponsor activityare
consistentlygrowing in the
Cyber Security space
Overall deal making volume &
medianvaluations have been
trending upward in the past
decade
119 129 125
156 152 147
169 182 188
298 287
0
50
100
150
200
250
300
350
400
450
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Merger & Acquistions Secondary Buyout Buyout IPO Other
Deal Count by Exit
*Sources: Pitchbook
Number of Deals
21. 21
Cyber Security Market Trends Expansion, Innovation, and Investment
*Sources: Statistia
Energy & Utilities,
7.83%
Healthcare, 11.92%
BFSI, 32.61%
IT & Telecom,
9.82%
Retail & Wholesale,
8.23%
Government,
24.10%
Other, 5.50%
Energy & Utilities Healthcare BFSI IT & Telecom
Retail & Wholesale Government Other
Credit Card Fraud,
4.07%
IdentityTheft,
4.95%
Investment Fraud,
5.41%
Tech Support,
5.76%
Extortion, 6.98%
Non-Payment /
Non-Delivery,
9.16%
Personal Data
Breach, 10.43%
PhishingScams,
53.24%
Credit Card Fraud Identity Theft Investment Fraud
Tech Support Extortion Non-Payment / Non-Delivery
Personal Data Breach Phishing Scams
Estimated Cost of CyberCrime ($B USD) as of Sep 2023
Share of Cyber Attack Types(2022)
Cyber Security Revenue Industry Share (2023)
$610
$13,820
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
2016 2018 2020 2022 2024 2026 2028
Data privacy
regulations are
becoming more
stringent,
prompting
organizations to
invest in
cybersecurity to
comply with
new regulations
The increasing adoption of digital
technologies is expanding the attack
surface and making organizations
more vulnerable to cyberattacks
Cybersecurity
market is ruled
by a few main
players, but
there is still
room for VC-
backed startups
with innovative
solutions to
disrupt market
Cyberattacks are becoming more
sophisticated and costly, driving
demand for cybersecurity solutions
Rising IoT gadgets, fueled by remote
work, create more opps for
cybercrime
Growth Opportunities
Increasing Threat
23. 23
Industry Overview
Dealflow drivers
Capital Investment & Deal Count IT Services MarketSize (Trillions USD)
The cybersecurity mesh concept essentially seeks to develop security protocols that can
prevent exposure at every single entry point of a network
As more data is processed and stored each day, it is becoming difficult to process data that is
located in far-off data centers. Edge computing looks to solve this by using smart objects and
network gateways to process and store data near the end user’s location
ICD predicted that in 2021, 80% of enterprises would switch to a cloud-based IT infrastructure
twice as fast as they did before COVID-19
$61
$48
$346
$55
$208
$65
Vmware Dell Broadcom
Arista Networks Cisco Huawei
Market Share (Market Cap in Bn USD)
Cybersecurity
Data Storage
Cloud
24. 24
Cisco Company Overview Basic Financial Overview
*Sources: CapIQ
$36,709
$39,005
$35,978 $36,014 $38,018
$43,142
$12,621
$12,899
$13,323 $13,856 $13,539
$13,804
$49,330
$51,904
$49,301 $49,870 $51,557
$56,946
2018 2019 2020 2021 2022 2023
Product Service Total Revenue
$25,569
$20,473
$14,583
$11,526
$8,915 $8,391
$14,909 $18,613
$18,458 $18,564 $19,302
$20,412
2018 2019 2020 2021 2022 2023
Debt Balance EBITDA
Cisco Company
Overview
Inception
Business
Segments and
End Markets
Product Lines
Cisco specializes in designing & selling networking
hardware, software & telecommunications equipment
Founded in 1984 by 2 Stanford University computer
scientists pioneering Local Area Network concept
1. Product Segment: Infrastructure Platform,
Applications, Security and Other Products
2. Services: maintenance and support
Sell across Americas, AMEA and APJC
Cisco Catalyst Switches (creating & managing LANs)
Cisco Routers, ASA, Webex, Meraki, UCS, IP Phones
Short-term
Assets
$ 10.1 Billion in Cash & Cash Equivalents
$ 16.0 Billion in Investments
$29,105
$5,306
$4,052
$3,859
$811
Secure, Agile Networks Internet for the Future
Collaboration End-to-End Security
Optimized Application Experiences Other Products
2023 Revenue Breakout(in $M) by Geography and Product
Steady Revenue Growth Driven by Product Segment
Strong Balance Sheet (Low Debt & Growing EBITDA)
Company
Overview
$33,447
$15,135
$8,417
Americas
EMEA
APJC
Figures in $M
Figures in $M
26. 26
Cisco Acquiring Splunk Overview Deal breakout , Synergies, and Other Conditions
• Cisco and Splunk Inc. announced the acquisition of Splunk in late September this year (21st September, 2023)
• Cisco acquires Splunk for $157 per share in cash (31% premium to company's last closing share price), which is approximately $28 billion in equity value
• The deal is expected to close by the end of the 3rd quarter of the calendar year 2024, subject to regulatory approval and other customary closing conditions,
including approval by Splunk shareholders
• Tidal Partners LLC is acting as the financial advisor to Cisco, and Qatalyst Partners and Morgan Stanley & Co. LLC are acting as financial advisors to Splunk
Revenue
Synergies
Cost
Synergies
Financial
Synergies
Operational
Synergy
• Cross-Selling: Tap into Cisco's vast customer base
• Bundling: Integrate Splunk's software with Cisco's
offerings
• Market Reach: Utilize Cisco's global footprint for
Splunk's expansion
• Operational Efficiencies: Eliminate redundancies post-
integration
• R&D Consolidation: Streamline overlapping projects
• Supply Chain: Leverage Cisco'sprocurement processes
for cost savings
• Borrowing Rates: Benefit from Cisco's stronger financial
stature
• Capital Allocation: Use Cisco's resources forSplunk's
strategic initiatives
• Unified Solutions:Integrate Splunk's analytics into Cisco's
products
• Best Practices: Share operational & development strategies
• Talent & Expertise: Combine the expertise of both
companies for innovation
Cisco Splunk Deal Breakdown
Potential Synergy Opportunities
*Sources: Splunk., Cisco,CapIQ
-80.00%
-60.00%
-40.00%
-20.00%
0.00%
20.00%
40.00%
60.00%
80.00%
Cisco Systems, Inc. (CSCO) Splunk Inc. (SPLK)
Tracking CSCO Vs. SPLK Share Price Percentage Change (%)
Deal Announcement
(21st Sep)
27. 27
Splunk Overview Financial Breakout
35.30%
25.19%
39.51%
2022
39.89%
18.48%
41.63%
2023
• Positive Revenue Growth: There's a consistent upward trajectory in
top-line growth, especially a significant increase in FY23
• Improved Cost Management: The gross margin rebounded in FY23,
indicating better cost management relative to revenue
• Stable Operating Efficiency: Despite the rise in revenue, operating
expenses are stable FY23, suggesting improved operational efficiency
• Continued Profitability Concerns: Even though losses have
narrowed, the company has yet to achieve profitability
• Potential Solvency Issues: The negative equity in FY23 signals
potential financial challenges and solvency concerns for the company
Company
Overview
Splunk
Overview
Inception
Main Services
and Business
Activities
Main
Customers
Software company that analyzes machine-generated big
data for insights to improve operational intelligence &
security
Founded in 2003 by Rob Das and Eric Swan. The
current CEO of Splunk as of 2022 is Gary Steele
Focuses on capturing, indexing, and analyzing
machine-generated big data to deliver operational
intelligence, security insights, and business analytics
to organizations
Cisco, IBM, Adobe, Porsche, Slack, Visa, Dominos,
Lenovo, BookMyShow, Telenor, Okta, Papa John’s
Key Takeawaysfrom FY21 to FY23 Financials
End Markets
IT Services, Computer Software, Financial Services,
Cybersecurity, Artificial Intelligence Related Software
*Sources: CapIQ, SplunkInc. 10-K
Revenue Breakout 2022 - 2023
Income Statement Historical:
($M USD) FY21 FY22 FY23
Total Revenue: 2,229.4 2,673.7 3,653.7
Revenue Growth (5.5%) 19.9% 36.7%
(-) Cost of Sales (547.3) (734.0) (816.0)
Gross Margin: 1,682.0 1,939.7 2,837.7
Gross Margin % 75.4% 72.5% 77.7%
(-) Total Operating Expenses: (2,462.2) (3,086.5) (3,063.2)
Operating Income (EBIT): (780.2) (1,146.8) (225.5)
EBIT Margin: (35.0%) (42.9%) (6.2%)
EBITDA (686.5) (1,047.7) (126.0)
EBITDA Margin: (30.8%) (39.2%) (3.4%)
Net Income to Parent: (908.0) (1,339.1) (277.9)
Net Income Margin: (40.7%) (50.1%) (7.6%)
Balance Sheet
Total Assets: 5,868.5 5,790.9 6,343.9
Total Liabilities: 4,274.5 5,568.1 6,454.4
Total Equity: 1,594.0 222.8 (110.5)
KPIs
Return On Assets: N/A -23.12% -4.38%
28. 28
Cisco & Splunk Deal Overview
Strategic
Rationale
Successful
Acquisition/
Integration
Impact on
Cybersecurity &
AI
Deal
Implications
Overview Description
Significant
Revenue
Opportunities &
Margin Expansion
▪ Expand security portfolio, improve observability capabilities for more complete view of IT infrastructure
▪ Use Splunk’s data analytics and data in SEIM market to improve own product suite
▪ Splunk aid Cisco’s expansion into hybrid, multi-cloud markets and transition to a more recurring revenue
business
▪ Combined expertise to develop AI-powered security & observability solutions (LLM Models for Gen AI
use)
▪ Splunk aid Cisco’s transition (expected to add $4 billion in recurring rev)
Splunk Complements zero trust portfolio (Splunk’ SOAR or ES acts as Policy Decision Point that Cisco lack)
Obstacles still
Ahead
▪ The ThousandEyes, App Dynamics, and Splunk Observability cloud require skillful product management
to craft together into a coherent product roadmap that services the cloud and on-premises needs of the
combined company’s customer base
▪ Address skepticism from clients that Cisco will degrade quality of Splunk’s SIEM
▪ Learn from previous failures of security acquisitions (Monterey Networks and Pirelli Optical) in 1999 and
recent success of Duo Security acquisition (was able to maintain brand independence
▪ Deal is still subjected to regulatory approval
▪ Validates importance of AI in Cyber Security (Cisco’s $28 billion purchase price indicate belief AI is
essential to Cyber Security in the future)
▪ Enhance Cisco’s strategic position in the AI Cyber Security market. With Splunk's AI capabilities, Cisco
will be able to offer a more comprehensive suite of security solutions to its customers
▪ Spur other large tech companies (Microsoft, Google, and IBM) to invest/increase investment in AI Cyber
Security
▪ Suggest growing importance of startups in the AI Cyber Security market
▪ While funding for Cyber Security startups is down overall, funding for AI Cyber Security startups remains
strong despite decreased funding for Cybersecurity startups-> AI Cyber Security is a key growth area for
the Cyber Security market
Indications of AI
Market
Attractiveness
29. 29
Splunk’s Public Comp’s Forward Multiples
6.4 x
10.8 x
5.6 x
2.7 x
9.4 x
4.0 x
6.7 x
5.4 x
8.3 x
4.9 x
2.6 x
7.9 x
3.5 x
6.0 x
0.0 x
2.0 x
4.0 x
6.0 x
8.0 x
10.0 x
12.0 x
TEV
/
Revenue
Multiples
EV / FY24 Revenue EV / FY25 Revenue
57.5 x
50.7 x
28.5 x
10.7 x
34.9 x
23.5 x
27.8 x
43.2 x
36.3 x
22.0 x
10.1 x
29.4 x
18.2 x
24.0 x
0.0 x
10.0 x
20.0 x
30.0 x
40.0 x
50.0 x
60.0 x
70.0 x
TEV
/
EBITDA
Multiples
EV/FY25 EBITDA
45.9 x 48.1 x
13.5 x
43.2 x
29.5 x
34.4 x
50.8 x
34.0 x
37.7 x
12.8 x
36.6 x
24.3 x
29.5 x
0.0 x
10.0 x
20.0 x
30.0 x
40.0 x
50.0 x
60.0 x
P
/
E
Multiples
EV / Revenue Multiples for 2024 & 2025
EV / EBITDA Multiples for 2024 & 2025
P/E Multiples for 2024 & 2025
*Sources: CapIQ
71.1 x
30. 30
Discounted Cash Flow for Splunk
*Sources: Splunk10K
Historical: Projected:
Splunk' FCF Projections: Units: FY21 FY22 FY23 FY24 FY25 FY26 FY27 FY28
Revenue: $ M $2,229.4 $2,673.7 $3,653.7 $3,959.0 $4,421.0 $4,952.0 $5,759.0 $6,526.0
Revenue Growth: % – 19.9% 36.7% 8.4% 11.7% 12.0% 16.3% 13.3%
(-) Cost of Sales: $ M (547.3) (734.0) (816.0) (884.2) (987.4) (1,105.9) (1,286.2) (1,457.5)
(-) Operating Expenses: $ M (2,462.2) (3,086.5) (3,073.2) (2,236.2) (2,450.2) (2,684.7) (2,862.3) (3,241.2)
Operating Income (EBIT): $ M (85.0) (220.2) 644.1 838.6 983.4 1,161.3 1,610.6 1,827.3
Operating Margin: % (3.8%) (8.2%) 17.6% 21.2% 22.2% 23.5% 28.0% 28.0%
(-) Taxes, Excluding Effect of Interest: $ M 17.0 44.0 (128.8) (167.7) (196.7) (232.3) (322.1) (365.5)
Net Operating Profit After Taxes
(NOPAT):
$ M (68.0) (176.2) 515.2 670.9 786.8 929.1 1,288.4 1,461.8
Adjustments for Non-Cash Charges:
(+) Depreciation & Amortization: $ M 99.5 99.1 99.5 95.5 96.0 113.9 132.5 143.6
% Revenue: % 4.5% 3.7% 2.7% 2.4% 2.2% 2.3% 2.3% 2.2%
(-) Deferred Income Taxes: $ M (33.0) (76.0) (34.0) (53.6) (69.1) (6.2) (5.8) (8.0)
% Income Taxes: % 476.1% 415.0% 274.0% (274.0%) (268.0%) (20.0%) (15.0%) (15.0%)
(+/-) Changes In Operating Assets &
Liabilities:
$ M 1,010.0 223.8 (296.4) (815.4) (910.5) (1,019.9) (1,186.1) (1,344.1)
% Revenue: % 45.3% 8.4% 8.1% 20.6% 20.6% 20.6% 20.6% 20.6%
% Change in Revenue: % – 50.4% (30.2%) (267.1%) (197.1%) (192.1%) (147.0%) (175.2%)
(-) Capital Expenditures: $ M (37.1) (10.7) (13.6) (16.6) (22.9) (36.3) (23.0) (28.0)
% Revenue: % 1.7% 0.4% 0.4% 0.4% 0.5% 0.7% 0.4% 0.4%
Unlevered Free Cash Flow: $ M 971.3 60.1 270.7 (119.2) (119.7) (19.4) 206.0 225.3
Growth Rate: % – (93.8%) 350.3% (144.0%) 0.5% (83.8%) (1159.6%) 9.4%
Terminal Value - Multiples Method:
Median CY25 TEV / EBITDA of Comps: 25.7 x
Median CY25 TEV / EBITDA of Target: 24.0 x
BaselineTerminal EBITDA Multiple: 25.0 x
BaselineTerminal Value: $ 49,271.3
Implied Terminal FCF Growth Rate: 8.3%
(+) PV of Terminal Value: 32,383.0
(+) PV of UFCFs: 312.8
Implied Enterprise Value: 32,695.7
% of Implied TEV from Terminal Value: 99.0%
(+) Cash & Investments: (2,757.4)
(+) Net Operating Losses: -
(-) Debt & Finance
Leases: -
(-) Preferred Stock: -
(-) OperatingLeases: -
(-) NoncontrollingInterests: 0.6
(-) Unfunded Pensions: -
Implied Equity Value: 29,938.9
Diluted Shares Outstanding: 168.520
Implied Share Price from DCF: $ 177.66
Premium / (Discount) to Current: 20.5%
Unlevered Free Cash Flow Calculations Terminal Value & Share Price
31. 31
$160.54
$166.80
$133.67
$140.94
$141.66
$156.33
$37.22
$117.16
$120.04
$123.59
$229.50
$194.23
$187.11
$203.77
$243.10
$280.75
$61.62
$207.22
$219.08
$242.62
$- $50.00 $100.00 $150.00 $200.00 $250.00 $300.00
Variable
Per Share Value
Assumptions
Trading
Comps
EV /
Revenue
LTM
Comp Set:
1.Elastic N.V
2.Datadog, Inc
3.New Relic, Inc
4.IBMInc
5.Dynatrace, Inc
6.PagerDuty, Inc
CY 24
CY 25
Trading
Comps
EV /
EBITDA
LTM
CY 24
CY 25
Trading
Comps
P / E
LTM
CY 24
CY 25
M&A Comps
LTM EV /
Revenue
• Used 3 Transactions in the Last 12
Months in the Software Space
DCF (No
Synergies)
Multiple
Method
• WACC of 8.76% - 11.96%
• EBITDA Multiples of 24.9x – 27.0x
Cisco's offer price for Splunk ($157) is fair according to Public Comps, Precedent Comps & our DCF
CurrentSharePrice:$147.49 Offer Price:$157.0
*Sources: Splunk10K.
Additional notes: precedent comps used transactions of SilverLake Technology Management andQualtrics International, Cloud Software GroupInc & Citrix Systems, TPGCapital & New Relic Inc