Technology has been growing rapidly, especially with the development of artificial intelligence and cognitive technologies. These technologies ensure that the machines not only emulate human activities but also human decision-making to a certain extent. Such cognitive technologies have applications wide-ranging from virtual digital assistants, fraud detection, algorithmic trading, drug discovery to automated driving. The increased use cases of AI have gained significant traction across firms, especially in financial services.
Moreover, investment banking firms have deployed RPA for back-office operations that involve structured tasks. However, investment research automation is still at a very nascent stage in investment banks and research firms.
Need for Research Automation
Research functions are facing an urgent need to overhaul their operating models because of multiple reasons. First, there are exponentially higher expectations from what research should deliver. Analysts are more challenged than ever to provide timely and insightful research reports to their clients.
Falling revenues
The teams carrying out investment research are under tremendous pressure considering the reduction in commissions due to the advent of electronic trading and discount brokerage firms. Moreover, in the western markets, the AUM fees earned by investment managers are falling due to the shift from passive to active investing. Moreover, with the introduction of newer regulations like the MiFID II, buy-side firms have to explicitly pay for sell-side research leading to a decoupling of research and trading activities. Lastly, with the increasing number of independent research firms, competition from proprietary data providers has gone up.
All this has led to a wave of consolidation due to falling revenues and increasing costs.
Improving technology
The development of AI has bolstered rapid technological innovation. As per IDC, $52 billion is the estimated worldwide spending in AI. The entire gamut of AI can extract, organize, make sense and report information. AI and cloud computing can potentially disrupt the current operating models for research and investment management. The investment management firms that can extract data from non-traditional locations are more likely to be advantageous due to the competitive edge and increased demand for their services.
Therefore, the firms that do not use these technologies are likelier to lag their competitive peers who adopt AI in research.
Now that we have seen why AI or automation is rising, it is pertinent to see how firms can use AI and automation in the investment research world.
How will automation help in Investment Management World?
The potential use cases of new-age technology in the process of investment automation are classified as follows:
1) Portfolio Management
In portfolio management, AI and machine learning tools track and identify signals on price movements and make use of vast data available. Machine learning tools work on similar principles as existing analytical tools used in systematic investing. The aim is to identify signals from data on which predictions relating to price level or volatility can be made, over various time horizons, to generate higher and uncorrelated returns.
2) Research
The use cases of intelligent automation vary levels of complexity and the value-chain process of investment research.
a. Intelligent Search
The intelligent searches allow analysts to search, navigate, extract, set alerts, analyze fillings with other critical data points. Analysts can now search for many companies within seconds. Deep learning and NLP have made it been possible for machines to understand semantics.
b. Intelligent Answering
Financial assistants use advanced search algorithms and ML to enable analysts to perform quantitative analysis on market data by analyzing relationships amongst events. It can now only take few minutes what earlier took about hours of research.
c. Big data analytics
Big data analytics analyzes SEC filings, conference calls, news releases, and investors’ transcripts. This analysis also identifies investment flags.
d. Knowledge Discovery
A few companies like JP Morgan have also developed an engine that will identify the best positions to be offered to clients based on current financial holdings, historical data, and market conditions. These engines make use of machine learning to suggest predictive recommendations.
e. Intelligent opinion mining
Quantitative market data is leveraged by categorizing and analyzing millions of tweets. This analysis understands the psychology of the investment crowd.
f. Data extraction
Bots can extract financial data from spreadsheets and digital reports in various formats like HTML, XBRL, or PDF – either on an ad-hoc or scheduled basis. It can also populate excel models for cash-flow analysis, revenue and profit projections, equity valuations, and drawing graphs.
g. Maintenance
According to a CRISIL GP&A study, research analysts spend 45% of their time on maintenance research, modifying and enhancing Excel models, or regularly updating data or data sources. A significant part of this work is rule-based, requiring access and processing of structured data, and is one of the first sets of tasks to be automated (rule-based RPA).
h. Report generation
Natural Language Generation (NLG) drafts sections of the report that an analyst can edit to make it ready for publication.
i. Valuation and risk analytics
Cognitive automation can spot and categorize risks (strategic, compliance, financial, operational, and reputational risks) by analyzing text in public domain data and management reports. The risk can be quantified into a risk premium and can offer valuable input to investment decision-making (to help create portfolios that diversify risk and generate alpha returns).
j. Unstructured data extraction
Bots using machine learning (ML) and natural language processing (NLP) can access letters to shareholders, press releases, social media posts, job boards, news, and customer reviews to identify and extract the information relevant to the firm. This capability can be extended using semantic analytics that uses conceptual models such as ontologies, thesauri, fuzzy logic, predictive modeling, and deep learning algorithms to provide the user with more relevant and accurate results.
h. Insights and sentiment analysis
Using NLP and machine learning, bots can help build a knowledge base and generate insights across information sets. For instance, to understand the public sentiment about a new product launched by the firm, bots can analyze customer comments across the web to determine overall sentiment.
Benefits of Investment Research Automation
Technology adoption will help in protecting profit margin
As per CRISIL GR&A, an average research team spends nearly 45% of its time on maintenance research like modeling, maintenance report, and data management. With automation, as per estimates, only 22.5% of the time would be incurred on these activities.
A detailed break-up of how automation will impact the P&L is as below. The bear case refers to the ‘No automation’ scenario; the base case refers to the ‘Investment automation’ scenario.
Keeping 1) savings in employees’ expenses and 2) lower time spent on research in mind, a cost of efficiency of 22.5% is likely to be expected in the investment banking firms with automation.
Considering that the headcount of employees would be saved and lower time would be spent on research, there would be a significant cost of efficiency of 22.5% which is likely to be expected in the investment banking firms on the adoption of automation.
Challenges
While we have seen how automation and technologies will bring a dynamic change in research, it would also be prudent to see some of the challenges that the investment management firms would face:
a. Research Culture
Research firms have been making use of PowerPoint, Word & Excel to perform more structured research activities. Research analysts may resist the adoption of newer intelligent platforms to automate such structured tasks. This reluctance may change may lead to slower adoption.
b. Heavy Investment
Although the companies are aware of the first-mover advantage that the automation in investment research brings in. However, the CFOs would still be reluctant to make this investment considering that the concept has not been tested much. Investment also leads to the rise of obsolescence risk as newer and newer technologies keep on developing with time.
c. Data related
Research firms have to extract data from non-traditional sources in order to differentiate their investment research and come up with differentiated research. Therefore, they have to nurture cross-functional teams to be able to integrate with domain expertise and technological expertise to clear, validate and convert into information that can be fed into artificial intelligence.
d. Fear of job loss
There may also be a threat to their jobs as the adoption of this technology may create jobs for tech-exposed analysts and take the jobs of traditional analysts. This may also lead to reluctance on their part to go ahead with the adoption of automation in investment research.
Investment research is definitely on the cusp of a technological revenue led by AI and cognitive technologies. Early adopters of advanced technology will be able to stay ahead of the game and withstand technology, regulation, and investment industry changes. However, the incumbents will have to take care of the above-mentioned challenges to be able to overcome and implement AI technology-based research automation solutions.
Drop a comment or write us an email at info@pirimidtech.com for any feedback about this article. Do check out our portfolio in building Robo advisory, Large Scale Trading Systems, Algo Trading, Stock Sentiments, Price Trends Forecasting, Backtesting frameworks, Credit Model, Open Banking, etc. on our website. Connect with us to discover how our Fintech expertise can help you build cutting-edge solutions powered by AI/ML.
This is a guest post by Yash Surana, who is a CA and an MBA (Finance) from SPJIMR, Mumbai, and loves writing about finance, strategy, startups & Personal Finance.