- Invesco study reveals that the artificial intelligence (AI) revolution is already underway: half of systematic investors have integrated AI into the investment process and the majority (75%) expect AI to match or exceed importance of traditional investment analysis within a decade.
- AI is mostly used to understand market trends and to optimize portfolio allocations; investors see potential in testing investment strategies and monitoring and adjusting trading positions in real-time.
- 41% of respondents are using natural language processing for sentiment analytics, with three-quarters expecting to use it in future.
- Most investors believe systematic strategies helped them navigate challenging market conditions in 2022.
- Systematic investing is evolving: investors are broadening their systematic toolkits by using more diverse strategies.
- ‘Growth’ becomes an established factor.
London, 30 October 2023 –Invesco today released the findings of its eighth annual Invesco Global Systematic Investing Study. The Invesco Global Systematic Investing Study (link here) is an evolution of the Invesco Global Factor Investing Study, published annually since 2016. The reposition this year reflects the changes within the quantitative investing world, and the use of quantitative methods beyond just factors. The study, which is based on the views of 130 institutional and wholesale investors that collectively manage $22.5 trillion in assets, also finds a growing consensus that the systematic toolkit can help investors navigate key challenges, such as volatile markets and imperfect data.
The report found that half of systematic investors have already integrated artificial intelligence (AI) into their investment process, and reveals a widespread expectation that AI tools will transform portfolio management in the years to come. The majority (62%) anticipate that, within a decade, AI will be as important as traditional investment analysis and 13% expect it to become more important.
The AI revolution already underway
Systematic investors are already using AI across a range of core functions.
Respondents reported harnessing AI to better understand the market environment and identify macroeconomic turning points: (46%) are using AI to identify patterns in market behaviour, and (38%) are using it for portfolio allocations and risk management. Investors appreciate AI’s ability to help mitigate human biases and forecast the unexpected.
Investors expect the use of AI to grow significantly in the coming years. While a significant minority (29%) already use it to develop and test investment strategies, the vast majority (76%) anticipate doing this in future, and while (20%) currently use it to monitor and adjust investments positions in real-time, more than half (55%) expect to do so moving forward.
Wholesale investors identified improved risk management as the main benefit of AI, cited by (76%) of respondents, followed by the flexibility to adapt to changing market conditions (65%). However, challenges remain: wholesale respondents cited the cost of implementation (64%) and the complexity and interpretability of AI models (61%) as the main obstacles to adoption (Figures 2 and 3).
“Among wholesale investors, we found a concern around the potential for AI-driven portfolio strategies to overshadow traditional models”, said Bernhard Langer, CIO, Quantitative Strategies at Invesco. “There is a sense that AI-driven models will be attractive to investors moving forward, particularly younger ones, meaning firms must adapt quickly.”
Institutional investors instead see accurate and timely insights (78%) as the most compelling benefit of AI, followed by improved risk management (74%) and increased efficiency and automation (68%). Their primary concerns are complexity (78%) and data quality and completeness (51%).
“The key challenge for institutional investors is stakeholder management. Investors need to be able to explain and justify the use of AI models as their stakeholders are wary of ‘black box’ solutions”, continued Langer. “The regulatory landscape surrounding the use of AI and decision accountability also remains ambiguous.”
The rise of natural language processing tools
Investors have embraced natural language processing (NLP) tools, which have been harnessed for a range of operations, such as summarising and digesting whitepapers, converting recommendations into accessible language for sales teams, and modifying communication tonality for different client groups.
NLP models have also been deployed in the investment process. (41%) of respondents are using NLP for sentiment analytics, and around three-quarters (73%) expect to do so in the future. Several investors reported searching online social channels to uncover prevailing market narratives around firms, measuring frequency of mentions and context, providing valuable insight for assessing risks and making short-term trading decisions (Figure 4).
APAC and North America lead the way
However, Invesco’s study found significant regional variations in attitudes towards AI and NLP, with investors in EMEA markedly more sceptical than their APAC and North America counterparts.
The majority (51%) of EMEA investors believe that AI will still be less important than traditional analysis methods in ten years’ time, versus just (10%) in North America and (7%) in APAC. Conversely, just (4%) of EMEA investors believe AI will supplant traditional analysis methods in that period, with much higher numbers observed in both North America (19%) and APAC (20%).
Moreover, North America and APAC investors are currently far more likely to be using AI in the investment process. APAC investors are twice as likely as EMEA investors to be using AI to identify patterns in market behaviour, and more than three times as likely to be using AI to adjust investment positions in real time. EMEA investors trail North America and APAC investors in each aspect of AI adoption.
The growing systematic toolkit helps investors tame markets
Factor investing has historically been the cornerstone of systematic investing, but Invesco’s study reveals a far larger toolkit of systematic strategies that have helped investors navigate the key challenges of recent years.
Tools to decipher the macroeconomic environment have become especially important, and the ability of systematic approaches to help mitigate market risks was a key theme in this year’s study: the majority (63%) of investors agreed that systematic strategies helped them manage market volatility in the past year. Moreover, nearly(60%) of respondents said that the new higher inflation market regime was supportive of the systematic approach, with only (6%) of institutional investors and (10%) of wholesale investors disagreeing.
For three-quarters of respondents, dynamic asset allocation has become a core component of their approach, helping them to rebalance and adjust their portfolios in response to the market environment. Systematic tools have helped investors identify and characterise the underlying macroeconomic regime, allowing them to make inferences about its impact on different asset classes, factors, regions, and sectors.
“Recent challenges have prompted investors to question how they navigate unexpected obstacles”, said Langer. “Respondents spoke of expanding beyond factors to better understand markets and identify when certain asset classes tend to outperform others”.
Bridging the ESG data gap
However, the usefulness of systematic approaches is not limited to the macroeconomic picture; respondents have commended systematic strategies as an antidote to the challenges around ESG, particularly bridging the ‘data gap’.
Invesco’s study found around two-thirds of respondents are using systematic strategies to incorporate ESG into their portfolios, and systematic tools have become useful for helping investors decode ESG variables and metrics, which can have a meaningful impact on performance.
Around half of respondents agree that systematic investing can help to apply ESG when data is scarce, and many noted that they were using systematic tools to reconcile the inconsistencies between ratings agencies and develop company scores from raw data.
“There is a low correlation between different ESG ratings agencies, which is of course a much less mature market than credit ratings. So we found investors were turning to systematic models to boost the quality of available data”, said Langer.
Beyond traditional asset classes and factors
Invesco’s study also found a growing consensus that the systematic approach can be applied across a broader range of asset classes than previously thought.
Systematic models are now well-embedded within fixed income and equities, but higher yields, coupled with a shift from quantitative easing, has meant that conventional macroeconomic considerations have returned to the fore in determining returns across various countries and sectors. This has boosted the appeal of systematic strategies for commodities and currencies: while only a quarter currently target commodities this way, (59%) view this as a focal point moving forward.
The new macroeconomic environment has also prompted investors to rethink conventional wisdom about what constitutes a factor.
Notably, four in five respondents now recognise ‘growth’ as a standalone factor, challenging traditional academic views which contended that ‘growth’ was difficult to define precisely. Investors do not see growth as the opposite of value, or vice versa; rather, as distinct and in some cases complementary factors, as evidenced by the rise of nuanced and blended factors like ‘growth at a reasonable price’.
