Art of The Possible: Viability of The Blockchain…
This guide will shed light on the different uses of AI in the Asset Management industry, by analyzing the impact on different aspects such as ESG, data and reporting.
A common theme found in most articles on Artificial Intelligence (AI) in the asset management industry is that active asset managers will benefit from AI technology because it will aid in generating alpha.
This is driven by intense competition with other active managers and competition with passive investment strategies. However, a better approach is to deploy AI throughout both active and passive asset management industries and across each part of the organization. While interest remains very high in AI, leaders face intense headwinds, such as how to present a business case to obtain funding approval regarding such an abstract and complex technology. Challenges include making buy vs. build decisions, calculating ROI, handling complex data-driven implementations, and managing user adoption.
This guide will review the use of AI in different areas including:
Natural Language Processing (NLP) is arguably the easiest technology in the AI world for people to understand and deploy. There is a wide variety of established software that can be deployed. NLP looks for “keywords” across large data sets (e.g., emails, CRM notes, websites, SEC filings, etc.) that can help with identifying the next best action that can assist with decision-making.
CRMs are using NLP, combined with predictive analytics, to determine the “next best action”. This is a game changer where the CRM flips from passively storing information to actively directing the user’s attention to the next best action to take. Leveraging AI technology, the CRM suggests which steps should likely be taken based on a given situation.
Sales teams typically follow the same “milk runs” when generating new business. Over time, their sales trips tend to look similar where they fall into a comfortable and highly repetitive travel pattern visiting the same people again and again regardless of the results. Opportunities are commonly lost as the salesperson stays in their comfort zone and ignores the latest insights gathered by the CRM. This includes not only gaining NLP insights from CRM emails and notes but is also combined with machine learning of data gathered from areas such as marketing, product research, sales reports, and customer service. AI can find missed opportunities in the data, optimizing travel time and costs, so that the salesperson maximizes their time on the road.
Marketing departments are using NLP to look for signals as to what is working and not working well for asset managers. Machine Learning (ML) is deployed to predict which website visitor activities are ripe for follow-ups and to ignore those that are not. There are common patterns that can be observed that signal a prospect’s level of interest in the product or service they are researching. They are using ML to assign these promising leads within the CRM to the right persons for high touch follow-ups or no touch automated interactions.
Asset management firms cannot afford to overlook the value and implementation of AI technology in all business units to effectively run & sustain operations in the evolving social, economic, and regulatory environments we are living in today. A study conducted by Jobscan found that as of 2022, at least 98% of Fortune 500 companies use AI technology for talent acquisition purposes. The GVR Group estimates the global HR software market size will exceed $10 billion by the end of 2022.
Due to social, economic, and regulatory changes, business leaders have invested more resources in HR Management as a result of the adoption of high-performance technologies that only a portion of today’s workforce is capable of implementing and managing. AI technology has helped reduce costs, improve strategies in talent acquisition & employee retention, transformed the workplace into an agile environment, and offers career development opportunities to a larger employee demographic. It has also helped companies reshape their culture which in turn helps both employees and the company fulfil operations. Rebuilding companies’ operative models leveraging AI technology has become necessary to remain competitive, align company cultures with its desired workforce, and remain competitive.
Like any business going through a digital transformation, the same applies to Asset Management firms. Unlike traditional businesses, asset managers benefit from AI technology in a more complex and different way than traditional organizations due to their ability to influence, reshape global markets, and impact people’s sentiments around how they invest. AI technology’s ability to predict capital market changes, investing behaviors, and asset future performance has influenced asset managers’ leadership into investing internally in advanced technologies and in the restructuring of internal human capital management strategies to attract skilled employees who can implement, oversee, and manage these technologies.
Leading businesses from cross-business sectors are transforming their operating models where technology is at the forefront of all initiatives. The long-term success of asset management firms will depend on the success of their HR Management initiatives, Technology Investments, IT Security, Governance, Risk Controls & Compliance. An organization’s effective human capital management can predict its future performance. As technology reduces the need for employing as many individuals as previously needed for specific business functions, people with more qualifications are needed to effectively implement and manage new technologies. Implementing AI technologies like NLP require the restructuring of asset managers’ system & operative models, which redefine the meaning and weight of the technology.
AI and NLP support talent acquisition goals in countless ways, including the efficient screening of large volumes of job applications to an accurate and automatic profile analysis of a candidate’s personality and employment skills. Also, by assessing internal communications, sentiment analysis is being applied to reduce employee turnover and prevent employers from being understaffed. It’s worth noting that NLP maximizes its value by detecting potential employee lawsuits against the employer and potential employee conflicts.
Before experiencing the digitalization era, financiers didn’t know the value of data and how far technology could go using it. Gen Z, for instance, prioritizes mental stimulation through learning, knows how to multitask, and understands that technology comes naturally to them. Having all new generations understand the value of data in today’s fast-evolving world, indicates that the recruitment of talent at an Asset Management firm could define its future success.
Using NLP/G engines, asset management firms can automatically create client reporting, investor performance documentation, and industry-specific analyses, deliver account statements, provide other reports with quality insights faster and save money in the process too. High-performing technology closes compliance gaps and promotes alignment within an organization and with evolving regulatory requirements.
Before AI, asset management firms generated reports for regulators and investors using manual processes, which ended up reflecting inaccurate information that in turn breached regulatory requirements, leading to penalties and reputational damage.
Asset managers face significant challenges in generating superior returns due to the large amount of widely available information that is impossible to digest within a timeframe that provides an edge over the competition. Additionally, many quant funds have developed advanced statistical models that seek to exploit any opportunity arising from market anomalies, valuation outliers, or new information that is made public. In this context, AI algorithms can assist fund managers in providing new stock-picking ideas by analyzing data faster and quickly structuring qualitative information.
In the last few years, AI applied to stock picking has been increasingly researched both by asset managers and finance academics. Several papers have demonstrated the capacity of Al algorithms to generate alpha. Different techniques have been studied, but Random Forests (RF), Support Vector Machines (SVM), and Neural Networks (NN) are among the most widely adopted. RF and SVM seem to be well adapted to analyzing vast amounts of quantitative data that can then be translated into stock picks, while NN can be used to quickly highlight key points in verbal or textual information such as company conference calls.
Today, Al is mostly adopted by quantitative funds or asset managers with strong quantitative expertise, even though their core investment process is not purely quant. They generally rely on Al for assistance rather than full delegation of the investment strategy (something that is not that infrequent with other quant strategies such as factor investing). At the same time, active asset managers, that do not rely on quantitative processes for their core strategy, seem to shy away from the advantages of Al.
Sia Partners believes that Al can be adopted by asset managers to a larger extent than it is today. Al techniques adapted to financial data and textual information (news, conference calls, etc.) could help fund managers absorb more information at a lower cost than when relying on broker research.
Investors are increasingly demanding top financial firms disclose their environmental, social, and governance (ESG) practices. To make more informed decisions, utilizing NLP to reduce the complexities of the language has been beneficial for investors. According to a Harvard Law School Forum on Governance, ESG practices positively correspond with profitability, while negatively correspond with volatility. Furthermore, firms with a high ESG score tend to be more diversified than firms that do not prioritize ESG. These findings are a key indicator that when firms can prioritize ESG, they can support the goals of their clients by maximizing profitability.
NLP is applied using text classification and sentiment analysis. This can be particularly helpful when parsing through lengthy articles and resources by pulling out keywords and phrases that indicate a certain sentiment. Below are areas across ESG where AI solutions can be utilized.
|Pollution||Diversity & Inclusion||Lobbying|
|Waste Management||Fair Labor Practices||Corruption|
|Carbon Emissions||Internal Policies||C-Suite Diversity|
As the financial industry deals with scaling and consistency problems resulting from an ever-growing and more demanding customer base, data that can be easily accessed and interpreted are becoming less of a benefit and more of a requirement. A clear example of this trend has been the move from legacy systems to automated, cloud-based solutions. This revolution, however, does not come without its challenges. The most notable among them is addressing dirty data. “Dirty data” often refers to removed data, corrupted data, or the replication/reformatting of data across different company systems, rendering it unusable, inconsistent, or slow to access.
What makes cleaning data so challenging lies largely in how differently the data is stored/labeled across the company. It’s no wonder data management can feel overwhelming when having to clean this for a company with millions or even billions of data points that may be continually growing.
In many cases, the data itself is not dirty but is instead inconsistent. For example, a client’s legal first name for tax purposes may be Joseph, for their client statements they prefer them to be addressed to Joe, but on the phone, they prefer to be called Joey by their account rep. Data management who is looking to reconcile, replicate, or reformat data then not only needs to make these connections but identify every instance in which this is done with a client. To make matters worse, finding errors in this process is almost as hard as the reconciliation itself since it requires a person to look at data management’s decisions and notice subtle differences between the cleaned data and the correct source data.
Also, some asset management firms say. “I would rather have some dirty data than risk losing precious information or paying to get it cleaned”. This response may make sense to some managers in a vacuum. After all, data cleaning is expensive and labor-intensive. However, most companies who don’t clean their data, lose, on average, 15% to 20% of their revenue from this according to MIT Sloan.
Does this mean all financial firms are doomed to either deal with the risk of inconsistent data or pay the dirty data tax? not necessarily. Many of these problems have been addressed, all that is left is to apply the solutions.
Data-oriented programming languages (i.e., SQL and Python) can be used to develop models for future cleanings by training them on only a handful of mappings done by hand. Think of this as reconciling phone numbers across the business. One group could use the format (xxx) xxx-xxxx, while another could use the format xxx.xxx.xxxx, and the end product needs to look like xxx-xxx-xxxx. In the case of removed or corrupted data, similar methods of training can teach a computer to follow pre-made rules like replacing empty fields or removing data before a certain date. A model can be trained to recognize the punctuation and correctly format the data as the human eye would. Similarly, User Acceptance Testing is much easier as the model can be told what the result is supposed to look and act like, which can be compared against what it does look/act like.
For more complicated cleaning procedures, definitions for each term could be interpreted with natural language processing mentioned earlier in this paper. Natural language processing is already being used for mapping electronic health records and could easily be applied to data with clear, accessible definitions, like financial data.
Finally, since the rules developed during an initial cleaning can be applied to new data, this process only needs to happen once.
As technology has allowed data volumes to grow exponentially so has the complexity of analyzing and managing risks faced by financial institutions and asset managers. While mathematical models and conventional rule-based algorithms are undoubtedly helpful in monitoring these risks, they are generally computationally heavy and relatively static. In this regard, AI can bring greater efficiency and even new ways to manage risks that only humans could once effectively monitor.
Quantitative models have been widely adopted for monitoring both market and credit risks. Generally, these models try to assess the likelihood of heavy losses, either because of extreme market movements or counterparty defaults. However, even state-of-art models may struggle with the nonlinear nature of economic shocks and their impact on markets and counterparties. In this regard, AI models underpinned by SVM or RF can help asset managers complement their analysis with new perspectives based on non-linear models that will try to predict both non-normal market returns and counterparty credit risks.
Asset managers not only face exogenous risks, but they must also make sure that their internal compliance procedures are followed and are up to date. Regulatory texts have been growing in complexity and volume. To be aware of the latest developments, asset managers can use AI to scan new regulations, create alerts, and draft summaries for key stakeholders in the firm. Likewise, asset managers can use AI to scan transactions made by their traders and fund managers to detect suspicious activities and complement the approaches that have already been implemented, like eliminating false positives. As an example, scholars from the Universities of Alberta and Chile successfully tested several algorithms to detect market manipulations (identified as SEC litigations containing references to “manipulation”, “marking the close”, “9(a)” and “10(b)”, sections of the Securities Exchange Act). See the chart below for more details. Similarly, NLP can help investment firms reduce the risk of misconduct by scanning both oral and written communications and by detecting unusual or encrypted discussions, rather than fishing for inappropriate words.
Human decisions are heavily influenced by emotions, experiences, psychology, the five senses, and an evolving set of norms. Until computers are buyers of goods and services, humans will make buying decisions based on these many influences. For example, just looking at the senses, can a robot pick fresh fruit by vision alone without the sense of smell, taste, or touch? Not yet at least.
While AI may help avoid repeating past mistakes, how well can it predict the future? Black swan events are where the most damage is done in the marketplace and are not easily foreseen. Most of these events are previously known but are statistically insignificant until a trigger event occurs.
Yet, the ultimate value proposition that AI offers is that it can do what no person can do and that is looking at millions, if not billions, of pieces of data and revealing actionable patterns. Just like the new James Webb telescope is revealing mysteries of the universe, AI acts similarly in that it can reveal patterns in a whole universe of unseen data much larger than any human can comprehend and does so completely devoid of emotion or bias. AI provides valuable insights to assist humans with daily functions whether at home or work.
Sia Partners is a next-generation management consulting firm. We offer a unique blend of AI and design capabilities, augmenting traditional consulting to deliver superior value to our clients. Leveraging AI, Consulting 4.0, is the augmentation of Sia Partners’ business advisory services with automation and artificial intelligence.
Heka is the ecosystem of AI solutions developed by Sia Partners. These solutions inherited from years of development, provide our customers with the industrial tools to address recurring issues to which data science provides a convincing response.
At the heart of the Consulting 4.0 strategy, Heka is Sia Partners' industrial artificial intelligence platform. Consisting of all the technical building blocks required to carry out Data Science projects, it allows the rapid deployment of Machine Learning or Deep Learning projects of all types: voice or image recognition, Natural Language Processing, etc.
Heka guarantees the time to market from idea to Proof of Concept (PoC) and from PoC to industrialization, as well as the scalability and robustness of our AI solutions.
Sia Partners offers AI accelerators as our customers' challenges are our source of inspiration and advanced AI technologies are our driving force. By combining these two elements, we are constantly inventing new AI components: a catalog of data sets, algorithms, and PoC that stimulate ideation and accelerate our AI projects.
Our Platform as a Service is made up of all the technical building blocks necessary to carry out AI projects, our Heka platform allows us to structure the way our data scientists work. It is the guarantee of solid code and scalable solutions to support our customers' projects over time.
Sia Partner’s Wealth and Asset Management practice provides industry-leading solutions to solve the most complex problems. Our consultants cover a wide array of expertise across the front, middle and back offices. We incorporate ESG, CSR, and our Greenly product to track and report our client’s carbon footprint.