Did you miss a session at the Data Summit? Watch On-Demand Here.
There is not a single industry that has been left untouched by the transformative impact of artificial intelligence technology in the last decade — financial services is no exception. The financial sector is well-known for seeking every possible edge to maximize its profits — thus, using machine learning and artificial intelligence was a no-brainer.
A plethora of use cases is leveraging the power of artificial intelligence (AI) — from fraud detection, risk assessment, improving customer satisfaction, increasing accounting and transactional automation to algorithmic trading.
What was traditionally a people-heavy industry with loads of analysts and money managers, financial services has slowly transformed into a lean technology-heavy behemoth. As a result, we are looking at augmented human intelligence using AI, resulting in greater efficiency, reduced costs for banking institutions and new offerings to consumers.
According to the OECD report on AI, ML, and Big Data in finance, global spending on AI is forecast to double for 2020-24, growing from $50 billion to more than $110 billion over four years.
Through its extensive financial inclusion efforts over the past decade and increased digitization of the economy, India is sitting on incredibly rich data. In the coming years, this data will be used to glean insights to provide targeted services and products to consumers. In addition, the growing number of fintech companies in India is ensuring the financial inclusion of every Indian to have access to capital and services at more incredible speed and convenience.
In capital markets, AI is taking over an increasingly more significant chunk of trade executions. What started as a trend following in the 1980s, traders and hedge funds soon utilized highly sophisticated algorithms and rule engines to execute trades — popularly known as algorithmic trading. Today, AI-driven algorithmic trading is being used to conceive trade ideas and trade executions. The high-frequency trading industry relies heavily on automated trade executions provided by ML models using techniques like mean reversion, anomaly detection, and various deep learning techniques to capture complex underlying patterns.
According to a 2020 JPMorgan study, over 60% of trades over $10 million were executed using algorithms. Moreover, the algorithmic trading market is expected to grow by $4 billion by 2024, bringing the total volume to $19 billion.
With the vast amount of opportunities for application in finance, AI also faces several challenges.
- AI is often perceived as a black box because users tend not to understand or explain why an AI model suggests or predicts a particular outcome. This challenge opens up the need for regulatory and governance frameworks for AI adopters to ensure no bias or discrimination is trained into a model. For example, imagine an AI biasing against a particular demographic of the population based on their gender. Data bias resulting in unfair discrimination will be antithetical to the financial inclusion goal of banks and institutions. Hence, explainable AI is gaining greater prominence to ensure human oversight and judgment.
- AI models continuously learn and refine their predictions on new data. However, it suffers from tail risk from black swan events like COVID-19, where the learnings of the ML models drift because of one-time skewed data. Such unforeseen circumstances not being captured by data undermine ML models’ predictive accuracy and degrade performance. Therefore, for all its technological and computing prowess, AI still requires a human-in-the-loop for many use cases. These are areas of active research for the AI community to solve over the coming decade.
Banks and financial institutions have continuously adopted technology to stay relevant and offer improved services to their customers. In the AI age, finance and banking will have become AI-first rather than use AI technology on their periphery. With the correct implementation, they can improve human decision-making and reduce risk, unlocking a trillion-dollar opportunity for this industry.
Sandeep Sudarshan is a senior manager in Ericsson R&D — Global AI Accelerator (GAIA).
Welcome to the VentureBeat community!
DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.
If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.
You might even consider contributing an article of your own!