An Overview of Technology Media and Telecom (TMT) - Part 1 of 2
Mod Note: Each day we'll be posting the top WSO forum posts of 2014. This one was originally posted on 6/29/14 and ranks #20 for the year by total silver banana count. You can see all our top ranked content here.
Given the positive response to our post on FIG, we are starting our TMT overview. Part 1 will cover Semiconductors, Telecom, Hardware and Internet. The second piece will cover Networking, Software and other technology.
Technology Media and Telecom (TMT) is an interesting space particularly for those that enjoy seeing valuations that can range from "WTF was that" (internet) to a classic DCF model (older technology companies). With that in mind, it is best to break the sector into smaller sub-segments to understand how and why certain companies obtain higher multiples than others.
While there are hundreds of ways to break down the sector, we'll go with the following subsectors: 1) Semiconductors, 2) Telecom, 3) Hardware, 4) Internet, 5) Networking, 6) Software and 7) other components.
Before we dive into each subsector, we should note that technology focuses less on various equations (which FIG relies on heavily as you can see from our previous post) so don't be surprised if you don't see a lot of calculations.
With that said… what are the key points to understanding tech?
1) Total Addressable Market (TAM): For each company you need to understand how big the opportunity is. It can be as simple as knowing how many Laptops and Desktops are shipped world wide to as large as "how many internet users are there on the planet". Using this same example, the PC market (desktops and notebooks only) could be a ~$200 billion revenue opportunity with total PC shipments of roughly 320 million… In comparison? If you look for the Internet, you can get to a TAM of $14 trillion… We don't have to explain why everyone fights about a firm's Total Addressable Market, it matters… a LOT. Now you understand why bankers fight tooth and nail over positioning a company within a specific sector/comparable group in Technology.
2) Growth rates: In lock step with the above point, you also want to know if the TAM is growing or shrinking. Piggy backing on the same example above you would quickly understand that it would be better to have exposure to Internet for growth because users continue to increase while the number of desktops and notebooks are flat to down as people move to their mobile phones and tablets. If the PC market suddenly sees innovation and the TAM is expected to grow to $300+ billion, you can bet anything that multiples will expand for companies with PC exposure (HPQ as an example). Similarly, if the number of internet users suddenly doubles for a specific company, say Facebook doubles its user base, get ready to see the stock move up as well.
3) A Real Technology Lead: This is where the engineers and techies step in. Lets say everything checks out so far, you have a product in a sector that is growing and has a large TAM… Now the rubber meets the road. Is this product differentiated? You should pound that question into your head after you've done a great job looking at parts 1 and 2. If you have a product that can be easily copied by a larger competitor because you have no Intellectual Property well… you are screwed (half-joke tone).
Side Note for Tech Nerds:
Differentiation is vague. We're using Intellectual Property as the example since it is the easiest to relate to but it can mean a lot of things. It can also mean they have an extremely large install base and were a great first mover (think WhatsApp /SnapChat/ Uber).
We're not going to use this post as a time to debate the valuation of any of these since they do have specific technologies associated with them, but to give you an idea WhatsApp was able to process 27 BILLION messages in a single day with JUST 55 employees. You'd have to be a tech nerd to understand how crazy this is because the stick rate on low-end international phones has to be sky high.
Tangent - remember that differentiation/technology lead matters a lot and just because you see a high price tag doesn't mean it was always a bad deal (remember Google acquiring YouTube?). In addition, we strongly doubt anyone reading this is smarter than Mark Zuck in terms of understanding tech, that includes the authors of this post, so if you see an expensive acquisition try to understand the technology before jumping to any conclusions.
We'll go into specific valuations for each subsector but the main ones you should know if someone asked you inside an elevator are as follows: 1) EV/Sales for companies with high growth and no profits yet – think companies you would take public, 2) P/E – main benchmark valuation methodology for medium to larger more establish companies, 3) FCF multiples, for large cap tech, think IBM/MSFT/AAPL/ORCL etc of the world 4) EV/EBITDA usually for internet based companies once they are relatively well established and 5) DCF: for the large cap high cash flow names.
With the introduction out of the way go ahead and jump to your sector of interest and take a glance.
Sub Sector Analysis
1) Semiconductor Overview (INTC, AMD, NVDA, ARMH, BRCM, TXN)
Semiconductors can be explained in many different ways, however, to boil it all down semis are the integrated circuits which act as the critical electronic components needed in system designs. Semis can serve in various functions from microprocessors, to memory, to analog, to logic to discretes and sensors. In general, semis account for ~30-35% of the BOM in electronic systems (think servers, smartphones, computers) etc. The total market was just over $300bil in 2013 with growth generally pegged to WW GDP growth.
What Matters in the space? i) Inventory Cycles – the space has traditionally been traded off of inventory cycles, mimicking ~6:2 ratio of builds to subsequent correction. Investors want to own through the early build quarters (6) and sell through the late build-early correcting quarters (2). ii) Return of FCF: The growth phase is transitioning into a more mature era, and slower unit demand is leading to excess capacity and broader industry underutilization, which we expect to persist through 2016. Given this backdrop, lower capital investments are needed as today's capacity is sufficient to meet demand. Consequently, we are likely to see greater FCF generation in the industry with companies endorsing shareholder-friendly capital-allocation strategies. iii) Newer technologies: Rising technical standards (i.e. LTE) are demanding greater complexities out of solutions where system design is increasingly challenged. Semi companies that are solving these problems are poised to benefit from both a rising unit and ASP perspective – a trend (upwards unit and ASP slope) that is most appealing to semi investors looking for revenue growth.
Main Industries to understand from a high level.
Computing: In 2014 the market is estimated at $100 billion, -9% Y/Y accounting for 32% of total semiconductor revenues. Impacted companies: Intel, Marvell, Micron, SanDisk, Texas Instruments, NVIDIA and Broadcom.
Consumer: In 2014 market will be ~$55 billion, +5% Y/Y, accounting for 17.5% of total semiconductor revenues. In general, the Consumer market benefited from rising content trends, including more wireless connectivity ("the internet of things", TVs, household appliances), higher penetration of general consumer electronics in emerging markets, mix-shift to faster internal processing (embedded SoCs) and rising touch-enabled devices across multiple platforms. Impacted companies: Broadcom, Marvell, Cypress, Freescale, Texas Instruments, Analog Devices.
Semiconductor Wireless: In 2014 the market is ~$75 billion, +9% Y/Y accounting for 25% of total semiconductor revenues. While the growth in handsets and tablets are slowing at the high end, the market is seeing considerable uptake in the mid-range and lower end, which have become increasingly feature-rich.
Semiconductor Wired: In 2014 the market will be ~$15 billion, +2% Y/Y accounting for 4.5% of total semiconductor revenues. Within the data center, 100GE adoption is beginning to materially ramp as Ethernet content also benefitted from growth trends. Other components that are slated to benefit are network and multi-core processors.
Auto: In 2014 the market is $25+ billion, +31% Y/Y, accounting for 9.0% of total semiconductor revenues. The Automotive market stands to benefit from higher units (industry expectations +4% Y/Y) as well as semiconductor content. In 2014, total unit growth is expected to be ~4% growing to 85 million driven by increasing wealth in emerging markets. The content increases are largely due to 'infotainment' and safety systems which are benefitting analog ICs and MCUs.
Theme #1: Enterprise Data: NAND Use to Broaden into Higher Value Applications
Theme #2: The Rise of Sector FCF
Theme #3: RF Integration
Valuation: Traditionally, the semiconductor industry has valued the space off price-to-earnings (P/E) multiples and EV/EBITDA multiples, as operating profits were always paramount in evaluating the effectiveness of on-going operations. This valuation method was prudent given the growth nature of the industry. However, the growth phase is transitioning into a more mature era, and slower unit demand (particularly in trailing edge) is leading to excess capacity and broader industry underutilization, which should persist through 2016. Given this backdrop, lower capital investments are needed as today's capacity is sufficient to meet demand. Consequently, we are likely to see greater FCF generation in the industry with companies endorsing shareholder-friendly capital-allocation strategies. Through this dynamic we should see five main characteristics unfold: i) lower capex as a % of sales; ii) rising industry FCF/share; iii) semiconductor leadership in FCF growth within the S&P500 sectors; iv) increasing emphasis to drive FCF vs. EPS; and v) rising P/FCF multiples across the industry, where the sector is trading at about six turns below the S&P500.
Financial Statement Dynamics: For semiconductor companies, the three most important criteria to follow are those impacting revenues, gross margins and CapEx cycles.
Revenues: As previously stated, industry revenue is generally pegged with GDP, and perhaps up to 5% on a forward basis. Revenues are impacted by ASPs (on average ~$0.45/component), units shipped (which may include distribution sell-through), mix-shift impacts (which affects ASPs) and product cycles. Mix shifts could be impacted by a move to higher technology standards (think 3G to LTE), increasing integration (think an integrated baseband + applications processors in mobile) or increasing software functionality (think GPU processing into big-data and cloud applications). Regarding product cycles, it is important to take into account seasonality which could lead to greater than expected volatility in the revenue line (think AAPL iPhone and Samsung Galaxy s4 ramps).
Gross Margins: Industry gross margins are ~45-50% range. In addition to revenues (see above), the moving parts of the gross margins are wafer costs, depreciation costs (GAAP basis), and cost efficiencies (think integration). In addition, foreign exchange may play a role if costs are overseas.
CapEx: Semi companies are either operate their own fabs (think Intel, Texas Instruments, and Samsung) or fab-less (majority of industry). A fab is where a semi company will design and fabricate its own integrated circuits. The cost of new fabs range from $2-$10bil (where costs are recognized using SL depreciation, usually over 10years). Thus CapEx cycles are important in assessing FCFs. The major foundries in the industry are TSMC, Global Foundries, UMC and IBM.
2) Telecom Overview (T, VZ, AMX)
This space is on the borderline of technology as well but we'll include a short summary of the space for high level educational purposes. Everyone on these boards is familiar with the major telecom companies who make up the space IE: AT&T, Sprint, America Movil etc. The reason we consider it on the "border" is when most people think of tech they think of IBM/AAPL/MSFT etc, not major CAPEX intensive companies like T/VZ. We did not want to bucket this space under networking (CSCO, JNPR etc.) as it would be confusing and incorrect.
Main Items to Understand: The answer here is quite clear. You want to know how many post paid and prepaid adds a company is acquiring. Naturally, post paid adds are more important. What you would like to see is - AT&T for example - adding multiple post paid subscribers (recurring monthly revenue) and to a lesser extent seeing the Company sell a few of the cheaper junk prepaid phones to help with cash flows a tad.
Quick Sub Segment: Keeping it simple here again, there are two main segments to look at Wireless/Wireline. You can pull up a recent filing by T to pull the majority of the information necessary to understand the two sectors. In short wireless is going to be your phone plan (could include tablet plan – if you're talking about gigs of data purchased instead of WiFi only) and wireline is going to be your TV/Internet subscription.
Top Numbers to Know:
Smartphones: Roughly 1 billion smartphones in 2013/2014 This is expected to grow in the at ~20%.
Tablets: Looking at 200M+ units (annual) as tablets replace PCs in the corporate/consumer world ("cannibalize" is the term they use) arguably. Arguably growth in the 20%+ range
Internet Users: About 2.5-3.0 billion
Financial Statement Dynamics: For the first time we're going to lead with a DCF/dividend valuation method and not have it as a joke comment. In general a DCF/dividend model can be used as part of the valuation because we are talking about enormous carriers that will unlikely go bankrupt and have been established for a multitude of years. Assuming user/subscription growth is solid, looking for 1) DCF on FCF, 2) EBITDA multiples, and 3) P/E Mutiples
3) Computer Hardware Overview
Many of you will be familiar with the hardware names (HPQ, AAPL, IBM etc.). In this space your larger players are essentially selling a hardware product such as an iPhone, Server or PC and layering on software/services to keep you coming back for more and generating recurring revenue (emphasis on operating systems such as iOS/Android etc.). Notice we are already mixing software with hardware causing a bit of confusion… but lets focus on the hardware side of the business and what you should know.
What Matters in the space? As noted, at the top of the post the focus is going to be on growth and a technology lead. In this case you can start with 1) growth which will be related to how many units or "volume" of the product you can sell [simplistically this would be iPhone units for a company like AAPL], 2) TAM expansion [another simplistic example would be how the smartphone market will grow in the future] and 3) hardware focuses a lot on margins, because higher margins will be a proof point that their technology is leading edge.
Before moving on, to expand on the last point, think of it like this. If Company A has 10% operating margins but Company B has 40% operating margins and they sell the exact same product and the exact same unit volume… There must be some sort of technological/operational edge because they are charging a much higher price than Company A. So in short, margins are eyed keenly in the hardware space, hence all the news articles speculating on gross margins at IBM, AAPL's iPhone/iPad and all of the Build Of Materials or "BOM" break downs you see on technology blogs.
Main Industries to understand from a high level, with example companies in brackets
PCs [HPQ, Lenovo]: As stated above, you're looking at a ~320M unit shipment industry with revenue of $200B (annual). As a rule of thumb the space is expected to be flat to slightly down as tablets become more and more widely deployed
Servers [IBM/HPQ]: You're looking at ~9-10M units with total revenue in the $50B+ range (annual). This space is also seeing contraction/flat unit expectations as it has matured. A signal? IBM's exit of its x86 business.
Smartphones [AAPL/BBRY/Samsung]: Lots of growth here. Depending on how you slice the market (again remember TAM slicing is important) you are looking at 1+ billion units and revenue of $250B+ (annual). More importantly? This is expected to grow in the future with expectations for annual unit growth in the 20%+ range.
Tablets [AAPL/MSFT/Samsung]: Again growth here. You guys are familiar with the major players, to give the jist $50B+ revs, looking at 200M+ units (annual). As tablets replace PCs in the corporate/consumer world ("cannibalize" is the term they use) arguably Tablets will be the biggest growth item within the "PC space". Notably, we have only included desktops and notebooks in the first bullet as PCs to highlight the difference. Arguably growth in the 20%+ range
Storage [EMC/ORCL/Hitachi - Private]: Here is where we get more tech focused and don't want to explain the difference between Network Attached Storage (NAS) and Storage Area Networks (SAN)… Have we confused you yet? Essentially all the stuff you create, all the words on this screen, the nudes you send on SnapChat are all eventually screen "shotted" and saved somewhere. IE - Storage. This is put into a data center and industry analysts try to analyze both 1) how much data is stored in terms of Terabytes and 2) how much revenue will be generated from the purchase of storage items. These numbers are honestly mind boggling, you're looking at ~40+ Zettabytes of storage by 2020 in a ~$25 billion market. Again don't want to get into the technicals because this space is highly confusing for a beginner but improvements to storage "efficiency" ie how much can be stored on a single device offsets a lot of the growth, so call it low to mid single digit long term growth.
Notably, we could include more subsectors here, for example… by splitting the phone market into smart phones and low end cheaper phones without internet access. We could also begin breaking up each segment, such as business only PC sales (ie: how you're reading this post) versus consumer PC sales (ie: how you watch stuff that is blocked on your computer at work) to show growth rates in an enterprise versus the regular consumer/retail market. For an overview of hardware, the above units should be just fine.
Valuation: For the hardware sector you're looking at the same metrics we outlined for tech in general, however you should be more familiar with 1) P/E metrics both on a GAAP and non-GAAP basis, 2) DCF for Large Cap hardware, 3) Sum of the Parts and 4) EV/Sales for growth hardware.
P/E: This is self explanatory, however the point is earnings growth is eyed a bit more in the space as companies are expected to grow EPS faster than sales (assuming we're talking about larger companies)
DCF: Again focused on large cap tech, as your EPS becomes larger, companies look for quality earnings. As a rule of thumb GAAP net income growth should begin to mirror FCF growth, so a DCF becomes more viable at this point.
Sum of the Parts: As you can see, many of these hardware companies will inevitably have software and services as line items. Sticking with Apple again, their iTunes would be an example of non-hardware related sales which should be valued differently from the sale of the actual hardware item (ie: it can be considered more of a software or services line item). Here you're valuing each piece with a different multiple, add them up and see if it reflects the Company as a whole.
EV/Sales: When you have no profits and you're operating at a loss… You better grow like a weed. This is where EV/sales becomes crucial. The idea is that if your company is growing at 50 or even 100%+ that you have a major technology lead, a major company could then become fearful of you and buy your entire company. This is where you see the "wtf multiples" come into play and sales multiples are much more common for companies that are going public (TWTR as a clear example).
Financial Statement Dynamics: For hardware, as noted above you'll usually see revenue breakouts by segment. This is called a "bottoms up analysis" where you look at each segment and give a detailed look at how it is going to grow in the future. Sticking with AAPL since it is going to be easy for readers to understand, you would model out revenue for iPhones and iPads differently and you would also have specific unit volume estimates for iPhones and iPads as well, since they will unlikely grow at the same growth rate over the next 12-24 months.
In addition, if you think of a Company like IBM, you would have to look at the margin structure by business line as well. Considering that the server sales (hardware) would have lower margins due to cost associated with it, but their software product line would naturally have significantly higher margins. So an alternative way to model would be Revenue and Margins by segment, to build up to a consolidated corporate gross/operating margin line item.
Finally, from a balance sheet and cash flow perspective. Since it takes a lot of cash to manage inventories of all of these products and new machinery needs to be made to test and develop new products, CAPEX (cash flow) and Inventory (balance sheet) changes are eyed.
This is far and away the easiest sector to understand in terms of what a company does and what the value proposition is. Negatively? It is also quite difficult to value as growth rates can climb or sink at a moments notice. This is the Ferrari of Tech, high risk high reward. If you want to take a look at some disasters look no further than GRPN or ZNGA… A turnaround? FB… Wild success? OPEN, recently acquired.
Since the space is growing at rapid rates lets go ahead and look at the high level trends that people need to be aware of:
More Users, More Money: Roughly 2.5-3.0B internet users on a global basis. This is going to benefit the space as more users means more revenue. This is relatively simple to understand and industry analysts attempt to size the total internet business which leads to numbers in the trillions of dollars ($4T+)
Move to Mobile: Simply put, people are interacting more with their mobile devices than they are with legacy desktops and notebooks. This moves in lock step with the above sector analysis on hardware, the growth is in tablets and smartphones. Therefore? Investors/bankers are looking at what percentage of revenue companies generate from mobile sales versus desktop/notebook. So if you're looking to sell your company, or take it public, you should have a clear explanation of a mobile strategy to help monetize your company. This will help you understand how a company like Facebook sees its valuation move up as mobile is a higher % of revenue and help you understand why Uber receives such a high multiple (all mobile revenue).
E-Commerce: Here is a space that can mix a bit with consumer bankers/investors as well. If your company sells products both online and in retail stores, the different businesses may need slightly different valuations. A good example of a clear internet e-commerce business? Gilt. We all remember Zappos as well (acquired by Amazon). This is incredibly simple, but as sizing preferences are established people will shop more online instead of at retail stores (doesn't take a genius to figure out how that will help margins!).
Social: More and more people are purchasing based on the preferences of their friends. IE: if your two best friends all buy XYZ product you're much more likely to buy it on their recommendation. Enter… Facebook, twitter, etc. This space is also seeing a lot of the issues talked about above come into play (move to mobile, people looking for increases in users etc.). A long post on Facebook is already up so no need to go further, you can check it out here.
High Level Numbers: Since this space is growing rapidly, the numbers are more difficult to pin down when you compare it to the other technology sectors (hardware, semis, components, etc.) but lets go ahead and write down some back of the envelope numbers:
- Internet Users: about 2.5-3.0 billion
- Internet Value: $4-8 trillion (obscure high level, McKinsey values the internet at ~$8 trillion – 2011 data)
- Total Internet Advertising Market: ~$50B+ annual
- Mobile Payment Market: $70B+ annual
- Online Gaming Market : $30B+ annual
Valuation: Here is where the fun begins… Valuing internet stocks. Since a lot of these hyper growth companies are going to be driven by users and many don't turn a profit you can look at 1) Sales growth, 2) User Growth, 3) EBITDA metrics, 4) P/Es and 5) FCF. As you move to larger and larger companies you would go to valuation metrics 3-5 (google as an example) but many of these companies do not have high (or negative) earnings so sales becomes the primary metrics along with user growth. No need to take a deeper look here since we'll be debating for days on the most important metric for each stock.
Financial Statement Dynamics: The long story short here is Growth. You're tracking more metrics that will impact the top line such as MAUs (Monthly Average Users) and use the user growth rate as a leading indicator for future revenue growth.
Conclusion so far?
If you want to have a solid career in tech, of the four items listed above you're likely best off in internet as it stands today. Yes there is going to be debate about what sector to join but essentially you want to join a sector where there is a lot of growth. Internet and Software would be the two biggest ones. We'll cover software in part 2.