Murata Manufacturing (6981): The Best Way To Play MLCC Demand Surge

, ,

Murata is the world’s largest producer of multi-layer ceramic chips (MLCCs) with over 40% market share. The company cranks out over 150 billion MLCC pieces per-month. The next closest peer, Samsung Electric (SEMCO) produces 100B pieces per-month.

Here’s why we’re interested in Murata now: The company’s main revenue source (MLCCs) will see a demand explosion from two sources: electric vehicles and 5G telecommunications.

The company owns the entire manufacturing process from start to finish, vertically integrating the supply of raw materials, sheet-casting, sintering, processing and finishing and inspection/packaging. They do all this in-house, allowing them to process, create and deliver customer orders faster than competition.

Murata’s dominant market share position allows it to invest more money in R&D than its competitors while keeping R&D expenses a lower percentage of total revenue. As an example, Murata will spend $1B on R&D this year compared to SEMCO’s (MLCC competitor) $416M R&D investment.

It also doesn’t hurt that the company’s long-term stock chart is tantalizingly bullish (see below):

These facts below should get anyone excited about the coming MLCC demand boom (emphasis mine):

    • The required number of automotive MLCCs increased from 3,000 units in 2012 to 8,000 units in 2018
    • Battery EVs (BEV) require much more MLCCs than internal combustion cars. They are expected to require around 30,000 MLCCs.
    • TESLA’s Model 3 contains over 9,000 MLCCs and their Model S & X each have over 10,000 MLCCs.
    • Murata predicts the MLCC usage of smartphones in 2024 will be 1.5 times more than that in 2019.
    • For instance, 5G smartphones which support sub-6GHz frequency band will need 10~15% MLCC more than those of 4G smartphones; and 5G smartphones which support mmWave will need 30~35% MLCC more than those of 4G smartphones. In estimation, each 5G smartphone will need more than 1000 units MLCC in the future.

It’s clear MLCCs will be vastly more important in five years than they are now. And we’re going to need a heck of a lot more of ‘em. In turn, MLCC producers, most notably the world’s largest, should be worth significantly more five years from now.

If that’s not convincing enough, The Global and China Multi-Layer Ceramic Capacitor (MLCC) Industry Report (2019 to 2025) expects MLCC supply to reach 6.1T/year by 2025. That’s six times higher than today’s production levels!

This essay will review the following:

    • What is an MLCC and Why Is It Important?
    • Murata’s Founding & Culture of Innovation
    • Murata’s Growth Story: Betting on EV & 5G

By the end of the essay, you’ll know what an MLCC is, why it’s important, where future demand will surface, and why Murata is the best-positioned company to capture that demand.

And before I forget, because a number of you asked for it we went ahead and created a monthly option for the Collective. You can find that here. This order form will only be live for another day or two so if you’d like to join our group make sure to sign up soon!

Now back to our regularly scheduled programming.

What Is An MLCC and Why Are They Important?

Multilayer Ceramic Capacitors (MLCC) are devices that store energy in the form of an electric field between layers of ceramic and metal material. Think of MLCCs like an Italian sub. Each layer consists of alternating genoa salami and capicola. MLCCs are also used to differentiate between high/low frequencies.

They were born out of necessity in Germany during the 1920s. The Germans ran out of the material, Mica. Luckily, they had spare porcelain (ceramic derivative) and tried it. The rest as they say is history.

MLCCs main benefits include the best high-frequency performance of any capacitor, as well as better stability in high temperatures. They also have a high voltage threshold, meaning they can withstand heavy electrostatic discharge without damage. Both of these features come in handy with electric vehicles (we’ll discuss that later).

The main disadvantage is smaller capacitance (the ability of a system to store an electric charge) per volume. MLCCs present a tradeoff between higher capacity for temperatures and voltage at the expense of electrical charge storage.

Two classes of MLCCs: Class I and II

Class 1 capacitors work best when high stability and low loss is required.

Class 2 capacitors have a higher capacitance per volume with thermal stability of typically 15%, so are better suited for less sensitive applications (source: JJSManufacturing)

They come in surface-mounted and leaded versions. Surface mounted means they “mount” surface boards (like CPUs and processors). Leaded capacitors have an epoxy coating and wires to assist in conduction.

There’s an MLCC In Every Device You Use

Chances are you use devices that use MLCCs every single day. Ceramic capacitors are the most produced capacitors in the world. Manufacturers crank out 1 Trillion of these microscopic devices every year.

They’re used in everything from smartphones to EV batteries to bluetooths and camera sensors.

Point to an electronic product or device and odds are, you’ll find an MLCC in it.

Why Are MLCCs Important?

MLCCs are inexpensive, extremely reliable and can withstand high voltages, frequencies and temperatures. This makes them the go-to capacitor for consumer electronics, 5G telecommunications and EV components.

They’re also incredibly scalable, with the smallest MLCC the size of a grain of sand.

As more devices need higher computing power at the edge, OEMs will look for the smallest MLCCs available. The smaller the MLCC the smaller OEMs can make their devices (see: wearable smart watches).

But perhaps the biggest reason why MLCCs are important is their ability to perform in EV batteries/cars and 5G smartphones and base stations. These two categories will drive MLCC growth for the next 5-10 years.

There’s one company that dominates the MLCC market: Murata Manufacturing. But before we understand where Murata sits in the current market, we need to understand the history of the company and its cultural fabric.

The Birth of Murata Manufacturing: The Story of How One Man Turned $102K Into A $14B Empire

Akira Murata founded the company in 1944 as a personal venture. At the start it was merely a small, family-run operation.

The company originally specialized in basic capacitors to solve the needs of a small number of customers.

But Murata wanted more. He wanted to help more customers and find new business. In 1947, he asked one of his school professors, Mr. Tanaka about a new ceramic product called barium titanate. Before discovering barium titanate, Murata used titanium-oxide. Titanium-oxide boasted a dielectric constant of 70-100. Barium titanate on the other hand had a whopping 1,000 – 10,000 constant.

Murata reorganized the company three years later with $102K of his own money (adjusted for inflation). It was also at this time Murata built his first factory outside Kyoto. The story on how he found the land to build the factory is fantastic.

In 1950, Murata received a call from a former work colleague at the Kyoto Research Institute, Mr. Senda. Senda happened to be the director of a ceramic laboratory. Senda showed Murata a large plot of land with massive pottery-stone reserves (i.e., ceramics).

This moment sparked Murata’s vision of factories, employees and subsidiaries across Japan.

Today, Murata generates over $14B in revenue and $2B in operating income. All that from a $102K investment. That’s a cool 19,607% return on Murata’s initial capital. Not bad.

So how did Murata go from a single factory to 77,00 employees and 150B capacitor production per month? To answer that question we have to understand the company’s culture and philosophy.

Murata’s Founding Philosophy: Pursuit of Originality & Moving First

Electronic component manufacturers are commodity businesses. They win by offering the lowest priced product to their customers. In doing so, they hope to gain market share at the expense of operating profit margins. Make it up in volume, right?

Murata’s competitive advantage is simple: they offer products that their competitors don’t. They’ve done that since their founding in 1950. The company notes in its 2019 report: “Murata offers products that competitors do not offer, to people that need them. This marked the start of Murata’s ‘pursuit of originality,’ which has led to originality across all aspects of our business, including our technology development capabilities, manufacturing capabilities, networks and organizational cooperation to integrate these elements.”

These strong competitive advantages took over half a century to solidify. And they were built on the back of solving a customer’s specific problem. Akira was the master of solving specific problems. He did things that didn’t scale (custom-building parts vs. mass production) to win the trust and adoration of customers.

This Pursuit of Originality is best shown in the company’s R&D segment, monozukuri manufacturing process and dogged determination to do everything in-house.

By keeping everything in-house, Akira could tinker with new raw materials, diagnose different manufacturing processes and create unique orders for specific customers. All the while keeping costs down and completion times quick.

Murata’s early success allowed it to scale quickly while reinvesting most of its profits into R&D to develop new products. This paved the way for Murata’s First Mover Advantage. Something it’s kept since its founding.

A History of Moving First

Murata’s history is earmarked by first-mover advantage. From its beginnings in 1940, the company recognized technological shifts and developed products for those rising industries. Murata was first-to-market with Japan’s first mass produced temperature compensating barium titanate ceramic capacitor for radios. That was five years after the company’s founding. Yet they saw the technological shift happening in radio during WWII.

The company executed during the 50s and 60s. They developed ceramic capacitors and semiconductors for black-and-white TV in the 50s and commercialized ceramic filters during the color-TV boom in the 60s.

Next came the 70s with the CB transceiver boom in the US and expanded use of audio-visual equipment in cars. Murata launched GIGAFIL dielectric filters for microwaves, commercialized ceramic resonators, chip ferrite beads and multi-layer LC filters during the decade.

The 90s and early 2000s gave rise to the Internet, miniaturization of mobile phones and spread of PCs. This led Murata to commercialize bluetooth modules, develop 0.4×0.2mm multilayer ceramic capacitors and MEMS gyro sensors.

Today, Murata continues to innovate and develop market-leading technology.

Murata’s scale is one of the many competitive advantages it has over its peers. It’s helped the company deliver 16.5% CAGR over the last decade (362.5% return).

As we’ll see in the final section, Murata is the best-positioned company to capture the positive MLCC demand shock from Electric Vehicles and 5G telecommunications technology.

Murata Today, Murata Tomorrow

Despite operating in a cyclical industry, Murata’s financial results are impressively consistent.

Since 2004 the company’s averaged 35%+ Gross Margins, 12-18% Operating Margins and reported only one losing year, 2008. During that time they’ve grown revenue from $4B to $14B, operating income from $539M to $2.24B and FCF from ~$400M to ~$900M.

Given those stats, one would expect to pay a lofty multiple of EBIT for the entire business. Not quite. At the time of writing, Murata trades roughly 18x NTM EBIT and 3x revenues.

Murata relies on three competitive advantages to keep its top-dog spot in an otherwise commoditized industry:

1. They anticipate market changes and customer needs better than competitors

This goes back to Murata’s early days of solving the needs of its initial customers. The company still embodies this tradition by focusing on their customers’ future problems, and how they can develop technologies to meet those needs.

Murata can do this better than any company in its industry. It’s 92 subsidiaries and thousands of customers provide tremendous insight other companies can’t get. With all these data points, Murata can spot trends and potential product failures before anyone else.

2. Continuous R&D investment enables new product development and IP accumulation

Murata has increased their R&D spend every year going back to 2004. In fact, if trends continue, they’ll invest over $1B in R&D in 2021. This is a massive competitive advantage against its competitors. The sheer dollar amount of R&D capital allows Murata to develop more products and fail more often in search of what works (see graph below).

For comparison, Samsung’s Electronics business (A009150) invested $419M in R&D last year.

Murata generates (on average) 36% of its revenue from new products. And on average the company spends ~6% of net sales on R&D. That’s important because while total R&D investment looks daunting ($1B), it still represents <10% of total revenue.

TDK is the only other company that compares to Murata’s R&D spending (TDK actually spends >$1B in R&D).

Part of this R&D strategy is to accumulate IP, specifically patents. As of 2018 the company had over 20.500 patent applications globally. That’s good for 29th highest in the world and 10th highest in Japan. It’s tough to compete with a company that has that many patents.

3. Strong Monozukuri Capabilities Enable Timely Supply

Translated literally, monozukuri means “making of things” or “production” in Japanese. But it goes deeper than that. Monozukuri encompasses, “the synthesis of technological prowess, know-how and spirit of Japan’s manufacturing process.”

Murata owns the entire manufacturing process from start to finish. The company vertically integrates the supply of raw materials, sheet-casting, sintering, processing and finishing and inspection/packaging.

Here’s Murata’s Director of Components discussing the power of Monozukuri (emphasis mine):

“A major factor in acquiring this high market share is the fact that we can complete everything from development to manufacturing internally. In other words, everything from ceramic material selection to production facilities and manufacturing process technology is taken care of by our own internal framework. As a result, customer requests can be quickly led forward into development, and products can be supplied at lower cost due to various cost reduction options.”

Murata Today: The Clear Leader in Filters and Capacitors

Murata competes on four major products (see global market share %):

    • Chip multilayer ceramic capacitors (40%): Discussed above
    • SAW filters (50%): Extract only the required portion of a radio signal. Key devices in high frequency circuits.
    • Noise suppression products EMI suppression filters (35%): Eliminate external noise and protect delicate electronic circuits
    • Connectivity Modules (55%): Allow various devices to access the Internet via radio signals

The company generates 36.5% of their sales from capacitors, 27.1% from Communications Modules, 25% from “Other”, 8% from Piezoelectric Components and 2.6% from Power supplies and other modules. Broken down by industry the company generates 48% of its revenues from Communications, 16.4% from Automotive Electronics, 15.8% from Computers & Peripherals and 15% from Home & Others.

From a regional standpoint, Murata earned 50.5% of its revenues from China, 16.5% from Asia, 15% from the Americas, 9% from Japan and 8% from Europe.

Murata’s MLCC segment grew 27% YoY and is the biggest factor in the company’s continued success. Let’s see how they plan to capture the coming demand shock.

Murata Tomorrow: A Management Team Pointed In The Right Direction

It’s worth reiterating that the only thing that matters for Murata’s future success is their ability to capture the growing demand for MLCCs in the Automotive and Communications markets via EV and 5G.

Management knows this, too. Both Murata’s President and Director of Components stress the importance of EV and 5G in their 2019 Annual Report.

Here’s what President Tsuneo Murata (the founder’s son) had to say about the automotive industry (emphasis mine):

We are also seeing changes in the automotive industry that will have a significant impact on the world of electronics. The potential of semiconductors and communication functions will increase through electrification and automated driving, and automobiles are expected to move closer to being considered electronic or communication devices just like smartphones.

Murata’s Director of Components, Toru Inoue echoed Tsuneo’s sentiment on its two key markets (emphasis mine):

“For 5G, it is naturally important to consider which applications will become mainstream as data volumes increase significantly, but the evolution of smartphones and all kinds of wearable devices will continue to centralize on becoming smaller and thinner and featuring increased functionality, and capacitors will be required to accommodate larger capacities … In addition, EVs, V2X and automated driving require high reliability in products that will never fail even in harsh environments such as high temperatures, high humidity, high voltage and high currents.”

Before diving into valuation, let’s summarize the bull thesis for EV and 5G.

Bull Thesis: Autonomous Driving

Murata’s MLCCs will meet heightened demands as powertrains move to EV and Autonomous Driving levels increase. The company expects Mild HEV (Hybrid/Electric) powertrains to reach ~15-20M units by 2024. At the same time, Autonomous Driving levels should increase from zero to a majority at Level 1 and Level 2.

This is important because higher levels of EV and Autonomous Driving mean more MLCCs in the car (see graph).

More importantly, automotive sales don’t need to be stellar for Murata to experience automotive MLCC growth (see below):

Imagine how valuable Murata’s Automotive MLCC segment will be when nearly every car on the road is fully electric with Level 3 Autonomous Driving capabilities.

Bull Thesis: 5G Communications

TSR estimates that by 2025 the world will have nearly 1B 5G smartphones in circulation. And remember, 5G smartphones need more MLCCs than LTE and “dumb” phones.

2020 MLCC Demand and Supply Analysis notes that (emphasis mine), “5G smartphones which support sub-6GHz frequency band will need 10~15% MLCC more than those of 4G smartphones; and 5G smartphones which support mmWave will need 30~35% MLCC more than those of 4G smartphones. In estimation, each 5G smartphone will need more than 1000 units MLCC in the future.

That’s a lot of future MLCC demand.

Murata notes the reasons for increased MLCC demand in 5G phones, including:

    • More frequency bands
    • Higher frequencies
    • Advanced communication technologies
    • More sensors and cameras
    • Larger batteries

The company has the components to meet this demand (see below):

As the largest producer of MLCC’s in the world, Murata is well-positioned to supply the needs of EV/Autonomous vehicles and 5G. But how could they fail? If we look back in five years and we’re wrong, where would the errors lie?


There’s a few major risks to Murata’s future success:

    • Rise in Chinese-owned MLCC production
    • Increased demand from alternative capacitors like aluminum polymer and tantalum due to high lead times in MLCC supply
    • Inability to correctly anticipate changing customer needs/landscape resulting in poor R&D investment (i.e., burning cash)
    • Increased losses in the company’s lithium-ion battery business (small part of the company’s operations)
    • Economic recession (or COVID-related scare) that delays demand for 5G and EV/Autonomous vehicle investment


Murata is a durable, consistent company that cranks out 30%+ gross margins and 12-17%+ EBIT margins like clockwork.

The company generates a 23% Operating Cushion, which isn’t bad. Yet more than 100% of that operating cushion is eaten by working capital. Last year inventory alone sucked 21% of the company’s Operating Cushion, followed by 18% from accounts receivables.

With supply gluts in the past, Murata should experience improved inventory/receivables management. This in turn will help it turn FCF positive over the next five years.

Today’s stock price assumes a few things on an EBITDA exit basis over the next five years:

    • Average revenue growth of 5%
    • 37% average gross margin
    • 16% average EBIT margin
    • 25% average EBITDA margin

These above assumptions get us $18B in revenues, $4.74B in EBITDA and $2.24B in after-tax operating earnings. A 15x 2025 EBITDA multiple at a 10% discount rate gets us a little over $43B in shareholder value. Which is where we stand today.

If the future demand from EV/Autonomous Driving and 5G are real and materialize, the above revenue growth rate is too low. What happens if we assume Murata grows 13%/year for the next five years?

In that world we get $25.9B in revenue, $6.77B in EBITDA and $3.2B in after-tax profits. Sticking with our 15x multiple on EBITDA gets us $62.5B in shareholder value (~42% higher than current price).

Concluding Thoughts

Today’s price assumes a low-growth environment for a company whose industry tailwinds are just getting started. The market doesn’t realize this, of course, because it’s terrible at predicting exponential growth. Who knows, maybe 13%/year top-line growth is too low? We don’t know.

What we do know is the future looks bright for Murata’s core products and markets. And Mr. Market doesn’t seem to think that.

memory chip

DRAM & NAND: Betting On The Semiconductor Supercycle

, , ,

“Digital technology, pervasively, is getting embedded in every place: every thing, every person, every walk of life is being fundamentally shaped by digital technology — it is happening in our homes, our work, our places of entertainment. It’s amazing to think of a world as a computer. I think that’s the right metaphor for us as we go forward.” – Satya Nadella, Microsoft CEO

If the world is a computer, storage and memory are the hydrogen and carbon molecules needed to sustain life. Because computers without memory are empty screens. Unable to perform the functions that give rise to technologies like AI and Deep Learning. 

Yet our world is moving closer and closer towards this computer-first reality. Smart home penetration is a perfect barometer. 48 million smart home devices entered new homes in 2019. This $27B market should grow near 21% CAGR for the next three-to-five years. And that’s just one example. 

As AI technology improves so does the number of potential applications. A few markets ripe for AI penetration include: 

  • Automotive 
  • Consumer Electronics
  • IT & Telecommunication
  • Medical 

But these markets won’t experience innovation without proper storage and memory technologies. 

Enter semiconductors. 

Last week we discussed why AI will turn semiconductors into a secular growth powerhouse. You can read that here. One major consequence of this new demand driver is the insatiable need for more data. 

McKinsey notes in their article, Artificial-intelligence hardware: New opportunities for semiconductor companies, the intensity of AI data storage needs (emphasis mine):

AI applications generate vast volumes of data—about 80 exabytes per year, which is expected to increase to 845 exabytes by 2025. In addition, developers are now using more data in AI and DL training, which also increases storage requirements. These shifts could lead to annual growth of 25 to 30 percent from 2017 to 2025 for storage—the highest rate of all segments we examined.”

This week we’re focusing on public companies aimed at capturing this demand and turning it into record profits and margins. 

But before we dive into specific companies, let’s break down the differences between memory and storage. 

DRAM & NAND: Know The Difference

It’s helpful to think of memory and storage in two ways: 

  1. DRAM is the memory used to store code for algorithms, processes, etc.
  2. NAND is the memory used to store data for pictures, music, etc.

DRAM: Short-Term Memory

Memory uses DRAM (dynamic random access memory) to perform its functions. DRAM is a volatile memory, meaning it stores memory when a device is on. But when you power off, so strip the memory. 

When you think of DRAM, think of your main computer processor and graphics cards. DRAM is also used in portable gaming devices and video game consoles. There’s a few key advantages of using DRAM (via 

  • Can be deleted and refreshed while running a program
  • Smaller size
  • Higher storage capacity 
  • 100x faster than NAND

That said, there are drawbacks to DRAM memory, such as: 

  • Data requires constant refreshing
  • Complex manufacturing process
  • Volatile memory

DRAM demand is here to stay thanks to autonomous driving, video game consoles and AI applications. Micron (MU) CEO Sanjay Mehrotra said (emphasis mine), “AI servers will require six times the amount of DRAM and twice the amount of SSDs compared with standard servers.”

ATPInc wrote a great article on the massive AI-induced DRAM demand, saying (emphasis mine), “As AI workloads continue to grow, hyperscale data centers require more and more memory. In the first quarter of the year, DRAM supply remained tight mainly due to the massive construction projects of data centers, some of which are bigger than football fields.” 

Another ATPInc article highlighted the importance of DRAM in cloud computing technologies. The article reads (emphasis mine), “In recent years, the use of DRAM has been increasingly extending beyond the personal computer and consumer electronics sphere. Higher capacities and low latencies are among the driving factors why DRAM is figuring extensively in industrial applications such as smart factories, health care, military, automotive, networking systems and data centers.

Dram will also be critical to IoT bc of its low latency. For instance, According to Gartner, “driverless cars contain over 80 GB of DRAM versus 5.5 GB in PCs and 2.5 GB in handsets, exemplifying the sharp increase in the memory demands of these emerging technologies.”

More companies will use AI and shift storage centers to the cloud. This will inevitably lead to increased DRAM demand and a sustained DRAM upcycle. But DRAM isn’t the only memory chip experiencing the AI demand bump. 

NAND: Memory At The Edge  

Storage is our long-term memory. It allows computers and applications to store large datasets, which it can then retrieve information when needed. 

There’s a few main differences between NAND and DRAM: 

  • NAND doesn’t need power to keep data
  • Ideal for portable devices
  • Cost-effective per-byte with high storage capacity 
  • Easily replaceable

NAND is the most exciting memory/storage component of the semiconductor technology stack. We’ll see more devices use AI-based technology at the edge. This will increase demand for NAND memory chips, which operate best at the edge due to their reduced energy requirements, portability and ability to store massive amounts of data. 

An IndustryResearch study reaffirms this belief in their report, GLOBAL 3D NAND FLASH MEMORY MARKET REPORT, saying (emphasis mine): 

“The global 3D NAND Flash Memory market size is projected to reach USD 47800 million by 2026, from USD 15540 million in 2020, at a CAGR of 20.6% during the forecast period.”

Why are they projected to grow at such a high clip? AI Adoption. Check out this statistic from Eetasia (emphasis mine): 

“AI technologies are now set to be rapidly adopted in embedded systems: analyst firm IDC expects the market for AI-optimized processors for edge computing systems to grow at a compound annual rate of 65% in the years to 2023. But this move to adopt AI raises questions about the sustainability of embedded developers’ current approach to the provision of memory for code storage.”

Flash-based (NAND) memory chips accounted for 17% of the global storage market in 2018. Toshiba estimates that by 2025, NAND memory will account for 40% of global chip storage. Again, the massive demand increase is driven by AI-enabled devices that need compute and storage power at the edge. 

ElectronicSpecifier reiterates the impotence of NAND memory devices for the future of AI-enabled technologies (emphasis mine): 

“Consumer products, such as smartphones, tablets and cameras, along with industrial equipment and sensors, automotive systems and medical devices, all rely upon flash memory, often integrated alongside their processors, that stores both data and the code they execute. However, data centres find attraction in flash memory due to its near real-time response to read/write requests, and high data transfer rate. As demand for massive data processing for artificial intelligence (AI) and machine learning applications grows, so interest in flash-based storage will evolve in tandem.

Now we know DRAM, NAND and why we’ll see tremendous growth in both memory chips. Let’s review a few companies we can buy to express our bullish DRAM/NAND theory. 

Public Companies Dominating The Memory Market

The memory market is an oligopoly between five companies (market share in %): 

  • Samsung (005930/SSNLF): 35% 
  • Micron (MU): 16.5%
  • Western Digital (WDC): 15%
  • SK Hynix (000660): 9.5%
  • Intel (INTC): 8.5%

An oligopoly in a commoditized market means one thing: lowest-cost competitor wins. In semiconductor language that means more storage space on less surface area. Yet the race towards smaller chips and lower prices resulted in a 46% price collapse in the NAND market in 2019. 

Thanks to the demand drivers, all these companies’ charts are setting up for very bullish moves. 

Let’s review our top three ideas in this space. 

I’ll provide a general description of the business, the bull case and what the charts are saying. Stick with us next week where we dive deep into an individual name that we love in this space.

Micron (MU): Our Favorite All-Around Memory Play

Business Description: Micron Technology, Inc. manufactures and sells memory and storage solutions worldwide. The company operates through four segments: Compute and Networking Business Unit, Mobile Business Unit, Storage Business Unit, and Embedded Business Unit. It offers memory and storage technologies, including DRAM, NAND, NOR Flash, and 3D XPoint memory under the Micron, Crucial, and Ballistix brands, as well as private labels. –

Bull Case: 

  • MU is one of the cheapest NAND companies in public markets (2.3x EV/Sales & 5x EV/EBITDA)
  • NAND memory accounts for 25-30% of total revenue
  • 3rd largest DRAM manufacturer in the world
  • 10%+ ROC over 30 years
  • CEO an industry leader in non-volatile memory (NAND)
  • Given memory chip price collapse, there’s fewer suppliers catering to more customers (demand/supply imbalance)
  • Rise in demand should offset the natural decline in ASP (average sale price) of NAND chips
  • Fewer players will result in focus on profit stabilization, not cost cutting and margin compression

Note: Alex sent out an in-depth MU write-up to Collective members this week. The full report is only available to Collective members


  • 5YR Average Growth Rate: 9.56%
  • 5YR Average EBITDA Margin: 44%
  • Capex as % of Revneue 5YR Average: 36%
  • 2024 EV/EBITDA Multiple: 9x

The above assumptions get us ~$87/share by 2025. That’s 71% upside from the current stock price. 

Tape Reading: 

MU currently has a pattern within a pattern. Both bullish. The daily chart (shown below) reveals a bullish inverse H&S pattern: 

daily chart of MU

MU Daily Chart

Now let’s zoom out further to the monthly time frame. The monthly time frame shows a coiling symmetrical triangle ready to propel higher: 

MU Monthly Chart

MU Monthly Chart

The stock is also above the 20MA pointing to further bullish sentiment. 

Samsung Electronics (005930/SSNLF): Our Favorite International Play

Business Description: Samsung Electronics Co., Ltd. engages in the consumer electronics, information technology and mobile communications, and device solutions businesses worldwide. It offers mobile phones, tablets, wearables, virtual reality, and audio products; TVs, and home theaters; OLED and LCD panels; laptops, computers, chrome books, HDM, memory and system LSI products, monitors, and printers; and home appliances, such as refrigerators, air conditioners, ovens, air purifiers, cooktops and hoods, microwaves, dishwashers, washers, dryers, vacuum cleaners, and heating products, as well as TV and home theater accessories. It also provides security and monitoring, trackers, Wi-Fi routers, hubs, sensors, outlets, and buttons. In addition, the company is involved in the technology and venture capital investment businesses; manufacture of semiconductor equipment and components; and provision of repair services for electronic devices. – 

Bull Case: 

  • Samsung is the cheapest memory chip manufacturer in the world by quantitative metrics (1.23x EV/Sales & 4.67x EV/EBITDA)
  • The company will invest a mind-numbing $115B into its semiconductor business over the next decade
  • It commands top market share in NAND memory chip production
  • The company’s stock rises and falls based off the results of two divisions: smartphones and semiconductors
  • William Keating notes Samsung’s semiconductor prowess (emphasis mine): “Samsung is one of only two companies in the world to have demonstrated volume production on a 7nm process, with an impressive and credible roadmap all the way to 3nm. We believe that the true value of this capability will still take some years to be realised.”
  • Samsung is one of the few companies with the supply capabilities to service the upcoming positive demand shock, as such, it will garner more of the operating profits
  • The company might produce the world’s first 160-layer NAND memory chip
  • Low expectations for revenue growth present low-risk upside optionality 


  • 5YR Average Growth Rate: 2%
  • 5YR Average EBITDA Margin: 30%
  • Capex as % of Revenue 5YR Average: 13%
  • 2024 EV/EBITDA Multiple: 8x

The above assumptions lead to $66B in 2024 EBITDA. Applying our 8x multiple gets us $528B in Enterprise Value. Add back cash and subtract debt and you get $550B in shareholder value. That’s 56% upside versus today’s market cap of $325B. 

Tape Reading: 

Samsung’s tape looks very strong on the monthly time frame. The stock is breaking out of a three-year Cup & Handle pattern: 

samsung monthly chart

Samsung Cup & Handle on Monthly Chart

A breakout above the neckline would also signal new all-time highs. 

Western Digital (WDC): Our Contrarian, Turnaround Play

Business Description: Western Digital Corporation develops, manufactures, and sells data storage devices and solutions. It offers client devices, including hard disk drives (HDDs) and solid state drives (SSDs) for computing devices, such as desktop and notebook personal computers (PCs), smart video systems, gaming consoles, and set top boxes; flash-based embedded storage products for mobile phones, tablets, notebook PCs, and other portable and wearable devices, as well as automotive, Internet of Things, industrial, and connected home applications; flash-based memory wafers; and embedded storage solutions and flash products. – 

Bull Case:

  • Trades 1.18x EV/Sales and <6x EV/EBITDA
  • Second largest manufacturer of SSD hard drives (behind Samsung)
  • 1,400+ patents around NAND and SSD technologies
  • Gross Margin will expand to 2016-2018 levels of ~33%
  • Operating Margin will expand to 12-15% 
  • Long-term support at current stock price
  • Company repays debt, buys back stock and pays a dividend (5%)
  • Company will deleverage as it moves through the more bullish part of the cycle


  • 5YR Average Growth Rate: 4% (GDP-type growth)
  • 5YR Average EBITDA Margin: 15% (vs. historical 13-21% estimate)
  • Capex as % of Revenue 5YR Average: 4.2% 
  • 2024 EV/EBITDA Multiple: 8x

The above assumptions lead to $20B in 2024 revenue and $3.55B in EBITDA. Applying our 8x multiple gets us $28.4B in Enterprise Value. Subtract $7B in net debt and you’re left with $21.4B in shareholder value, or $70/share. That’s an 84% upside from current prices. 

Tape Reading: 

WDC’s a picture-perfect bottom-picker’s stock. Price is currently anchored to a long-term support level of $38/share. You can clearly see the support when viewing a monthly chart (see below): 

WDC monthly chart

WDC Monthly Chart

The current price offers a great reward/risk trade set-up. Investors can buy at the close of this month’s candle with a stop below support (around $33/share). 

Risks To Companies’ Bull Thesis

There’s a two main risks we’re worried about with our bullish NAND thesis: 

  • Competition from Chinese chip manufacturers

China’s investing $100B to bring chip development to the mainland. China chip success would mean less revenue from companies like MU, which generate roughly 50% of their revenues from China. 

New Chinese entrants would also likely result in price wars and lower margins. Currently, the oligopoly is best served to stabilize prices and enjoy generally high profit margins for everyone. 

  • Global Macro Slowdown

As the general economy goes, so do semiconductor companies. A global slowdown would reduce semiconductor orders and R&D.

AI & Semiconductors: Understanding The Next Super-Cycle


“Mechanisms of artificial intelligence provide real-time fast decision-making based on the analysis of huge amounts of information, which gives tremendous advantages in quality and effectiveness … If someone can provide a monopoly in the field of artificial intelligence, then the consequences are clear to all of us: he will rule the world.” – Vladimir Putin

AI is the Snitch in the geopolitical game of Quidditch. Control the Snitch, control the game. As Putin notes, AI offers endless potential; both good and bad.

Look no further than China, who committed $150B in its effort to lead the world in AI technology.

Abishur Prakash of Scientific American put it bluntly (emphasis mine):

“As nations compete around AI, they are part of the biggest battle for global power since World War II. Except, this battle is not about land or resources. It is about data, defense and economy. And, ultimately, how these variables give a nation more control over the world.”

Semiconductors are the basic building block to any artificial intelligence application. These small pieces of hardware enable AI algorithms to store, run, and test more data.

The rise of AI has flipped the semiconductor industry from a cyclical sector into a secular growth powerhouse. A few statistics point to this secular growth story:

    • AI-based semiconductors will grow 18%/year over the next 3+ years (5x greater than non-AI chips)
    • AI-specific memory chips command 300% higher prices than standard memory chips
    • By 2025, AI-related semiconductor technology will account for 20% total revenue

Semiconductor companies will capitalize on these trends by capturing more of the AI technology stack. Under previous technological shifts (smartphones, cloud computing, etc.) semiconductors took smaller portions of the revenue pie (10-20% in most cases).

Here’s the main thing that matters: McKinsey estimates that in an AI-led growth cycle, semiconductors can capture 40-50% of the revenue pie. Remember, AI-based technology will represent roughly 20% of all revenue by 2025. This percentage should increase over time. That means semiconductors will generate 40-50% of a much larger TAM.

This essay will cover three topics:

    1. The AI Technology Stack & Semiconductor Positioning
    2. How AI Is Changing Semiconductor Hardware Demand & A Few Examples
    3. Why Investors Should Care

Let’s dive in.

The AI Technology Stack: Semiconductors Will Capture More Share

There’s nine layers in the AI Technology stack (see on right). Semiconductors live at the bottom in the hardware layer. Traditionally, this layer meant commoditized products, lower margins and less profitability.

Anyways, these nine layers fixate on two goals:

    1. Train data
    2. Make inferences from that data

Semiconductors focus on two parts of the stack: head nodes and accelerators. These are the neurons and synapses of the neural network brain.

Your Roadblock is Our Opportunity

According to McKinsey, head nodes and accelerators are the top bottleneck issue for AI developers that want to improve their programs (emphasis mine):

“When developers are trying to improve training and inference, they often encounter roadblocks related to the hardware layer, which includes storage, memory, logic, and networking.”

You can’t run AI programs without these two components. It’s like a human brain. No amount of data matters if you don’t have enough synapses (nodes) to relay information to the brain (accelerator).

It’s this exact reason why semiconductors will switch from a commodity product to a differentiated piece of hardware. Companies will pay more for specialized, AI-specific hardware that makes their programs run more efficiently with less data and computing power.

Why? Because that’s where the roadblock is. It’s not at the software level or any other point in the AI-technology stack. The developer with the best hardware wins.

To understand the dynamics of hardware demand, let’s look at the actual components involved in these AI programs.

The Four Types of AI Chips

AI developers generally have four types of chips at their disposal:

    • Application Specific Integrated Circuits (ASIC)
    • Central Processing Units (CPU)
    • Graphics Processing Units (GPU)
    • Field-Programmable Gate Array (FPGA)

AI algorithms are increasing the demand for ASIC chips compared to all others. That’s the one we want to pay attention to. Why is this? Allerin offers a few reasons (emphasis mine)

“An ASIC chip is a chip designed to carry out specific computer operations. Hence, ASIC chips are now being developed to specifically test and train AI algorithms. ASICs can be used to run a specific and narrow AI algorithm function. The chips can handle the workload in parallelism. Hence, AI algorithms can be accelerated faster on an ASIC chip. The only major tech giant investing in ASIC chips is Google. Google announced its second-generation ASIC chips called the Tensor Processing Unit (TPU), based on TensorFlow. The TPUs are designed and optimized only for artificial intelligence, machine learning, and deep learning tasks.”

The increased demand will allow semiconductor companies to venture up the technology stack as they build customizable ASIC chips for their OEM partners. We’ll cover that in more detail in the next section.

Now that we know what the AI technology stack is and where semiconductors fall in that stack, it’s time to examine how AI is reshaping semiconductor hardware demand for that all-important last layer.

How AI Is Reshaping Semiconductor Hardware Demand

Gavin Baker’s content was at the top of my reading list when I started my semiconductor/AI research. Baker’s interview with the Market NZZ offers clues at the future of semiconductors, the importance of AI and demand shocks.

Gavin mentioned three separate demand shocks in the semiconductor market:

    • Virtualization helped utilize servers more efficiently
    • Cloud Computing further increased server utilization
    • Smartphones cannibalized PC market growth

Now, thanks to artificial intelligence, semiconductors are finally experiencing a positive demand shock. Here’s Gavin’s take (emphasis mine):

“Now, we have a positive demand shock in the form of artificial intelligence … because data quantity is so important for quality, is much more semiconductor intensive.”

He later expands on this thought, saying (emphasis mine), “AI as a demand driver is just getting going. When humans write software, they understand that we want to limit the amount of computational resources. They write software in very elegant ways to minimize compute and memory. Artificial intelligence is none of that. AI is all about semiconductor brute force.

Semiconductor brute force. AI doesn’t care about computational resources. It cares about maximizing and optimizing its algorithm to make better decisions. What’s needed in that environment? More storage, more capacity and higher memory bandwidths.

McKinsey’s semiconductor/AI research reaffirms Gavin’s thoughts. In their article, Artificial-intelligence hardware: New opportunities for semiconductor companies, McKinsey notes how much data storage AI applications will require (emphasis mine):

“AI applications generate vast volumes of data—about 80 exabytes per year, which is expected to increase to 845 exabytes by 2025. In addition, developers are now using more data in AI and DL training, which also increases storage requirements. These shifts could lead to annual growth of 25 to 30 percent from 2017 to 2025 for storage—the highest rate of all segments we examined.”

Deep learning and neural networks are driving most of the data demand. McKinsey explains, saying (emphasis mine), “AI applications have high memory-bandwidth requirements since computing layers within deep neural networks must pass input data to thousands of cores as quickly as possible.

The changes in demand drivers mainly affect four categories within the semiconductor space: compute, memory, storage and networking. Let’s see what the future will look like for these categories in an AI-first world (data from McKinsey).

Compute: Deep Learning Takes Share in Data Centers & Edge

There’s two areas we should consider when analyzing the computing space:

    • Data centers
    • The ‘Edge’

Both spaces have seen tremendous growth in training and inference hardware. This makes sense as all ML/DL need to train and test their algorithms. Data centers will see double the demand for training and 4-5x the demand for inference hardware.

The edge poses even greater demand needs. In 2017, Edge inference and training markets combined for a total of <$0.3B. By 2025, those markets will grow to a combined $5-$6B! That’s the power of AI-driven demand shocks.

ASICs chips look ready to fulfill this increased demand. McKinsey estimates that by 2025, ASICs chips will command 50% of data-center usage and 70% of edge application usage.

Memory: High-Bandwidth & On-Chip Will Dominate

AI is shifting memory demands away from DRAM (dynamic random access memory) to two other sources:

    • High-bandwidth memory (HBM)
    • On-chip memory (OCM)

HBM let’s AI algorithms process more data, faster. This no-brainer solution led two of the largest technology companies (NVDA and GOOGL) to switch their preference to HBM vs. DRAM. That switch comes at a cost as HBM is 3x more expensive than DRAM. But again, in an AI-based world, these chips become differentiators — not commodities.

AI is also changing the way memory is stored at the edge. More companies are switching to on-chip memory devices as a way to save both space and money. Consider GOOGL’s TPU. Developing that one ASIC saved the company billions of dollars in opening up new data center storage units.

Now that it’s easier to store memory for edge devices to generate inferences, it’s sparked a rise in edge-based inference projects. Paul McClellan, writer for elaborates on this phenomenon (emphasis mine):

“As AI algorithms moved out into edge devices such as smartphones and smart speakers, there was a desire to do more of the inference at the edge. The power and delay inherent in uploading all the raw data to the cloud was a problem, as are privacy issues. But inference at the edge requires orders of magnitude less power than a data center server or big GPU. In turn, that has led to a proliferation of edge inference chip projects.”

Looking out 5-10 years, on-chip should command a much higher percentage of the total memory market as more companies move to edge computing/inference.

Storage: Training Will Dominate Storage Needs

When you’re training an algorithm you need all the data you can get. Yet at the inference level, the data requirements drop off significantly.

This fluidity in storage needs creates an opportunity for new storage technologies. Specifically, new forms of non-volatile memory (NVM). Think of NVM as long-term memory. It’s the anniversaries, birthdays, and ticker of your worst-performing stock.

McKinsey sees storage demand increasing 25-30%/year for the next five years as AI-based programs consume more data at the edge and in data centers.

Networking: Speed Kills — Avoid Bottlenecking

Networking is extremely important in an AI-first world. Networks connect multiple servers to each other to perform the important task of training a model. Slow network speeds reduce the AI’s ability to train its model.

This can have devastating effects in autonomous driving, where edge programs need to make decisions in real-time. Staying with autonomous driving, McKinsey estimates it takes roughly 140 servers to reach 97% accuracy in obstacle detection. Network speed could mean the difference between a vehicle slamming a guard rail or cruising home unscathed.

McKinsey sees two viable options for potential AI networking issues:

AI In Action: What We’re Excited About

We now know a few things about AI. We know what the technology stack looks like. We know where semiconductors fit in that stack and how AI is reshaping the hardware demand cycle. We also know the four main areas of semiconductor hardware that will see disruption and innovation from AI-based technologies.

Let’s cover a few real-life areas where we’re excited about the future of AI application.

The below use cases were taken from this devprojournal article.

Application 1: Education

AI has many applications in the underearning education industry. AI can tackle most of the monotonous tasks a teacher performs. This leaves the teacher free to do what he does best, teach kids.

There are also opportunities for AI-enabled tutoring. Companies like Content Technologies use AI to turn large, complicated textbooks into easy-to-read flashcards, note packets and practice exams.

COVID-19 has also accelerated the role online learning plays in a person’s education. Carnegie University developed its own software to provide remedial classes to incoming freshman and sophomores — saving billions of dollars in tuition fees.

Application 2: Healthcare

Healthcare is arguably the most popular space for AI-based interventions. There’s a few that excite us, including:

    • Tumor/disease diagnosis
    • Symptom checkers
    • Clinical study pop-ups based on radiology image results
    • Medication trackers

You can find 28 more examples of AI in healthcare here.

Application 3: Finance

The finance companies that win over the next 5-10 years will employ AI-first business models. Applications vary from fraud detection, loan processing, credit risk assessments and advertising.

Cardlytics (CDLX), a Macro Ops portfolio holding, does just that. Using AI to sift through millions of card/bank transactions, CDLX is able to provide marketers with advertising channels and consumers with deals from brand partners.

These three applications bring us to our final section: why you should care.

Why Should We Care About Semiconductors?

Semiconductors and AI are in the early stages of an exponential, long-term growth trend. We don’t know which companies will win in the end. But what we do know is that this is one of the best spots to fish over the next decade.

Why? The market is horrible at pricing long-term exponential growth. It’s too short-term oriented to see the wave forming way out in the ocean. Mr. Market’s focused on next quarter and next year. Not on the major shifts happening over the next five years.

Mr. Market’s ignorance is our opportunity. His short time-frame is our alpha. By understanding the drivers of this long-term secular growth trend, we can position ourselves to invest in these areas now.

McKinsey estimates that the AI-spurred semiconductor market will grow ~18%/year for the next five years (see right). Things get better when you look at specific areas like storage, which will grow near 30%/year for the next five years.

The bull thesis is simple:

    • AI-based technologies will represent larger and larger portions of total semiconductor revenue
    • Semiconductors should get 40-50% of this growing AI-specific revenue base (as opposed to 10-20% in other cycles)
    • The shift from cyclical to secular growth will result in smoother revenues and earnings, which will lead to higher multiples.

Next week we’ll explore these high-growth semiconductor spaces in greater detail. We’ll reveal why industries like storage are set to pop and what companies we should invest in to capture those trends.

We’re excited about this space and the long-term growth prospects. Moreover, some of the long-term charts for these semiconductor companies are setting up for much higher prices. And we’ve got our eyes on under-the-radar companies taking advantage of these new technologies with interesting business models.

See you next week!

Super Stock Case Study: Cintas Corp (CTAS)

, ,

Cintas Corp (CTAS) provides corporate identity uniforms and related business services primarily in North America, Latin America, Europe, and Asia. It operates through Uniform Rental and Facility Services and First Aid and Safety Services segments. The company rents and services uniforms and other garments, including flame resistant clothing, mats, mops and shop towels, and other ancillary items; and provides restroom cleaning services and supplies, and carpet and tile cleaning services, as well as sells uniforms. It also offers first aid and safety services, and fire protection products and services.

Cintas is a boring business.

But this boring business has returned over 1,000% for investors since 2009?

Our CTAS case study dives into the early days of the company. We examine initial investor sentiment. What did the company do well? What were investors’ priorities? We place ourselves in the shoes of those first investors. Those that saw what others could only recognize in hindsight.

We’ll review past newspaper articles and investor write-ups. The bull and bear cases.

Consider this case study a time-travel machine.

Our goal: to know what really made these super stocks soar.

Analyzing past winners won’t guarantee future success. But it does paint a picture of what a winner might look like and hopefully after doing a number of these we can tease out some commonalities that’ll help us identify super stocks down the road.

So let’s dive into the wonderful world of uniform renting, cleaning, and laundering.

The Early Days: Acme Industrial Laundry Company

Richard “Doc” Farmer founded Cintas in 1929 under the name Acme Industrial Laundry Company. Doc got his start collecting chemical-soaked rags from various factories. He would then wash and re-sell the rags to factories for a fee.

By 1956, Doc Farmer’s grandson Richard “Dick” Farmer joined the family business. Fresh with ideas from an undergraduate degree, Dick Farmer grew sales from $300K in 1959 to $847K in 1963. Dick assumed the role of CEO in 1968.

Farmer then drew up a new business plan for the company: Open uniform rental plants across the United States.

Acme opened its first uniform rental plant in Cleveland, OH in 1968. In 1972 they changed the name to Cintas. Then in 1983 Cintas went public and traded on the OTC (Over-The-Counter) market.

Let’s learn how CTAS became today’s Super Stock.

Lesson 1: Cintas’ Fast (and Early!) Growth

CTAS hit the ground running after its IPO. A 1990 newspaper article highlights CTAS’ success as a public company (right):

During its seven years on the market, CTAS put up:

    • 24% compounded annual earnings growth
    • 24% compounded annual sales growth
    • 20% average ROE

Growth, growth and more growth. That was the key to CTAS’ early success. The company’s strategy was working. And investors could see it on the top and bottom-line.

Where did that growth come from? Expansion into new and existing markets. By 1990, CTAS commanded 10% share in the uniform rental market. The more uniform rental plants CTAS could fit on the map, the more money they made. This snippet from a 1995 interview outlines CTAS’ core business strategy (emphasis mine):

In good economic times, uniform companies grow by increasing the size of existing accounts and convincing companies that have never rented workers’ uniforms that it’s cost-efficient and enhances their corporate image, says analyst Jim Stoeffel of Smith Barney. In bad times, the big companies grow more through acquisition, because more of the 700 or so small uniform-rental companies become willing to sell their businesses for lower prices, says analyst Craig.”

Between its IPO and 1990, CTAS expanded to 37 new markets and operated in 26 states.

Moreover, the company’s story made sense. Here was a boring business servicing a high-demand, highly fragmented market. The roll-up strategy worked. The business did near 17% EBIT margins and had a long runway for growth. Insiders owned over half the company. They had little debt on the balance sheet.

CTAS outlines the bull thesis in their 1996 Annual Report:

Is it a stretch to say that screenshot is all the due diligence needed to invest in the company in 1996?

It’s easy to see the growth potential in hindsight. But what were the analysts thinking at the time? Here’s a snippet from a 1991 Kiplinger’s Personal Finance Magazine article (emphasis mine):

“Because Cintas operates in only three-fourths of the largest markets, brokerage firm Alex. Brown & Sons considers its potential for growth by expansion to be “substantial.” Analyst Sally Smith estimates long-term earnings growth at 15% to 20% a year. Similarly, Shearson Lehman Brothers picks Cintas to outperform the market over the coming year. It cites the company’s ability to sustain sales and earnings growth of existing outlets while spending on expansion. Shearson estimates $1.73 per share in earnings for fiscal 1992, versus $1.47 in fiscal ’91, and projects a possible share price of $52 to $53 by spring.”

Revenue and earnings growth are staples of super stocks. CTAS had both in spades. The company expanded market share and by 1996 was generating $730M in annual revenue. Read this clip from CTAS’ 1996 Annual Report (emphasis mine):

“When Cintas went public in August 1983, we were comparable in size to the other public companies in our industry.  Since then, Cintas has grown at a faster rate and now is the largest public company – almost twice as large as the next largest company in the industry.

And that’s the most important question: why was CTAS able to grow twice as large as the next largest company in the industry? There’s a few reasons for this success, much of which we’ll cover later in this piece.

    1. Aggressive Expansion: The company expanded into more territories faster than its competition. This created a flywheel effect. More facilities allowed the company to service more companies in their local areas. It also helped them command better pricing from its suppliers. Each new distribution center brought the company closer to its customers, which further reduced costs.
    2. Leader in Acquisitions: The uniform rental business was loaded with mom and pop operations. And nobody rolled-up more of these companies than CTAS. The company’s consistently strong balance sheet allowed them to buy in any market cycle. So, while their competition shored up operations to conserve their (little) cash, CTAS got aggressive.
    3. Higher Spending On Facility Improvement: More facilities meant more customers, which meant more revenue. The increased revenue allowed CTAS to invest heavily in optimizing the performance of their facilities. The company was constantly investing in making their facilities run better and sport the latest technology.

CTAS created a snowball effect. And at that point the goal was simple: expand into as many territories as possible. This snowball helped take CTAS from <5% market share to 17% by 1996.

All investors had to do was follow the number of new facilities CTAS opened, and confirm that their existing facilities were operating profitably.

But it’s not that easy. Many investors sat idle as CTAS made 20%+ annual returns. Why? The answer to that question lies in our second lesson.

Lesson 2: Super Stocks Always Look Expensive

Super Stocks will always look expensive on a short-term valuation basis. There’s intuitive logic behind this idea. You don’t buy an Armani suit for the same price as an off-brand Target sport coat. One is clearly higher quality.

You get what you pay for. Super Stocks look expensive because they’re higher quality businesses. You shouldn’t expect to pay Five Below prices for something that belongs behind a glass display case.

CTAS was no exception. In the previous 1995 article, portfolio manager Mark Fuller described CTAS’ valuation (emphasis mine):

“Cintas, which earned $1.12 a share in fiscal (May) 1994, traded recently at 363/4, a seemingly pricey 33 times trailing earnings. “Seemingly” is the operative word, however. The p/e ratio has been sky-high throughout Cintas’s decade as a public company, with 28 as its historical mid-range. ‘“Some of our best investments look sort of expensive,“’ says Mark Fuller, a portfolio manager at William Blair Investment Management.”

Let’s focus on the phrase “sort of expensive”. This implies a few things. First, it suggests that it’s not a traditional “value” stock. You wouldn’t have found CTAS on a deep value scan. Yet at the same time the price never reached astronomical levels. It’s this in-between valuation that propels Super Stocks higher.

And I get it. A 38x P/E multiple assumes a 2.60% forward return. But that’s without considering actual earnings growth. CTAS traded at 38x earnings in 1994. But they were growing those earnings over 25% annually. That’s a 1.52x PEG (Price-to-Earnings Growth). Again, not bad.

Looking expensive comes at a cost. It’s around this time that short sellers come out of the woodwork. The hope, of course, is selling short a company trading above their estimate of “fair multiple”. CTAS fell victim to various short reports. Check out this one below from a 1999 Streetwalker report:

Here’s their reasoning for shorting CTAS at 38x earnings:

    • “Back out the acquisitions, though, and you have a much more stagnant-look-ing company”
    • “The biggest growth in employment is coming from technology, service and healthcare jobs, many of which don’t require uniforms.”

There’s one major problem with the Streetwalker report. It didn’t actually address why the company couldn’t continue to grow over the next five-to-ten years. The main crux of the argument was “it’s at 38x earnings and grows via acquisitions.”

Moreover, CTAS outlined their acquisition strategy from the beginning. It was always a large part of their growth plans. Sam Rovit and Catherine Lemire described CTAS’ strategy in their 2003 Forethought letter:

Here’s the best part of that snippet: “Since the 1960s, Cintas has supplemented its organic growth with a steady diet of small acquisitions.” Knowing this, Streetwalker’s report falls mute.

By 2004, CTAS owned 30% of the uniform market. Their second largest competitor, Aramark, couldn’t compete. The former chemical rag washer generated 60% higher EBIT margins than Aramark in 2004. With competitive advantages like that, It makes sense the company traded at near 40x earnings.

The above is a monthly chart of CTAS stock price since its IPO in 1984. Notice the stock’s torrid gains from 1984-1999. It wasn’t until the Dotcom Bubble that CTAS experienced its first true test/trading range.

Lesson 3: Insider Ownership & A Boring Business

CTAS had ample skin in the game from the top down. Doc Farmer owned 52% of the company during his stint as chairman. Early CTAS management didn’t rely on salaries and bonuses to generate wealth. They needed operational success and stock price appreciation. Most of CTAS’ management’s net worth was in their stock ownership.

On top of that, the company reduced its share count from 171M in 2004 to 104M in 2020.

The uniform rental business is a boring business. A 1991 article titled, Unappetizing suggestions for tasty stock profits, defined CTAS’ business as, “dull, dull and dull.” Twitter user @PermanentCap shared similar thoughts (see right).

It’s part of what made CTAS such a great company. Boring industries usually have loads of mom-and-pop shops. Family-run businesses with no exit plan because their son Jeremy wants to develop software, not clean clothes.

Boring industries are the perfect hunting grounds for a CTAS roll-up strategy.

Lesson 4: Low Debt

CTAS used little debt. The 1996 Annual report shows a Debt/Capital ratio of 22.5%. In fact, CTAS stayed around that range for most of its decade price advance. It wasn’t until 2017 that we saw Debt/Capital breach the 50% mark.

While the Dotcom bubble showered markets with lackluster balance sheets, CTAS stood tall. The company’s balance sheet strength allowed them to buy businesses in any market cycle. What a huge advantage! Again, it all goes back to CTAS’ core strategy: buy and expand into new and existing markets.

Wrapping Up: Yesterday’s Winners Provide Today’s Clues

CTAS possessed a few key attributes that made it one of the highest returning stocks of the decade:

    • Early and fast revenue/earnings growth
    • Industry-leading EBIT margins
    • Boring industry with small competitors (apt for roll-up)
    • Heavily incentivized management team
    • Little debt

There’s a reason why early and fast revenue/earnings growth is the number one attribute listed. Earnings and revenue growth drive long-term stock price performance. If a company is consistently growing earnings and exceeding expectations, the stock will eventually reflect that growth.

Check out CTAS’ revenue growth since 2009:

Like we mentioned, revenue growth isn’t enough. We need earnings per share growth. Here’s what CTAS’ EPS looked like during the same time frame:

CTAS increased it’s EPS from $1.59/share in 2004 to $8.36/share in 2020. But to achieve Super Stock status, we need one more thing: multiple expansion.

Here’s a chart of CTAS’ P/E ratio since 2009:

Next let’s look at EV/Sales from the same time frame:

Again we see the same story. 1x sales to 5.42x sales in ten years.Next let’s look at EV/Sales from the same time frame:

Revenue growth + Earnings growth x Multiple Expansion = Super Stock

Not every Super Stock will possess these exact attributes. But it gives you a few ingredients needed for long-term success. A few more filters to screen potential investments.

Macro Ops Composite System (MOCS) Analysis

We’re stoked to bring you our new Macro Ops Composite System (or MOCS, for short). MOCS allows us to get a full picture of a company’s health at a point in time. Think of it like taking the temperature of a patient. A doctor starts there, and dives deeper. We shamelessly cloned much of this system from our friend Brian Feroldi.

The MOCS analyzes the following categories:

    • Financials
    • Moat
    • Potential
    • Customers
    • Company-specific factors
    • Management & Culture
    • Stock Price

Each category has a different weighting. For example, financials score between 0 – 5. Moats score between 0 and 15. Etcetera, etc. This reinforces what we want to look for in investments.

Those seven categories give us our Pre-Garbage Bin Score. From here, we run each company through our Garbage Bin Test (GBT).

The GBT reduces a company’s overall score for negative characteristics like:

    • Increased share dilution
    • Customer concentration
    • Outside forces
    • Large share price decline
    • Complicated financials
    • Currency risk

These criteria subtract from the total score. The most you can lose in the Pre-Garbage Bin Test is -44.

The highest rating a company can get is 100. The lowest, -44. The higher the better. When it comes to making the final investment decision, the scoring goes:

    • >80: Why haven’t we bought (or bought more)?
    • 65 – 79: We should invest
    • <60: Avoid for now

Cintas (CTAS) MOCS Score

CTAS scored an 85 before the GBT. We deducted 9 points from the GBT for a total score of 76. That scores CTAS in the 76% percentile of top companies. And it falls under the “We should invest” range above.

Macro Ops Collective members can access the MOCS score in the Operators dashboard.

Reading The Tape: Cintas Chart Analysis

We know the fundamental story behind CTAS’ success. Now let’s hit the tape and analyze the company’s stock chart. We’re using the weekly time frame and going all the way back to the IPO.

We’ll reveal chart patterns that provided an optimal, low-risk entry point.

Let’s get after it!

Chart 1: 1983 – 1987

Chart 2: 1989 – 1991

Chart 3: 1994 – 1997

Chart 4: 1999 – 2001

Chart 5: 2013 – 2015

Chart 6: 2016 – 2018

Accrued Revenue: The Ultimate Cash-Sucker

, ,

Hope you had a great week! Last week we discussed deferred revenue and why we should understand it. This week we’re examining its evil cousin: accrued revenue. There’s a lot in common. But here’s the key difference:

    • Deferred Revenue = job not complete but cash received
    • Accrued Revenue = job complete but cash not received

Alright let’s dive in.

What Is “Accrued Revenue”?

Accrued revenue occurs when a company completes a service or delivers a product but hasn’t received payment for that product or service. Let’s use our lemonade stand as an example.

If we sell a glass of $1 lemonade to a customer that promises to pay us tomorrow, that’s $1 of accrued revenue. We’ve completed the task of delivering our goods to the customer. Now we wait for reimbursement.

Why does this happen? You can thank GAAP accounting. GAAP accounting states that a company must recognize revenue at the time it’s earned. Not when the company receives the cash.

This is an important distinction. Most companies perform services and provide goods without accepting the cash up front. Think of manufacturing companies. Usually these companies supply parts and widgets on a Net-30 basis.

In other words, the customer (who receives the widget) has 30 days to pay for that widget.

You might know accrued revenue as another name, accounts receivable.

Mechanics of Accrued Revenue on Financial Statements

Accrued revenue hits the financial statement in two ways: the balance sheet and income statement. As soon as a sale takes place, the company recognizes that sale as revenue on the income statement.

If we don’t receive cash at the point of sale, we have to also record an increase in the accounts receivables account on the balance sheet.

When the customer pays for the goods or service, we reduce that dollar amount on the accounts receivable account and increase the company’s cash amount on the balance sheet.

The Importance of Accrued Revenue

Understanding accrued revs is important because it offers us clues to the underlying health of a business. It allows us to spot cash-flow issues before they reach an earnings transcript or analyst write-up.

Remember: when in doubt, follow the cash

Here’s what you need to know about accrued revenue analysis:

    • Rising accrued revenues = bad sign
    • Shrinking accrued revenues = good sign

Rising accrued levels means the company’s having trouble collecting cash for goods and services it already performed.

Lower accrued revenues means the company’s collecting more cash from goods and services than it’s recording.

We want to buy businesses that reduce accrued revenues over time and avoid those that grow such balances. Doing so will save us a lot of money and stress over time.

If you have any questions feel free to reach out.

GoodRX (GDRX) S-1 Breakdown Analysis

, ,
GoodRX (GDRX) filed their S-1 earlier this week. I read It so you don’t have to (but you should). Here’s a thread on what I found interesting, fascinating and down-right incredible from the company. I’m starting at zero. Follow along here

GDRX Facts & Figures

  • #1 most downloaded medical app
  • 4.9M Monthly Active Users
  • 80%+ Repeat Activity
  • $20B+ in Consumer Savings
  • 150B daily pricing data points
  • 4 platform offerings
  • Est. Market Cap: ~$9.9B

Business Overview

Mission: To help Americans get the healthcare they need at a price they can afford.
So far it’s working (really) well.
The company estimates 18M of their customers could NOT have afforded to fill their Rx without the company’s savings tools.

How GoodRX Makes Money

Receives fees from partners, which is mostly Pharmacy Benefit Managers (PBMs) when customer uses GDRX code.
Fees are % of fees that partners earn OR a fixed payment per transaction.
Recurring nature to GDRX model as code is saved to consumer profile. 

Financial Results

  • GMV via prescription offering: $2.5B
  • Compounded annual revenue growth rate: 57% since 2016
  • Generated $388M Revenue in 2019
  • Generated $66M Net Income in 2019
  • 2019 Adj. EBITDA: $160M

Solving Healthcare Consumer Issues

GoodRX notes 5 major healthcare consumer “lacks” in its S-1:
  • Lack of Consumer-focused solutions
  • Lack of Affordability
  • Lack of Transparency
  • Lack of Access to Care
  • Lack of Resources for Healthcare pros

GDRX Total Addressable Market

GoodRX estimates their TAM around $800B. That’s a HUGE number.
Here’s how it breaks down:
  • $524B Prescription Care
  • $30B Pharma manufacturer solutions
  • $250B Telehealth
Initial Surprises: Telehealth is nearly 32% of TAM

The GoodRX Value Proposition

It’s the coveted win-win-win:
– Consumers: Simpler, more affordable Rxs
– Healthcare Pros: Increased medication adherence and greater price transparency (also links w/ EHR)
– Healthcare Co’s: Reach & provide affordable solutions (Rxs) to customers

What Makes GoodRX Different

There’s six strengths that reinforce GoodRX’s powerful network effects:
  1. Leading platform
  2. Trusted Brand
  3. Scaled & Growing Network
  4. Consumer-focus
  5. Extensible Platform
  6. Cash generative monetization model
See image for descriptions…

Analyzing Income Statement

– GDRX grew revenue from $99M in 2016 to $388M in 2019 (crazy growth)
– Biggest operating expense currently: SG&A, which was 46% of revenues last year
– Operating Margin: 36% (real nice)
– Pre-Tax Earnings: $83M (21% margin)
– EPS grew from -$0.11 to $0.19 in four years (w/ growing share count)
– 2019 EPS of $0.19 is computed using weighted average shares post-IPO.
– Six-month ended YoY: $0.09 vs. $0.15 in 2020 on $15M more income
– SBC: <1% of revenues

Analyzing Balance Sheet

– $126M in actual cash
– If you adjust for the pro-forma IPO, they get nearly $800M in cash
– Total Debt (incl. LT debt): $700M
– Total Est. Capitalization: $1.078B
– Financed Biz via Cash from Ops (crazy, right?)
See breakdown below …
(Debt & Contractual Obligations)
  • <1YR: $41M
  • 1-3YR: $85.6M
  • 3-5YR: $82.7M
  • >5YR: $711M

GDRX Key Operating Metrics

One of the most important KPI’s to measure for GDRX is Monthly Active Users (MAU).
GDRX’s MAU trend is absolutely incredible:
– 2016: 718k
– 2017: 1.28M
– 2018: 2.02M
– 2019: 3.18M
– 2020: 4.88M

Where Will Future Growth Come From?

GDRX outlines 3 ways to grow revs outside Rx codes:
– Subscription offerings: Gold, Kroger Savings
– Pharma Manufacturing Solutions: Provide low-cost solutions to expensive brand-name meds
– Telehealth: Online visits / marketplace

Recap: How To Track GDRX Bull Thesis

– Monitor MAU growth & Repeat Activity
– Size + Strength of Healthcare Partner Network
– Growth of Platform & Telehealth

Meet The Founders (Letter Analysis)

– Making healthcare easy is *crazy* hard
– Consumers (insured or not) needed tools to help
– Reduce cost of nearly every generic drug by >70%
– Prices are less than typical insurance

Thinking About GDRX Valuation

At the current estimated IPO price and shares, GDRX will trade roughly:
– 25x 2019 Revenues
– 62x 2019 EBITDA
– 119x 2019 Pre-tax Profits
They’re also growing 57% compounded since 2016.

Deferred Revenue: Don’t Miss The Next SaaS Winner

, ,

Hope you’ve had a great week! We paused our accounting series to highlight our Collective membership service. Doors closed last Sunday. If you didn’t get in but remain interested, shoot me an email.

I know, you missed your weekly dose of accounting knowledge.

But we’re back baby! This week we’re covering Deferred Revenue. It’s an important topic in the age of SaaS business models. If we don’t understand it, we’ll pass on an amazing business.

Let’s dive in.

What Is “Deferred Revenue (DR)”?

Deferred revenue is money a company receives before delivering those goods or services. Think of it like an advancement payment. Like we mentioned above, deferred revenue is common in the SaaS space. But older, more blue-collar industries also use deferred revenue.

Let’s use lawn care as an example. A lawn care company might need a 50% down payment before they even get to your property. They haven’t performed any service, yet have your money.

That’s deferred revenue.

Where Does It Sit on The Financial Statement?

Now it’s time to get into the weeds. Since deferred revenue isn’t “earned” in the traditional sense — we can’t show it on the income statement. So where does it hide? The liabilities section of the balance sheet.

At first glance, that sounds weird. Why should potential income live as a liability? Think about it this way. It’s like an IOU with another company for future services. You are liable as a company to perform those services or deliver those products.

If we don’t, we’ll lose that revenue and face a potential lawsuit. Let’s look at an example of this with a name we’re currently digging into at the Collective.

I’ve highlighted the DR section on the liabilities side (note: also called “unearned revenue”):

Now that we know where it sits, let’s learn why it’s important to track.

Why Should We Study Deferred Revenue?

There’s two reasons we should study this. First, we can plot the long-term growth of a company’s DR. We don’t want a dramatic increase in DR. A high DR balance signals underlying issues. Does the business have problems bringing products to market? Is there something wrong with distribution?

Second, it helps us value SaaS businesses that might show little in current revenues. If we ignore deferred revenue, we could miss a good software business. Accounting for DRe allows us to model what the company may earn once it’s collected on those revenues.

Next week we’ll look at Accrued Revenue.

If you have any questions feel free to reach out.

Other Income Analysis: Spot Danger Before You Invest

, ,

Last we discussed R&D costs and whether we should expense or capitalize such investments. You can check that out here.

This week we’re shifting gears and focusing on a reader’s request: Other Income (OI, for this article)

I love this topic and can’t wait to dive in.

Let’s go!

What Is “Other Income”?

Remove line items on an income statement and you see nothing but revenues, expenses and bottom-line income. If that were the case, we’d have no hierarchy of income streams.

But GAAP does separate the income statement by line items. And we do have a hierarchy of income streams.

This is where we see the term other income. What is OI? In short, its earnings generated outside normal business operations. The keyphrase here is normal business operations.

Such earnings include:

    • Interest
    • Gains on investments
    • Sale of long-term fixed assets
    • One-time gains on credits

It’s one thing to know this exists. But why does it matter?

Why Should We Care?

It’s important to know where and how a company generates profits. The more a business earns outside its normal operations, the less reliable the income.

If a company can’t generate profits from its core business we shouldn’t buy it. But many investors buy companies that look cheap thanks to one-time other income gains.

Why do they buy them? Quantitative value screens. Think about it. Screens don’t discern between “OI” and “core income”. All they see is the bottom-line number. A company that generated a large OI gain will look cheap on that basis.

“OI” Example: Facebook (FB)

Let’s use FB as an example for OI. Check out FB’s last two years’ income statement numbers (via TIKR):

The line items highlighted in blue are sources of OI.

You can see that it doesn’t matter in FB’s case. They generated nearly $24B in core operating income in 2019. $800M in OI is a simple rounding error.

But that’s not the point. The point is to train your eye to look for these line items. Because not every company looks like FB.

How To Gauge Other Income Severity

OI analysis is important. It tells us whether a company generates most of its earnings from its core business. Or if other (unreliable) streams of income bolster the bottom-line.

Here’s a quick heuristic to gauge whether a company’s OI is something to worry about.

OI Ratio: “Other Income” / EBIT

This ratio reveals how much of a company’s total operating income comes from its other earnings sources. Higher ratios signal less reliable income streams.

But again, everything comes with a caveat.

The best way to run this ratio is over time on a rolling basis. We shouldn’t punish companies for true one-time asset sales or restructurings.

Examine this ratio over the last five years. What’s the trend? Is it consistent? These questions help pinpoint exactly how the company truly makes their nut.

If you have any questions feel free to reach out.

Capitalizing R&D: Learn What GAAP Fails To Teach You

, ,

Last we discussed Accounts Payable and the role they play in cash flow efficiency. We learned that one company’s A/R is another company’s A/P. This week we’re analyzing one of the more complex accounting problems: capitalizing research & development (R&D) costs.

We’ll learn the difference between capitalizing vs. expensing, why it matters and how it changes margins, profits and returns.

Why We Should Capitalize R&D Costs

US GAAP accounting requires companies to treat R&D as expenses on the income statement (P&L). That biotech company spending $1M developing a new drug? Under GAAP that’s a $1M expense, not an asset.

This distorts profits, margins and ROA calculations for many tech/R&D heavy firms. That’s not ideal. Think about it. Companies that leverage R&D to generate revenue and cash have periods of massive R&D investment.

If you don’t adjust for R&D, the income statement looks like a bouncy ball on cocaine.

Investments in R&D support the long-term cash-flow generation of a company. They’re operating assets. As such, we should capitalize them on the balance sheet.

Capitalizing R&D is also a more conservative estimation of a company’s returns and profitability for two reasons:

    1. Capitalizing increases a company’s total assets (which could reduce ROA/ROE)
    2. Capitalizing keeps an amortization expense on the income statement

Let’s learn how to do that.

Capitalizing R&D Investments (Step-By-Step): Boeing, Inc. (BA)

We’re using Boeing (BA) as our example. Remember, we’re keeping things simple with this example. You can get more complex (see here), but that’s not our goal.

Step 1: Find Amortization Duration 

The first step in capitalizing R&D is to determine the useful life of a company’s R&D assets. The longer the usefulness, the longer the amortization period (i.e., how many years we expense). High-tech companies in ever-changing industries have low amortization periods.

BA makes airplanes that have long lifespans. We’ll use ten years.

Step 2: Calculate Value of Amortized R&D Asset

We calculate the amortized value of the R&D asset by using the straight-line amortization method. This means we amortize 1/10th of the total sum of the last ten years’ R&D expense. If we chose 7 years, we’d amortize 1/7th of the total sum of the last seven years’ R&D expense.

Then we calculate the annual amortization expense by dividing that year’s R&D by 1/10th (0.1).

Here’s what that looks like:

Step 3: Recalculate Book Value of Equity

We need to add our capitalized R&D value to tangible book value to get a better picture of the company. This is simple. Take Tangible Book Value and add R&D value.

Here’s what that looks like in 2018:

$339M (BV) + $3.27B (R&D value) = $4.059B

Step 4: Recalculate EBIT

Next we adjust EBIT to reflect our capitalized R&D investments. To do this, we take our current year EBIT, add current year R&D expense and subtract total amortization of R&D.

Here’s what that looks for BA in 2018:

$11.6B (EBIT) + $3.27B (R&D) – $3.71B (Amortization) = $11.16B Adjusted EBIT

Why does EBIT drop after adjusting for R&D investment? BA’s decreasing R&D investment (see above).

To Be Continued …

I know, capitalizing R&D is confusing. But it’s important if we want to understand the true economics of an investment-heavy business.

We’ll review this concept in more detail next week, don’t worry.

If you have any questions feel free to reach out.

Unity Software: Breaking Down S-1 Filing

, ,

Unity Software released their S-1 filing yesterday morning. I spent the afternoon digging into the filing pulling out what I thought interesting, fascinating and down-right awesome. Here’s the link to the S-1 if you want to follow along at home.

The below post is the unrolled version of my earlier Twitter thread. You can find that here. Let’s get to it.

Unity (prop ticker: $U) filed their S-1 this AM. I read the filing so you don’t have to (but you should). Here’s a thread on what I found interesting, fascinating and down-right incredible from the company. I’m starting from zero.

High-level Facts & Figures

These high-level stats impressed me:

  • 1.5M monthly active creators
  • 50%+ of all mobile/PC/console games made w/ Unity
  • 3B app downloads per month (!)
  • 15K new projects made per day
  • Creators span 190 countries

Business Motto: “We believe the world is a better place with more creators in it. Creators, ranging from game developers to artists, architects, automotive designers, filmmakers and others, use Unity to make their imaginations come to life.”

Power to the creators.

What Makes Unity Great?

Offering: $U provides set of software solutions to create, run and monetize real-time 2D/3D content for mobile, tablets, PCs, consoles and AR/VR.

End-Result: A fundamentally more engaging and immersive gaming experience. Instant adaptability.

What Makes Unity Great (pt.2)

$U offers tremendous scale. They can service huge development teams as well as individual creators.

Individuals can rapidly create and tweak new games (increases adoption)

Once built, $U games can run on 20+ platforms (zero barriers 2e)

A Two Solution Biz Model

Create Solutions: Lets users (artists, devs, engineers, etc.) create 2D/3D interactive, realtime content

Operate Solutions (more interesting): Offers cust. ability to grow & engage with user-base, as well as run/monetize content.

How Unity Monetizes Two Solutions

Create Solutions (CS): monthly subscription

Operate Solutions (OS): Rev-share and usage-based

The reason: Generate revs as builders develop content and ALSO generate revs as their game grows and becomes successful.


Create Solutions: 43% of Revs

Five plans customers can buy with Create Solutions:

  • Unity Plus
  • Unity Pro
  • Unity Enterprise
  • Unity Personal
  • Unity Student

Customers typically buy 1-3 yr subscriptions, billed monthly, quarterly or annually.

Operate Solutions: 57% of Revs

Revenues from this segment are lumpy (depending on popularity of games)

Unity relies on expanding their customer base to smooth revenue volatility

Aligned incentives: U generates more revenue as their customer’s game gains popularity

Let’s Talk About TAM, Shall We?

Unity sees their TAM around $29B across both gaming and other industries.

They’re looking beyond gaming and their current markets.

Future TAM could be (according to mgmt) “multiple times larger than the opportunity today.”

Rapid Revenue Growth, Shrinking Operating Losses

Unity grew revenue 42% YoY ($380M to $541M).

They’re currently losing money ($163M last year and $131M in 2018).

Net cash used in operating activities decreased from $81M in 2018 to $68M in 2019.

The Power of Real-Time 3D

Unity is banking on the power of Real-Time 3D to drive creative innovation, more engaged users and better games.

They break down the power of 3D:

  1. Interactive: Connect with content & other gamers
  2. Real-Time: 120 images per sec
  3. 3D: more real

Technology Driving Real-Time 3D Adoption

Rapidly evolving tech allows $U to create better, more realistic and interactive games than ever before.

This new tech relies on these four pillars:

  1. Computing power
  2. Platforms & Devices
  3. Distribution
  4. Connectivity

Unity Growth Strategies

$U has identified five ways they can grow in the future:

  1. Invest in product innovation
  2. Grow existing gaming customers
  3. Grow new gaming customers
  4. Grow beyond gaming
  5. Grow across global markets

See more below:

Diving Into Financial Data (What Stood Out)

In 2019, $U generated $541M in revenue and $423M in Gross Profit (>78% GM!!)

Shares Outstanding: 114M – Customers > $100K revenue: 600 in 2019 vs. 484 in 2018

Dollar-based Net Expansion Rate: 133% in 2019

Diving Into Financials (Balance Sheet)


  • Cash: $453M
  • Working Capital: $339M
  • Total Assets: $1.29B


  • Deferred Revs: $107M
  • Total Debt: $124M


  • Shareholders Equity: $647M

Interesting Graphs on Customers, Retentions Over Time

Check out these graphs on:

  • Customers > $100K over time
  • % of Revenue Rep by Customer > $100K over time
  • Dollar-based Net Expansion Rate over time
  • Customer Cohort Analysis

Letter From Unity CEO

I liked how the S-1 included a letter from $U CEO.

Here’s a couple interesting quotes from that letter:

Unity Reviews from Other Game Studios

The consensus: Unity allows game developers to rapidly create, test and curate new games. This in turn fosters a more creative environment, which leads to more potential blockbuster games.

Customer Diversification

The company has myriad customers using both its Creative, Operate and Architecture Solutions, including:


  • Arena of Valor
  • Iron Man VR
  • Pokemon GO


  • EA
  • Tencent
  • Ubisoft Mobile Games


  • Samsung
  • Skanska (Volvo)

Executive Management Comp

  • CEO: $8.4M total compensation (owns 3.4% of company)
  • CFO: $5.23M total compensation

Shareholder Split:

  • Sequoia Capital owns 24% of shares
  • Silver Lake Partners owns another 18.2%