China Radically Shifts Tech and Growth Focus…

‘I’ve been a long-term bear over the past several years on the China real estate sector and it has hugely underperformed. Investors are now fretting that the housing/fixed investment/leverage perpetual motion machine can’t be sustained much longer – property developers in China offer bond yields in the 7-10% range (despite wide dispersion in fundamentals – not all are pursuing the Evergrande shell game by diversifying into with EVs etc.).China’s housing market has proved remarkably impervious to endless policy tweaks meant to slow it down – home prices across the 70 major cities rose 4.8% in April y/y, the fastest monthly pace since last September which puts the pressure on the government to impose even more draconian measures as housing affordability is a very tangible and embarrassing manifestation of extreme inequality hanging over the Communist Party’s centenary this year. Ironically, it’s exacerbating the looming demographic implosion which may be the one force that can finally end the ‘concrete deposit box’ mentality. Indeed, housing costs on most social surveys are a key factor in the plummeting fertility rate… ‘ – ‘China Goes Childless…‘ Weekly Insight, May 21st 2021

Mark Zuckerberg’s cringeworthy 4th July stunt holding an American flag while hovering over a lake on an electric surfboard looked an act of hubris after Facebook won its FTC antitrust case and would never have been attempted by his jittery mainland peers.  There are broad parallels between the Biden Administration’s latest attempt to limit corporate concentration (as a factor in rising inequality/suppressing wage growth) in tech and China’s assault on its tech giants. The unconstrained agglomeration of data analytics and hence market share ultimately undermines growth and competitiveness. The difference is in the exercise of power, which as with the pandemic response last year is far more decisive – the US is debating regulation within an updated anti-trust framework for the network effect digital platform era (i.e. that looks beyond a very crude consumer welfare calculation), China is implementing one at breakneck speed. Monopolistic rent seeking and regulatory capture have become a global economic phenomenon far beyond tech, but it’s hard to see any other country being able to tackle the now widely accepted negative economic and social externalities as decisively. The investor pain of this realignment of political versus corporate power is being front loaded in mainland names, but will be protracted in the US and Europe…

Source: Entext

There is a pressing global need to rebalance power between state, network effect tech platforms that defy conventional consumer welfare based anti-trust analysis and consumers whose behavioural data is the ‘product’. China has hugely lagged even inadequate Western regulatory and anti-competitive laws and institutions – the legal framework is only now being built and anti-trust sanctions have extended far beyond tech e.g. logistics and concrete companies have all come under scrutiny/been fined for price fixing. There has been popular discontent with price discrimination practices, working conditions and with merchants being forced to pay for online traffic and their products copied by a generic version if successful. With the top-down policy focus now on the enterprise/industrial software space to leverage the 5G rollout and boost productivity, the fact that so much consumer data is siloed off in closed ecosystems is detrimental, while the sheer dominance of a handful of companies has slowed the number of new Chinese start-ups; any that show promise are swiftly crushed or acquired, similar to chaebol dominance in S. Korea or FAANG+ in the US.  

There is a level of long-term policy coherence that many foreign investors are overlooking. Firstly, as it achieved in 5G, China is aiming to set global standards in platform regulation and anti-trust enforcement.  Secondly, there is a ‘crowding out’ aspect – the policy focus for economic development is now on hardware (particularly semis) and enterprise software innovation, not incremental value extraction in consumer platforms. Reducing the attraction of the latter is therefore partly about shifting the incentive framework for Tencent, Alibaba etc., but they remain crucial to Beijing’s global value chain/digital currency ambitions and there will be closure on the regulatory deluge over the next 3-6 months (and a relative mean reversion trade versus US peers). The country has vertically integrated cleantech supply chains over the past decade to become a global leader in alternative energy (from lithium and cobalt to advanced batteries in EVs, polysilicon to PV modules in solar, rare earth metals for magnets to finished wind turbines etc.) and it would be dangerously complacent to underestimate its leapfrog potential, from quantum communications to RISC-V open source architecture and next generation compound semiconductors. It has used FDI and technology transfer to accumulate strategic process knowledge and then build indigenous national champions in one sector after another, although leading edge chip fabrication has unique bottlenecks via EUV lithography.

Demographics are the clearly the main driver for the demolition of the online education sector, and China’s ominous population trajectory was highlighted back in that may note (including the highly suspect official fertility rate), and the focus in the latest five-year plan is less on GDP than population growth. There is now a degree of official panic about falling birth rates and a determination to make child rearing more affordable. Once the first regulations came out about limiting tutoring hours, it was a clear warning about the direction of travel. I ran a China education thematic stock basket in 2019, but closed it on regulatory tail risks, which even then were growing after repeated official warnings. Every shopping mall I’ve ever visited in China has an upper floor dedicated to Western sounding tutoring outfits, from kindergarten math schools to cramming exhausted teenagers for the brutal university entrance exams. With about 80% of children ages 6-18 enrolled in some form of extracurricular academic tutoring, those malls will now face empty units. Indeed, they were often the busiest floors, particularly at weekends and in the evenings, but perhaps world leading levels of childhood myopia might improve with more time spend outdoors.

Source: NBS, Entext

Beijing has decided that allowing households to follow the Korean model and end up spending 12-14% of disposable income on private education services was socially regressive, as well as exacerbating the fertility implosion covered earlier in the year given the soaring expense of rearing a child. The $120-150bn revenue sector will likely see revenues shrink 60-70% with massive lay-offs, not least among foreign tutors who have been singled out as unwelcome. Again, the US faces similar issues with private education costs running out of control and creating economic distortions (including delaying household formation), albeit the problem is more at college level where fees and housing are now $75-100k annually and student debt has topped $2trn. Cheaper housing via Singapore style HDB blocks or developer build to rent schemes is another policy priority to boost cut child rearing costs rather than speculative ‘concrete deposit boxes left empty in their millions. That backdrop is important to the latest duration mismatch/margin call panic on Evergrande’s ~$90bn debt pile; we’ve already seen a sharp slowdown in the real estate market in Q2, from bank funding to floorspace sold and started.

The big question is whether it has made itself too big to fail – real estate investment across the whole supply chain is 25% plus of Chinese GDP while real estate loans also make up around 28% of outstanding bank loans. Recently, China’s financial regulator instructed banks to conduct a stress test around an Evergrande failure so the endgame to this saga should finally be getting close. We will soon discover whether Beijing is willing to let it implode as a cautionary tale to the rest of the sector and a means to end moral hazard. A liquidation process that is more or less orderly would ultimately be a net positive for credit risk pricing on mainland capital markets. The implication of recent events is that China is now on a very decisive policy path to reduce inequality, boost CCP legitimacy and underpin technological decoupling from the US and hence the geopolitical challenge to it across wider Asia.

AI Led Services Automation Remains Key Theme…

“…the ability to control the whole flow of products from the warehouse to the end customer…that gives us a lot of ability not only to control the flow of the product, but also flow of information…I think you’ll see it, too, as a customer where you’re starting to get more precise estimates of delivery. You’ll get notes that say, hey, you’re eight stops away from your delivery, et cetera….we see a lot of benefit from that…because we pretty much have perfect information between the order placement allocation to warehouses where we’re going to pick and box up the product and send it on its way. So lots of advantages. We are continuing to invest, and we’ll see a large investment in this area through 2021 as well.” – CFO post Q1 results

‘With employees at Amazon’s fulfilment centres protesting poor working conditions during the pandemic, the tailwind for automation hardware looks strong – better conditions/wages are a pragmatic interim measure. The global market for warehouse/logistics automation was $53bn last year but could double by mid-decade with retailers following the Amazon/Ocado lead post pandemic. Amazon already has more than 200,000 mobile robots working in its warehouses, mostly to deliver products from shelves to workers packing items in boxes for shipment, eliminating the need for workers to leave the conveyor system. The next stage of industrial automation for Amazon involves robotic systems that pick and pack items, combining AI software, computer vision sensors and suction/grasping arms. As for Uber, humans are the weakest link in the retail business model, and further automating their tasks will be a priority after this, not only for Amazon but across the grocery sector, much of which still relies on workers picking goods off retail outlet shelves to put in a van. That worked (at a high margin cost) in the pandemic to allow rapid scaling of grocery deliveries by limiting store opening hours to allow them to become part-time fulfilment centres. However, once this passes every food retailer board will be forced to face up to the classic incumbent ‘creative destruction’ dilemma and reinvent their online business models.’ Weekly Insight, 7th May 2020

‘…COVID may be accelerating a profound change that gets robotics to the upside of that J-curve. Plug-and-play” systems are much more mundane than the advanced robots that dominate the headlines. Most don’t even rely on artificial intelligence. But they embody sophisticated new automation that can be put to practical use with relatively modest preparation and disruption. Plug-and-play systems include automated guided vehicles, which use laser-based LIDAR sensors to navigate around factory or warehouse floors. They also include computer-controlled conveyors and sorting machines and even automated baggers. Brynjolfsson and Beane say the surging demand for plug-and-play systems signals a potentially profound shift toward robots that are highly sophisticated on the inside but appear simple on the outside. As robot manufacturers catch on to the rising demand for plug-and-play robots, spurred in part by the COVID pandemic, they will come closer to delivering the real promise of robot productivity. “There’s an instructive lesson here from the introduction of electricity,” Brynjolfsson says. “When factories became electrified [at the start of the 1900s], they didn’t have any significant productivity increase for 30 years. It only happened when they realized that electricity allowed them to change from having one huge steam-powered motor to having smaller motors in everything. Then you had a doubling or tripling of productivity. The same is true today of advanced robotics.” From the Stanford Human Centred AI blog

  • We have long been highlighting the prospect of radical service sector automation over the next decade, replicating that in manufacturing from the 1990s onward, a trend with profound political implications. That has been expressed over the past few years in a thematic global stock basket including names like Cognex and Rockwell in the US, Shenzhen Inovance in China, Daifuku, Yaskawa and Keyence in Japan, Delta Electronics in Taiwan etc., comprising makers of various automation vision/motion sensors, components such as actuators and full systems. The long held view has been that automation capex growth would shift from robots in safety cages performing specific functions on factory production lines to ‘cobots’ working alongside humans in fulfilment warehousing and logistics depots.

‘Automation as a Service’ Via Cloud Toolkits Will Be Transformational…

Source: Entext

  • The ‘automation of consumption’ theme, including the rising use of robots across the logistics and fulfilment sectors has proved a good bet and is still in its early stages, with precise geolocation and developments in self-driving technology converging with robotics and accelerating the uptake of cobots, autonomous delivery vehicles and drones etc. Indeed, because the underlying technologies overlap, following the spate of M&A from commerce, car and robot companies will increasingly merge – Honda and Hyundai (which bought US walking robot maker Boston Dynamics last year) are early examples of ‘systems integration’ among auto OEMs. Amazon and Alibaba now extensively use robots developed in-house to bring the shelves to the pickers, who stand stationary instead of walking around the warehouse, with the average worker picking 3x as many items in the automated system.
  • These residual human roles will likely be automated by end decade, as challenges like engineering ‘haptic feedback’ into robot grippers are solved. Food delivery has boomed over the past year but its preparation is now migrating to industrial scale, automated ‘dark kitchens’ that can prepare thousands of meals daily across a range of cuisines. The DoorDash acquisition of food service robotics start-up Chowbotics (which replaces humans in customised salad preparation) is a sign of things to come across the sector, with dozens of companies aiming to automate specific kitchen functions such as flipping burgers in fast food outlets. The company also entered into a delivery automation JV with GM and acquired self-driving start-up Scotty Labs. Delivery robots and even drones are going to become ubiquitous by end decade and transform the economics of order fulfilment – Starship has developed small 45kg pedestrian robots that can carry items within a 6km radius with a cargo bay can only be opened with the recipient’s smartphone and are similar in design to Amazon’s Scout and FedEx’s Roxo same-day delivery robots.

Robotic Process Automation (RPA) Gaining Traction Across Professional Services…

Perhaps the most compelling reason for thinking a disruption lies ahead, though, is a realisation on the part of many professionals themselves. Over the past five years, I have heard doctors worry about the implausibility of seeing patients online, been told by teachers that students cannot learn properly unless they are in the same room and listened to lawyers insist that court work could never be done virtually. And yet, in a matter of weeks, telemedicine, online learning and virtual courts have become the norm. The main barrier to transformation in the professions was never technological. Many of these technologies have been around for years. The main barriers were cultural. Most people are resistant to new ways of working, white-collar workers particularly so. With the Covid-19 crisis, those cultural barriers have largely disappeared, leaving many preferences for traditional ways of working looking more like indulgences.’  Daniel Susskind, fellow in economics at Balliol College, Oxford and the author of ‘A World Without Work’

  • While the retail supply chain is now rapidly automating, even more profoundly, ‘deep automation’ is now heading to offices. Back in 2003 when AI had largely been abandoned as unworkable, the economists David Autor, Frank Levy and Richard Murnane published a study on technological change noting that machines perform tasks, a narrower unit of work than most human jobs. They classified these task as “routine” or “nonroutine” (constructing a financial model is typically routine, cleaning a hotel room involves a series of non-routine tasks and therefore hard to automate). This approach helped explain why technology could transform the nature of many jobs without eliminating them in large numbers and also why a ‘barbell’ jobs market developed, with low-paid and high-paid jobs resilient while the middle was hollowed out. In his book ‘A World Without Work’, Oxford University economist Daniel Susskind argued that the second part of that analysis is now becoming questionable as reinforcement learning software encroaches upon non-routine skilled tasks.
  • He points out that what Google DeepMind’s AI supercomputers can do today will likely be achievable on laptops and even phones using edge computing within 10-15 years. The Artificial Intelligence Index project, based at Stanford University, which tracks a wide variety of benchmarks, highlights rapid progress at symbolic achievements from playing poker to translation, speech recognition, and classifying diseases such as skin cancer and diabetes via image pattern recognition. As machine learning adapts to ever more tasks, the reorganization of human work in response will be wrenching across the cognitive distribution curve. Only fringe US Presidential (and now more plausible NYC mayor) candidate Andrew Yang has focused on this key policy challenge, but it will likely redefine politics over the course of this decade and have profound investment implications.
  • Over the next several years, language models will likely become far more general purpose, encompassing an unimaginable range of problem types. Being able to have a world described through language and rendered as an image or video, or even asking text-based questions about the world with answers based on a system’s understanding of our nuanced reality sounds like science fiction but is now within our grasp by mid-decade. Selene is NVIDIA’s own AI supercomputer used for experimental and production operations. These services have been valuable for NVIDIA, but first in entertainment and gaming where text can be used to feed into image or video synthesis (to create worlds from description) and then more broadly to corporate applications, companies globally will have a toolkit to replicate a far broader range of insight previously assumed unique to human intelligence. Language modelling is one of the most expensive done in machine learning – we are at the point of asking the language model to solve a new problem that it’s never been before and it is generating an accurate answer by contextualising, a true (if at this stage, expensive) breakthrough.
  • The potential applications are endless – Amazon now has a cloud service called Connect to automate customer service call centers for other companies with natural language chatbots, interactive voice response, and automated customer voice authentication. The service is integrated with Amazon’s Lex AI that powers the automatic speech recognition and natural language understanding in Amazon Alexa, offers a “Wisdom” feature that searches across independent databases like Salesforce and ServiceNow repositories to find answers for customer queries significantly faster than any human agent, analyses customer sentiment in real time and can automatically alert supervisors with issues during calls. It looks like sales of Connect jumped 150% to roughly $175m last year – this looks likely to be a billion-dollar software market within 2-3 years.
  • Susskind quoted above makes a good point – the past year of remote, virtual working has been an experiment allowing service companies to reimagine the workplace. Ever more sophisticated AI has changed everything, starting with the definition of “routine” work, which has vastly expanded, making it harder for displaced from low-skilled jobs to retrain for more challenging roles. Indeed, it’s been a long-standing theme that office space per capita would decline (with process automation in insurance, banking etc. the structural tailwind). The full implications of the ‘proof of concept’ trial forced upon us by the pandemic will be profound and already being felt in tech and finance hubs like NY and San Francisco, which risk seeing their tax bases and ability to fund services erode as employees flee to low-income tax states like Florida and Texas. The implementation of AI in the professional service sectors is straightforward wherever tasks are well defined and there is a sufficient library of labelled data available to train a neural net to acquire a useable function e.g., automating standard legal contracts.
  • Rapidly advancing computer vision algorithms and neural networks allow the software to “see” desktops and understand relevant documents, using AI to understand patterns in the data and how best to break the process down into replicable steps. RPA start-ups have been building automation toolkits so that IT departments with limited coding experience can input their own processes and get those automated – cloud based ‘automation as a service’ to boost white collar productivity is coming.
  • The approach is particularly applicable in any industry with repetitive, heavily rule-based process-oriented tasks, such as insurance or accounting e.g. patient records in healthcare, preparing invoices in financial services, digitizing legal documents, or automating repeat searches (files, priors, data to cite) for insurance claim representatives.  Across the service sectors, robotic process automation (RPA) combined with AI software is emerging as one of the fastest growth new enterprise software markets. An RPA tool can be triggered manually or automatically, move or populate data between prescribed locations, document audit trails, conduct calculations, perform actions, and trigger automatic responses. It automates rote tasks, handing them off from humans to software algorithms. 
  • UiPath raised $1.3bn last month via a SPAC listing and is trading at a $37bn valuation (>10x more than in 2018 and approaching a third of SAP’s market value). Its automation clients range from Google and GE to Equifax, HP, and McDonald’s. UK AIM listed Blue Prism Kryon and FortressIQ in the US have also raised capital to expand their product ranges while Microsoft acquired Softomotive last year (now part of Microsoft Power Platform). The UIPath approach starts with generating a map of workflows to recommend what to automate, and then moves into the data-gathering process e.g., taking screenshots as an employee goes through a task. It’s on track for an IPO this year – start-up competitors include Automation Anywhere, which raised $290m from investors including SoftBank in 2019. 
  • In terms of sector automation, mid-back office admin work looks to be where warehousing and fulfilment was five years ago – i.e. on the cusp of being radically reshaped as humans are increasingly left doing only those tasks the machines can’t analyse, model and replicate. Even as we face multiple labour market bottlenecks over the next couple of years in the post-pandemic recovery (notably in blue collar sectors from truck driving to manufacturing and construction, where Millennials seem unwilling to replace retiring Boomers), on a five year plus view a major dislocation looms as the cognitive threshold for a middle class life gets pushed ever higher by highly scalable automation breakthroughs. In that context, Andrew Yang may not have all the answers, but he’s been asking many of the right questions…and portfolios should retain thematic exposure to the secular AI/sensor investment opportunity.

Hacking Shock Latest US ‘Global Risk Manager’ Failure…

‘In broader terms, the FAA response (to the Boeing 737 scandal) underlines yet again that the US now risks losing its post WW2 leadership role as the global default ‘risk manager’ in everything from aviation to disease control (the CDC) and ultimately even trade and finance (via SWIFT) because of a steady erosion of credibility and capacity under a dysfunctional administration, as highlighted in the latest book by Michael Lewis, ‘The Fifth Risk: Undoing Democracy’. That lack of competence/capacity will raise the stakes when an inevitable international crisis occurs…’ Weekly Insight, 19th March 2019

‘CrowdStrike’s strength is in the high-growth endpoint security market but also offers services like threat intelligence and cyber-attack response services. Identity management specialist Okta is another cloud-based security software player also now worth over $15bn after a spectacular run (up 5x since early 2018). The shift to cloud-based architecture in the cybersecurity space means that even SMEs will be able to benefit from pre-emptive machine learning led threat intelligence and quarantine. Aside from this cloud shift now underway, blockchain technology could potentially help enhance cyber-defence as the architecture helps prevent malicious activities via consensus mechanisms and to detect data tampering. Blockchains resolve the ‘lack of trust’ problem between counterparties and owing to their distributed nature, blockchains provide no ‘hackable’ entrance or a central point of failure and thus inherently provide more security when compared with legacy database-driven transactional structures. ‘- Weekly Insight, July 22nd 2019

The point made back in that note in March last year about a dangerous lack of US ‘competence/capacity’ has sadly proved valid this year – the disastrous ‘regulatory capture’ failures at the FAA (and China was the first country to ground the 737 MAX) have been repeated across multiple other key Washington agencies (including most disastrously the CDC) which were once the global standard. We end the year with the US accidentally discovering that it has suffered the most audacious hacking attack ever (which seems to have begun at least a year ago – FireEye’s blog covers what we know at this stage well). Investigations into the hack are ongoing, but the failure of national security agencies, like the CDC’s costly failure to roll out effective Covid tests earlier this year, highlights chronic underinvestment in technocratic capacity and a huge degree of complacency.

Yet again, the US has shown stunning ineptitude in exercising its historic role as global reinsurer/risk manager, the topic of Michael Lewis’ timely book. This goes far deeper than the dysfunction of the Trump White House and can’t be readily fixed by the appointment of a more competent and mainstream cabinet. The ‘anti-government’ (ex-sacrosanct military) ideology which has defined Washington politics since the 1990s has hollowed out most key institutions with chronic underinvestment, and has seen government R&D spend on basic science fall to record lows versus GDP. A generation of retiring public servants with decades of expertise have simply not been replaced with people of the same calibre, given Congressionally mandated headcount and budget constraints leading to an accelerating ‘brain drain’ to the private sector among technical specialists in particular. There are certainly now many areas which can for the first time be outsourced to the private sector such as NASA/Pentagon satellite launches, but the capacity to prepare contingency plans to manage tail risks and effectively regulate the destabilising perverse incentives of capitalism remain core government functions.

US market oversight has been eroded to such an extent that the US Treasury market came close to imploding at the depths of the pandemic deleveraging panic – indeed pricing dysfunction and moral hazard have been a key aspect of this year’s events. Liquidity disappeared in bond markets in March as highly leveraged hedge fund basis swap trades collapsed and ‘one click’ bond ETF liquidity exacerbated the fallout. We saw similar chaos in the US oil futures market a month later, after Saudi and Russia briefly abandoned OPEC+ and a small group of UK traders helped crash the expiring physical delivery WTI contract into negative levels.

China’s long-term strategic goal is to become a leading force in setting global standards (as it has successfully in 5G and Asian trade, with a huge EU investment deal pending despite US efforts to derail it) and to make its capital markets competitive with the US in terms of depth of liquidity, breadth of asset classes/products and regulatory credibility, a key precursor to RMB internationalisation. One of the ironies of this year is that US markets have become characterized by a narrative driven retail frenzy, just as mainland ones have become less so on relentless domestic and foreign institution inflows (indeed, ex-China EM portfolio flows have been negative this year).

While there clearly remains a long way to go, Beijing has continued the liberalization of its markets this year and is now leading in pre-emptive policies to ensure long-term stability (the crackdown on Ant Financial should be seen in that context, specifically consumer credit growth as other platforms copied its model). It’s also leading in payments innovations like the digital RMB (likely launching within 12-18mths) and the blockchain infrastructure to manage it. Indeed China is the first country to pose a threat to the US in software (already consumer, with for instance TikTok clearly superior as an engagement engine to US social media peers), because it culturally has no problem accepting the iterative, launch to market and patch errors model it entails in a way perfectionist Japan and Germany never could. That comes at a time when US regulators are behind the curve and allowing dangerous levels of speculation to develop via single stock options etc.

Pandemic tail risk was regularly highlighted by epidemiologists and healthcare policy makers but not remotely reflected in government preparedness ex Asia. We’ve covered cybersecurity tail risk over recent years as an attractive investment theme via stocks such as CrowdStrike, and we are clearly getting closer to a nation state level cyber conflict after a series of escalations in recent years. After this event, issues like cyber security and data privacy can no longer be left largely to the private sector – we will need new oversight and governance structures for the tech giants built around national security priorities as much as consumer data privacy rights. 

US officials are still working to understand the full consequences of the hack, which is a hugely successful attempt to spy on internal communications and steal critical national security information (it hasn’t been a destructive attack that damaged or shut down computer systems, as some major cyberattacks have done in the past, but clearly makes that potential future escalation far easier). This doesn’t look great news for the rollout of the ‘Internet of Things’ model in the US (accelerating via 5G in firewalled China, with Beijing pushing Alibaba/Tencent to develop relevant enterprise software), which inherently means a massive increase in the number of potential hacking entry endpoints via industrial and logistics infrastructure etc.

Indeed, the implications of this event may prove net bearish for the software sector ex security next year – US networks have become dangerously open and interconnected and as with healthcare post Covid, rebuilding resilience will become a long-term policy priority. The full ramifications will become clearer in Q1, but after the shambolic pandemic response, it should be sobering amid the current hubris in markets that the US has found itself so astonishingly vulnerable to brilliantly engineered malware. Threats which could disrupt our entire digital infrastructure at any moment will have to be taken as seriously as viral ones. Perhaps the Russians should reverse their cyber intelligence division into one of the exploding list of Wall Street SPACs, but I suspect the Chinese will make them an offer they can’t refuse first.

Peak US Exceptionalism?

‘The right thing for the Fed to do is to maximize economic performance. The right thing for the president and Congress to do is to adjust the tax system to make a fairer economy. We should have higher corporate tax rates, full taxation of carried interest and capital gains, close loopholes that allow capital gains to escape taxation, tax penalties on leveraged buybacks and limit corporate interest deductibility.’ Former Treasury Secretary, Larry Summers interviewed recently

The points made in a note entitled ‘Fed Firepower Overwhelms Fundamentals and IB Consensus…’ back in early April are now widely accepted and remain key. Policymakers were attempting to make everyone ‘whole’ on their pandemic losses, from furloughed workers to mismanaged hedge funds and until bond markets stage a mutiny, will continue to do so – there was no point overthinking it. The past couple of months is a reminder that the intuitive notion that markets and economies track together is hugely misleading and gives economists a deluded claim to be able to help (rather than usually hinder) the investment process.

There is almost no correlation between US stock returns and real GDP growth contemporaneously, and it remains modest even a year ahead. Unlike say last summer, when PMI inventory/sales data etc. were giving a lead to a growth reacceleration and risk on signal, in the current situation, there has been little practical gain in poring over diffusion indices – granular data on consumer activity and corporate outlook statements have been more useful guides.

There has been too much ‘the pandemic changes everything’ analysis, but we’ve highlighted the underlying strength in the US economy coming into this crisis e.g. in housing, with its household formation as well as mortgage cost underpinnings. The MBA Purchase Applications Index is another ‘V’, soaring from 180 in April to 325 in June, up 20% y/y, while median home listing prices have jumped from -1% in April to +8% on Redfin data.  Whether the pandemic has boosted per capita housing demand on a sustained basis (urban apartments to suburban houses with dual work from home space) remains to be seen, but regardless Millennials who have married and started families later than previous generations for a variety of reasons (not least student debt) will be buying houses at a faster pace. Starts should rebound strongly in H2, and resurgent lumber prices bear watching as a signal – it would also add support to the copper rally. On Google trends data, new vehicle related search traffic has turned positive y/y for the first time since February, confirmed by the Mannheim used car index rebound since May.

Often the headline numbers have been misleading in any case e.g in relation to US unemployment and the perverse incentives of workers and businesses to game the system. That macro data will become relevant again as the ‘bungee jump’ distortions abate and economies reopen further to confirm or refute investor assumptions on the speed of the potential earnings recovery, as in China right now. The cross asset return expectations distribution curve is still skewed to deflationary outcomes, which looks dangerous as redistribution and reshoring become dominant political themes, against the backdrop of fiscal and monetary policy convergence. Assuming there is no ‘fiscal cliff’ in July and most support measures are extended (adding another 4-5% of GDP in stimulus), bond investors risk a shock by mid-late 2021 if the labour market re-tightens surprisingly fast in a post vaccine world and inflationary pressures accordingly return.

The consensus has capitulated since May on views including that we were likely to see a retest of the March lows, that this was ‘just a bear market rally’, that negative US rates were even a possibility worth discounting etc. The ‘Great Depression’ economic analogues and technicians overlaying 1930 price charts on 2020 ones have been shelved. Aside from the historic scale of global stimulus, three factors have been pivotal to sparking the risk-on frenzy: firstly, the rapid rebound in Chinese industrial activity, reflected in mainland oil, copper and iron ore demand, to feed resurgent infrastructure spending and housing starts. Excavator sales have also jumped in Q2, and the GPS tracked utilisation rate for Komatsu excavators in China has hit the highest since late 2018. Secondly the reduction in Eurozone tail risk via breakthrough debt mutualisation proposals, aggressive German fiscal reflation and expanded ECB peripheral buying and thirdly the unemployment insurance/PPP transfer windfall in the US which has offset the household earned income impact of the lockdown, and exaggerated the scale of the (still severe) employment downturn.

So far, reopening in Europe and Asia is proceeding remarkably well despite inevitable regional/city level clusters, the US clearly less so as sensible public health advice gets politicised and dragged into the culture wars. Consensus expected 2021 EPS is about $164, putting the 2021 forward PER at 20x, versus 18x at the February high – the Fed trillions are surely worth a couple of points of multiple expansion, but now as the data improves, things get more nuanced from a policy perspective. An optimistic forecast at this point might be 2021 EPS of $175 (i.e. the 2020 EPS expectation pre lockdown) but a growing risk for US equities after record outperformance relative to ROW is that the Democrats sweep to power and push through a progressive agenda, from tech regulation to tougher ‘gig economy’ labour legislation.

The partial rollback of the 2017 tax cuts Biden has endorsed would take $20 or so off whatever EPS number you project for next year. The other key issue for relative US performance into next year is belated anti-trust scrutiny of the tech sector. So far, fitful attempts at a pro-cyclical value rotation have faltered on Covid headlines, despite the overextension of tech growth/the quality factor and the risk that we have seen a ‘front loading’ in Q2 of secular online transition trends. At historic EV/sales and FCF multiple software sector valuations, any disappointment would trigger a deep selloff given crowded positioning. Fifty state attorneys general are probing Google’s digital advertising business practices, alongside a similar probe being led by the Department of Justice, which marks the first serious attempt (with bipartisan support) to overhaul anti-trust law and curtail the inherently winner takes all dynamic of web platforms. In the EU, Apple faces new investigations extending from its app download duopoly with Google to NFC payments.

The systematic tax arbitrage of digital IP (notably via Ireland) which has suppressed effective tax rates for US tech (and pharma) over the past decade faces OECD scrutiny and an inevitable move to some form of local revenue based inferred minimum tax. While the Trump administration is willing to risk a transatlantic trade war on the issue, a Biden one seeking multilateral consensus almost certainly won’t. That earnings tailwind would likely drive a rotation out of large cap tech, which dominates most portfolios (including fast growing ESG ones) and has reached a record share of US and global indices. Meanwhile, a successful vaccine this autumn would clearly be a game changer for the economic outlook beyond 2021 and justify a more procyclical, value oriented portfolio stance. Among several promising candidates, Astra Zeneca’s vaccine partnership with the Oxford Jenner Institute and the Gates Foundation bears close watching, which is heading into Phase 3 trials in Brazil and is scaling production capacity to 2bn doses, succeeds this autumn.

We’ll know in September…what will that do, after the astonishing monetary/fiscal stimulus seen globally, to the growth and inflation outlook beyond next year? We would enter a classic end of cycle blow off phase, because the fiscal and capex spending drags which had delayed it are gone – politics now matters more than anything, and tacking inequality and climate change in the wake of the pandemic implies a generational regime shift for asset markets. Rather that worrying about ‘V’ shapes and trying to second guess epidemiologists, the bigger questions right now for asset allocators are whether the US is approaching a peak share of MSCI AC and if we are beginning a shift to a structurally inflationary environment. Both look reasonable bets to us…

Product Cycles Imply Looming PMI Inflection Point…

‘I even think the President has recognized it’s not cost effective to push the tariffs anymore, and if he wants to continue pressure on China, which I think he probably does, I think this would be a good path to pursue. I do think we reached the end of the road with regard to cost-benefit analysis on the tariff side. In fact, I think a lot of the tariffs that have been recently suggested on the last tranches aren’t going to take effect…this is a matter, in my mind, of consumer protection…” Economist Larry Lindsay, former Fed governor and director of the National Economic Council under President GW Bush, speaking recently

In the course of recent client meetings in London and Asia, I’ve reiterated the message contained in the notes since late August  – tariffs had outlived their usefulness for the US as a trade policy tool given the negative feedback via consumer and business.  Even if this ‘phase one’ trade agreement hadn’t been announced, it seemed doubtful the December hike would have been implemented and further US containment measures will likely shift toward technology and capital flow restrictions, with a narrower market impact. China and the US acted in their narrow self-interest and therefore this ‘non-deal’ which has attracted much dismissive comment stands a good chance of surviving to a Trump-Xi photo opportunity in Chile.

As usual, sentiment will follow price action into year end, and extremely defensive positioning is fitfully unwinding and that will accelerate if we see a second derivative improvement in the macro data through year-end. The market has been  braced for more bad news and hugging long duration bonds and their equity expression via quality/low-vol equities as an expensive comfort blanket. In fact, we’ve seen a strong rally over the past six weeks in high global beta, deep cyclical markets from the DAX to KOSPI and Topix. For students of behavioural finance, the next few BoAML institutional surveys will be fascinating to read after the ‘hunkered in the bunker’ fearful tone of recent ones. The dollar index reacted predictably to trade relief,  crucial to EM participating in a wider risk appetite reversal and for the White House is as important as selling more pork and soybeans (even if the FX ‘stability’ aspect of the deal is a theatrical stunt).

To sustain the recent rotation toward cyclical value, we need a rebound in two key industrial sectors we’ve been structurally bearish on since 2016/17 – autos and smartphones. In both cases, a technology shift (diesel to hybrid electric in Europe/Euro 6 equivalent in China, 4 to 5G in mobile) and the rapid rise of a second-hand market in China have dented new sales. Those postponed consumption/extended replacement cycle headwinds are now abating – mainstream German car brands have electrified, and buyers of ICE cars worried about residual values can now choose a VW e-Golf over a Tesla.

Indeed, the Ifo survey of German auto sector confidence is showing signs of stabilisation, while annualised output is also likely bottoming at just over 5m units. Last month, EU demand for new passenger cars increased by 14.5% to 1.2m units – the growth is certainly flattered by a low base following the introduction of a new emissions testing regime last year, but its striking that four of the five major EU markets saw double-digit gains. Over the first nine months of 2019, new car registrations were down 1.6% y/y but that should be turning positive into early 2020 – much of the slump in demand has been due to a technical industry transition creating consumer confusion over residual values etc. Tighter emissions standards have also been a drag in China, which adopted the local equivalent of Euro 6 for ICE engines in June this summer – sales have begun to recover since.

As for smartphones, while 5G handsets are still only about 1% of Chinese sales, with 40 cities fully networked by mid-2020 and operators offering 30-40% discounts on 3000-4500 RMB handsets (taking them closer to 4G prices), that proportion should surge toward double digits through H1. Given this upgrade cycle and flattering base effects, the overall market which has been falling 5% or so y/y in recent months should turn strongly positive by mid-2020. Pre-registrations for 5G service are approaching 10m users, even with just a handful of handsets currently available (although dozens more Chinese designs will be launched by mid next year).

Making China the biggest and fastest 5G deployment is a key technology priority for Beijing, not only for the direct benefits in terms of stimulating the service economy with related new products but the scale economies it affords to then dominate the global market for handset and infrastructure hardware. It’s critical to dig beneath the headlines to understand the role of say weaker semi pricing in the slump in Korean exports by value and the resurgence in the Chinese current account surplus YTD, as much as the role of Boeing’s 737 fiasco in US export/durables data. A semi restocking cycle into Q1 now seems plausible and the divergence between chip equipment names like ASML and the commodity chipmakers in Asia, who have lagged US peers YTD, should close.

Against this backdrop of a nascent cyclical recovery in key global sectors which have been a drag on PMI and trade data, you can be constructive on risk exposure at this point while utterly realistic about the low odds of any comprehensive trade settlement over the next year. The weakness seen in global PMI and other industrial data over the past 12-18mths was as much about a transition in key technology product cycles as it was trade uncertainty…that will be true of any ‘surprise’ rebound in H1 20 also.






ECB Risks Igniting Currency War…

‘We have not had a rules-based international monetary system since President Nixon ended the Bretton Woods agreement in August 1971. Today there are compelling reasons—political, economic, and strategic—for President Trump to initiate the establishment of a new international monetary system. The current monetary regime permits governments to knowingly distort exchange rates under the guise of national monetary autonomy while paying lip service to avoiding trade protectionism. We make America great again by making America’s money great again.’ Fed nominee Judy Shelton making her job pitch via a Cato Institute essay last year

Bond markets have been caught up in aggressive front running of ECB easing, the greater fool theory being practiced on a scale of trillions. The big question now is whether markets are over discounting the impact of Fed/ECB easing in DM sovereign and credit markets – the move in peripheral debt underlines the ‘never mind the price, we can flip it to the ECB’ frenzy. This may be framed in terms of anchoring inflation expectations, but also via FX markets implies using the ECB balance sheet to grab a bigger share of weak global demand. Shelton’s views above that this is a form of de factor competitive devaluation are becoming more mainstream in Washington; it’s ominous that Trump is taking an unhealthy interest in ECB policy in his Twitter feed.

Given the current extremely low absolute level of inflation expectations and rates versus prior QE programs, diminishing returns are inevitable. Similarly, we are approaching a limit (even with tiered deposit rates) on how negative rates can go without the adverse consequences via the banking and insurance/pensions sectors overwhelming any lending or wealth effect impetus. Some central bank researchers (even at the BoJ) have belatedly begun focusing on the ‘reversal rate’ (the policy rate at which accommodative monetary policy becomes contractionary for lending). Christine Lagarde’s beach reading  should include Capitalism Without Capital: The Rise of the Intangible Economy, exploring the dematerializing nature of incremental growth, a key theme in our research.

Traditional economics is focused on managing resources in conditions of scarcity and high fixed costs, when abundance and zero or at least negligible marginal cost models are beginning to predominate. The DGSE models beloved of central banks ignore these huge structural shifts and hence the risk of unintended negative feedback loops. The ECB already faced a tricky technical challenge if it wishes to maintain an orderly, liquid Eurozone bond market. Further QE will involve difficult political trade-offs for EU governments touching on the essence of EMU; the clear danger of a divergence with US monetary policy as a catalyst for a transatlantic political confrontation may reinforce German intransigence.

For instance, buying fewer German and Dutch bonds, where the current 33% limit looms, and more Italian and French would add about €700bn to QE capacity, but the current split is based on capital contributions to the ECB. Changing it implies in the mind of a Berlin taxi driver a transfer from industrious German savers to spendthrift Italians. If the Fed simply delivers a 25bps ‘insurance cut’ In July or September, late cycle dollar resilience into 2020 will become a key risk to US growth, earnings and Trump’s re-election hopes. We’ve seen a concerted effort begin to bully currency ‘manipulators’, with wider EM Asia now (e.g. Vietnam steel tariffs) being dragged into that category. With the USD at its highest in REER terms since 2002, the US trade deficit has grown 6.4% y/y in the first five months of the year.

The Treasury has recently tightened its criteria for assessing possible currency “manipulation” in its latest Foreign Exchange Report. Of course, much of the Chinese deficit with the US is in reality a disguised Korean and Taiwanese one via reprocessing imports of high value components (notably semiconductors from TSMC and Samsung etc.) but as supply chains move out of China the true bilateral trade picture will become more obvious. The number of major US trading partners liable for scrutiny will extend well beyond China to include large current account surplus Asian countries like Thailand and Vietnam as well as Taiwan and Korea.

The Commerce Department has also proposed a regulation that would make currency “undervaluation” a countervailable subsidy, while the Treasury also cut the current account surplus threshold for triggering closer scrutiny from 3 to 2% of GDP. With the US bilateral trade balances reduced to a form of Trumpian national P&L, and the ex-oil deficit at a record, bond markets risk turmoil unless the Fed and ECB policy paths re-converge. If the White House can’t bully the Fed into matching ECB easing as an exercise in FX management, it will likely attempt via trade intervention to limit the extent of that easing.

Germany with (stalling) exports at almost 50% of GDP and domestic content in its US made cars at half Japanese levels is acutely vulnerable – pushing a further 10-20bps off Bund yields hardly justifies the growth shock risk of a full blown transatlantic tariff war. There is now a real danger that consensus hopes for the scale and impact of ECB action are overblown and yields face mean reversion volatility this summer. It may be constrained not only by its likely impotence in terms of boosting inflation expectations but the risk that asset buying proves entirely counterproductive if the US trade focus shifts from China to Europe.  If buying more BTPs implies Americans buying fewer BMWs, Lagarde will need all her famed political skills to keep Germany on board…

Huawei Quarantine Accelerates Tech ‘Balkanization’ Trend

“The three core issues of concern to China are the cancellation of all tariffs, that trade purchases be in line with reality, and that the text of the agreement be revised so as to be more balanced. These issues must be resolved. This trade friction has made us more aware that we have shortcomings in terms of economic structure, quality of development, and core technologies. [We must] enhance the sense of urgency in accelerating indigenous innovation and resolve the bottleneck imposed by insufficient mastery of core technologies. CCP Propaganda Department Commentary on the US confrontation in People’s Daily newspaper last Friday

We have regularly covered the rise of China’s tech sector over the past decade and the complacency in Silicon Valley which has only recently been shattered by Huawei’s threatened dominance of the global 5G rollout. The ‘balkanization’ trend already evident in software ecosystems from messaging to payments will now extend to hardware throughout the supply chain – there will be a China centric physical internet architecture over the next decade centred on Huawei as much as a software one around companies like Alibaba, Tencent and Bytedance with the big ‘neutral’ EM markets like India a key battleground.

In the near term, the notion that China can be quarantined as a technology supplier looks naïve – alternative vendors like Nokia and Ericsson make 5G hardware with Chinese partners in mainland factories for sale globally, leaving their equipment vulnerable to malicious tampering – will they have to relocate production to ‘safe’ locations to sell into the US? If so, the global rollout will be significantly delayed…

Huawei has been rapidly vertically integrating with its own chip making subsidiary, HiSilicon, producing highly advanced 5G designs, albeit fabricated (as for Apple and Google) by TSMC. As a Chinese tech VC contact told me in Shanghai last October, cut China off from TSMC and it would be casus belli like the Japanese oil embargo in 1940. If the US is serious about confronting China’s competitive threat on a sustained basis, it needs to boost government R&D spend on basic science (instead it’s been cut), deepen the capacity of domestic tech supply chains via education investment and targeted tax credits or cutting off broader Chinese access to advanced GPU chips and postgraduate student access to STEM courses at colleges such as MIT and Stanford. Those moves would be expensive and disruptive for the US economy and are still possible future steps if relations become even more antagonistic.

China offers uniquely low (marginal) consumer electronics assembly costs plus high-volume flexibility. You can never replicate the mainland factory dormitory model in the US or even Korea and Taiwan. The increasingly well educated rural migrants filling Foxconn factories are the ultimate ‘on demand’ workforce – manufacturing elsewhere would mean higher assembly costs/decreased flexibility thanks to the constraints of current generation assembly line robotics.

Even with high levels of automation, a fully US assembled iPhone would likely cost about 40% more at the factory gate (with Apple gross margins at ~37%, the current tariffs imply a price rise of about 15% to offset the impact, unless the RMB plunged toward 7.5 versus the USD). For instance, Quanta Computer, the largest laptop maker in the world with clients from Apple to Dell, warned last week that the logistical costs of shifting consumer electronics production out of China could prove as expensive as the tariffs themselves. 

Automation is part of the answer to offshoring from China but changes the business model – assembly by Chinese migrant workers is a marginal cost for a factory owner; robots are a fixed cost, although the assembly of an additional unit has zero marginal costs (excluding overheads like maintenance etc.). New AI software will help but it’s still expensive and time consuming to program factory robots to perform multiple tasks.

The implication of this shift from marginal to fixed costs is that there is a heavy incentive to stick with a specific design: any change requires significant capital investment to update the robotic assembly line – the flexibility of the entire consumer electronics sector will deteriorate, with higher inventories and fixed overheads once it loses China’s unique attributes as a global production base – given wafer thin margins, consumer prices will trend higher even if a face saving deal could avoid further tariffs.

Relocating the supply chain for smartphones or laptops will be hugely disruptive, wherever the destination and it certainly won’t be the US – the much-hyped Foxconn factory to make TV screens in Wisconsin has become a fiasco, despite $4bn in tax breaks and subsidies. The lack of a skilled manufacturing workforce will be a key constraint on US re-shoring, as much as logistics considerations – flying components in from Taiwan/S. Korea to assemble phones in the US makes little sense.

The effort to lobotomise Huawei looks ill thought out and will have generated intense lobbying by the most adversely impacted US suppliers – there will likely be some nuance in the implementation. However, the endgame is now clear and US tech companies will have to mitigate ongoing compliance risks by reducing exposure to Chinese SOEs. We’ve seen a few companies like Go-Pro begin relocation (to Mexico in this case), but this will now accelerate while China will race to become a ‘full stack’ technology power by mastering semiconductor fabrication and an indigenous mobile operating system.

In the case of both Iran and Huawei, the US has unilaterally exercised its global power over international bank payments systems and key technologies like the Android OS to exercise brute force geopolitical leverage – the lesson drawn by many in Europe and Asia is that alternative architectures are now needed. Ultimately, China which remains the only country apart from the US to understand the critical value of ‘platform’ software and is catching up rapidly in AI and quantum computing research, will become an even more formidable competitor to the US tech giants. The Xi 2025 plan, whose ambitions triggered panic in Washington from the Pentagon to Congress, will now be implemented sooner, by any and all means possible.


Google Takes Gaming From Console to Cloud

‘…with EA’s cloud gaming service and Microsoft’s plan to build a streaming, subscription-based service. Faster average broadband speeds mean lower latency is achievable – although graphic-heavy games may require the equivalent of the 5G network rolling out initially in Asia. The EA system uses essentially the same content delivery technology as used by Netflix and it is betting that the technological shifts that upended the music and movie industries are now coming to games, which looks shrewd. The advantage of owning the distribution platform which is the essence of the Tencent/NetEase strategy is clear in the mobile games space…the industry is slowly evolving toward subscription-based cloud gaming – game ownership, whether via disc or download, will become redundant and gaming publishers will boost recurring revenue. Of course, the flipside is that missing out on a key title launch can reverberate through lower earnings trends for several years – the quality of earnings improves for sector winners, but the volatility is higher for losers so sector performance dispersion will rise sharply.’ –  Weekly Insight, September 18th 2018

“Games are no longer episodic consumptive media for most people — they are now really the basis of new massive online communities that are a lot more like social networks in the way that they function and that we invest spend time in them. So much happening in the pattern of consumption of this medium that no one is paying attention to. You can’t put humans in a space like [Fortnite] for that long and not have massive, profound social consequences. We are just not seeing them yet. I’m almost a little bit ashamed as a gamer myself not to have seen just how big this is. This is staggeringly large.” Improbable co-founder Herman Narula interviewed recently

He’s right about the potential for massive multiplayer games to become the ultimate social network, and it behoves every equity portfolio manager to understand how fast this sector is now evolving. I’ve been a secular bull of the global gaming sector as a key play on the dematerialisation of consumption and the rise of the digital ‘metaverse’ in which real money is spent on acquiring virtual utility and status. As immersive technology becomes more sophisticated (particularly as AR/VR becomes practical via edge computing), more and more incremental consumption will shift from tangible ‘stuff’ to virtual experiences.

We remain long a global gaming exposure stock basket which is up almost 20% YTD and the point I made in that note last September was that the video gaming industry, which looks set to become the dominant form of entertainment for under 30s (directly playing, watching on video and other forms of live engagement such as esports), was inevitably set to follow the wider media transition to an on-demand subscription model. The legacy model of creating a hardware platform, such as Sony’s PlayStation and Nintendo’s Switch, and then charging publishers for the right to access it, is now under relentless pressure. The Japanese companies have responded by creating subscription services and offering content other than games, but Google’s push into the industry is a seminal moment. The search giant however not only has to provide a smooth lag-free experience, but also convince key game title publishers to port their content.

As Nintendo prepares to launch two new models of its Switch console (which generates over 80% of total revenue) the risk of annual sales peaking at just under 18m units is growing as the industry undergoes a paradigm shift. As cloud subscription-based streaming goes mainstream, the company’s real value may lie in its well-known game franchises, such as Pokémon licensed to the web giants hungry for content. The US web giants are likely to follow Tencent in vertically integrating to obtain content. The Netflix analogy is relevant – we are set to see a bidding war for game IP and an M&A surge for publishers over the next couple of years.

On its earnings call last month, Sony’s CFO said cloud gaming could pose a threat to PlayStation in the next five years but that the company believed that would take a much longer time horizon – that looks dangerously complacent. Google has set the bar at producing images in ultra-HD (also known as 4K) which is ahead of today’s consoles and at a fast-enough frame rate to make movements appear smooth (Google is matching the 60 frames per second of the latest Xbox and PlayStation consoles). Google’s data centres are among the world’s most powerful – to render images smoothly for Stadia, Google has built optimised new server “blades” that can be slotted into racks in its data centres.

These are embedded with custom-designed GPUs produced by AMD and in terms of raw computing power these generate over 10 teraflops to the individual user versus Microsoft’s Xbox One X at about 6 teraflops (though the next generation of consoles will likely double that). A key technological challenge will be to eliminate the latency, or lag, that can slow the response time when a player presses a button on the game controller. Google claims more than 7,500 nodes on its global computing network, increasing the chance that players will be close enough to see a fast response time but users outside large cities could find themselves many miles from a Google node. The new game controller is connected through Wi-Fi and communicates directly with Google’s data centre, but it will rely on local network operators for the household connection. Google must be betting that its latest video compression algorithms will be able to deliver the seamless quality of service required to be competitive.

Recent progress in this field has been impressive. For instance, Microsoft’s new Project Zipline compression algorithm is fast enough to compress data while it’s being written to an SSD or uploaded from an IoT device and can deliver up to 96% compression of the original data on the Microsoft internal network. The reason it’s so fast and so efficient is that it uses a custom hardware accelerator to look for many times more patterns than compression algorithms can usually handle; data that matches any of those patterns gets replaced by a reference to the pattern, taking up much less space.

The long-term implications for the hardware storage market of this AI pattern recognition/compression software approach are clearly adverse – the optimisation of slack which defines the tech sector is coming to data transmission and storage hardware. Google said it recommends broadband speeds of at least 25mbps to deliver HD images (known as 1080p). Some potential customers won’t have enough bandwidth. It also limits playing games on mobile devices, unless they are connected over Wi-Fi to wired networks. Future upgrades to mobile networks will help, but tellingly Google itself made no mention of the potential of 5G networks, which in the US/Europe won’t go mainstream until the mid-2020s given the gradual planned telco rollout.

As highlighted in that sector note last year, viewing activity on Twitch or Huya offers a useful insight into overall game popularity – watching expert players live-stream on these platforms correlates highly with game download/in app purchase activity. While Fortnite still has the largest number of Twitch channels, its share of viewers have been declining in recent months, down from a peak of 19% to 9%, being displaced the Apex Legends, which accounted for over 20% of viewers after its recent launch before settling at around 6-7% viewership while the PUBG share of viewers has been steady. Apex Legends has had the best launch month of any free-to-play game in history, having generated $92m last month from in-app spending. Fortnite (which is constantly updated) has been impacted, but recently reached over 250m players since launch.

While new game approvals in China have resumed, Tencent has yet to get permission to monetise Fortnite or the mobile version of PUBG. Sales from the Value-Added Services unit, which includes online games and messaging, climbed 9% to 43.7bn RMB but costs surged 43% y/y as investment in content and server financial technology surged. Tencent plans to introduce a new category of sales when it next reports to break down specific categories in the “others” revenue section in its financial statement. Tencent’s one-off costs contributed to a 19 ppt drop in operating margins to 20% but as a result its biggest revenue growth now comes from cloud services, where sales more than doubled last year, to 11% of revenues.

This is a crucial infrastructure investment for Tencent, not just to diversify away from gaming/WeChat revenue dependence but also to create the infrastructure to launch its own cloud gaming system (which is apparently already in beta testing). The web giants will likely demand exclusivity/original publisher content to drive subscription package differentiation as Microsoft, Tencent and Google launch competing cloud platforms within 12-18mths. As elsewhere across the media sector under the impact of streaming, content creators are in the sweet spot…

China Consumption Shifting From Property to Pensions…

One of the biggest issues for global investors over the next 3-5 years is that China may be running out of prospective home buyers as the prime household formation age cohort shrinks, creating a secular slowdown in a sector which directly generates 13-15% of GDP and in its wider spillover impact closer to 22-25%. Consumption in China doesn’t neatly fit the Western ‘life cycle’ model of neoclassical economics (e.g. housing tends to be bought in advance of marriage/household formation and left unfinished and empty until that point) but the structural demographic shift will transform housing demand as much as the shape of monetary and fiscal policy. Beijing’s efforts since 2015 to boost the birth-rate by easing restrictions on family size always looked likely to fail, given the ever-rising costs of child rearing and fertility trends across developed urban Asia.

Japan’s experience in recent decades indicates that when rapid growth begins to slow in an economy with very high corporate and household savings driving fixed investment, demand can prove extremely difficult to manage, particularly when demographic decline sets in simultaneously. This is particularly true if the deliberate promotion of credit growth and asset price bubbles has been part of the mechanism used to sustain demand. The tactical stance has been overweight China and AXJ despite still poor earnings and macro momentum but structural growth constraints are becoming binding as rising debt and declining demographics interact to radically change policy trade-offs, while the US is now intent on blocking or at least slowing significantly the technology upgrade path.

The real story behind China’s well documented economic imbalances is not just about structurally weak consumption versus investment as a share of GDP but also a large-scale net transfer of savings from abroad (and particularly the US) to the mainland corporate sector, a process which the White House, with broad bipartisan political support now seems committed to ending, whether a new trade deal is concluded or not. China’s plan to move up the value chain rapidly by ‘acquiring’ foreign IP to boost productivity as the workforce and investment intensity declines will now be much harder to achieve, even if it doesn’t in the worst case lose access to advanced semiconductor imports.

The country has an ongoing growth tailwind from urbanization (currently at 59% on official data, but by global standards likely 5-6 points higher), a remarkably advanced digital economy and payment systems, first world transport infrastructure in and between the major cities (and soon communication via 5G), and a growing number of globally competitive companies in mid-tech industrial and consumer markets. However, its economic dynamism will face a growing demographic drag as resource are diverted (whether public or private) to fund rising healthcare and pension costs.

Births last year dropped by 2m to 15.2m, and the median age will reach 48 by 2050, or about 10 years older than the US now. The total number of working-age adults (aged 16-59) fell by 0.44 % to just over 897m while the growth in the pool of rural migrant workers fell sharply, rising just 0.6% to 288m, down from 1.7% in 2017. The national old-age dependency ratio is already at 15%, and twice that in some depopulating peripheral provinces. The population on official data grew by 5.3m last year or 0.38% and Beijing has estimated that China’s population won’t peak until 2029 at around 1.44bn, but some demographers believe that tipping point has already arrived.

The ‘extensive’ growth model of adding more workers and capital becomes untenable as demographic decline starts. So too is the highly expansionary monetary policy that saw the PBoC balance sheet and M2/GDP ratio explode. That never generated much consumer inflation, as it was largely sterilised via housing which absorbed excess migrant labour and industrial capacity as a concrete inflation sink isn’t sustainable much longer. At the same time, the sterilisation of exporter dollar earnings and build-up of net foreign assets on the PBoC balance sheet and Treasury buying by SAFE is also winding down, as FX reserve growth peaks peak.

US demands to eliminate the bilateral deficit simply hasten the existing trend toward a current account deficit. Whatever the exact demographic glidepath, China is going to have to employ its human capital much more efficiently over the next couple of decades and refocus on intensive, productivity led growth. Global investors are going to have to adjust to the perpetual motion machine that drove global capital flows from the late 1990s not just stalling but going into reverse i.e. China will likely become a substantial net portfolio capital importer over the next decade, as it needs to fund soaring fiscal deficits, just as aging households begin running down savings…

With 12% more males than females in the 15-29 age cohort, having an apartment boosts prospects for marriage, and that factor as well as migrants buying properties to retire to in their home provinces helps explain much of the 45-50m ‘empty’ apartments that generate scare headlines. In China, housing has taken on the role of a ‘dowry’ for male offspring that gold has traditionally for female ones in India, but the 34m overall male surplus will rise and create a growing pool of involuntarily unmarried men (so-called “bare branches”). The home ownership rate among young Chinese households is consequently very high, thanks to help from parents who in a major city will have built up huge housing equity. Demographics will clearly begin to impact this cultural support for real estate, as the number of 20something males enters steep decline.

The 20-29 age cohort, the main source of new demand for housing, will have declined by about 80m by the mid-2020s from its peak in 2012/13 and the proportionate decline is similar for the teenage cohort behind them, slowing pre-emptive parental demand. A rising divorce rate and proportion of never married (in the case of many males, not by choice but economic circumstance) will offer some support to housing demand, but alongside a dramatic slowdown in migration, the overall fundamental demand picture will deteriorate materially.

Pension coverage is now relatively high for a country at China’s income level but the income generating assets to fund defined benefits are hugely inadequate and this is where boosting private assets and returns becomes critical to maintaining systemic solvency. The pension funding deficit covered by central government is likely to reach over $150bn annually by next year. Deeper capital markets (including ultimately access by foreign mutual funds with local distribution partners) supported by pension savings inflows are a key part of the wider reform agenda. Current contribution rates for state/SOE pensions are far too low; private pensions and employer annuities (i.e. the addressable market for the insurers) are just over a quarter of total assets with the basic pension/national social security fund comprising the balance.

That ratio will gradually shift over the next decade in favour of private assets – the public pension system’s 43% replacement rate (ratio of annual benefits to final salary) implies a significant cash flow deficit will open up that could amount to over $1.5trn within a decade. Insurers will be a key part of the funding solution – weak capital markets and regulatory changes slowing premium growth have been a drag but they remain a key China exposure as inadequate social provision, from healthcare to pensions, is funded directly by individuals.

Total pension assets are just over 10% of GDP compared to 35% in Korea and HK. Assets will have to grow dramatically over the next decade to close the funding gap and the private share of total pension assets (currently sub 30%) will become dominant. Beijing already has the fallout from local government deleveraging and SOE restructuring to absorb which will see central government debt/GDP double to 70-80% over the next 3-5 years from the currently reported 37% – bailing out the pension system as well simply isn’t realistic. Assets managed by Chinese insurers have already reached over $2.6trn, even as new policy premium sales have slowed since 2016.   Solvency rules are now closer to international norms – capital requirements had been based on simple metrics of size but will now vary in line with how quickly policies turn over and how premiums are invested. Firms that rely excessively on short duration policies or invest heavily in equities must hold a much bigger capital cushion.

The slide in bond yields and A-shares has inevitably hit investment returns (as evidenced by the recent China Life profit warning), but offsetting this is an improving competitive landscape. The restrictions on wealth management product issuance (bank WMPs were offering yields of about 5.4% a year ago versus an average guaranteed rate offered by universal insurance products 30-70 bps lower) has seen retail investors return to insurers. Even with foreign firms likely to grow their share from the current low 5% base as the market opens, investors seem too bearish on life insurer growth prospects, with the key stocks on sub 1x price to embedded value multiples. As with education, investors in the asset gathering/private pension theme have to see through regulatory volatility to focus on the secular tailwind for revenue growth.

‘Peak Smartphone’ Slams Tech Hardware

‘…we are reaching ‘peak smartphone’ and the upgrade cycle is lengthening as new features prove less than compelling – the mobile app landscape has certainly become stagnant. Heavy discounting of the Samsung S8 ($150 plus) is apparent already in the US and the recent profit warning by the UK’s largest phone retailer highlighted a consumer behavioural shift. While the hype cycle will intensify ahead of this month’s launch of a premium advanced OLED screen model with facial recognition security, if the queues outside Apple shops for the new iPhones prove surprisingly short (or short lived), watch out below.’ Weekly Insight, September 6th, 2017

‘The ‘Peak Smartphone’ theme justified an underweight stance back in Q4 across the most heavily exposure Asian supply chain stocks, with DDR4 DRAM prices down about 20% from its January peak and NAND flash down a third. Many DRAM bulls put premature faith in AI and AVs as incremental demand drivers, certainly valid medium term, but not on a scale to offset smartphone/PV weakness this year while for chip makers the end of the crypto mining frenzy is starting to impact. TSMC said this week that sales will rise by only a ‘high single-digit’ percentage in USD terms, down from a previous (reduced) projection of 10%.’ Weekly Insight, July 20th 2018

The smartphone saturation risk we covered over a year ago has gone mainstream and rippled through the global hardware supply chain, with NAND flash prices crashing this year, DRAM also now peaking while production volumes for premium new phone models have disappointed across the sector. It’s hard to believe that most tech analysts missed this pretty obvious inflection point, when even the IMF noticed the industry had topped out as an Asian growth driver (smartphone shipments were about a sixth of global trade growth last year), but mindless extrapolation of the prevailing trend remains the default setting of bottom up analysts.

Smartphones have seen little innovation beyond camera quality and biometric security since 2016; the lack of compelling new hardware features (or new apps requiring them), and the growing ‘digital detox’ trend as awareness of the adverse productivity impact of notification addiction rises have both dragged on demand growth. Tighter network upgrade policies as well as limited ‘must have’ innovation have seen US and European consumers replacing phones every 2.5 plus years, a rise of 7-8 months since early 2016.

As we highlighted a year ago, the refurbished market has seen explosive growth and was the fastest growth segment last year, reaching over 140m units and is still growing at a mid-late teens pace. Three-year old ‘as new’ iPhones and Samsungs have been flooding emerging markets at typically a third of the price but nearly all the functionality of new. The weak outlook was confirmed by several Apple suppliers this week including Lumentum, Japan Display and UK chipmaker IQE this week – Apple as a ’luxury’ tech brand has finally succumbed to wider sector dynamics, although higher ASPs and the shift to services mean that the component supply chain takes a bigger hit.

Until 5G rolls out at scale from 2020 (and compelling new use cases beyond watching Netflix on the move have yet to appear), it’s hard to see the wireless sector regaining much impetus. While Samsung and Huawei plan foldable screen launches next year, initial volumes are likely to be modest at a price point above $1,000. Valuations now certainly look far more reasonable and expectations more realistic, but the downgrade cycle looks set to run into Q1 19. Our view remains to focus exposure  on stocks with high exposure to the nascent autonomous vehicle/automation sensor and emerging consumer segments such as smart speakers.

However, there will be an opportunity in the 5G rollout in niche areas such as high-end optical chips/dark fibre as well as smart antennae. Chinese producers will have a head start and gain critical mass, as the country embarks on the telecom equivalent of the hugely impressive high-speed rail network, which I used several times on a research visit last month. There will certainly be opportunities in China for specialist foreign vendors like Nokia, but the ‘full strength’ 5G being implemented via a new national core network is a critical component in the effort to drive domestic suppliers up the hardware value chain, in both phones and wireless infrastructure.