Welcome to Stuff I Thought About Last Week, a collection of topics on tech, innovation, science, the digital economic transition, the finance industry, bovine VR, and whatever else made me think last week. Please grab me on Twitter with any thoughts or feedback.

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)

In today’s post: NZS Capital partners with Jupiter Asset Management; debt fuels Chinese ecommerce; Google’s risk of a lost decade; non-ergodicity goes mainstream in the journal Nature Physics; concentrated funds underperform while gaining share; cows wearing VR headsets; and lots more below...

Stuff about Innovation and Technology
NZS Partners with Jupiter Asset Management 🎉
NZS Capital is excited to announce our partnership with Jupiter Asset Management, an investment company based in London with £45.9B (~$55B) in assets as of 9/30/19 across equity and fixed-income strategies. Jupiter is home to talented and genuinely active investors, and we’re looking forward to launching with them early next year! More from the press release:

“Jupiter is pleased to announce that it will take a minority stake in US-based investment firm NZS Capital and enter into a strategic partnership with the company in Q1 2020. NZS Capital was co-founded by fund managers Brad Slingerlend and Brinton Johns who worked together for 14 years at Janus Capital prior to setting up on their own. They both have long and successful track records of managing global equities portfolios on behalf of institutional and retail investors.
As part of this partnership, Jupiter will be the exclusive distributor of NZS’s global growth unconstrained-styled investment portfolios. This includes launching Jupiter-branded mutual funds based on these strategies in the future. Initially, the priority will be to target institutional investors, including those in the US where Jupiter will begin to have dedicated sales resource for the first time.”

Bovine VR: Ignorance is Bliss
Cows could produce more milk if they are living in a virtual pasture using VR headsets (the link has photos!) tuned to their eyes. Or, at the very least, the VR reduces anxiety in preliminary trials. So, in the bovine Matrix: “There is no...grass?”

AWS to Power Fox Content
The CTO and President of Digital at Fox, Paul Cheesbrough, posted a great recap (Twitter thread) of Amazon’s big user conference AWS re:Invent. Paul is one of the most forward-thinking tech leaders I know, and he was at AWS to announce a big streaming infrastructure deal for Fox’s content. AWS Media Services will power Fox’s production facilities and stream Fox’s sports, news, and other video content. It's interesting to roll this partnership into the future and ponder how NFL broadcast rights might enter into the mix. Notably, Fox will run AWS’ Outpost and Local Zone technology in their own facilities as the cloud moves back on premises for data-intensive applications.

Bringing High Fashion to a Hotel Near You
Rent the Runway, which offers access to designer clothing, is partnering with W Hotels to make clothes available at four hotels. For $69, guests can pick four outfits to wear during their stay. Broadly, this bundle of apparel and travel is an interesting example of two separate businesses combining to create a compelling value for consumers – I think we’ll see more disconnected services bundle together in the future.

Zero-Rate Borrowing Behind Singles’ Day Growth
Alibaba’s Ant Financial played an unexpected role on Singles’ Day (Alibaba’s big one-day ecommerce event) by extending credit via its Huabei lending division. According to an external estimate, half of Singles’ Day purchases were made using credit at a zero-interest rate. Historically, credit has not been a driver of Singles’ Day sales (according to the US-China Economic Review Commission’s December report, page 8 of the PDF). Unrelated, but in the same report, is an update on the aggressive but successful 5G roll out in China, which is well ahead of the US.

Back to Basic Cable
The geniuses behind South Park are at it again with another episode – which appears ripped from the pages of SITALWeek – titled “Basic Cable.” The shift to streaming has become a broken value proposition of fumbling multiple apps, passwords, user interfaces, restrictions, lack of universal search, etc. Now, the pendulum has swung back to where basic cable, when bundled with broadband, will give most folks in the US more for less. We’re back on Comcast video in our house - it’s a better UI that integrates Netflix and Prime Video, plus we’re saving a lot of money. South Park Season 23 Episode 9, “Basic Cable,” lampoons the disarray of streaming apps while also poking fun at the cable companies AND ripping on AT&T/HBO for overpaying for streaming rights...like the $500M they paid for South Park! Here’s a short clip (with NSFW language, depending on where you work – sorry, it might require a VPN if you’re outside the US).

Google Chrysalis?
The Google founders stepped down from their management roles at Alphabet and Google X this week. Pattern recognition is useless in most cases, but the situation seems somewhat analogous to the lost decade and paralysis that Microsoft faced. As an outsider, I can only speculate: Google's heightened regulatory risks, combined with a tense internal cultural moment (at least according to various press leaks), and a series of game-changing paradigm shifts on the horizon with AI, augmented reality, etc. feels like Microsoft circa 2000 when Gates became chairman and the company subsequently missed the shift to mobile. It took a visionary leader, Satya Nadella, to steer Microsoft to new heights. By all accounts, Sundar Pichai could be that type of leader for Google (he’s certainly not a Steve Ballmer!), but it’s likely to be a cultural turning point requiring emergence of a new identity. The next few years at Google could be marked by an inability to stack new S-curves due to the forward regulatory fear that accompanies regulatory capture (which is likely to cement their current monopolies) and a possible self-imposed breakup of the company – spinning out properties like Waymo and YouTube. The range of outcomes widens when a culture shifts away from that established by its founders.

TikTok’s Ad Ambitions
TikTok is targeting to build a big-brand ad business with the help of a former Facebook head of advertising. This FT article discusses TikTok’s new office in the former WhatsApp building and their poaching of several FB advertising execs. Efforts will focus on marketing products featured in short videos. But, brands are hesitant to jump aboard given the regulatory and broad US scrutiny of TikTok. For example, a lawsuit in California alleges that TikTok is syphoning off data, including biometric data from uploaded video, and sending it to servers in China as well as sharing with Alibaba, Tencent, and Baidu. This lawsuit joins a chorus of accusations (which may or may not be true) against the Chinese company as it faces scrutiny from CFIUS as to whether it can operate in the US under Chinese ownership. Meanwhile, the musical-meme app continues to win the time and hearts of America’s teens. I suspect the efforts to make it independent won’t be enough and, ultimately, it will need to completely change hands via sale to an American buyer (Snap or Twitter make the most sense because Facebook or Google would likely face regulatory scrutiny themselves).

Look Out!
Speaking of apps putting you at risk: hospital visits and serious injuries are spiking from staring at TikTok and other addictive phone apps. This report suggests that there are possibly upwards of 76,000 head and neck injuries from trips and falls related to looking at phones instead of what’s in front of you. Be careful out there – SITALWeek does not assume responsibility for falls while reading on your phone!

Semiconductor News Roundup: Imagination’s “GPU of Everything”, China’s Recruiting Efforts, Light-Speed Computation, and Shortcomings – and Promise – of FPGAs
Imagination has launched the “GPU of Everything” targeted at distributed AI and graphics workloads for any IoT, smartphone, or connected device outside the data center. Imagination use to provide all of Apple’s iPhone/iPad graphics processors until Apple insourced. That move sent Imagination on a path that involved buying MIPS (a processor licensing company that competes with ARM) and ultimately selling to a Chinese private equity fund called Canyon Bridge. Traditionally, Imagination has had very good technology, and they now have their sights set on the rest of the market that’s not already covered by a vertically-integrated solution (Softbank’s ARM has its own GPU engine, as does Qualcomm).

3000 engineers have left Taiwan to join China’s domestic effort to build semiconductor independence. This isn’t a new trend, having started 20 years ago, but there have been a series of high-profile departures from TSM and UMC over the last few years according to this Nikkei article.

Researchers have developed the first nano-scale chip that harnesses both electrons and photons. The electro-optical device tries to solve the problem of designing a single device capable of accommodating two types of wavelengths: light has a much longer wavelength than electrons, therefore requiring bigger chips. Their workaround: “They created a design which allowed them to compress light into a nano-sized volume through what is known as surface plasmon polariton.” (I don’t know what that means either!)

FPGAs are struggling to gain share in the AI market for a few important reasons. One of the biggest problems remains the dearth of programmers who are familiar with their unique languages compared to other products like NVIDIA (Cuda or C++) or new ASICs like Graphcore (C++). FPGAs also have to run all at once alongside an x86 server chip, whereas AI ASICs can turn parts of the chip on and off, thus conserving power. While GPUs from NVIDIA remain dominant in machine learning, this article discusses the potential role that FPGAs, from companies like Flex Logix and Microchip, might play in the more heterogeneous inference market.

Miscellaneous Stuff
Ole Peters’ Non-Ergodic Economic Model Gets Well-Deserved Attention
Traditional economic theory has no utility and zero basis in reality, in large part because economists have always assumed that the expected outcome for an event is equivalent to the average of all outcomes. However, at the individual level, all that matters is the path through time – a point thoroughly ignored by economists. This glaring oversight has led to a failure of economic theory for hundreds of years. Even behavioral economics, which aims to improve upon the theory, has shortcomings for the same reasons. However, there’s a better way to do the math: model using non-ergodic systems! We’ve been praising (and implementing) Ole Peters’ groundbreaking theory ever since we wrote about it in Complexity Investing back in 2014. Peters published a detailed article in Nature Physics this week, or you can also read this approachable synopsis of the article posted by the Santa Fe Institute:
“Instead of averaging wealth across parallel possibilities, Peters advocates an approach that models how an individual's wealth evolves along a single path through time. In a disarmingly simple example, he randomly multiplies the player's total wealth by either 150% or 60% depending on the coin toss. That player lives with the gain or loss of each round, carrying it with them to the next turn. As the play time increases, Peters' model reveals an array of individual trajectories. They all follow unique paths. And in contrast to the classical conception, all paths eventually plummet downward. In other words, the approach reveals a fray of exponential losses where the classical conception would show a single exponential gain.”
(SITALWeek #197 and SITALWeek #202 have a little more color on ergodicity as well.)

The Universe? It’s Just a Math Equation
Physicist Max Tegmark was on Sean Carroll’s podcast this week to talk about his view of the universe as a purely mathematical system. Max and Sean have a back-and-forth discussion of some nuanced arguments about the multiverse and possible simulated reality. It’s an approachable conversation; and, if you like it, I recommend Max’s book Our Mathematical Universe, which requires a little more effort on the part of the reader, but is filled with interesting cosmological concepts.

Through the Looking Glass
A mirror doesn’t flip images left or right (or up or down), but it does flip depth as a result of specular reflection, which means things that are further from us (and closer to the mirror) appear larger than an object right next to us (and farther from the mirror). That’s what creates the illusion that a mirror is a window into another place. Confused? Check out this short video from Minute Physics on YouTube.

Self-Storage Boom: Two Possible Explanations
Back in STALWeek #217, I posted a popular story from the WSJ on the 5x increase in new self-storage buildings in the US over the last four years; I suggested it might be a result of free money looking for a place to invest rather than actual demand for outsourced closet space. Since then, I’ve been looking for alternate explanations. The first I’ve come up with relates to a topic I’ve discussed recently: people aren’t moving. Average time spent in an owned home is 13 years, double what it was 20 years ago. We clean house when we move, but when we sit still, junk piles up around us, and renting a storage unit avoids the emotionally/mentally taxing effort of critical stuff “editing.” The second explanation is related to the first, and it’s simple demographics: People over 65 are a growing percentage of the population, and they are staying in homes longer and are likely downsizing when they do move. This interesting NPR interview with the author of a book on second-hand goods discusses how only 1/3 of items donated to Goodwill are sold and reasons to buy second-hand instead of new.

False Advertising Improves Reef Health
Piping the sound of a healthy coral reef into a dying one attracts fish necessary to repair it.

Stuff about Geopolitics, Economics, and the Finance Industry
Concentrated Portfolio Construction Exacerbates Cognitive Bias Pitfalls
The number of active mutual funds with under 35 holdings more than doubled from 149 in 2007 to 310 today (making up just over 9% of all active equity funds). Assets across the concentrated funds have tripled to $161B. The shift in portfolio construction seems to be an effort to differentiate active management from cheaper passive alternatives, which tend to hold more positions. However, the results are distinctly underwhelming: the concentrated funds are lagging the S&P 500 by ~6% annualized over the last decade (total performance of 244% compared to 322% since 2008). Running concentrated positions requires an especially strong control for cognitive bias risks. Position sizing, “conviction,” ego, and the brain’s incredible ability to make us buy into our own BS all conspire to introduce unnecessary risk into concentrated portfolios. (Here at NZS Capital, we have a few solutions for these bias problems. 😉)

Active Fundholders Incur Greater Tax Burdens as Money Shifts to Passives
As investors move from active to passives, they leave the remaining active mutual fund owners holding a higher tax burden. As funds sell shares to redeem investors exiting, the remaining owners pay higher than their fair share of taxes – highlighting the tax inefficiencies of traditional US mutual funds compared to ETFs (which workaround the double-taxation problem).

Blackstone CEO’s Visions of Grandeur
PE giant Blackstone has $554B in assets under management, but another $148B in commitments to invest in real estate and other private businesses. The CEO recently commented – while on his book tour – that smaller PE shops will see inferior returns as they compete with each other for deals while bigger firms can buy bigger deals with less competition (the article cites 8,000 shops investing $4T). The logic here isn’t at all clear to me – why would larger assets have better returns or fewer bidders in today’s free money environment?

Calpers’ Alternative Reality of 7%+ Returns
The $387B California pension Calpers has cut external public equity investments from $30B to $5B, reduced the number of management firms from 17 to three, and virtually eliminated emerging managers (firms with less than $2B under management). The firings will save $100M of the $119M spent on outside managers. It’s not clear if that $100M is net of what Calpers will spend to manage this money, or if the money is even staying in the public equities category. Maybe it will be going toward more of those levered, private assets that the Calpers CEO is seeking? Hopefully she has double checked those EBITDA assumptions (see next paragraph) and she understands the moral hazard of some PE investments as Joseph Stiglitz recently pointed out.

Corporate Debt at New High, with Loans based on “Fake EBITDA”?
$2.43T of corporate debt has been issued so far this year, surpassing 2017’s record with a few weeks yet to go. Aggregate corporate debt is now near $10T – 47% of the US economy (with some comments from our friend and brilliant bond investor Gibson Smith in the WP article!) However, concerns are building over the quality of many of the levered deals funded over the last few years: “More than half the companies that were part of a leveraged buyout in 2016 missed their earnings projections by more than 25% last year” according to Alan Waxman of TPG Sixth Street Partners who also warned of “fake EBITDA” underlying many loan assumptions.

SEC to NYSE: “No”
That was quick: the SEC has already shut down – without comment – the NYSE’s request to allow companies to raise primary capital in direct listings. This response is a blow to VCs hoping to cash out fast in IPOs. Here is a great down in the weeds analysis on direct listings if you're interested in this topic from Non-GAAP's newsletter.
✌