NetEase (NTES)

Profitability and Tech is always an interesting question. In analyzing the US tech giants, generally the business model, such as in the case of Google and Facebook, came as a result of getting to scale quickly, reinforcing network effects with very little focus on initial cash flow generation. Monetization would come only later as the platform has already established itself as a gateway monopoly on its users attention span. I won’t bore readers with further analysis on these tech aggregators, Ben Thompson over at Stratchery already does a fine job covering this topic.

But what about companies that had intense competition and a lack of ready venture capital? how do companies get the capital to build themselves to scale? What if they don’t have an effective monopoly on consumer attention span? What do they use to stem the cash burn? In thinking about these questions, I came about two type of cash engines that large “tech” firms and in fact others before have used to power themselves into scale in the US and China tech ecosystem. The two engines are essentially the positive working capital cycle and video games.

The WC Cash Engine
Following the playbook of every retail giant before, from Sears to Walmart: with scale, one significant advantage is that you now have enough bargaining power to pay suppliers 30 days while your customers pay you upfront, this working capital cycle than can generate the free cash flow needed for you to get to scale so long as you grow by heaps and bounds. Amazon, Alibaba and JD essentially were able to get to scale of this business model. Amazon has  been self funding since the early 2000s though operating cash flow generation (of about USD10 billion/year today). The cash gets spent on reinforcing its scale through logistics build out as well as to funding new ventures such as AWS. To the same extent, this is how the ecommerce platform Alibaba and JD have also built themselves up to scale.

The Video Game Cash Engine
To study Tencent is to study video games. For most of its history, Tencent was playing second fiddle to portals such as Sohu and Sina. For close to a decade, its QQ messenger platform was largely an AIM clone with no effective monetization. Yet the company was an early pioneer in terms of microtransactions on gaming platforms. Its video game segment was highly profitable and it engaged in a decade of both acquiring and developing its own video games. The cash flows generated by this segment went towards the development of the Wechat platform when the mobile wave arrived in the 2010s.

Existing Cash Engines
In studying Amazon and Tencent, we see that both companies had different but strong cash engines. The engine had two purpose, one which was to make them into quasi venture capital funds, as they both were able to deploy capital into ancillary businesses which become billion dollar enterprises (AWS in the case of Amazon and Wechat in the case of Tencent). Another was that the cash flow provided a cushion in period of distress, such in the aftermath of the dot-com cycle for Amazon and in the case of Tencent, keep it relevant in the face of strong competition. As a result, history has shown these cash engines create both upside optionality and downside protection from a risk reward perspective.

As we are interested in outperforming investments, are there any second tier incumbents today who have similarly comparable cash flows models which created embedded upside option value and downside protection? The two that comes to mind today are JD and Netease. JD is essentially playing the same game as Amazon, it generated Q2 TTM FCF of ~USD4.5 billion, compared against ~USD10 for Amazon, the cash is recycled into building out logistics centers, automation, acquisitions (Dada, BitAuto, etc.), all with a single minded purpose of getting to massive scale. This was already a long term position of mine and Richard over at Value Ventures already has done some very high quality coverage on JD. But another cash engine often neglected in reference to Chinese TMT is Netease.

NetEase Financial Profile
As my thesis is centered around the cash engine, it is important to go through the financial profile of NetEase. The company today has a ~USD40 billion market cap. It has ~USD6 billion of cash on the balance sheet and has consistently generated USD1-2 billion in FCF every year (more than USD2.2 in 2016 and on track for ~USD2 billion in 2017).  As a result, the company is trading at a very reasonable 15x EV/FCF.

The History of Change
Typically, capital markets discount Chinese companies relative to their U.S. peers as a result of a focus on governance, VIE structuring worries, country risk and perceived notion of copycating without any real innovation. What I found as a technology watcher over the years is that there are some real positive points to Chinese companies outside of growth. One example is that Chinese management tend to be more mentally flexible and adaptive to change in a way that one simply does not see in the U.S. Outside of the notable example of Apple, there are only a handful of companies that were able to right the ship when industry/market changes have been eroding the original core business model. Netease’s history demonstrates one of one of the most successful pivots in modern corporate history.

Following a stint as an engineer for U.S. based Sybase, the then 28 year old William Ding created 163.com as a free email service based on the RMB500,000 he had earned for developing a telecommunication billing software. Initially setup in 1997 as a hotmail clone when China only had 600,000 internet users, the company quickly began to copy the yahoo model of internet portal aggregation which included a search engine, news, chat services and web hosting. By 1999, with 6.7 million internet users, the web was taking off and Netease was generating close to 4.2 million page views and had 1.7 million registered users for its e-mail services.

The Dot-Com Era(1997-2002)
By 1999, the country had just opened its technology sector to foreign capital and Netease was able to raise USD70 million and list on Nasdaq together with the two other popular portals, Sina and Sohu. However by 2001, the Nasdaq bubble had burst and the shares in Netease had fallen by over 95%, from a peak of USD3.4/share to 17 cents/share. To add to its troubles, there was discovered accounting irregularities that led to the resignation of the then CEO King Lai and COO Susan Chen. William, having briefly become a paper billionaire only to fall back into obscurity, took back control of the company and over the objection of his board, took his company into a direction that would bring it back into relevance.

The PC Era (2002-2009)
In 2001, the company developed Fantasy Westward Journey, one of the first desktop based MMORPG in China. Inspired by the 16th century Chinese novel, Journey to the West, the game became an instant hit, quickly scaling user base by more than doubling every year. By 2005, the game had 25 million registered users and was the largest and most profitable game in China at the time. Buoyed by the successful pivot into game development, Netease began to develop a host of other titles such as Tianxia III, Heroes of Tang Dynasty Zero and Ghost II. Seeing the success of the MMORPG genre in China, Netease began its partnership with Blizzard for the which it is best known for, operating World of Warcraft in China, which at its peak had 20 million users, accounting for approximately half of all World of Warcraft players globally.  By 2007-2009, the company was generating USD300-500 million in net revenues and earning half of that in profits. However, it wasn’t until the Mobile era did the company really hit its strides.

The Mobile Era (2010-2017)
By 2010, the company was dominant in the Chinese PC gaming space, but the market had started to Change. Smart phones were starting to become popular and the market for mobile games had opened up. At its start, the market was fragmented, with casual games dominating the charts such as Happy Element’s Anipop and Happy Fish and Idreamsky’s distributing foreign titles such Temple Run, Subway Surfer and Fruit Ninja. Tencent had an early lead into mobile gaming, leading with titles Gunz Dash, Speed Up, Fight the Landlord and as time went on furthered its dominance with titles  such as Legend of Mir, We MOBA and Happy Lord. By 2015-2016, Tencent was dominating the top 10 charts of Android and iOS games which seemed further solidified with the purchase of Supercell in 2016 for USD8.6 billion. Netease on the other hand was playing catch-up. By 2014, the company had developed mobile versions of its popular PC games, such as Fantasy Westward Journey and Ghost Story, which remain hits to this day. In 2016, the Company generated a another home-run with the Japanese Heian-era themed RPG Onmyoji.  Today, the company is neck and neck with Tencent, with 4 of the top 10 mobile games by revenue in China, same as Tencent.

Gaming Going Forward
More Recently, Tencent has again taken back the top chart with its King of Glory title, a mobile version of the globally popular Dota styled game League of Legend made by Riot Games, which was acquired by Tencent in 2011. The game became a sensation overnight in China, with over 160 million monthly active players and 80 million daily active users. By May 2017, the game had become the highest grossing mobile game in the world. The game is on track to generate USD7 billion in 2017 revenues and is expected to contribute to 50% of Tencent’s 2017 mobile gaming revenue or 20% of its overall 2017 revenues.  But as is always in a highly competitive and attention driven market, Netease is has made a comeback in recent months with the development of Wilderness Survival(荒野行动) or Rules of Survival, as it is known in NA, a copy of the latest global hit PlayerUnknown’s Battleground (PUBG), unseating Tencent’s dominant MOBA from the monthly #1 charts. As we can see, Netease’s capabilities as well as ability to partner with foreign partners to operate their titles has made it every bit as competitive as Tencent, despite engaging in less blockbuster style acquisition of studios.

The Art of Capturing the Consumer Surplus
As I have talked about in my previous gaming sector primer, one of the beauty of the economics of gaming is the ability to price segregate. This is the so-called freemium model in which one can acquire lots of customers cheaply early on by providing the game for free and then creating in-game purchase only items which can make one more powerful, allowing the capture of so called “Whales” who can spend indiscriminately. Everyone is quite familiar with this model when it comes to gaming, and it has been receiving strong push-backs from the consumer base recently.

It turns out that there is great art and subtlety in the balance between monetization and user experience. It is important not to too empower the higher spenders which turn-off the rest of the user base as well as making small/non-spenders still feel like they are progressing. Companies that have managed this balance well are for example the popular Japanese mobile game Puzzles and Dragons as well as League of Legend, and Counterstrike:Source Zero, where generally the monetization is largely cosmetic and there is still a large skilled component to gameplay. U.S. game developers are late to this gaming, having primarily built themselves around a movie-like model and as a result have been too hamfisted in terms of shoving monetization down people’s throat, such as EA’s Real Racing 3 and the new Battlefront II. 

Chinese developers, such as Tencent and Netease, have had decades of experience now with managing large and dynamic user-bases and striking the right balance between monetization and user experience, simply because the propensity to spend was quite low in China in the early years and gaming in China is quite social, which require very good community management and a user feed-back loop. In the latest call with Netease, William stressed that monetization of Minecraft would come later, despite the massive 30 million customer base achieved in just two quarters, in order to build up the mod ecosystem and only then would the company think about monetization. I believe this is the right call, gaming companies need to think more long-term about games as building, managing ecosystems that keep its occupants happy. As such I am confident that the Chinese incumbents are thinking very systematically and quite adapted to the new realities of gaming monetization.

In fact I will even further and suggest that lessons learned by Tencent in managing the ecosystems of its games is what created the DNA that allowed it to build, manage and built on top of the world’s largest/greatest social network platform. The types of lessons learned from managing large scaled games is akin to running a small country with the ability to run economic and social experiments. Thus the types of insights gained into user psychology, utility and network structure are invaluable to the building of future consumer facing businesses. In fact, the pivot to video games for NetEase likely came as the result of an early tracking and understanding of consumer preferences and spending patterns. As such, I believe there is significant upside optionality in NetEease in various other industries/segments that are being completely discounted.

In a future article on Tencent, I will expand the idea that managing video games is much like running a country, why this may result in the creation of a 2.0 version of the East India Company and how this may create a post-political world run by technology companies. For a preview, think about the popular approval rating today of Amazon versus the current U.S. political system.

Upside Optionality in E-commerce and Food
On top of gaming, Netease also runs a fast growing foreign branded B2C platform Kaola.com. Since its launch in 2015, Kaola has become the largest cross-border e-commerce player in China, with about USD1 billion in GMV in 2016 and growing at 80-100%, on target to hit ~USD2 billion GMV in 2017. Netease is making a big push to ramp-up the Kaola platform, promising to purchase USD11 billion worth of inventory over the next three years from Japan, Europe and the US. With a market leadership position in the segment, it should not be surprising if Kaola is able to achieve USD8-10 billion in GMV by 2020, an impressive feat for a company with little e-commerce exposure just two years ago.

In addition, Netease is engaged in the pork farming business. The company created the subsidary Weiyang Farms for the raising of high end pigs in China using high quality feeds and sustainable methods. The company raised its series A from Dianping, Sinovation and JD totaling USD23.5 million. Pork is an important staple for the Chinese diet and recent auctions by the company have shown the ability to price high end pork at the same level as Kobe beef. While the premise may seem ridiculous to people, I believe the demand is there very large demand for organic, high quality pork in China and that the opportunity though still nascent, could become quite large in the coming years.

Valuations
I think in the case of high cash flow businesses with little expected growth, the market typically prices the company like a utility, typically somewhere between 8-12x FCF. For Netease, the utility approach would value the business at around USD29 billion, or at a 33% discount to the current market price. In a sense, the market is not pricing in a huge amount of upside to the business. But this seems extremely conservative for a company that has been able to successfully pivot to gaming and transition to mobile over the past 20 years, delivering returns to share holders in excess of 35%, compounded for the past 20 years. This goes back to my previous write-up on Nintendo and the influence of short term cycles on pricing. Wall Street rarely ever look at the bigger picture and narrowly focuses on short term quarterly earnings and outlooks, and as such cyclical hit driven businesses such as gaming will inevitable be undervalued in periods between the big hits, even if the hits are every couple of years. This is especially relevant as Onmyoji is tapering off in China and is only just now expanding into North America.

With that in mind, I want to approach it from a more long-termed perspective. On a very conservative basis, the gaming market in China is growing by 9.4% per annum, according to Newzoo. So assuming Tencent and Netease each continue to hold between 35-45% market share, the underlying gaming business of Netease should be worth USD40-70 billion by 2020. What is not being priced in is the Growth of Koala and Weiyang, which in my mind should be worth USD6-14 billion and USD1-3 billion respectively. This combined with the cash hoard of USD10-13 billion should put the value of of the business at between 60-100 billion market cap. On a risk adjusted basis, this is expected to generate a 1.9x CoC creturn and a IRR of 23% for three years, a respectable return over any index.

At the current price of around USD300-310/share, trading at around 20x TTM PE and 15x 2017 FCF,  while offering a 1.3% dividend yield, I consider the stock very attractive on a risk adjusted basis.

 

Gaming Sector Primer

One of the businesses you learn very quickly are the best compounders to own and pass on to ones children historically have been liquor distributors and cigarette producers. There is something to the low price inelastic segment coupled with an addictive product that produces natural scaled monopolies either on the product or the distribution side that simply allows for high ROE over extended periods of time. I believe this was the genius of Buffett’s investment in Coca Cola. Coca Cola’s business model was about creating distribution across the globe, to the point where it was easier to find a bottle of Coca Cola than a bottle of water in far flung places. With this massive distribution network, it was easy to simply acquire new brands or create new products and gain impressive topline and margin growth from pushing various products through your large horizontal channel.

One of the challenges with this horizontal distribution model is that it does not enable the capture of consumer surplus. Take Coca Cola for example, it already has deep reaches throughout emerging markets, it is hard to see how it can replicate its growth to the same extent going forward, as there simply aren’t another two billion people in China and India to introduce to the brand. Another potential source for growth is capturing a greater share of the consumer wallet share. Here again, you run into certain problems with certain product categories, such as Coca Cola. Even if the consumer was in love with your product and was willing to drink 10 sugary drinks a day, you are at most capturing a few dollar in daily value despite the potentially high utility. This may not be a problem today, but 20 years from now, when your average consumer doubles or triples in income, it is highly unlikely their spending per product will scale relative to their income going forward. This has been the reason as to why Alchohol has been a better play than products like Coca Cola for this very reason, in that you are more easily able to price segregate by consumer and as a result be able to capture more of the consumer surplus.

Gaming Sector
The gaming sector is THE most attractive industry I see today due to:
– Relative low price point and inelasticity
– Highly addictive products
– Cyclical defensive
– Ability for some to capture consumer surplus
– Consistent high margins and ROE

The downsides are:
– Difficulty in developing new IP
– high rate of reinvestment

In many ways, gaming has many of the same characteristics as the other addictive sin products. The products have high utility function relative to the price paid. Prices are generally low and relatively inelastic for the average gamer. Games could be made to be highly addictive due to structured progression relative to the chaotic world people operate in. Throughout the GFC, video game company revenues shared the same robustness as tobacco and alcohol to the economic cycle. The sector seem poised to perform well regardless of future outcomes. Whether the world becomes a post-scarcity world in which 90% of the people will spend their time writing holodeck programs or experiencing holodecks or whether we live in a world where 80% of the population are living on universal income because of automation, escapism will be in ever increasing demand. What I do know is that we are unlikely uninvent mass media and information technology, which continues to create winner take all market in what Nassim Taleb calls extremistan from mediocrestan, resulting in greater search from forms of alternate escapism and achievements.

However, this is where the similarity ends. Unlike tobacco and alcohol, which have had the same recipe for hundreds if not thousands of years, video games must innovate in a timely matter less it becomes stale and monotonous, requiring large amounts of re-investments.  In addition, building a new IP is incredibly hard, as per my previous discussions on Nintendo. In many ways, the prevailing view on the sector is to look at it in the same way as the Movie industry. Traditional PC and console gaming, just like in movies, have largely relied on fixed pricing per consumer. Consumers are expected to pay largely the same amount regardless of how much utility he/she can derive from the product. This dynamic results in an inability to capture consumer surplus and forces the producer to grow sales by focusing on expanding horizontal distribution, which is often times very expensive, leading to a strong focus on reusing IP instead of creating new ones.

The Mobile Gaming Model
Something changed in the business of gaming with the advent of mobile gaming. In order to attract a large base of customers early, developers largely started with a premium model and quickly realized that the 80:20 rules applied when it comes to capturing consumer surplus. It turns out only a small portion of the gaming population of a product or service generates an outsized portion of revenues. This discovery is now being replicated by PC and console games, with most PC and console companies looking for ways to introduce micro-transactions and aspects of gambling into their latest offerings. This model is incredibly powerful from a investment perspective and the trade off to high reinvestment cost and IP risk is worth the risk. Imagine being able to sell $5 worth of coca cola to someone and $5000 worth to someone else, suddenly the need to build a massive horizontal distribution platform is terribly wasteful and your efforts are better spent capturing a smaller share of premium clients, which is made possible by the internet and ubiquitous mobile phones. To some extent, I think this is also the beauty of companies like Amazon and Facebook, in that they are enablers of businesses to capture specific consumer surpluses in a focused way while the incumbents CPG companies are locked into a horizontal distribution model which are ripe for disruption.

China Market

Source: Newzoo

Today, China is the largest gaming market in the world, and it is largely dominated by mobile, unlike the US and Europe, which is spread evenly between Console, PC and Mobile. As I mentioned earlier, the best play for the west is really on IP as a result of the fragmented distribution channels, which narrows down the number of potential investments. As a result, I have made my bet on Nintendo and will divert my focus onto the China market.

In many ways, being the last mover has tremendous advantages for emerging markets such as China in that they can more readily leapfrog inefficient business models, whether it is land lines being replaced by mobile or brick and mortar being replaced by Tmall and JD. Today the gaming sector in China is dominated by mobile due to the ubiquitous nature of phones compared to the lack of platform penetration of console and to a lesser extent PC games. As a result, the two Chinese companies which dominate the gaming space, Tencent and Netease, which together control arond 80% market share, were able to leapfrog the traditional console/pc game market and focus largely on mobile and are in my opinion global leaders at the art of capturing consumer surpluses.

In my next couple of articles, I will take a deep dive into Netease and Tencent.

 

Secular Stagnation in the Developed World

“What is it that you believe in that almost nobody else believes in?”    – Peter Thiel

I think one of the most important thing that nobody really believes in is that we are entering a protracted period of low growth, driven by demographics, slowing technological change and a miss-attribution of productivity. Despite the early preview that we have seen from Japan, the belief in stagnation has very little following among academic economists, although the idea is percolating through the mainsteam through the likes of Tyler Cowen and  Larry Summers.

The Lie That Benefits Everyone

A world of increasing growth is one that almost everyone benefits. Companies can comfortably keep their current business plans. Investors can continue to enjoy an increase in earnings which justify ever higher asset values. Consumers and workers benefit from increasing incomes, improving goods and services. The public sector can continue to take on ever large debts, which is erodes in real terms over time by growth and large pension liabilities can easily be managed (see Investors above).

The Evidence for Growth

“During my lifetime, here’s a guy that was born in 1930 … my lifetime real GDP per capita went up six for one. Well, to think of an economy that already was among the top in the world, having in one person’s lifetime a six-for-one change in real output per capita, that’s staggering. That’s never happened in the history of the world.”    – Warren Buffett

It is true that the U.S. economy has grown remarkably since the 1930s, and most economic models assume a forward rate of real growth of 2+% going forward. However this was not always a straight forward paths. During the great depression, there were talks of prolonged economic stagnation due to falling birthrates and very high savings, but growth bounced back strongly in most countries following the end of WWII. For someone who was born in the 1890s and lived to the 1970s, the world would’ve changed far beyond what he could have imagined. From a transportation perspective, we went from horses to cars, from being grounded to widespread air travel and the moon landing. In the home, we went from wood stoves to widespread electrification and the rise in television and radio. Large businesses and corporations were adapting mainframe computers to improve organizational management and solving complex problems. This is essentially the power of compounding and a exponential function. If one was to extrapolate from the amount of progress that had been made from the 40 year period from 1930s to the 1970s, one would logically assume that we would all be vacationing on mars and have solved aging over the next 40. Which turned out to be far too optimistic.

The Evidence for Stagnation

“We wanted flying cars, instead we got 140 characters.”    – Peter Thiel

The GDP per capita of the U.S. in 1950 was around $9,500 on a inflation adjusted real dollar basis, by 1980, this had increased to around $18,600, an approximate two fold increase, representing a real growth rate of about 1.4%.  However, from 1980 to 2010, the GDP per capita grew to $30,500, representing a 1.6x increase, or a 1% growth rate, at a notable decrease from the first 40 years.

From a historical perspective, even a 1% growth rate is a large increase in improvement. For most of human history, growth was non-existent, economic power was largely a function of population growth, which was sustained by agricultural improvements. Power was held by kings and wealth and status was largely concentrated in the form of land granted to favored individuals.

Something changed starting in the 1800s and continues today. The industrial revolution rapidly changed the way wealth and power was concentrated. For the first time in human history, wealth was not simply a product of merchandise arbitrage and a residue of power, but was driven by a marriage between innovation and capital accumulation. This process started in the 1800s and really reached its zenith in the 1900s, where GDP per capital in the western world saw a 10 fold increase. But a story often forgotten was that the western human population increased rapidly over this period as well, tripling in the 1800s and again in the 1900s.

Demographic Slowdown

One of the easiest things to predict in economics is demographics. The number of 20 year olds in 2037 is set in stone, baring immigration, one just has to look at the number of births today. in the 1950s, the US population was 150 million, which roughly doubled going into 2000. Think about going forward 50 years, it is highly unlikely that the growth rate is likely going forward. Native birth rates are at below replacement levels and immigration is the primary source of population increases. Given the rise in protectionism and populism, it seems unlikely that immigration will continue at the same pace. According to the IMF, By 2050, we will end up with a population of between 350 – 400 million. If one thinks about GDP as a component of productivity and population growth, population growth was driving somewhere around .80-1.2% of real GDP growth. From 1900 – 1950, population growth averaged around 1.4% per year. From 1950-2000, it averaged around 1.2%. By the 2000s, population growth averaged only around 0.75%. If we extrapolate things forward, it is very possible that we may see growth rates of only 0.5% for the next 20 years.

Productivity

I suspect that we have significantly overestimated individual productivity. A major labor trend since the 1930s has been the rising participation of women in the labor force. In a single breadwinner family, a women’s labor such as home child rearing and homemaking was not part of GDP numbers. As more and more women entered into the workforce and had to hire nannies/babysitters and order more takeouts, these numbers started flowing into GDP, creating what is essentially overstated growth. I suspect somewhere around 30-50% of the per capita growth probably came from this shift. If one looks at trend among high income household, there is actually a strong preference for wives to spend their full time raising children instead of pursing careers. So if this overstatement was driving a large part of individual productivity numbers, it is highly likely that the adjusted real GDP per capita was increasing only at somewhere between 0.5-0.7% over the past 40 years.

Additionally, there are signs that even this low productivity number is falling. If one thinks about it in terms of technology and innovation, it is increasingly concentrated around Silicon Valley. We have far more scientists and engineers than we ever had in the 50s and 60s, but if we measure things in terms of new inventions, the productivity levels are now at abysmal levels. There are fewer companies being formed today than anytime over the past 40 years. In large parts of the U.S. there is a hollowing out of industry, a decline in real income and a sense of desperation that fed into a Trump victory.

But it wasn’t always like this. In the 1950s, there were dozens of Silicon Valleys. Southern California was seeing a renaissance in aerospace technologies. Chicago was the home to many great electronic manufacturers employing millions and Detroit was the center of automotive technology, employing millions of engineers, designers and workers. Over time, the number of hubs of innovation have shrunk, and by the early 2000s, we are left with the twin hubs of Finance (NY) and Information Technology (Silicon Valley). After the GFC, it may be that we are left with only Silicon Valley.

Growth Going Forward

If one adds the declining population growth of 0.5% and a GDP per capita increase of only 0.5-0.7%, we end up with a real growth rate of somewhere around 1.0-1.2%, a remarkable decrease from prior decades of over 2.0%. This number is corroborated by Robert Gordon in his TED talk. The implications are huge, this level of  growth no longer implied a 50% increase in living standard every 20 year (or a generation). The debt crisis we see in the developed world today is a direct consequence of this lack of growth.

Return to Feudalism

It is very possible that the industrial revolution was a one time event in human history and that we will revert back to a sort of semi-stagnant world. Or it may be that we will have a prolonged period of stagnation (similar to the dark ages) before reaching another industrial revolution like event.

Either way, the short term implications are dire. It is highly unlikely that we can expect significantly higher company earnings going forward (as companies are already telling us through buying back shares instead of investing), it is unlikely we will see sustained rise in REAL interest rates. More importantly, it is unlikely that we will see the significant improvements to living standards that have historically made this country the optimistic beacon of hope to the world.

What does a neo-feudal America look like? While the concept may seem strange to a lot of people, actually I would like to suggest that it unsurprisingly looks similar to now. High and rising asset values, increasingly expensive and elitist education as young people compete for fewer high earning jobs, increasing wealth inequality and a out of touch upper-middle class. But we have seen this before, in fact, this was the norm for much of human history.

In ancient feudal societies, power came from the king/emperor, and nobles received wealth, power and titles from the king. The major assets were property based or monetary (gold/silver) and wealth largely came from inheritance from previous generations and either earned through trades or gained through grants by the king. For your average commoner however, land ownership was forever out of reach relative to the price of labor. It wasn’t until the industrial revolution that all of this changed. Labor started becoming more important relative to property and inheritance. The likes of Carnegie and Ford came from relatively austere background but were able to grow large fortunes through their intellect and entrepreneurship. We saw this at its very extreme in the 1950s, where a high school educated worker could get a factory job and comfortably own a house, two cars and have a stay-at-home wife. But now we are going back to the era where family fortunes are becoming more important. Asset prices continue to rise relative to income, making owners of capital wealthier at the cost of new potential owners. The dynamic is similar to the feudalism of old, except now productive assets include a mix of financial assets instead of just land. One can see this in its rawest form in places like China today, where residents in Beijing and Shanghai acquired 3-4 housing unit through their work units for cheap (effectively getting noble status from the emperor) in the 90s and make far more on rent and price appreciation today than ever working. The education system today has become a simple proxy for wealth for the privileged and a cruel indentured servitude for the multitude. The only channel left in terms of social mobility seems to be studying Computer Science (see above on Silicon Valley) at a high tier state university these days which can secure a relatively high paying job without going into unimaginable debt.

Living in Stagnation Land

While most commentators focus on policy maker levers, inevitable debt defaults or inflation that may come about as a result of the unwound in high leverage built up in pursuit of growth which won’t come, my belief is that central bankers can probably suppress the volatility in such a way where the debt simply gets nationalized over time (see Japan which now owns over half of its national debt). Either way, it is incredibly naive to presume that anyone has any optics on how it will play out. Regardless of what policy options are selected, I don’t believe the underlying malaise is solvable, as a result the investment angle is relatively straight forward.

The compression in nominal interest rates that has lit the fire on asset prices starting in the 80s and drove much of asset price appreciation will still be ongoing. After all, in a 1% growth world, why should risk-free rates be anything but zero on a real basis? Why should inflation be anything but hovering slightly around zero. In such a world, wouldn’t a 4% corporate bond yield and a 8% equity return still be a bargain? What would happen if inflation fall to zero? Wouldn’t a 2.5% 10 year gov bond be quite attractive? What would happen to equities that are yielding 10% if people demand it only 5%, couldn’t equities still double?

In this stagnant world, you are probably not getting paid enough to take on new risks and there is very little compression left on the fixed income side. Chances are you would be quite happy getting a 3% dividend and marginal earnings yield from Proctor and Gamble and so will the millions of boomers who will have to de-risk their portfolio in the coming years.

Essentially there are two ends of the investment spectrum that offer the most opportunities: Monopolistic businesses which can generate decent ROE and are nearly impossible to disrupt and large technology conglomerates that are eroding traditional business models through information technology as it is the only disruptive sector left.

This then leads one to conclude that an optimal U.S. based portfolio involves holding a large amount of something like Berkshire Hathaway and a few technology disruptors like Amazon. Additionally, as the market forward prices in more of this bimodal distribution, there will be opportunities to grab discounted companies which people perceive as strongly threatened by incremental innovation but are actually fairly robust to it.

Technology Primer

I started by looking at the S&P500 since the early 1900s to understand if there was a mental model I could use to outperform the index. What I found was that starting in the late 1800s, most of the companies that ended up creating enormous shareholder value started as technology companies. Dupont was a technology company in the 1800s, Edison Electric was a technology company in the 1900s, GM was a technology company in the 1920s. If one looks at the market capitalization of companies today, many of the largest are today’s technology companies: Google, Apple, Amazon, Microsoft have market capitalizations which eclipse many old traditional industries.

There is clearly something about technology companies that allows them to generate enormous shareholder value and create wealth for society, afterall, if the stock index was still dominated by the likes of American Cotton Oil and American Sugar Refining like in the 1900s, we would all be living in a different and much poorer world. The challenge with investing in technology companies is that for every Edison Electric there is a Zenith Radio. for every IBM there is a Sinclaire Rearch and for every Apple there is a Blackberry. There had to be a way to analyze or put into context why some companies last longer than others and why some managed to generate returns over a thousand fold for its investors while others fizzile and pop with the latest turn in the economic cycle.

I agree with Peter Thiel’s thesis that technology advancement is probably slowing down and primarily concentrated around information technology as a result of regulations around the “world of stuff”. I believe this will continue to be the trend of the next 20-30 years, with some exceptions. I believe historically, there were massive innovation happening across multiple disciplines from the 1800s-1950s, where technological progress branched  out like a tree and built on itself. Since the 1970s or so, this branch narrowed  into only a few sectors, primarily focused in information technology.

So I wanted to strip out the massive amount innovations that have taken place around agriculture, communications, transportation, energy and petrochemicals over the past 100 years. It is helpful to think backwards if you will, focusing on just the age of computers and back into the late 1800s, into the birth of the electronic age. When we focus on just the computer led revolution in the 1950s and electricity revolution in the 1870s, we start seeing some interesting trends.

Information Technology Cycles

The following is my attempt at describing from a very high level the history of computing and informational technology. What I found was that technology moves both in cycles of proliferation and consolidation and that there is a directional theme.

Electronic Age
Electricity/Scaled Power (1875-1900): Platform competition and consolidation phase: Edison vs Tesla. Enterprise facing, originally to convince government and factory owners to replace physical labor.
(Important companies: GE and Westinghouse)

Electronics (1900-1925): Electricity prices fall as infrastructure rollout, creating a proliferation of electronic hobbyists and startups. Hardware dominates first half, platform dominates next half. Consumer Facing.
(Important companies: RCA, Zenith, Galvin (Motorola), most went belly up, upstream and downstream do well)

Electronics Consolidation (1925-1950): The electronics sector consolidates, with content platform operators (software) dominating, and winners usually emerge from previously niche targets.
(Important companies: AT&T, NBC, IBM)

Computing/Software Age
Computing/Scaled Electronic (1950-1975): Platform consolidation: IBM and the seven dwarfs. Enterprise facing, originally to convince government and large scale enterprise to replace  basic mental labor.
(Important companies: IBM)

From Hardware to Software (1975-2000): Computing price falls, creating a proliferation of hobbyists, IT startups and companies. Lots of PC makers who go bust, software and internet dominate in the second half.
(Important companies: Intel, Microsoft, Apple, Dell, most early PC makers go belly up, up and downstream do well)

Software Consolidation (2000-2025): Sectors like Retail, information and social consolidate. Hardware continues to fall away to software as content platform start to dominate, winner is emerging across multiple sub verticals.
(Important companies: Amazon, Apple, Alibaba, Tencent)

Cloud/AI Age
Machine Learning (ML)/Smart Software (2025-2050): Platform consolidation phase. Enterprise facing, convince companies this is the future of advanced mental labor. It will be about the use of AI at the industrial scale being adapted by corporations and governments for various analytical, resource management and decision making processes.
(Candidate Important companies: AWS, Microsoft, Palantir, companies yet to exist in China)

Specialized AI (2050-2075): Machine learning/AI becomes cheaper, creating a proliferation of AI hobbyists and developers, high level programing focused, consumer facing. This will be about the proliferation of AI in our everyday lives. Just as how starting in the 1990s, the story of technology was one with the integration of computers and software to our daily lives, to some extent, outsourcing the memory/logic processing components of our lives to computers. Now, we will be calling upon multiple AIs to do many deal of the functions of our lives in perhaps faster input times (using neural interfaces). Companies and products could be formed in seconds from simply thinking through the process and outsourcing it to smart software who functionally construct its various parts in second to minutes depending on how quickly you can direct and scale your digital minions to do your bidding.
(Candidate important companies: most will go bust, up and downstream will likely do well.

General AI (2075-2100): Consolidation of Software/AI. It will be about the web linkages of human/AI integration. Whether this looks like some sort of neural net where we live in a web of information buffeted by our web of AI advisors.

The average core technology cycle seems to be approximately 75 years, with 25 years of infrastructure platform competition and roll-out, 25 years of new venture creation most of which gets washed out, and 25 years of industry consolidation.

The initial phase of the cycle is dominated by 1-3 players fighting it out in a standards war that then gets rolled out as the foundational infrastructure of the next 50 years:

In the initial phase, the general theme is that the companies start off as enterprise and government facing. In the historical electronics and computer cycle, there has been two such periods, the roll out of electrical grid by GE/Westinghouse and the roll out of mainframe computers by IBM. In the case of the AC vs DC standard war between Edison and Tesla, both men through their respective backers, GE and Westinghouse, to invent and supply standardized power equipment and build power plant, the initial consumers were local government to replace previous gas lamp systems with electrical street lamps and to factories. The new technology was seen as cheaper and more efficient than previously gas operated lambs operated by cities and more importantly, electricity delivered by wire is seen as much more efficient than having onsite power for factory operators, allowing for the displacement of onsite steam engines and boilers. The birth of the mainframe computer was similar, although less competitive. Due to an earlier lead and first mover position, IBM was able to dominate the market over its competitors, making them irrelevant. The mainframe computer was sold to large corporations and government defense agencies for the calculation of complex formulas that had previously required the employment of hundreds of staff. The two initial cycles produce two monopolistic businesses that are still DOW components today, GE and IBM, which set the standard for a drop in the price of electricity and computing that would benefit the next phase of the cycle.

The proliferation phase, the general theme is that there is a cheapening of the previous platform technology, electricity and computing power, which enables hobbyists to take advantage and start building new products in the electronics and PC space. This is characterized by period of high new firm creation, high levels of creativity and a move from enterprise focus to the consumer. We saw it in the 1900-1920s with the birth of the electronics industry, there were dozens of radio makers, appliance makers and various electronics manufacturing startups. During this time, there was heavy competition from the host of new hardware businesses, with most profits coming from IP licensing. A similar analogy could be made of the PC era of the 1980s-2000s, there was a lot of hardware makers, it was highly competitive, and the winners of the era were primarily focusing on software (ie Microsoft).

The consolidation phase, In the case of the 1920s, the commoditization of hardware created an opportunity for content providers. History is littered with names of large scale manufacturer that are now defuct, names like Zenith, admiral and dozens of electronics makers which use to hover mid-western industrial hubs such as Chicago. The content providers in Radio and TV became more dominant. Through the current consolidation phase, starting in 2000, we see less new venture creation and generally a focus on the content/distribution platform side of consumer technology. Amazon, Netflix, Google, Apple are key examples of the shift.

What it means today,

I would argue that we are in the consolidation part of the cycle until 2020-2030, where it will continue to be about shrinking number of consumer technology players and the dominance of content over infrastructure. This explains the phenomenal increase of FANG stocks relative to the rest of the market. As a result, for the next 5-10 years, if one can pick good entry points for Amazon (I will discuss my dislike of Google in a future post), one will be successful, and the company will likely last for a long time. It remains to be seen if Apple can create a bigger ecosystem. Contrarian value plays probably revolve around Chinese ADRs which are out of favor with the US public.

From 2025 onward, the most exciting prospect is the idea of a repeat of the infrastructure buildout that happened at the onset of the electricity and computing age. I believe that the new phase will be targeting enterprise and government and will be a winner take all fight similar to the few cycles. The previous cycles revolved around scaled mechanical energy and scaled linear mental processing, I believe the next phase will revolve around advanced mental processing. I think that corporations and government while benefiting from scale, are beholden to inefficient decision making due to complicated management structuring. The automation of complex management process that can analyze of complex large data and deploy resources more efficiently will become invaluable to large corporate/government clients. Companies that have potential in this space are big data players with big client channels like AWS and Palantir or emergent players which have yet to develop in the US or China. It is also possible that will come from a spinout team from one of current tech giants.

It is a serious mistake to continue to look for consumer plays in the technology space. I believe that the sector is consolidating and played out. New entrants will find it nearly impossible to survive at the downturn of the next cycle and current incumbents will solidity their positioning. The next batch of potentially interesting companies will likely come from the enterprise space. While Wall Street analysts continues to be focused on AWS and the fight over IaaS and PaaS, the most interesting segment to watch is in advanced analytics and resource allocation tools, which is ripe for innovation. There are lots of fat in the middle management layer of corporations and governments. I strongly believe that this space is where the next once in a generation infrastructure player (like GE and IBM) can offer an impressive solution which will fill the niche. The early identification of this company/companies will generate significant out performance relative to any market benchmark.

 

Long Term Compounding

A large part of what allowed Warren Buffet to generate high rates of return over time was investing through gearing (of his insurance floats) in higher ROE businesses that were able to sustain their high economic returns over decades. What makes these businesses sustain high profitability over time? Warren likes to call it moat, business schools call it competitive advantage, in a sense it is all about various forms of monopoly power.

I think there are two sorts of businesses that tend to maintain their monopoly power over long periods of time: price makers and distributors.

Price Makers

The key here is price inelasticity. These companies tend to follow into two buckets, low price consumer goods which provide high utility (such as coca cola) or expensive industrial component makers which are critical to the supply chain. These businesses tend to survive through economic cycles as they maintain their pricing power. Key challenges for consumer businesses in this segment is keeping up with changes in consumer tastes. Key challenges for industrial price makers is investing in R&D to maintain their positioning to thwart off competition.

In a classic porter’s five forces model, these companies benefit from low supplier and customer bargaining power by its very nature and must invest to prevent new entrants and substitution.

Distributors

Distributors are the classic network effect businesses. A virtuous cycle of increasing customer base and increasing supplier bases allow the distributor to sit on increasing scale economics which it can capture for itself or pass on to suppliers/customers to build up its moat. Typically these businesses are capex light and can generate negative working capital, which can then be lent to suppliers or customers on a selective basis to capture the most attractive segments. Key challenges for these businesses is the ease of substitution. Uber may be a large scaled platform, but all it takes is for consumers to hit two buttons switch to Lyft.  In many verticals, traditional distributors are being displaced by online platforms such as Amazon and Alibaba, but many hold out due to the nature of their products and services being provided.

Similar to price markers, these companies also benefit from low supplier and customer bargaining power. Threats to these businesses come from industry shifts and threat from new technology.

Nintendo (NTDOY)

Nintendo provides an unique binary outcome opportunity which provides an attractive risk adjusted return due to the asymmetry of the payoff.

Unlike its western counterpart in Disney, Nintendo is relatively disliked by Wallstreet and investors. For the past decade or so the company has flopped on new platform releases and failed to make significant headway on mobile (with falling interests in Pokemon Go and a flop in Mario run). But there is substantial brand value and cash on the balance sheet of the company which provides a substantial margin of safety. In addition, there is upside optionality in the release of the switch platform and large potential IP capitalization opportunities in mobile which is being underestimated by current investors,

Liquidation Value

The company has over USD9 billion in net cash, no debt and a substantial IP portfolio.  The company owns and controls valuable franchises in Mario, Pokemon, Zelda, Metroid, Animal Crossing, Starfox, Donkey Kong, Kirby, Pikman, etc.

The IP created by Nintendo over the past 50 years simply have unprecedented reach and are nearly impossible to replicate. Surveys have been done which show that Pikachu and Mario are more recognizable characters than mickey mouse. The huge popularity of Pokemon Go across multiple geographies in 2016 certainly support this story.

The most valuable IP owned by Nintendo is Pokemon, which is 1/3 controlled by Nintendo and 2/3 by two other companies (Gamefreak and Creature Inc). Nintendo has an undisclosed stake in Creature Inc and most analysts assume that Nintendo have a defacto control over Pokemon Inc, implying ownership at slightly greater than 50%.

In a recent License Global! report, the Pokemon company disclosed that it managed to generated USD2.1 billion in revenue last year. With marginal overheads, we can assume it generates anywhere between USD1.5-1.8 billion in free cash flow. Valued at a reasonable 10x FCF would put the pokemon company at between USD15-18 billion, or which Nintendo’s stake is somewhere around USD7-9 billion.

To estimate the IP value of the remaining Nintendo IP portfolio, we can look at some other recent IP paid:

• 2015 Activision Acquisition of Kings Digital/Candy Crush: USD5.9 billion
• 2012 Disney Acquisition of Lucasfilm/Star Wars: USD4 billion
• 2009 Disney Acquisiton of Marvel: USD4 billion

It is very reasonable for the remainder of the Nintendo IP to be worth atleast USD4 billion, with Mario and Link being the most valued IP of the pool. If we then estimate the total IP of Nintendo to be in the USD11-13 billion range. The company should have a liquidation value (Cash + IP) of around USD21 billion.

At Nintendo’s current valuation of around USD29 billion, investors value the company at around a 38% premium to the company’s liquidation value. Given that any acquisition will require a significant premium over the cash + IP value, it stands to reason that USD21 billion represents a bottom estimate for a floor of value for the company.

Cyclical Business

It should be noted that Nintendo stock pricing is highly correlated to the sales of its popular consoles.

To some extent, the company is similar to Pixar prior to the Disney Acquisition.

As Michael Burry corrected surmised in 2000, Wallstreet has a poor record of valuing hits driven businesses like Pixar, where shares tanked between the releases of Toy Story 2 in 1999 and Monster Inc in 2001. The intermediate gap in which revenue is low while production/operation costs remained steady creates volatility in earnings which caused short term investors to exit on mass despite healthy underlying fundamentals.

At its peak in 2007-2009, Nintendo’s revenue more than tripled and EBIT increased five fold from sales of the Wii console.

From 2010 – 2016, with sales from the Wii falling off, and a largely unsuccessful Wii U launch, revenues have hovered in a steady state of USD5-6 billion. Share prices hit their lows in 2013-2014, when a historically stable (2000-2010) EBITDA margin of 23-30% went into negative territory from the write-down of inventories of Wii-U and higher than normal R&D expenses as Nintendo ramped up preparation for the NX console. By 2015-2016, EBITDA margin increased to 8-12%, which is still far below historic levels as a result of falling sales from a lack of major game launches.

Upside Potential 

It is quite difficult to forecast any financial/cashflow measures off a hits driven business, when margins and FCF could differ significantly dependent on levels of sales, which is why it may be helpful to think in simple heuristics. If the new console, the Switch, will replicate the success of the Wii, it is possible for the share price to hit USD70-80 billion market cap space. This is discounting any upside value of further monetization of the mobile space.

At a downside of USD21 billion in liquidation value and a upside of USD75 billion, the current market cap of USD29 billion reflect the market pricing a 15% chance that the Switch will be a hit, assuming also no further mobile potential. I believe this is an incredibly conservative assumption. While it is true that in recent years, the game market has bifercated between hardcore console gamers who demand the latest technology (xbox and ps) and casual gamers who have migrated to the smartphone/pad platform, the Wii proved that there is a market for casual party games in the console format, especially if it is user friendly. The success of the 3DS platform shows there remains a large market for dedicated handheld gaming devices. If we assign a simple one third chance that the Switch will become a hit, the shares are trading currently at a 22% discount.

Another thing to note is how differently Nintendo is positioning itself relative to Sony and Microsoft in its marketing strategy. There is no focus on having the best graphics and specs. If you look at the commercials for the Switch, there were no kids anywhere to be found, just 20-30 year olds. This gaming machine is marketed as a nostalgia machine for Millennials, which isn’t being fully appreciated. How many of us wish we could go back to carefree elementary school days of playing your n64 after school, or to the highschool/college days of playing the Wii with your buddies for hours on end. Now you are toughing it out in the real world, loaded with student debt, struggling to save for a down payment amidst soaring home prices, dealing with marriages/relationships, potentially having first kids. Gosh, don’t you wish you could just go back to the days when you could play Mario Cart with your buddies all over again. Boomers had their cars and I suspect Nintendo is probably the equivalent for millennials. In addition, the relatively poor sales numbers for Xbox one (26 million units) and even Playstation 4 (47 million units), relative to previous generations is a sign there may be a large opening here.

Even if Console becomes a dead end for Nintendo, mobile remains a huge underdeveloped potential source of revenue for the company. Of the top 10 best selling games of all times across all platforms, 5 were games developed by Nintendo. If Nintendo were to shift its focus to mobile gaming apps, there is a high probability that it could dominate the mobile gaming market, given its widespread IP recognition. If a underdeveloped version of Pokemon managed to rank #4 in the top grossing app in the U.S., generating USD2.1 billion in revenues, it is easy to see any number of Nintendo IPs generating similar sales figures. If Candy Crush could be valued at USD6 billion for generating USD2 billion in annual revenues, it is easy to see 1-2 Nintendo games consistently in the top 10 charts, generating  USD5-16 billion of valuation uplift.

Conclusion

The recent move by Nintendo in terms of releasing Mario Run has convinced me that management have changed their perception towards exclusivity of the gaming IP. I have no doubt Nintendo will stumble initially moving into the mobile market but there is a very good chance that with its IP it can generate significant hit games over the next 2-3 years. Given the strong smart phone penetration (more than 2 bn globally compared against 100 mn for the 3ds), it is a much higher margin business with potentially more reach.

As a result, On a risk adjusted basis, assuming a 30% chance of a success of its next Gen console and eventual success at smartphone/tablet monetization, Nintendo looks relatively attractive.

On a upside basis, should the Switch prove to be very successful, together with mobile monetization, the shares could trade at 3-4x current valuations over the next 2-3 years, generating IRRs in excess of 100%. I believe that at current prices, Mr. Market is pricing in very little upside to a company with 125 years of operating history, a strong portfolio of gaming IPs and a track record for creating both successful consoles and titles. This looks to be an attractive investment opportunity.