SITALWeek #203
Stuff I thought about last week 7-28-19
Greetings – why have home repair services been such a hard nut to crack for Internet platforms? Alibaba’s new open-source processor and the transformative dynamics in semiconductors that most investors are missing; the misinformed use of “data” in investing; Hail Satan?, the documentary; how neurons are similar to galaxies; and, lots more below. As always, reply back with thoughts or grab me on Twitter.
Click here to SIGN UP for SITALWeek’s Sunday EMAIL (please note some ad blocking software may disrupt the signup form; if you have any issues or questions please email sitalweek@nzscapital.com)
Stuff about Innovation and Technology
Silicon Valley party planners rejoice and VCs groan as Masa rolls the dice again with the launch of the $108B Vision Fund 2, despite the unclear and difficult to measure success of Vision Fund I. Softbank reported at their annual shareholder meeting in May that the first Vision Fund’s limited partners achieved an annualized rate of return of 29%. It’s not certain what starting date was used in that calculation, but the fund was said to begin investing in Q1 2018. Over that rough time period the iShares US Tech Index (ticker: IYW) was up around 20%, which was achieved with trading liquidity, and without relying on the mark to market calculations of Vision Fund’s illiquid securities (for example, WeWork, which saw its value rise from $20B to $47B based on two investments from the Vision Fund during this time period). It’s not clear to me what Apple and Microsoft, who are backing the new Vision Fund, expect to get from the investment. It seems unlikely that Microsoft will convince developers at Vision Fund-backed startups to use Azure over AWS. In contrast, Microsoft’s $1B investment into OpenAI could do a lot to help it’s credibility with the machine learning and AI community. But, Microsoft might want to just put its money into its own stock, which is up over 50% since the first Vision Fund launched!
Denver International Airport has seen a 20% increase in passengers from 2015 to 2018, but cars parked at the airport by travelers are down from their 2016 peak. The airport is canceling plans for new parking garages. This is just one of many ways ride sharing is slowly changing infrastructure, and it’s just beginning. The rise of delivery, especially food, but also local retail items, is likely to cause a big change in infrastructure soon as well. Also of note in the article: Since 2015, Uber and Lyft rides to/from the airport are up from 5x to 10x vs. the number of cab rides, which are down by 1/3 over the same period. DIA had $185M in parking revenues in 2018 and made $9M from Uber and Lyft fees.
After I wrote last week on the rapid transformation of the quick-serve food industry, I didn’t have to wait long to read that Uber was more aggressively pursuing a bundling strategy. The company is testing a $25 monthly pass that includes free Eats delivery and bike rides along with discounts on car rides. Uber is also pursuing an interesting strategy of signing exclusive partners, like the Starbucks deal announced this week.
News this week that Amazon would partner with US real estate broker Realogy in the US to, amongst other things, provide home services vendors (plumbers, floor layers, etc.) to home shoppers reminds me of past failed attempts in this industry. Zillow launched Digs over five years ago to target remodel shoppers, and had much broader aspirations in the home-services market at the time. Speaking at the Raymond James conference on May 29th 2014, Zillow’s then-CFO had the following to say about the “$25-35B” opportunity: “It's still nascent and we are still iterating on the product, but we think long term we can have a really nice consumer-based home improvement, home services product in that space and have a lot of opportunity there as well.”
Well, that didn’t happen, and today this effort appears to be dormant at the company. In the home services industry, it has been very hard to create durable network effects with sustainable economics (i.e., good lifetime value to acquisition costs for both sides of the marketplace, consumers and service vendors). The main reason is leakage: once you find a plumber, both you and the plumber would prefer you just contact each other directly. Many attempts have been made – and I recognize people point to platforms that have reached some scale, like the merger that created ANGI Home Services – but I struggle to see a winning platform emerging. Once a small-business home service provider has a client base and word of mouth referral funnel, it’s hard to maintain their participation, even if you offer software to run their business and payment platforms. Further, many home service issues are one-time or every several years; with such low frequency, it is very hard for network-effect platforms to emerge (new and used car marketplaces face a similar issue, and it’s why I think Zillow’s advertising spend remained stubbornly high over the years, weighing on margins). Thumbtack, another startup in this space, recently raised $150M after rebooting their business. They describe some of the complexity of the industry in this blog post: “What worked for house cleaning fell apart when applied to wedding planners. What worked in urban centers with a lot of supply didn’t work in rural locations with smaller supply. Every change had a million variants and it took time to convince pros to try the new system.”
Redfin reported a 9% increase in prices for the most affordable homes in the US, and Zillow reported an accelerating increase in rents. It seems likely to me that we see renters’ rights and rent controls become a national political topic, possibly by 2020’s presidential election. Recently, on the Equity LifeStyle Properties conference call, one of the largest owners of mobile home and RV parks in the US answered this question rent control:
“There's been a lot of press lately about rent-control initiatives across the United States. We have 23 properties right now subject to mandated rent control primarily in California. But we've been actively opposing that rent control -- those rent-control ordinances for over 20 years...”
RISC-V is an open-source answer to the dying x86 processor, as workloads are increasingly heterogeneous with specific power requirements and functions. Open source is hard to control with trade tariffs and export bans, although Microsoft-owned GitHub, the largest software code repository used for many open-source projects, recently started blocking developers in countries subject to trade sanctions. I think the only outcome of these bans will be to create new code repositories that supercede borders – you can’t keep open-source code a secret. The open-source nature of RISC-V is no doubt feeding the design boom globally with special strength in China and India. Chinese Internet giant Alibaba just released its own RISC-V processor design blueprint targeted at 5G, AI, and autonomous driving. Alibaba's chip subsidiary is called Pingtouge, Mandarin for honey badger! Chinese chip makers would still need TSM in Taiwan or Samsung in Korea to make their leading edge RISC-V chips, and much of the design and manufacture depends on US and European companies, which underscores China's ongoing semiconductor vulnerabilities. At some point we also need to think about the implications for the mainstream MCU providers dependent on ARM and MIPS: Alibaba is open sourcing this chip design, which could have implications for margins system wide in the future. The flip side is that incumbent chip makers who have field application engineers and deep customer knowledge can also leverage RISC-V themselves without paying royalties to Softbanks's ARM. I wrote more about ARM, RISC-V, open source, and the mega trends around IoT+AI+5G that most investors are still missing in this post, updated yesterday, on the NZS Capital Blog.
Okta reports that 77% of their customers with Office 365 are also adopting Zoom and Slack. They also see best-of-breed app engagement as a sign that customers are operating with heterogeneous collaboration tools. This is great insight from Okta, which, it should be noted, has more forward looking customers than average. This doesn’t change my view that, in general, team-based collaboration tools remain niche markets in the giant office-collaboration space – I discussed that disappointment and possible reasons for it a couple of weeks ago here.
In Gartner’s annual cloud infrastructure report, they note the disingenuous “price drops” Amazon Web Services touts, pointing out that AWS’ Compute Service (it’s largest offering) hasn’t reduced prices in five years despite the big drop in component costs over that time! This thread on Twitter recaps the cautions Gartner had on each of the cloud platforms.
A UC Berkeley lab has designed a drone that can collapse to fly through small spaces. This brings back my nightmares from when I watched the thought provoking Slaughterbots video (where facial recognition and autonomous tech turn drones into devastatingly easy weapons; Black Mirror covered this concept in the episode “Hated in the Nation” as well).
Chinese tourism to the US continues to fall, but this interview with the CEO of China’s giant travel site Ctrip indicates that Jane remains bullish on China outbound travel. She also stresses the importance of making visas easier. Chinese tourists may opt to go to Italy, where they do not need visas, instead of Epcot. As a long time Disney and Seaworld investor, I keep one eye on attendance, which isn’t so hot this summer as reported by Theme Park Insider (and, also in this article about Orlando). Part of the reason is price increases, which were largely implemented to manage attendance and demand by Disney. I like the long-term trends of theme parks, so part of me secretly wishes for weak attendance to create buying opportunities.
Gallium is being increasingly used as a substrate for semiconductor manufacturing and is particularly well suited to 5G. China currently pulls 95% of the world’s Gallium out of the ground, but rising prices will make supply viable elsewhere. Soitec, which makes specialty semi substrates, is expanding into the GaN with an acquisition earlier this year.
Salesforce.com entered the Chinese market through an exclusive distribution partnership with Alibaba. Alibaba is the fourth largest cloud platform globally, and is largely popular only in Asia. This is an odd deal on multiple fronts. Salesforce indicates this deal is in response to multinational corporations looking for global support. Are large US companies wanting to run Salesforce through Alibaba’s data centers in China? That seems unlikely. The deal covers Taiwan (and Hong Kong), which seems like it could put Salesforce’s Taiwanese customers’ data in the hands of mainland China, with potential for the Chinese government to access it. Salesforce has a number of platform products that could contain sensitive data on customers, products, and more. Lastly, and this is perhaps the most puzzling part of the announcement: Marc Benioff is a long-time supporter of the Dalai Lama.
Google’s Deepmind unit used evolutionary, population-based training that combined hand-tuning with random searching to improve Waymo’s autonomous driving algorithms. Read about the results that created a 24% drop in false-positive identifications of objects like pedestrians on Deepmind’s blog post here.
Mixhalo raised funding to let you go to a rock concert live and then listen to it with your headphones on. I might just have to file this under “I don’t get it,” and move on.
Subscriber blackouts like the current one between AT&T (DirecTV) and CBS are increasingly common, and I’d argue they are now just events to help media companies acquire new subscribers for their direct-streaming apps. Searches for “CBS All Access” surged on Google last weekend, and, increasingly, legacy video distributors have no choice but to pay the fees. A better solution is for legacy video distributors like cable and satellite to collaborate with streaming apps by making them available on set-top boxes. Of course cable is in a much better position to benefit from this than satellite because of their ability to sell high speed data, which satellite can not match. Netflix discussed the upside to partnering on their last quarterly earnings call:
“But the bundles that we are doing are nice incremental accelerants to our acquisition, especially into a user population that may not be as tech-forward as the folks that sign up with us directly. They may not have a Smart TV connected to the Internet at home, they may not have an adapter product like a Roku or an Amazon Fire TV, but they do have a set-top box from their pay-TV operator, and if we can put the Netflix application in a nice seamless integration into that set-top box. If we can then include the actual subscription to Netflix as part of their pay-TV offering as a bundle, then it makes it super simple for those folks to be able to get the same kind of experience that our subscribers who sign up with us directly do and get the same benefits.”
As I’ve lamented in the past, what I still need is a re-bundled video experience with a common search and user interface; however, that won’t be happening anytime soon due to ongoing Balkanization of content. Related: Comcast, which owns NBC Universal and Sky, discussed their plans to launch their advertising-based streaming app next year; maybe the new NBC-Sky streaming app can grow its subscriber base faster by getting into a carriage fee argument with its parent company Comcast, and blacking itself out!
Toyota teams up with BYD and CATL in China for batteries and EVs as it senses Panasonic may not be able to meet demand. With BYD, the two companies will develop a Toyota-branded truck for the Chinesemarket.
Famous Birthdays, the site for all-things teen – with 20M users of its database of famous TikTok, YouTube, and Instagram stars – has 18 employees and has taken no VC money. They also have pages for non-human celebrities.
Miscellaneous Stuff
The fractal nature of our Universe extends from the large-scale structure of the Universe itself all the way down to the structure of neurons in human brains – that’s over a factor of a billion, billion, billion! Is this apophenia? Apophenia is a term all investors should know: it’s when you make mistaken connections between unrelated things. In this case, it turns out it’s not: there is real similarity between the brain and the Universe, and it’s a different type than the fractal dimensions we see in things like tree branches (and consciousness). There really isn’t anything practical in this analysis, it’s just cool!
Hail Satan? is a fascinating documentary on The Satanic Temple, which started as a way to demonstrate the importance of free speech and remind everyone that the US is a pluralistic nation, but turned into an international movement. It demonstrates a lot of beautiful concepts, and I enjoyed it, but you might hate it! The evil is almost never in the “witches,” but the witch hunt itself.
Anyone want to guess which company this line from Michael Pollan's book How to Change Your Mind is referring to!? “I know of one Bay Area company today that uses psychedelics in its management training.”I've got a few educated guesses on my short list (hint: they are mentioned so far in this week’s newsletter!).
Stuff about Geopolitics, Economics, and the Finance Industry
This article in Institutional Investor about the ridiculousness of portfolio managers claiming to use “AI” gives me a good reason to rant a bit about the use of data in stock selection and portfolio construction. I don’t believe “non-traditional” datasets, like point of sale or credit card transactions, retail foot traffic, web scraping, etc., are a means to get an edge in stock picking. And, unless you are a theoretical physicist, the more you model a complex system, paradoxically, the less informed you might be. Bill Miller often talks about three sources of advantage in investing: 1) Informational: I know something you don’t know; 2) Analytical: we know similar things but I analyze it differently; 3) Behavioral: we have the same information and analysis, but come to different conclusions about how to express a view in a portfolio. Data can cut across these advantages with varying degrees of danger. The most important thing to do is decide what your time horizon is. The longer the time horizon, the less useful “datapoints” and analysis are to your process. It doesn’t mean you can’t collect and review relevant data, and, like a good Bayesian, adjust your credence up or down – that’s still important as it relates to position sizing in a broader portfolio. Taking this to its extreme conclusion, the only asset managers that should use non-traditional datasets or “AI” for informational advantages are those quant hedge funds who have hundreds of PhDs and access to things like real-time satellite photos of Walmart’s parking lots trading with time horizons of nanoseconds. And, best of luck to them – that’s a complete waste of time, and will never be a long-term viable source of alpha because information is ultimately available to anyone who wants it at all times, and there is no way to know how much of that information is built into a stock price already. Increasingly, stock exchanges are selling trade data faster than your supercomputer can put a trade on! The 2nd source of outperformance, using data and AI to create an Analytical advantage, doesn’t generate many opportunities either. Modeling is important to understand a range of outcomes, but consistently predicting the future is impossible – no one can do it. We know this from lessons learned in the study of complex adaptive systems (the heart of our investing framework!). The 3rd source of advantage, Behavioral, is actually a ripe source for data analysis. Here the data are free: mining your own performance or analyst data for sources of cognitive bias that inhibit proper decision making is easy and worthwhile. It’s useful to know when you are likely to make mistakes and determine which teams are better or worse at getting decisions right or wrong. The biggest problem with data, and this is the same problem with modeling forecasts for companies or the economy, is that once you collect a lot of data and use it, you’re brain convinces you that all of the sudden you are smarter and less likely to make a mistake. However, the opposite is true. So, be smart with data, focus mostly on internal data, and play a game you can win based on your time horizon.
It strikes me that the congressional fear we’ve discussed over big tech getting into financials (such as Facebook’s Libra blockchain currency) is potentially very good for incumbents in the financial transaction value chain. Although I cannot think of a more consensus trade than the credit card networks, beloved by every card carrying growth investor in the world today, it seems like Congress is cementing the position of companies like Visa and Mastercard. On the other hand, can governments stop distributed, transactional currencies from existing if they aren’t attached to a company they can punish, such as Facebook? If a Libra-like effort takes off without equity owners or centralized structure, how could that be stopped? This article, which attempts to discredit non-government currencies, fails to produce any legitimate arguments.
What do the MY and MO ratios tell us about long-term stock market trends? The MY ratio is the ratio of 35-49 year olds compared to 20-34 year olds. This is a measure of roughly the number of people saving vs spending – as you get older, you tend to put more in the stock market on average. The MY is more important than the MO ratio, which is 35-49 versus older folks, who tend to be taking money out of the market in retirement. (It seems boomers have been staying invested in the market longer, which is one reason the MO ratio isn’t as useful.) The MY ratio bottomed in 2016 and is supposed to trend up for 16 years. Even the study’s author says not to rely to heavily on this prediction, which investing’s own tabloid magazine Barron’s covered (yeah, I know that’s a hypocritical description for someone to use who’s been quoted in the publication many times!). For one thing, we know the higher non-housing debt burden for millennials is keeping them from buying houses, and it seems like you would buy a house before investing in the stock market. I’d suggest much larger factors are at play, including the ever-shrinking number of shares to invest in as fewer companies go public and share buybacks continue, as well as what appear to be technology-driven, structurally-low rates. If there is indeed structurally more investing happening from demographics (increased numerator) combined with these other factors (decreasing denominator) for the next decade, well, you do the math...
As protests continue in Hong Kong, and China is likely to bring mainland military force to bear, the FT reports on why Hong Kong residents are so angry: in particular, younger folks are feeling a sense of hopelessness with the current system. It’s interesting how China’s information suppression extends all the way to mainland students studying abroad. Recently, on an Australian campus, protesting students from Hong Kong were attacked by mainland China students – to the applause of the Chinese consulate in Australia. One mainland China student defending the country said “China has absolute respect for Xinjiang and Hong Kong in terms of politics and human rights.” Australia’s Foreign Minister responded to the events: Senator Payne said the right to free speech and to peaceful and lawful protest is protected in Australia, even on contentious and sensitive issues. “The Australian Government expects all foreign diplomatic representatives to respect these rights,” she said in comments obtained by AAP. “The Government would be particularly concerned if any foreign diplomatic mission were to act in ways that could undermine such rights, including by encouraging disruptive or potentially violent behaviour."
Ireland joins the UK’s FCA and Norwegian regulators in calling out funds that claim to be active, but are too close to the index performance. I think fund trustees globally are falling down on their job, failing to assure funds are active and allowing high fees without performance elements. Not only are passive funds and ETFs coming for these “closet-indexing” managers, now regulators are coming for them as well.
VC deals in China are down 90% to just $2B in the 2nd quarter of 2019. I thought this podcast with Chinese VC and former tech exec Kai-Fu Lee by MIT scientist Lex Fridman was an interesting dialog on Chinese tech culture and investment, and how regional governments participate in the tech investment cycle.
Goldman Sachs sees traction with big companies for ESG money-market fund: “The $1.5 billion money-market fund helps corporate treasurers steer money to bond brokerages operated by minorities, women and veterans, reflecting a growing shift toward investing with environmental, social and governance, or ESG, principles.”
Disclaimers:
The content of this newsletter is my personal opinion as of the date published and are subject to change without notice and may not reflect the opinion of NZS Capital, LLC (“NZS”). This newsletter is simply an informal gathering of topics I’ve recently read and thought about. It generally covers topics related to the digitization of the global economy, technology and innovation, macro and geopolitics, as well as scientific progress, especially in the fields of cosmology and the brain. I will frequently state things in the newsletter that contradict my own views in order to be provocative. I often I try to make jokes, and they aren’t very funny – sorry.
I may include links to third-party websites as a convenience, and the inclusion of such links does not imply any endorsement, approval, investigation, verification or monitoring by NZS Capital, LLC (“NZS”). If you choose to visit the linked sites, you do so at your own risk, and you will be subject to such sites' terms of use and privacy policies, over which NZS Capital has no control. In no event will NZS be responsible for any information or content within the linked sites or your use of the linked sites.
Nothing in this newsletter should be construed as investment advice. The information contained herein is only as current as of the date indicated and may be superseded by subsequent market events or for other reasons. There is no guarantee that the information supplied is accurate, complete, or timely. Past performance is not a guarantee of future results.
Investing involves risk, including the possible loss of principal and fluctuation of value. Nothing contained in this newsletter is an offer to sell or solicit any investment services or securities. Initial Public Offerings (IPOs) are highly speculative investments and may be subject to lower liquidity and greater volatility. Special risks associated with IPOs include limited operating history, unseasoned trading, high turnover and non-repeatable performance.