Tomorrow’s Titans 2022: Alexander Hyll

Identifying paradigms with a quantamental approach

The Hedge Fund Journal
Originally published on 01 August 2022
  • Founder, CIO and CEO, Adaptive Hedge Fund Management, Linköping, Sweden

Structural shifts are a constant in an ever-transforming world, creating new opportunities and challenges for industries, businesses and individuals. Determined to capitalize on structural winners and losers from these shifts, Linköping-based manager Alexander Hyll launched Adaptive Paradigm Alpha in early 2020. Following more than two years of successful management and several hires, Hyll and his team are launching a fully authorised alternative investment fund with the same strategy and name in the third quarter of this year.

Ahead of founding Adaptive Hedge Fund Management, Hyll perceived two market gaps in the Nordic hedge fund space: “There were very few quantamental equity long/short funds that combined quantitative analysis with a fundamental overlay,” recalls Hyll. “But more importantly, even fewer offered a genuinely market neutral return profile with low correlations and volatility.”

His large cap global equity market neutral strategy, Adaptive Paradigm Alpha, has achieved target returns of 10-12% and volatility of no more than 3-4% during the first two years. Absolute return generation has been almost evenly balanced between long and short positions, and beta has been near zero. The largest drawdown has been just 0.73% using daily data and 0.31% on calendar monthly data.

Hyll explains finding strong causal links is paramount to the strategy. “While there may be distractions in the shorter term, we believe that the economy is operated by people who are mostly rational. This means that a certain stimulus or measurable trend has a given foreseeable effect. Based on this conviction, we look for measurable relationships and trends in the market where we can connect a cause and effect. We view connections between such relationships as paradigms.”

Paradigms might sometimes be long/short pairs, but they can also have three or four or more constituents including various hedges. “We have found that investing on a relative basis has several advantages compared to individual long or short positions. It really suits our strategy and goal well,” says Hyll. “A significant change to certain value chain will affect that landscape forever, and regardless of the direction of the general market, some companies will be preferred over others.”

Over 100 tail risks are monitored and those surpassing certain thresholds can be hedged with a mix of currencies, commodities, and sector ETFs. A paradigm has a soft target of 30% performance, but a hard stop at -10% loss. “The majority of our paradigms are structured using single name equity options, so the hard stop is built into the position from the start,” explains Hyll.

Proprietary data and quantamental process

Quantitative analysis is used to test and corroborate fundamental theses, including statistical regressions and clustering to gauge if the postulated linkages are working. Proprietary datasets are gathered from research institutes and web scraping. The data includes some traditional fundamental and macro data, but alternative data could be as much as 70% of the mix, with examples being companies’ carbon footprints, credit card spending and electrical networks in China. 

It is not easy to define an exact percentage split between quantitative and discretionary inputs because their relative importance varies between paradigms. On balance Hyll estimates that, “The process may be more quantitatively driven, with human discretion a sanity check for market risk or company risk outside the algorithm. Equally, in some cases quantitative analysis disproves a fundamental view”.

Believability and consensus driven culture

The team of five in-house includes two investment professionals, a COO, a quantitative analyst and a software coder. The general company board includes senior people with operational, risk and compliance expertise. Hyll has ultimate responsibility as CIO, but the firm’s believability culture creates a democratic and meritocratic forum for competing ideas. “For instance, I am more quantitative whereas co-portfolio manager Arin Kamangar is more fundamental, so his view carries more weight in fundamental matters. Arguments need to be backed up by evidence and so far, nearly every decision has been consensus-based after internal discussion and debate,” says Hyll. 

The culture is partly inspired by Bridgewater principles but has also been modified to fit the Adaptive culture: “Radical truth and transparency is intended to synthesize where we do not agree, create trust to accept feedback, and take personal prestige out of our decision making. We are all working towards common solutions and goals,” says Hyll.

Examples of paradigms

Idea generation for a global mandate can seem like a big task. Hyll explains that the team is structured in its research. “When researching paradigms, we find ideas within global megatrends such as new technology, ongoing socio-demographic change or sustainability, or in smaller more sector or geographic specific trends or disruptions.” 

A profitable US semiconductor paradigm shorting Nvidia versus AMD had several drivers. “As GPU architecture shrinks down to 3 nanometres, it is harder for Nvidia to maintain an advantage. Nvidia said it gained a temporary boost from taking a huge market share in crypto mining where its kernels were suitable, and others were not,” explains Hyll. 

A “smart farming” paradigm encapsulated two megatrends in one position. “In order to solve the increasing need for food for a growing world population, digital and automated precision farming tools are needed to increase crop yields. In addition, environmental goals require reduced use of fertilizers and pesticides,” points out Hyll. “We reviewed the status of the agricultural fleet in the US, and together with high crop prices and low interest rates we assessed conditions were favorable for an investment cycle.” This trade was exited after the fund’s long position was acquired by a global machinery company, but it could be revisited in some form in the future. “This trend is still ongoing, and we believe elevated crop input prices and drought will contribute to an increased awareness for agricultural efficiency,” remarks Hyll.

Style factor and Covid exposures 

The big picture mega trends might sound like they have a growth bias – and possibly would do for long only or long biased investors – but Adaptive’s paradigms contain hedges, most of them are smaller trends, and some of them are wagering on losers from disruption. “During our company selection, no style factor, whether it is growth, value, or cyclical, will make up an outsized part of the fund. We spend much time monitoring concentration risks, both when introducing a new paradigm to the portfolio, and also on an ongoing basis as new macro factors appear. Our balanced long/short attribution is a result of evenly distributing bullish and bearish biased themes,” says Hyll, and other style factors are managed ad hoc. Controlling Covid risks required some deliberate adjustments. Hyll reckons exposure to either ‘stay at home’ and ‘reopening or normalization’ themes was pretty much neutralized and balanced. “Macro themes come and go. The fund had positive performance during the first quarter of 2020 and 2022, both difficult periods, and that is something we are proud of,” says Hyll. 

Hit rates, time frames, and exit triggers

After going through a diligent analysis and thorough feasibility stage, around two ideas per quarter become paradigms and 90% of them are invested in eventually, but some may spend time on a watchlist before being activated. “If a paradigm adds to concentration risk it might need to await exits from other paradigms to find its place,” says Hyll. 

The intended holding period of a paradigm is 12-18 months but can be shorter if a profit target is met sooner or if it gets stopped out; the average life so far has been just under one year. 

Over the past year out of ten paradigms in the portfolio, two have hit their stop loss, and some others have been exited at smaller losses. “E-broadcasting was a theme where we relied on sell-side research that overestimated some companies’ sensitivity to the theme. We later found it was other companies that were benefiting from it. In response we made changes to the use of sell-side research, further investigating sensitivity factors as part of our primary research,” says Hyll.

Paradigms can also be exited if disqualifying risks, including ESG risks such as bad governance or environmental factors are identified; in one case there was implicit oil exposure that Adaptive did not want to hedge. 

ESG

Many paradigms have an ESG dimension. An energy efficiency and renovation paradigm is based both on the growth prospects of the companies, driven by EU support packages, and probable investor demand for the stocks with a clear sustainability profile, whether EU taxonomy aligned or not. Adaptive are monitoring paradigms’ alignment with the evolving EU taxonomy, but not targeting any specific percentage alignment. 

An opioid theme has identified alternatives with lower risk of addiction, misuse, overdose and potential to treat abstinence. “This has a social dimension though that is not the primary driver. The main thesis is based on the need for healthier alternatives less prone to overdoses at a time when one of the largest drugmakers has exited the market,” says Hyll.

Adaptive anticipates making disclosures under SFDR category 8 and are closely following how the legislation will be implemented. “We want to leave a sustainable financial footprint,” says Hyll. 

Carbon markets are being monitored, though Adaptive has not traded it yet.  

Leverage, beta forecasting and rebalancing 

Gross exposure, including delta-adjusted option exposures have historically been around 200%, which is one reason why the strategy has undershot its volatility target – and some factors are being recalibrated to raise volatility and returns closer to targets. Gross exposure is determined by a range of factors, including realized and implied volatility, recent performance, and the Kelly criterion: “We look at mathematics and game theory around optimal bet sizes,” says Hyll. 

Adaptive has developed a proprietary method of forecasting beta, partly to enhance hedging efficiency. It is based on both historical and forecast variables, which is logical when the investment philosophy seeks new patterns of correlation. “Historical volatility is compared with the estimated impact of correlations in relation to drivers,” says Hyll. Adaptive has also devised its own beta model, which blends elements of the CAPM and APT. “After starting research, we found some inspiration in earlier academic research, but decided to go in another direction. Our beta provides a better benchmark for hedges: they are 5-7% more effective using the in-house beta forecast than a standard benchmark.” The techniques were developed by the quantitative analysis team, which has support from one member of the investment advisory board, Professor Martin Singull, a specialist in multivariate statistics, who helps to validate models, methods, and applications of time series. 

The strategy allows limited breathing room for beta exposure, not to make deliberate directional bets, but rather to optimize the transaction costs of rebalancing. “Maintaining strict beta neutrality could be too expensive when asset volatility can move beta exposure significantly on a daily basis. Typically, we rebalance to keep beta within a range of +-5% and estimate that this rebalancing costs around 1% per year in more volatile periods,” says Hyll.

Dynamically optimizing option structures 

Most trades are expressed through option spreads, though the balance between options and cash equities varies with option market liquidity, which has deteriorated in 2022 according to Adaptive. 

Precise option structures also adapt to market conditions. Higher option costs can lead to using more out of the money strikes. The gap between implied and realized volatility on the benchmark further influences the balance between debit and credit option spreads. “We prefer to be long of premium but had a period of being biased to credit spreads. Trading costs are another factor we take into consideration. Implied volatility on options has also pushed us back toward credit spreads recently,” says Hyll.

Most of the options are plain vanilla, one to one ratio spreads, though other ratios can be used to offset vega or gamma. Exotics have not been used. Nearly all the options traded are exchange listed, though there could be an occasional OTC or swaption. 

Adaptive have backtested the use of options compared with cash equities, and found it adds an average of 1% per year, of which 90% came from net receipt of premium income and only 10% was from capped downside; it has been rare for the paradigms to gap through the hard stop of 10%, and this is another possible reason to review tightness of spreads. 

Service providers and vehicles

SEB is prime broker and custodian with Goldman Sachs as executing broker. A local Swedish administrator has initially been selected for a Swedish special fund. The Swedish domiciled fund will be marketed in Sweden, Norway, Finland and Switzerland. The strategy is estimated to have a capacity of $3-4 billion before meaningful alpha deterioration. There are plans to launch a Luxembourg fund, but no date is currently set.

Current soft commitments in seed round are from a hedge fund entrepreneur, companies, trusts, family and friends. Targeting a Q3 2022 launch of its special fund with early bird discounts for seed investors, Adaptive is currently in closing discussions with a large Nordic financial institution regarding participation in the seeding of the fund. Hyll states that, “This is the next step towards our goal of being a premiere institutional source of alpha”.

Full Report

Tomorrow’s Titans 2022