Over the past quarter, Huddlestock* was up +12.92% versus +7.45% for the NASDAQ 100**, +7.12% for the MSCI All Country World Index*** and 3.14% for the HFRX Equity Hedge Index****. Huddlestock was up +3.06% in January.
Several factors affected performance results in January including the bursting of the cryptocurrency bubble, the probable end of a bond bull market that started in 1981 and the best first week in the equity markets since 1987.
In this brief, we will be discussing the bifurcation that a switch to a rising interest rate environment is likely to cause. We will argue for increasing our tilt into emerging and frontier markets and will focus on one of the drivers of our secular tech-driven, bull market hypothesis, artificial intelligence.
* Estimated actual performance of all strategies/themes on the Huddlestock platform on an unleveraged basis.
** The NASDAQ 100 is a stock market index made up of 107 equity securities issued by 100 of the largest non-financial companies listed on the NASDAQ. It is a modified capitalization-weighted index.
*** The MSCI ACWI Index is designed to represent the performance of the full opportunity set of large- and mid-cap stocks across 23 developed and 24 emerging markets. It covers approximately 85% of the free float-adjusted market capitalization in each market.
**** The HFRX Global Hedge Fund Index is designed to be representative of the overall composition of the hedge fund universe. It is comprised of all eligible hedge fund strategies.
Emerging and Frontier Markets
The US has a more developed economic cycle than emerging markets and has an equity market that is priced at optimistically high valuations. Coupling this with the distinct possibility that the normalisation of quantitative easing, in the form of interest rate hikes, will be more aggressive than the market is currently expecting we have become somewhat cautious on US equities generally.
We are also expecting volatility in US equity markets to increase materially as the federal reserve pushes ahead with the difficult task of undoing past stimulus. This is particularly true at this stage of this cycle because most asset classes have become very sensitive to rising interest rates given how low they are.
There are some exceptions to this thinking however, and while it makes sense to move into assets that provide long exposure to normalisation (like banks), a very promising ‘4th Industrial revolution’ is also underway.
Leading technology companies with low debt and accelerating earnings involved in secular game-changers like artificial intelligence, industrial automation, cloud computing, autonomous vehicles and the internet-of-things are clearly also very promising.
The return of relatively stable commodity prices combined with additional demand in the US and Europe (supported by an uptick in private corporate expenditure) will continue to support growth.
The IMF now expects GDP in emerging economies to accelerate every year through 2021 and for developing economies to decelerate starting in 2018. Importantly, emerging markets are less vulnerable to higher US interest rates than they were in the past because many of these countries have healthier current account balances and have less USD reliant debt profiles.
Acuity Partners, a full-service investment bank in Sri Lanka, has joined us as a strategy vendor and will start providing you with local Sri Lankan investment opportunities in the coming weeks.
What does this mean for Investors?
Within US equity markets we are expecting higher volatility than we have seen over the past few years and a shift towards companies that will benefit from higher interest rates. We believe high-flying technology companies that aren’t overly-leveraged will continue to benefit because there is still a lot of money sitting on the sidelines waiting to be deployed.
This view will lead us to occasionally seek to hedge volatility, especially around events that reflect on whether the US economy is bumping up against its capacity constraints (tight labour market, wage growth, inflation, etc.). At this stage, we continue to believe that we are in the melt-up period of this expansion though that view is sensitive to the rapidity with which interest rates rise and the extent to which that surprises the market.
Artificial Intelligence and Machine Learning
Artificial intelligence (AI) and machine learning (ML) are among the most significant technological developments in recent history. We expect every industry to be materially affected by this technology over the coming years.
Here we want to look at some of the latest developments in the field, summarise some of the current applications of the technology (some of which can be quite hidden from view) and name some of the companies that have deployed significant resources in this area.
ML is dependent on the cumbersome process of feeding large amounts of data into algorithms to train them and allow them to learn. Recent advances in how neural networks (one of the cornerstone algorithms in ML) operate are making them able to learn based on far less data, without losing accuracy. This could have serious consequences for the hardware industry, especially those focused on developing ever more powerful processors designed to run today’s data intensive AI algorithms.
AI running on quantum computers that are accessible via the cloud would likely change the hardware landscape completely. We are seeing a race between Chinese and American technology giants to deliver AI online and compete to become the dominant player in what’s widely believed to be the next big computing paradigm.
Below is a non-exhaustive list of areas where AI/ML is being applied:
Security: Face detection is being used to improve security at public venues while online it is being used for monitoring, preventing and responding to cyber attacks.
Trading: High frequency traders use ML on proprietary ASICs chips to execute trades in picoseconds and beat their competitors in the race to exploit the ‘plumbing’ of the financial system. We have added a ML strategy to Huddlestock that looks to profit from the reflexive interaction between ETFs and their constituent stocks.
Healthcare: ML algorithms can process more information and more accurately detect patterns in solving diagnostic and prognostic problems than their human counterparts.
Marketing Personalization: ML is being used to better understand a customer’s needs to improve conversion rates and predict whether a new product will be successful.
Fraud Detection: ML is being used to distinguish between legitimate and fraudulent transactions.
Recommendations: Products/Movies/Music is recommended based on your tastes compared to that of millions of others.
Online Search: Search engines have stepped away from brute force search methods and indexing and are using ML to serve up the content you’re most likely to be searching for. It’s also being applied to being able to search images by content.
Language, conversation and chatbots: Natural language processing is being used in applications that stand in for customer service agents. It’s also being used on legal contracts, real-time translation, biometric security and speech recognition, which is in turn used for voice search.
Smart Cars and Cities: Autonomous vehicles and inter-networked devices across cities are going to fundamentally change transportation and logistics.
IoT and Big Data: Extracting valuable, structured information from vast amounts of unstructured data like that generated by the internet-of-things is ripe for ML. One of the visions we have at Huddlestock is that this information will be plugged straight into the platform to provide you with real-time investment ideas as the sensors themselves detect and extract relevant, actionable information.
There are many hundreds of companies busy with AI/ML, here’s a handful that have caught the attention at the MIT Technology Review:
What does this mean for Investors?
AI/ML will only become more important going forward and is likely to affect everyone and everything. At this stage we are interested in trying to anticipate which companies are going to dominate this new computing paradigm. The race to be a dominant player has already begun and will only intensify and as such we need to understand the effect that significant developments could have on other companies, such as hardware providers. Buying ETFs due to their simplicity will become progressively less effective as these companies start to disrupt each other as well.
We will either see funds flow out of simple and popular ETFs, as market participants realise that they provide only an illusion of diversification, or we’ll see continued inflows which we believe, combined with the uncertainty surrounding the impact of rising interest rates, will have a high likelihood of eventually causing (at worst) or amplifying (at best) the next equity bear market. As a result, we intend to start limiting our exposure to simple and popular ETFs and see a need to start tracking to what degree each investment idea is exposed to flows into or out of these products.
Algo-Chain, managed by Dr Allan Lane and his team, recently joined Huddlestock as a strategy vendor. They are focused on carefully selecting ETF investments based on their deep understanding of these products. Allan was formerly the head of ETF Product Research at BlackRock and the Global Head of Quantitative Research at RBS.
Pro Tip: Risk Constraints
Every month we’ll be focusing on delivering an insight that will help you get the most out of Huddlestock.
This month we would like to share with you some of the basics of picking and subscribing to a suitable strategy/theme.
The difference between a theme and a strategy is that themes tend to relate to transitory phenomena that focus on a limited number of companies while strategies are based on deeper philosophical and abstract considerations related to return generation, style rotation, portfolio construction and risk management. An example of a theme is the legalisation of marijuana theme while an example of a strategy is the ‘Carbon: Top picks strategy’ which postulates that carefully sized investments in companies with a low carbon footprint will outperform on average.
Three initial considerations in picking a strategy/theme relate to the type of strategy (i.e. long only, short only or long/short), the horizon and the risk level.
‘Long only’ strategies only deliver investment ideas that would make money if the price of the investment rises. ‘Short only’ strategies only deliver short ideas which make money when the price falls. A long/short strategy will deliver both types and could do so by delivering one of each (a pair) that have been specifically tailored to offset each other’s risks while delivering an outsized risk-adjusted return.
The horizon is an indication of how long the vendor is likely to keep your money invested in the idea before closing it. This number is only an indication since anticipating an exact holding period at the time that the investment idea is submitted is a nearly impossible task in most cases.
The risk level that we assign to a strategy is the result of the ongoing due diligence process we subject each strategy to. Besides objective criteria related to historical performance we also consider the experience of the vendor, the universe of investments they will be focussing on and how well their investment approach fits with the investment climate.
Once you have narrowed down the list of strategies to those that are most suitable to you it is useful to understand a strategy’s historical performance. We have done substantial research into which aspects of a track record are suggestive of skill (as opposed to luck). Skill has the property of being replicable through time, a property, which luck, by its very nature, cannot have. We report some strategy performance metrics which you can find in the ‘statistics’ section of each strategy. We have found that understanding a vendor’s behaviour when faced with a drawdown period is most indicative of skill.
For any feedback, please email us at email@example.com.
Michel van Tol, PhD
Chief Investment Officer
Download this brief as a PDF file here.