Artificial Intelligence (AL), a branch of computer science, enables machines to imitate intelligent human behaviour towards solving intricate problems. It is called artificial intelligence, as human intelligence is considered real intelligence. Machine Learning (ML), a powerful subset of AI, helps devices learn and improve from past experiences without the need for explicit programming. ML processes large amounts of data to extract actionable information that helps create a competitive edge, and it has proved to be a game-changer for the business world, steering aggressive growth, innovation and sustainability agendas across different verticals.

Given the blurred definitions of AI and ML in practice, it is not easy to accurately measure their progression in quantifiable terms. However, various surveys and studies from time to time have given us a fair idea of their rapid

in business and industry.

Machine Learning & Business Intelligence
The increasing volume and complexity of business data drives the commercial adoption of ML in business analytics, which has progressed by leaps and bounds from the glory days of the conventional Extract, Transform, Load (ETL) tools. ML has greatly enhanced business intelligence by processing and analysing large, complex datasets to identify patterns that otherwise stay undetected.

Simply put, ML-enabled pattern recognition is a ‘machine way’ of identifying data regularity – which is about its stability, consistency and symmetry- and classifying events based on input data. Thus, more than merely monitoring behaviour, analysing users’ actions reveals actionable insights about their behaviour which is invariably complex and varied.

The importance of predictive ML can be gauged from its success in different sectors. Today, many retailers are creating personalised product recommendations in line with buying patterns. Healthcare insurance providers are developing information-rich consumer profiles. Digital media houses are predicting the success of entertainment shows to make intelligent airing decisions. Food tech companies are personalising every customer’s landing page in line with their granular food preferences, which prompts the buyer towards a ‘buy’ based on the sheer delight of finding the desired recipe, ingredients, and cuisine style.

Machine Learning Applications in Financial services
ML opportunities abound in the financial space. ML can help provide banking customers, as also stock market investors, with curated pages highlighting the offerings suited to their needs or with the sectors and stocks of their choice; more importantly, it can guide them towards better banking and investing decisions through prudent choices in the light of prevalent market trends, realities and need of the customer.

It is pertinent to note the potent dual use of ML in trading and investing. It can be used for analysing stocks, as also for analysing investing behaviour of investors.

For stock analysis, AI ensures tremendous value addition: it collates clean data and crunches and classifies it to draw intelligent inferences through pattern recognition. In stock trading, it also helps minimise the post-execution impact on stock prices by splitting orders into smaller chunks, besides identifying arbitrage opportunities across diverse markets.

This snapshot of popular AI trading and investing tools gives an idea of a burgeoning market in the making.

Kavout Brainchild of ex-Google executives, has developed “K Score” – an AI-enhanced stock rating system using pattern recognition technology and a price forecasting engine.
Auquan Platform for asset managers to dig non-obvious connections, news, lookahead bias affecting investment decisions.
EquBot AI platform integrated with IBM Watson enabling faster data crunching, AI-made portfolios and sentiment analysis.
Blackbox Stocks Comes with a pre-market scanner to spot most active stocks and their degree of volatility.
Neurensic Now part of Trading Technologies, enables a continuous assessment of the compliance risk associated with complex trading behaviours.
Sigmoidal Unearths actionable patterns between securities and capital market expectations.

For investor behaviour analysis, AI goes beyond mere personalization to forecast how the said behaviour will influence business decisions. This churn brings to light invaluable information like, for instance, the actionable segmentation of customers into different groups (and thus targeted for different product offerings) based on their spending and saving patterns.

Today, many tech firms are studying tons of unstructured data sets and mining invaluable insights and patterns to evaluate the reliability of company guidance disclosures, the correlation between projections and performance, and the likelihood of growth upswings and downfalls. In addition, digital assistant providers are enabling guided conversations that simulate the “why” and “how” questions that a knowledgeable financial advisor is adept at asking and answering.

AI and ML in India…and the road ahead
According to a 2020 study by Analytics India and AnalytixLabs, 16% of the analytics revenues across all enterprises are attributed to advanced analytics, predictive modelling, and data science. Although this share is impressive, the fact remains that the AI market space in India is still at a nascent stage.

Undoubtedly, there is humungous scope for the use of AI in the future, beyond Behavioural Analytics, Robo Advisory, Stock Scoring, and Portfolio Diagnostics. Prospective areas include both stakeholder-specific (such as customer onboarding, self-service offerings, vendor management) and system-specific (such as risk management, anti-money laundering, fraud detection.)

As more and more users join the AI bandwagon, data will only grow in volumes, velocity, veracity, variety and value. The chart below (courtesy: German market data platform Statista) gives a fair idea of the big data revolution in the making, which will soar higher on the wings of AI and ML.

Yes Sec-Big data chart

AI and ML in investing: about human empowerment, not displacement

The capital market is enormously complex. Given that market data, feeds are the principal inputs for an algorithm; machines can miss scores of elusive opportunities that only the human brain is adept at spotting. AI has a serious limitation of not offering a long-term strategy purely based on the status quo or past information. Humans can enhance the probabilistic AI results, validating them in the light of intuition and discretion, like how a doctor studies an ultrasound image.

AI and ML are key enablers for enhancing trading and investment decisions. Competent financial advisors make intelligent use of the actionable AI and ML input which provides a comprehensive analysis of stocks/sectors and investor behaviour to create long-term wealth for different clients. This ultimate goal is achieved only through disciplined and diversified investments, in line with respective income profiles, risk appetites, available market opportunities, and applicable incentives like tax deductions and exemptions.

This article is authored by Gopinath Natarajan, Head Investments & Products, Yes Securities. The views are his own)

AI in investing is about human empowerment, not displacement

By ariox