‘Volatility Volcano’ Erupts…

Changes in the real Fed funds rate have historically led realised equity volatility by about two years, due to the lags between official rate moves and risk-taking (the Fed started hiking in December 2015, albeit at a glacial pace). Low realized volatility feeding into quant based theoretical models has always fuelled the intellectual hubris of finance PhDs (think LTCM etc.) and ultimately proved toxic. The pre-crisis period shows that misplaced correlation assumptions can lead to a far more benign assessment of overall asset risk than is prudent.’  Weekly Insight ‘Sitting on the Volatility Volcano…’ Oct 12th 2017

‘That issue of deteriorating market depth and the proliferation of highly correlated factor based strategies which are de facto short volatility/long equity beta (including long duration corporate credit as an equity proxy) will become a big story next year…the value at risk models are in this sustained low realized volatility environment at maximum exposure to (particularly US) equities. A shock to consensus positioning, be it from inflation, policy or politics could see an ‘air pocket’ liquidity event. Overall, it looks wise to look for ways to trade against to trade the prevailing bias that growth, inflation and interest rates are anchored permanently lower.’ Weekly Insight, 18th December 2017

Our view coming into 2017 was that market structure rather than macro was the biggest risk and to prepare for both higher rates and volatility. That meant underweighting rate duration, being long equity reflation winners and finding volatility hedges. The importance of this selloff is to signal a shift to more nuanced risk appetite after the simplistic ‘melt up’ hype. As highlighted in those Q4 notes referencing the ominous LTCM precedent, the quant/factor investing boom was vulnerable to a paradigm shift in its input variables. While the financial engineers tinkering with factor models suffer a reality check, it’s unlikely that the multi-year bull market is over with earnings growth momentum still accelerating in many markets. Most systematic momentum following funds which soared in January are now down YTD, but given limited leverage overall, this doesn’t look a systemic event like LTCM threatened to become.

However, the role of ultra-low interest rates as a discounting mechanism and low realised volatility as a driver of equity appetite for factor-based strategies such as risk parity has belatedly come into focus for the consensus. If Q1 16 was about investors stress testing portfolios for deflation risks, this time the adjustment is to a higher risk-free discount rate; the 2-year bond overtaking the S&P dividend yield last month was a cautionary signal. When we initiated a VIX long in our tactical portfolio in October, it was one of the most glaring anomalies across markets and generated a 2.6x return when we took profits in this week’s panic.

The length of this correction will be determined by whether a ‘buy the dip’ mentality prevails among an influx of new Millennial investors evident in recent statements from online brokers like TD Ameritrade, as much as whether 10-yr yields will top out at 3% term. It’s unclear if some have been buying stocks on their credit cards, as they clearly were crypto coins in Q4 according to MasterCard’s latest results call (and watch for a spike in card delinquencies in Q2).

The rise in long-term US rates so far is less about Fed policy or inflation expectations as it is a looser fiscal stance. As highlighted in that December note, the supply of US bonds will rise sharply this year at the same moment that demand potentially ebbs as real economy demand for capital in Europe/EM rises. Meanwhile, the Fed will buy $420bn fewer Treasuries than it did in 2017 and in 2019 will reduce purchases by another $600bn. It certainly helps that the ECB and Bank of Japan will be buying nearly all local sovereign issuance and the jump in yields may encourage some active multi-asset managers to re-weight fixed income over equities, as well as the automatic risk-parity rebalancing.

Exposure to key secular themes such as autonomous vehicles/robots, the shift to ‘biological software’ in the drug industry etc. should be opportunistically accumulated into weakness. Earnings momentum globally remains strong;  the scale of guidance upgrades in Japan which has a dominant position in several emerging technology supply chains such as Lidar, vision sensors, monoclonal antibodies and EV batteries should offer support once markets settle. Meanwhile, active investors can take some comfort from the humiliation of the quants whose share of the market has risen to unhealthy and potentially destabilising levels.


Oil Market Punishing Recency Bias

Overall, Peak Oil has now been replaced by Peak Oil Demand as a popular investment thesis, but timescale is as ever critical in investment. While the rise of electric vehicles is inexorable over the next 15-20 years as productivity drives a better power to weight and cost ratio for batteries, recently the popularity of SUV sales in both the US and EM implies a cyclical upswing in gasoline demand through end decade. Official statistics can often mislead based on poor sampling and opaque inventory accumulation.’ Macro Weekly Insight – May 18th 2017

The dramatic surge in oil since late summer has surprised the consensus and energy equities have in recent weeks been catching up with the crude move. The oil rally reflects the fundamental tightening we were writing about since late Q2. Crude futures moving into backwardation last summer saw oil tanker storage turn uneconomic and the popular ‘glut’ narrative which saw several high profile bank analysts predict $40-45/bl Brent this year is now shifting.

While global demand is still surprising to the upside amid unusually synchronized growth, and OPEC holding the line at least through mid-year, US shale supply ‘elasticity’ remains a key factor. By 2025, the growth in American oil production will equal that achieved by Saudi Arabia at the height of its expansion, the IEA has claimed in its annual World Energy Outlook, turning the US into a net exporter of fossil energy as shale liquids output reaches 13m bpd out of a US total approaching 17m. Recency bias or the behavioural tendency to apply undue weight to the latest data and simply extrapolate that trend is endemic in the forecasting game. History suggests that the IEA  (relied upon by IB analysts to derive their forecasts) has a very grubby crystal ball, having serially underestimated EM demand for instance – their shale forecasts don’t look plausible.

oil production

Source: US EIA

I wrote a note in July 2013 entitled ‘Are Oil Prices Ignoring a Technology Driven Supply Boost?’, highlighting that US output growth continued to surprise the consensus, while emerging market demand growth was already clearly peaking with Chinese fixed asset investment, but it took almost a year for the re-pricing of that supply shock to begin. Having been oblivious to shale’s impact back then, investors are being complacent as to its sustainability now, and the per rig output data for the key shale regions (which has been volatile recently) bears close watching as much as the overall rig count.

Consensus fears that shale DUCs (drilled uncompleted wells) will flood the market with supply again look unrealistic. When these wells are completed, it will be gradual and the natural decline in legacy shale production will be difficult to overcome – geology will likely dominate shale ‘manufacturing’ innovation. Production from the early Eagle Ford and Permian plays has been declining – it looks unlikely that net shale production will increase by more than a low single digit % in 2018.

We’re seeing a shift to capital discipline across a sector that generated negative free cashflow of $170bn over the past five years. While QE failed to ignite a wider capex recovery, it helped create an investment boom in tech and shale, but capital markets are growing impatient with a sector chronically unable to generate positive returns on investment. The days of shareholder value destructive shale output growth at any cost look to be over, as executive incentives evolve.

Offshore oil service stocks and rig builders remain unattractive versus onshore as global energy capex will likely remain depressed, aside from high IRR productivity led projects. Low marginal cost E&P plays look vulnerable to a round of M&A consolidation by the majors who have been replacing only about 60% of production with new reserves. If the consensus assumption that shale can reaccelerate output growth is overoptimistic, then OPEC will grow its effective market share again. Keeping 1.2 mb/d (plus nearly 0.6 m bpd from non-OPEC) offline for another year could push the market into a deficit situation, leading to accelerated inventory drawdowns and prices heading to the $80-90/bl range in H1 19. That’s a scenario worth stress testing in portfolios…


Tech ‘Second Movers’Begin to Perform

The right place this year has been in tech, above all for EM investors with the sector now over a quarter of the MSCI index and accounting for almost half of total performance YTD. The rise of ‘embedded intelligence’ as vision and other sensors enabled by AI software become ubiquitous has been a key theme of ours in recent years, and has attracted a belated consensus frenzy this year, with pure play stocks from Nvidia in the US to iFlytek in China soaring. However, the growth into value tilt we’ve suggested since the summer applies within as well as across sectors. There has been a pretty spectacular valuation arbitrage as new hardware technology and business models are adopted from retail to autos, among incumbents who sit on a fraction of the multiples of the perceived pure play tech leaders. We will see a similar trend in banking/insurance over the next couple of years as fintech startups broadly disappoint, but their innovations slash operating and customer acquisition costs for the more far sighted established names.

The global auto sector has rallied hard since the summer, partly because investors are waking up to the fact that several automakers have very undervalued IP in the EV/AV/transportation as a service space that offsets the wider industry stranded asset risk. This reflects the historical pattern as innovation diffuses – the re-rating opportunity from tech will increasingly be among incumbents adopting new technology to boost competitiveness, from Japanese banks like SBI experimenting with blockchain/crypto payments to car makers like GM, Toyota or Ford able to scale EV/AV roll out better than Tesla ever can. Selling digital advertising via click-bait news feeds is infinitely easier than mass manufacturing.

Indeed, the aggregation effects that fuelled social media/search and e-commerce platforms are not generally relevant in hardware, which tends to get commoditized rapidly. From an investment perspective we’ll have a multiplicity of overlapping technologies and ways to play their relative success, most of which will be in long established but reinventing blue chips. Disruption reflected in relative performance may well be driven more from within incumbent sectors (reflected in rising return dispersion) than from outside them.

WalMart has rallied strongly in recent months versus Amazon and GM versus Tesla as both start benefiting from the underestimated ‘second mover’ advantages of allying scale economies with new business models and technologies. GM or GE were not first movers in diesel locomotive engines in the 1930/40s but by rapidly buying up promising start-ups and harnessing them to vast engineering and balance sheet resources ended up dominating the transition from steam over the subsequent two decades, crushing smaller ‘first mover’ competitors along the way.

second movers

The parallels with GM’s pivotal Cruise Automation acquisition and Chevy Bolt platform being rolled out via Lyft are worth watching closely (with Ford attempting a similar reinvention) as are WalMart’s belated e-commerce shift via the Jet.com acquisition. The breathless Silicon Valley hype is ignoring a looming period of accelerated evolution within sectors that will radically change their leadership and where spotting the sector Dodos unable to rapidly evolve within a portfolio is critical.

As an example, being a value investor in tech is generally a losing proposition but Intel which in terms of chip sector evolution has become a ‘second mover’ by missing the GPU versus CPU shift has been racing to integrate a series of strategic acquisitions (notably Mobileye in autos and Nervana in AI chip design) to play catch up in this market – if it succeeds against low expectations, the huge valuation discount to the perceived AI market leaders will close and versus AMD relative performance has now turned decisively. Tech disruption won’t be the much feared extinction event for adaptable incumbents – the quality of management will be key alongside aggressive M&A to preempt new entrant threats, but there are growing signs that hugely rated tech insurgents are not  going to have it all their own way for much longer. Therein lies the investment opportunity…


Blockchain Deconstructs the Corporation…

JPM CEO Jamie Dimon recently declared that he would sack any of his traders playing bitcoin which he compared to Tulip mania, but given the broad volatility famine that has crushed Wall Street’s Q3 results, any half decent trader would surely be keeping their skills sharp in the wild world of cryptocurrencies. It’s important to differentiate between permissioned networks (e.g. across banks to settle securities) with a central authority and fully autonomous ‘permissionless’ ones such as that underpinning bitcoin, which threaten the role of banks as the apex of a longstanding financial hierarchy. Near term, blockchain offers banks, insurers and asset managers a margin windfall by slashing processing costs – long term, it threatens to undermine their gatekeeping role.

While I’ve been a sceptic of the recent ICO boom and the resulting profusion of new cryptocurrencies (supply certainly isn’t limited in the aggregate), but blockchain is probably comparable in long-term impact to the TCP/IP internet protocol in the 1990s. In other words, it could prove as significant an enabling technology in transactional terms as the original internet protocol was in transforming communications. By now, all investors are aware of the disruption risk to many incumbent business models from new technologies that shatter barriers to entry and compress margins and pricing power, but there is an even more fundamental question looming.

Could the firm, in the sense of the formal corporate entities that have evolved since Dutch spice expeditions to Asia gathered risk sharing investors in the 17th century, now become unbundled?  The information search comparative advantage of the firm versus consumer  or individual entrepreneur in classic microeconomic theory has narrowed dramatically while the cost of enforcing and supervising contract execution has tumbled via sharing economy models and soon the widespread use of digital ledgers.

As a reminder, blockchain is a distributed ledger technology, and uses a self-sustaining, peer-to-peer database to record and manage transactions in a decentralized manner. Verification of data is undertaken via complex algorithms and consensus among multiple systems, making it almost completely tamper-proof.  Data is transferred or stored using a series of blocks, each of which have a cryptographic hash protecting its contents. Any update or transaction on the data creates a new record in the form of another block, which is added to the existing blockchain.

The defining, radical characteristic of the digital economy in terms of traditional economics is its zero-marginal cost and we’ve now reached a point where something can be simultaneously digital and unique, without any tangible representation (bitcoin being a case in point). The bedrock of economic exchange is trust, and we now have technology that allows a reliable degree of trust to be established between total strangers in order to share resources, which has had profound implications for investors. That principle can be applied via the blockchain to anything from stock certificates to trust contracts, property title deeds etc. In that scenario, the elaborate hierarchy of trust we have built over several centuries via the legal system and complex compliance and oversight procedures becomes increasingly redundant.

This new distributed trust architecture could replicate much of the organization of a firm built on contracts, from incorporation to supply chain relationships to employee relations. If contracts can be automated, then what will happen to traditional firm hierarchical management structures, processes, and intermediaries like lawyers? Much of the corporate world is built on exploiting transactional ‘friction’ and information asymmetries between consumers and suppliers – the Internet over the past two decades has been a force for reducing these profit opportunities, from hotel room pricing to asset management. While crowdsourced trust between strangers via sharing economy platforms has been powerful in this respect, the advent of the blockchain over the next few years looks far more revolutionary.

As I’ve highlighted in many notes, online services and the advent of AI software are reducing slack and redundancy in the economic system – the internet from its inception has been about reducing search costs and price asymmetries between producers and consumers i.e. acting as a fundamentally deflationary force (a shift which central bankers still struggle to adapt their outdated equilibrium models to). Sharing economy platforms are zero marginal cost business models in terms of adding inventory (e.g.  advertising for Google/Facebook).

Google, Facebook, Twitter or Airbnb rely on the contributions of users as a means to generate value within their own platforms. Of course, in the existing sharing economy model, the value produced by participants is harvested by the platform owner with most of those contributing to the value production getting nothing beyond free access to services like social media, search or marketing their products.

Blockchain changes that because it facilitates the exchange of value in a secure and decentralized manner, without the need for an intermediary – to that extent, it’s a medium term threat to incumbent web giant business models, which are centralized rather than distributed and extract economic rent from often unwitting users. Digital ledgers and smart contracts reduce the capacity of the firm to enforce and exploit ‘trust arbitrage’. A far more fluid economic structure will develop over the next decade and beyond, with project focused associations of specialists coalescing and disbanding, their contractual obligations and financial entitlements will be delivered via blockchain.

Next year, we will begin to see significant blockchain and ‘smart contract’ deployment across the finance sector from maritime insurance contracts that rewrite themselves in real time to asset leasing and securities settlement. The logistics/freight forwarding business is another ripe for disruption and millions of mid-level admin jobs will be automated out of existence globally over the next few years as a result. While positive for early mover finance sector margins over the next few years, the implications for prime real estate look ominous, as office towers, which are simply warehouses of human inventory, begin to empty out.

Blockchain protocols make it technologically possible for a ‘Decentralized Autonomous Organization’ (DAO) of individuals with relevant skills to associate and organize for a specific task with the power to execute smart contracts between them and a client that can replicate many of the functions of a traditional corporate entity.  This shift marks the advent of a new generation of “dematerialized” organizations that do not require physical offices, assets, or even formal employees. Of course, this utopian vision won’t apply where regulatory compliance demands a centralized authority or significant capital/fixed investment is required but can certainly work across many ‘asset light’ service sectors from design to advertising and professional services, where freelancing is already common.

There is also a trade-off between blockchain network size versus transaction frequency. The blockchain is so far struggling to solve this trade-off between network size versus transactional frequency, and until that technical dilemma is resolved it’s hard to see this technology competing with existing payment networks like credit card networks or SWIFT on a global scale anytime soon (i.e. achieving several thousand transactions per second).

However, despite these and other limitations, the huge scale of capital now being invested in this area (some of it via the notorious ICOs) and surging interest from blue chip companies (particularly in Japan, as covered recently) suggest that every investor needs to reflect on the nature and implications of this potentially revolutionary technology. The current generation of cryptocurrencies may ultimately fall by the wayside, but the innovations that enabled them almost certainly won’t…

Asian Crisis Still Resonates…

There haven’t been any official events to mark the 20th anniversary of the Asian crisis, which erupted when Thailand freely floated the THB on July 2nd 1997 (having struggled to fund an 8% current account deficit, twice the average across EM Asia at the time). Within a year that had caused a domino effect ‘margin call’ across emerging markets, culminating in the collapse of Russian markets and the devaluation of the ruble in August 1998. Russia only accounted for just over 4% of world GDP in 1997, but was a major borrower of short-term capital and the contagion effects were rapid.

What was until then a regional crisis spread across GEM and then global markets, causing spreads and volatility to surge with a cross correlated violence that blew up value-at-risk models. That contributed to the collapse of $126bn AUM hedge fund LTCM by year-end, which led to a Fed engineered bailout. That would have been a potential shock on a par with Lehman’s implosion, but the fallout was pre-empted unlike in 2008 – LTCM had successfully hedged most of the risk from the Asian currency crisis and delivered a 17% after fees return in 1997 (after over 40% in each of the previous two years by applying huge leverage on its ‘relative value’ strategy).

Global capital flows since then have been shaped by these events, as Asian (and wider EM) governments committed to never allow themselves to become vulnerable to speculative attack again, driving mercantilist policies and a huge build-up in official reserves as a bulwark. Asia’s currency reserves stand at almost $6.3trn over half the global total, versus less than $1trn in 1996 as a portfolio flow ‘margin call’ loomed.

Those reserves were largely recycled into Treasuries, funding persistent US current account deficits and the steady deterioration of the country’s net international position or ‘net worth’. Aggregate Asian CA surpluses are back at pre-crisis levels of over $600bn (and with China’s surplus looking understated because of very questionable growth of the estimated tourism deficit as covered in a recent Fed analysis, probably well over $700bn).

I’ve just been to Budapest, where the raddled face of hedge fund legend George Soros stares down from government sponsored billboards, labelling him  as a scheming currency manipulator (he has accused his native country under its right-wing, anti-immigrant leadership of becoming a ‘mafia state’). It seems unlikely that for all his pseudo intellectual ‘reflexivity’ theories Soros could have anticipated the longer-term fallout from his currency attacks across Asia in those tumultuous months. One of those is that he helped end the era of swashbuckling macro hedge funds bending intimidated governments to their will – no hedge fund balance sheet can fight an aggressively deployed central bank one.

As we highlighted back in 2014/15 as the consensus abandoned EM assets, the move to floating exchange rate systems and high reserves versus FX debt made another crisis, which was widely feared at the time, highly unlikely – tumbling currencies acted as an EM macro pressure relief valve and forcibly closed current account deficits from Indonesia to Brazil, underpinning the ‘surprise’ rally in EM assets over the past 18mths. However, the sustainability of the capital recycling model built up since the late 1990s looks very doubtful – the US economy simply can’t absorb huge surpluses from both Asian and Europe any longer as the ‘consumer of last resort’.

The Asian crisis led indirectly to the 2008 global one, as the ‘savings glut’ being forced upon (a very receptive and overly deregulated) US suppressed funding costs, as explained by Ben Bernanke back in 2015. In the interim, the Fed by rapidly cutting rates in an overreaction to the LTCM debacle fuelled the Dotcom bubble – unintended consequences have been the defining feature of central bank policymaking over the past two decades. Meantime, is there another LTCM combination of hubris and leverage lurking within the global financial system after another extended period of low volatility? We’ll find out as the first central banks take tentative steps toward policy and balance sheet normalization in the next 12-18mths, but it’s probably not in Asia…

Easy Money Made in Tech?

The simplistic ‘Trumpflation’ trade has now almost fully reversed as expected in December when we suggested a non-consensus overweight in global tech, with the USD back at pre-election levels mid surreal chaos at the White House. The growing dysfunction in Washington has been offset by the broadly positive tone to Q1 corporate results as well as political tail risks in Europe receding until the Italian election next year – global earnings momentum continues to be led by Europe, Japan and HK/China while India was the weakest major market yet again in April and its re-rating looks unjustified.

Since late Q4, we’ve seen a steady value into growth rotation, as tech has become the biggest global overweight bet on institutional positioning surveys (notably in EM where active fund tech allocations have now overtaken financials) followed by banks (notably Eurozone, which have substantially outperformed US YTD) while defensives  are being shunned.

Tech has now become a momentum trade, although parallels to 1999/2000 are (with a handful of exceptions) unjustified, in that rising earnings expectations have broadly underpinned soaring stock prices as nearly every leading global web and hardware name has beaten (rising) consensus expectations, from Nvidia to Tencent. Investors and sell side analysts back in Q4 underestimated scale/aggregation effects for the global internet names and the strength of the demand/pricing environment for semiconductors, and underweight funds have had to close bearish bets.

A longstanding theme is that we’re seeing an explosion in tech innovation driven by ever cheaper data collection/analysis and that aggregate market earnings power will increasingly concentrate in this sector. That’s reflected in the S&P 500 IT sector forward margin on consensus IBES data reaching more than double that of rest of the market (9.7% vs. over 20%). From the 2009 trough at about 9%, the IT sector has seen a spectacular level of margin expansion, as well as revenue growth (over 10% forecast this year vs. 3.5% ex IT).

US equity margins ex IT remain well below pre-crisis trend and the tech weighting now at 22% as much as relative risk asset inflows explains much of US equity outperformance in recent years versus Europe/Japan. Nonetheless, the overall tech sector has moved from a consensus underweight to overweight since December while the risk of a deeper than expected Chinese slowdown in H2 amid signs that pricing in more commodity chip markets like DRAM is now peaking suggest it’s time to be much more selective. The growing popular pushback against perceived monopolistic Silicon Valley economic ‘rent seeking’ and corporate tax arbitrage is also becoming a medium-term risk.

The tech smartphone supply chain in Taiwan has seen a notable recent deterioration in analyst estimates, dragging the overall index revisions ratio negative. I’d expect memory chip names to see similar downgrades this month and next. The easy money has been made in the overweight Asian tech hardware trade which was a suggested allocation in Q4 and we would now be taking profits and reallocating toward unloved energy and rate sensitives.

Contrarian Bets Pay Off in Q1…

‘Wall Street has projected its wildest fantasies onto the largely blank canvas Trump provided, but this is politics reinvented as performance art. As the lack of policy coherence or even consistency becomes painfully clear, investors risk a degree of buyer’s remorse. Many of the key policy suggestions look contradictory, such as the border adjusted corporate tax system and its impact on boosting the already ‘too strong’ USD or the proposed tax cuts… political gridlock is now a risk if he also splits the Republican Party with what is fast looking more like a Silvio Berlusconi than Reagan style administration.’ Weekly Insight, 23rd Jan 2017

A key theme since December was that the consensus assumption of a rampant USD,  10-yr Treasury yields testing 3% and accelerating US growth looked as misguided as the deflation panic a year ago. The implosion of what we termed a month ago as the ‘incoherent mess’ healthcare proposal occurred during our recent Asian roadshow, further shaking complacency. The overweight bet in EM equities paid off, with a 13% return the best quarter in five years while a resumption of carry trade inflows boosted EM FX to its second-best quarter led by the 11% Mexican Peso rally (18% from its January low), distilling the broader Trump reassessment across markets.

Global growth has outperformed value as expected and our preference for North Asian cyclical exposure (notably tech) has been justified – Korean exports surged almost 14% last month led by semis while Samsung is one of several regional tech names hitting new highs. Chinese and European demand rather than US has been the key upside driver for Asian trade this year. India’s rally has been surprisingly strong and driven further multiple expansion on a surge in domestic mutual fund inflows YTD, but the credit and earnings growth backdrop remains very weak. Reflecting the rapid shift in sentiment, on IIF data net capital flows to EM were positive last month, with China seeing the first  inflow since 2014.

With the yield curve flattening again, US banks have now joined the reversal of crowded post-election consensus trades. The hope that feuding Republicans factions can now regroup and move swiftly on tax and regulatory reform looks optimistic – the CBO projected $839bn in savings on Medicaid spending over a decade as 14m beneficiaries were removed, a key offset for deep corporate tax cuts. The border adjustment tax was meant to be a revenue windfall to cut the headline rate, but is already mired in Congressional wrangling and the conservative Freedom Caucus looks set for a war of attrition ahead of next year’s mid-term elections.

The Washington swamp has drained Trump, whose ignorance of even basic policy details and lack of a core political base in Congress were always obvious weaknesses which will have to be addressed but this is not a man with any experience in building coalitions. Before tax reform can even be discussed seriously, the highly controversial budget has to be passed, with many Senators angry as brutal cuts to key departments and basic scientific research budgets. Our view remains that tax cuts will be both more modest and later than most expect.

With the dollar index testing four month lows, we have taken profits on the tactical short on USDJPY and the FTSE 350 mining sector, which has corrected with the sharp iron ore reversal. In contrast to further upside (particularly for surprises in Europe, recent US data looks soft. That ranges from the slump in gasoline and supermarket food sales (both branded and generic) to a broad downturn in bank credit growth.

With investment banks belatedly noticing the divergence between ‘soft’ survey and hard reported data, consensus inflation and growth expectations look vulnerable for H2. We’ve been highlighting rising stress in the auto loan market since Q4 and the auto market bears close watching in Q2 for deterioration in sales volumes, loan delinquencies and residual values.  New vehicle incentives are reaching new highs at about $3500 per unit while second hand values are now falling at an almost 8% y/y pace, as subprime repossessions spike after a 5% decline in 2016.

These signs of underlying US consumer weakness will be key to Q2 asset allocation, as will be the risk of China tightening policy more broadly to slow a still booming housing market. For the US consumer, while belated IRS tax refunds will help, the shock of absorbing 20% plus hikes in healthcare insurance costs will squeeze discretionary spending across low income households in H2. Growth is more likely to slow to about 1.5% rather than accelerate and we remain underweight US ex tech versus EM and Europe.

There are signs that the recessionary style freefall in gasoline demand is now abating, which drove US inventories to a record in February and has been a bigger factor in the crude selloff than shale output rebounding. If we see an OPEC output cap extension, as seems likely give the Saudi desperation to get the Aramco deal away next year, a bullish stance on energy should pay off and we’re now adding long global oil E&P exposure to the tactical portfolio. Downside on that trade would require  US gasoline sales to sustain the remarkably weak trend seen in Q1, in which case the wider economy is heading for a much deeper slowdown and the 10-year is more likely to test 2 than 3% by mid-year…

Tech Growth Rebounds as USD Reverses

“This year has been about deep value mean reversion, but next should see underlying secular growth themes reassert…” –  Weekly Insight, 5th December 2016

‘Poker has been one of the hardest games for AI to crack, because you see only partial information about the game state. There is no single optimal move. Instead, an AI player has to randomize its actions to make opponents uncertain when it is bluffing…’ Andrew Ng, chief scientist at Baidu

Having been tactically overweight value (resources and industrial cyclicals) all of last year, we suggested in December that the post US election selloff in global tech as investors belatedly rushed to make reflation bets looked like an attractive contrarian opportunity. Since then both Chinese and US large cap tech names have outperformed from Alibaba to Facebook, with even Tesla enjoying a sharp rally. The overwhelming consensus expected a rampant USD, but this wasn’t consistent with the Trump trade agenda – the dollar index has had its worst January in over a decade. We remain tactically long EM local currency debt and short USDJPY but have now closed the long Japan exporter position for a 24% gain – Japan is still well represented in our long industrial automation and Embedded Intelligence sensor stock baskets. The dollar selloff has certainly helped the US tech sector given its high overseas exposure, but superior earnings momentum/visibility remains the key support, underpinned by broadly above expectations Q4 results.

At 22x earnings, the S&P 500 premium rating looks reasonable given a series of multi-year product cycles – the shift to cloud computing, rapid progress in AI/machine learning and a steady rollout of internet of things sensors as well as augmented/virtual reality offer sequential and overlapping growth drivers through end decade. The latest results from Microsoft, Intel and Alphabet indicated that cloud computing is the most significant near-term driver of growth while the recent CES hardware event in Las Vegas was dominated by new AI applications.

Meantime, a key theme remains that portfolio diversification will increasingly be embedded within 15-20 tech ‘conglomerates’ as they disrupt adjacent legacy sectors with lagging productivity.  Amazon and Uber have both recently expanded into ocean freight and trucking logistics respectively. The key advantage these tech groups have is their ability to optimize slack capacity and automate scheduling via superior data algorithms. The latest developments in ‘reinforcement learning’ AI are remarkable, notably the Libratus program which has exceeded the much hyped achievement of Google’s AlphaGo by mastering the most complex form of poker. It learned the game from scratch using an algorithm called counterfactual regret minimization (which sounds like it might prove useful to many investors).

By playing at random initially, over several months of iterative ‘training’ (involving several trillion hands of poker) it reached a level where it could outwit elite human players by predicting their moves, playing a much wider range of bets and randomizing these bets  – all without ever first being given the rules of the game. As we have long highlighted, the real long-term shareholder value in technology investing is in the layers of proprietary software that bind networks together and aggregate their data flow. The rapid progress in the models used for AI application development and associated processing hardware amplifies the scale of that opportunity (and also boosts the value of niche hardware sensor makers collecting the newly valuable data).

It seems inevitable that well paid ‘pattern recognition’ type roles from insurance claims assessors to junior M&A lawyers, auditors and hospital radiographers face automation over the next decade by increasingly intelligent software. AI technology has leapt with astonishing speed from facial recognition to putting on a poker face, and as it evolves further investors and politicians will struggle to absorb the profound economic implications, but one will be a growing concentration of incremental corporate earnings and free cash flow growth within the technology sector.

Reflation Trade Becomes Consensus…

In a vintage year for contrarian asset allocation, the trajectory for markets has been to overreact to China deflation/devaluation risks, the shock of the misguided negative rates experiment (which served to confirm deflation fears and therefore backfired as a signalling tool) and what we termed the  ‘sideshow’ of Brexit. Overly pessimistic expectations were reset from Q2 as macro data such as PMIs and net earnings revisions across the MSCI AC index rebounded (led by EM), reflected in the turn in bond yields from July as ‘macro hypochondria’ peaked.

Positioning for the commodity cycle and EM bottoming in Q1 looked to us like a compelling bet, as did overweighting value as a style factor. Globally, value has outperformed by about 20 ppts since July, the second-best run versus ‘quality’ as a portfolio factor in almost 40 years. Our rotation from bond yield flattening to steepening sector exposure mid-year (including long Eurostoxx banks) also generated huge alpha in the tactical portfolio. If there was one issue this year which reflected a spectacular failure of nuanced analysis and caused global shock waves, it was the Q1 panic over Chinese FX reserves and the risk of a ‘shock devaluation’ which coincided with equally hysterical forecasts of oil going to $20 to generate a deflation panic rippling across markets.

Our view was that at least half the apparent capital flight was in fact rapid FX debt deleveraging and that rather than a deflationary shock, China was poised to inject an inflationary impetus to the global economy as the investment cycle turned. Ironically, the rebound in China’s fixed investment (particularly residential construction) has been weaker than we expected but the surge in industrial commodity prices even stronger, now that reflation is suddenly fashionable as an investment thesis. We stuck with a $50-60/bl end 2016 oil price target as market rebalancing began, making global E&P stocks and US HY energy debt a contrarian overweight. Meanwhile, the RMB correcting from very overvalued real effective levels was something to be welcomed, so long as orderly.

The US election has provided an alibi for sell side strategists to wipe the slate clean on an astonishing series of analytical errors and belatedly jump on the reflation trade. If the overwhelming investor consensus a year ago was for further EM/commodity downside and long quality/growth, this year it is long USD and reflation winners and positioning has shifted accordingly but simplistic extrapolation remains the default forecasting tool of most analysts. Indeed, almost comical herding behaviour continues to define markets from oil to the JPY (for instance, the yen has seen consensus targets versus the USD gyrate from 125 in mid-2015 to 90 by mid-2016 and now back to 125 again). Investors have been stampeding in and out of markets with record speed as the narrative and momentum shifts. With the rise in AUM of systematic trend following funds and the retail ETF hordes that ride their coattails, markets have never been more of an ‘echo chamber’ of confirmation bias  – remaining strictly agnostic is key to getting ahead of crucial inflection points.

The consensus was hugely misguided a year ago, and the impact of the Trump regime whether in terms of tax policy or geopolitics looks less benign than it now naively expects. Ultimately, if Trump is to attract cheers rather than jeers at the rallies he intends to continue holding for his rustbelt true believers, low-skilled labour has got to get a bigger share of the economic pie i.e. companies will have to divert cash flow from shareholder distributions to capex and wages and shift their capital structure from debt to equity.

Tax cuts will accordingly come with strings attached designed to incentivise investment and manufacturing re-shoring. The USD and Treasury yields are not a one-way bet given subtle negative feedback loops and both look extended, global miners ex gold now look relatively expensive and quality growth/defensives decent value again. While 2017 is unlikely to be quite the contrarian bonanza this year was, it would be wise to read those investment bank annual outlook notes with a healthy degree of scepticism. Or perhaps safer still to ignore them completely…






China Housing Booms on Land Shortage…

‘The property sector contributes about 15% of GDP directly, but up to twice that on a broad multiplier basis on HKMA estimates, while real estate accounts for about 70% of household assets. On a cyclical basis, the housing market has been improving since mid-Q1, despite the misguided Spanish/Irish property crash parallels drawn by some observers (ignoring the huge equity cushion, the lack of securitization or a secondary market to propagate panic selling etc.). Ongoing easing measures and liquidity displacement from equity markets will boost property in the leading Chinese cities further this summer. Ultimately, the government desperately needs the fixed investment cycle in real estate to stabilize…’ Macro Weekly, May 22nd 2015

‘Mainland developer destocking continues to tighten the market as sector lending picks up. Indeed, the chart showing cumulative real estate floor space started versus sold is a critical one right now – it shows the huge inventory build-up in recent years has peaked, ex the Tier-3 cities – it’s hard to see destocking going much further. Your view on global cyclicals/materials as well as Chinese momentum hinges on whether you believe that property investment is set to stage a rebound.’  Macro Weekly, Jan 11th 2016

It was clear by mid-2015 that a cyclical rebound in China’s property market was looming, driven by healthy fundamental demand as inventory levels peaked, yet this critical macro inflection point was bizarrely overlooked by most commentators. We’ve gone from a China ‘running out of FX reserves’ consensus narrative in January to a ‘running out of apartments’ one now i.e. the feared deflationary impulse has become an inflationary one for the global economy, even before housing investment catches up with the inventory cycle.

Indeed the surprise has been the relatively weak new build response, but the suggested Q1 overweight bet in domestic deep cyclicals exposed to the construction cycle as well as global materials/mining has outperformed substantially this year. The largest Chinese cities have joined many in the West from San Francisco to Dublin and London where soaring prices have failed to generate significant developer new supply, partly because of zoned land and funding shortages.

This latest property boom is therefore very different to the 2009-2013 one, when developers and local SOEs rather than households were the driving force and leveraging aggressively. Given the fevered price surge, which is reaching even third tier cities, it’s remarkable that new starts fell over 19% y/y in September. Rapidly shrinking inventories offer fundamental support, although the recent spectacular pace of official price growth (exceeding 30-40% y/y across several of the 15 largest cities) is clearly unsustainable.

The backdrop remains broadly positive – rapid urbanization (from an official but likely understated 56% in 2015) is spurring first-time housing demand; demand for housing upgrades will increase alongside rising incomes as about 40% of urban households are living in low-quality housing built before 2000 while there are more than 230m people aged 20-29 who will be first time buyers over the next decade. The affordability ratio nationwide was at the lowest end of the historical range at just over 7x income as of the end of 2015, making property relatively attractive as interest rates fell and A-shares imploded. The ratio of house prices to household disposable income in Tier-1 cities was just under 15x at the end of 2015 and has now risen to about 20x, approaching HK levels and well above London/NY.

However, these conventional metrics are distorted by the huge amount of space hoarded as ‘concrete deposit boxes’ by the top 1% of households (who own about 25% of vacant space) and substantial undeclared ‘grey income’ for the elite. An alternative ‘reality check’ is to look at housing wealth versus national income; aggregate Spanish residential property peaked at about 460% of GDP in 2007; Australia (and NZ), considered the frothiest global real estate markets alongside Canada (and a surge in Chinese buying has been a major factor in all three) are now worth over 350% of GDP.

The US peaked at just under 200% and is now 140%. China’s property market is worth 350-400% of GDP at current price levels. Land remains the key form of collateral within corporate balance sheets and the wider financial system, as with Japan in the 1980s and national average prices are up 5x since 2004 on the independent Wharton/NSU/ Tsinghua index, and 11x in Beijing. It’s important to note that Spain, the US and Ireland all saw a dramatic supply response to rising prices which alongside the collapse in credit availability helped crash the market 40-50% peak to trough post 2008. China doesn’t face the same series of shocks but real estate is now clearly overheated and calming the market is becoming a policy priority for Beijing.

The bottleneck is what has been historically been a rapid new build response in China capping price surges is land. In the first nine months of 2016, residential land supply in 100 medium and large cities dropped by 10% y/y, even as apartment sales soared. In Shanghai and Beijing, land supply through September dropped by 33% and 80% respectively y/y. The recent trend of developers bidding aggressively for premium land is directly due to low levels of zoned land for sale (itself partly a function of the ongoing anti-corruption drive, which has created local government paralysis in what has been a very lucrative activity for city level bureaucrats).

Significantly, the Shanghai government has just tightened regulation on the financing of land purchases – developers are not allowed to buy land using financing from banks, trusts etc. but from internal funds. Developers violating the new regulations would lose the deposit they make before auctions and will be banned from purchasing land in Shanghai for three years but this looks more a supply than demand side issue.  The only sustainable solution is significantly more land and new floor space supply over the next year, further boosting producer prices and cyclical growth momentum, even as housing prices peak.