Case studies: artificial intelligence in the financial services industry

Multi-asset investing has gained traction as investors seek other diversified solutions to meet their specific needs. HSBC has introduced a new approach to multi-asset investing that combines AI and big data. 

The AI-powered multi-asset index (AiMAX) family helps investors construct a diversified growth portfolio by using modern outlets, such as social media posts, and traditional financial sources, like financial statements. It uses historical data and continuously learns from new information to build portfolios. It also includes a diversified portfolio of exchange-traded funds (ETFs) that can generate consistent returns and rebalance weekly. This rules-based multi-asset strategy generates investment insights from data using IBM Watson. It aims to improve capital market assumption accuracy by using AI and big data to offer customers a better asset allocation strategy.

AiMAX invests across 15 asset classes, spanning five investable asset types, including developed bonds, inflation assets emerging markets, developed equities, cash index, and real assets. It rebalances the index portfolio every week in three steps. Step one uses AI to forecast the future returns of each asset class by analysing current and historical data points. In step two, portfolio testing is used to perform combination tests for all 15 investable asset classes based on the principles of modern portfolio theory, thus generating each portfolio’s forecasted return, volatility, and correlation. Finally, step three uses the solution to help investors select the portfolio with the highest forecasted return, and a portion of the index is rebalanced to reflect the new portfolio.

HSBC collaborated with start-up EquBot, using a data analytics processor from IBM Watson Discovery to develop the tool. HSBC claims that AiMAX would have performed above the S&P 500 Price Return Index with significantly less volatility and more diversification based on simulated and historical returns.

The Covid-19 pandemic altered daily routines globally. Lockdowns, travel restrictions, and social distancing encouraged people to stay at home and use online services, venturing out only if necessary. This also led financial institutions to find ways to reach customers remotely via mobile and internet channels, fulfilling tasks that would otherwise be conducted in person.

In March 2020, RBC launched an AI-powered mobile app that allowed customers to verify their identity digitally when opening a bank account. It includes digital government identity verification capabilities and is a part of RBC’s omnichannel technology stack. The app is available for customers at the bank’s branches and remotely through mobile apps and the bank’s website. The verification procedures comply with AML, know your client (KYC), and privacy laws in Canada.

The RBC app uses AI to compare a customer’s government ID with driver's licenses and passports data. It relies on nearfield communications (NFC) to verify e-passports. It contains templates for every form of government ID, and the AI system continuously scans them for defects in the format. The app uses ML that learns from prior scans to detect subtle differences that could indicate a counterfeit document. Customers scan their IDs using the app, which enters their personal information into their account profiles and the computers of bank advisers, creating a connection between the branch and the app.

For driver’s licenses and other government IDs, the app scanner uses the camera to upload information on the bank advisor’s screen in real-time. For e-passports, customers tap the chip on the phone’s card reader to extract the necessary information. The account opening process begins with the bank advisers sending an encrypted email to recognise customers, who then click on their IDs in the mobile app to initiate the verification process.

While completing the verification process, customers take a selfie that offers a 3D model of their face. The app compares it with the data from the e-passport or the photo scanned off the license to prevent fraud. The entire process occurs remotely, and opening a bank account takes less time compared to conventional sit-down processes at the bank. Moreover, it allows bank advisers to use their time for other activities that drive customer engagement. The app is available in almost 100 locations across Ontario and Quebec.

Conventional digital advisory services offer standard advice to all customers, which often does not appeal to everyone, since each client has various sources of income and spending habits. DBS Bank has employed AI and a data analytics-driven solution to develop intelligent banking features to help customers improve savings with hyper-personalised insights and recommendations directly from their mobile banking app.

The solution provides advice based on each customer’s saving accounts, investment portfolios, and spending habits to help them grow their wealth and manage finances. The solution also provides services such as flagging uncommon transactions in a buyer’s accounts and showing if a customer can automate bill payments based on their account balance. On the wealth management side, the technology can also advise stock preferences to help clients purchase company shares in a specific sector based on their portfolio. It can also notify them about suitable foreign exchange charges in case they deal with international money transfers.

Additionally, in 2021, DBS introduced an AI-powered digital investment advisory feature to its NAV Planner, a digital advisory tool on the DBS app. This feature assesses individuals' investment risk profiles and gives tailored suggestions. DBS claims to have created over 100 AI and ML algorithms that analyse customer data points to generate nudges that offer personalised product recommendations and celebrate customers' milestones.