Commentaries and insightful analyses on the world of finance, technology and IT.

« April 2017 | Main | June 2017 »

May 31, 2017

5 ways in which artificial intelligence is redefining banking

Artificial Intelligence (AI) has become an oft heard buzzword in the financial services industry. Be it improving customer interactions, analyzing millions of data points in seconds or detecting fraud. As I attended Infosys Confluence 2017 in San Francisco last week, and interacted with several banking and financial services leaders, one thing became stark clear- AI in the banking industry is no longer about pilot projects or enthusiastic experiments. It has established its value as a technology that can significantly improve the banking experience in the near future.

As AI adoption gathers pace, several aspects of banking are set for a makeover. Here are the 5 key areas that I think will be most significantly impacted with the rise of AI:

The banker bots: The bots are everywhere. And they are redefining the way banks are delivering customer experience. Be it Swedebank's Nina or Mizuho Bank's Pepper, virtual assistants have made their way right up to the customers in the banking ecosystem. As more and more banks adopt chat-bots, the technology behind them- Artificial Intelligence- is set to learn, evolve and become more agile and efficient. As a result, we are likely to see even more bots becoming bankers.

Catching the fraudsters: PayPal, which processed $235 billion in payments last year from over 4 billion transactions by more than 170 million customers, uses a deep-learning system based on AI to detect fraud. Not only does the system flag unusual transactions, it also profiles these frauds as a "feature," or a rule that can be applied in real-time to stop purchases that fit this profile. This has helped keep PayPal's fraud rate remarkably low, at 0.32 percent of revenue--a figure far better than the average of 1.32 percent that merchants see, according to a study by LexisNexis. As AI adoption gains traction, more and more banks are likely to utilize AI technologies for detecting and combating fraud.

Smartening the back-office: While the customer facing side of AI technologies is well elucidated, we often miss out on the impact that AI is having in the back-office operations of banks. AI is removing thousands of man-hours from banks' sheets by reviewing loan agreements, identifying repayment patterns and bringing in Robotic Process Automation (RPA) to populate data entry and increase processing speed, especially for structured data. This part, according to us, can be real game changer for banks in saving costs and increasing efficiencies in the near future.

Making data-based, real-time recommendations: Banks have often struggled to make relevant recommendations to their customers about their products and services. The model that traditional banks have been following is creating standard recommendations for a set of customers and flashing them at various point-of-contacts, without any real targeting metrics. AI engines are set to change all that. These self-learning systems are cultivating user data based on their behavioral patterns, banking history and in some cases even their public profiles to make suitable recommendations to users. In recent times, banks have been utilizing these recommendation engines as a key tool to upsell and drive incremental revenue.

Bringing Fintech innovations to customers: A lot of AI innovation is happening in the Fintech space. Companies like Anki and are reducing human intervention in customer interactions. These innovations are propelling banks to integrate specialist third party services from niche start-ups in a very flexible way and bring these services to their customers. New tools are facilitating integration and cognitive agents are making it faster to train and activate a customer facing agent to sell these services to all clients. Thus, collaboration, rather than competition, is booming between banks and Fintechs, thanks to technologies like AI. While banks have the edge of an already established customer base, and their ability to scale offerings quickly, Fintechs are bringing in the innovation factor to the banking party!

As AI adoption accelerates, we see more advanced use cases emerging for banks, such as identifying opportunities from data and actively suggesting intelligent, dynamic policy changes. The speed and scalability of cognitive technologies will result in a slew of growth opportunities for banks which incorporate these approaches into their strategy.

May 19, 2017

Artificial Intelligence in Capital Markets: a tale of opportunities and Fintech collaborations

According to a recently released IDC spending guide, worldwide Cognitive Systems and Artificial Intelligence revenues are forecast to surge past USD 47Billion in 2020. According to another research firm, Opimas Research, in 2017, financial firms alone will spend more than USD 1.5 billion on artificial intelligence (AI) related technologies and by 2021, USD 2.8 billion, representing an increase of a whopping 75%.

Capital Markets, like every other space, is seeing a surge in technological solutions that are coming of age and delivering increased performance- especially in areas of advanced analytics, real time trade-processing platforms and improved regulatory compliance. These technological solutions in a way, have come as a panacea for Capital Market firms, which are grappling with increasing compliance costs and shrinking bottom lines.

Thanks to these macroeconomic factors, Capital Market firms are increasingly looking towards advanced technology solutions like AI to increase employee efficiency and aid faster decision making.

As a technology, AI already has established use cases in areas of client relationship management, trade execution, reconciliations, transaction reporting, tax operations, and several other areas.

We see initial implementations like robotic process automation for reduced manual errors and improving process speeds by automation of repeatable IT tasks. Even as the Capital Market firms explore more advanced use cases for AI, a few areas that we are seeing pilot adoptions include speech recognition, machine learning platforms and virtual agents.

However, the siloed legacy infrastructure that most of the capital market institutions have, coupled with lack of a cohesive, top-driven automation strategy are acting as impediments in the way of effective AI implementation. In such a scenario, many Capital Market firms are taking the route of small, targeted phases of adoption, which can scale in sync with their IT infrastructure.

Another aspect that we find interesting, is the alternate route that these organizations are looking to leverage for bringing in AI and other automation technologies into their ecosystem -- partnerships with Fintech players. Capital Market firms, much like banks, are partnering with Fintech players for things like AI driven post-trade processing platforms and advanced analytics. The most prominent model of these engagements as of now is via accelerators and labs, and the next phase can bring in investments and acquisitions.

As Capital Markets gradually move up the AI value chain, we can expect more Fintech collaborations, advanced use cases for areas like fraud detection and prevention and anti-money laundering activities.

Keep watching this space for the latest in banking technology and trends!

May 8, 2017

ATMs - celebrating a remarkable half-century milestone

Automated teller machines (ATMs) are celebrating their 50th birthday this year. The first ATM was invented by John Shepherd-Barron on 27 June, 1967, and was first installed in Enfield Town branch in north London. Within 50 years, it has evolved from a simple cash dispensing machine to an all-in-one digital banking outpost. Simultaneously, the ATM Industry Association (ATMIA) is also completing 20 years. So, this year marks dual celebrations for both 50 and 20 year milestones!

There were three million ATMs operating worldwide as of 2016. By 2020, the number is expected to reach four million. In fact, the world is going to witness around 37 percent growth in ATM installations between 2014 and 2020. This includes 50 percent growth in ATM cash withdrawals despite the presence of online banking.

It is no surprise that ATMs are the preferred banking channel, allowing customers to directly access their money. During India's demonetization drive, ATMs were the most visited machines. Most of us had to queue in front of them to get the first glimpse of the new 500 and 2,000 rupee notes.

Over the last five decades, ATMs have transformed from cash withdrawal machines to self-service channels, complementing branch, mobile, and Internet banking.

Today's ATMs are providing a wide range of banking services, as follows:

1.Card-less - With growing demand for near-field communication (NFC), personal identification number (PIN) and plastic cards are slowly becoming redundant. Several banks across the globe now support the use of smartphone for cash withdrawal from an ATM using NFC technology. Customers can download banking applications on their smartphones and wave their smartphones to get instant access to their money and other services. Such card-less ATMs can reduce customer data-related fraud which account for 30 percent or US$2 billion losses annually. Many leading banks across the globe have started launching an emergency cash service on card-less ATMs.

2.Multiple functionality - Thanks to rapidly evolving technology and smart innovations, ATMs now perform multiple functions including simple cash dispensation, instant loan provision, live video advise from tellers, among others. Banks are rapidly modernizing and innovating their ATM banking interfaces with simplified touch-screen user interface and advanced cybersecurity protection. The remote ATM monitoring software with advanced analytics are attempting to reduce the downtime of machine and operational costs, while improving utilization and customer satisfaction.

3.Cash recycling - These make cash deposited by customers for cash dispense operations by other customers thereby reducing the cost to banks for cash deliveries to ATMs. Chinese ATM vendor, GRGBanking has estimated that cash-recycling machines can decrease the daily cost of ATM operation by 18 to 25 percent, and could save as much as US$948,000 per 100 ATMs annually. This green solution is widely accepted by Japan -- in fact, 100 percent of new ATM installations in Japan are cash recyclers.

4.Biometric authentication - According to a survey, 40 percent of banking customers feel PIN is not a safe way to withdraw cash at ATMs. Biometrics are emerging as more reliable method of authenticating user's identity. With biometrics, banks can ensure only authorized customers who go through the fingerprint, palm and finger vein patterns, or iris scans can get access to their accounts.

The recent biometrics technique of customer authentication includes using customer's heartbeat rhythms, being tested by Canada's largest bank.

Let us browse the following key innovations:

1.Face-recognition ATMs: Customers can be authenticated based on facial recognition. Some leading ATM manufacturers have started offering this mechanism.

2.Loan-dispensation ATMs: Based on predictive analytics, banks can cross-sell through ATMs. The user interface can be personalized to offer most frequent transactions and next best offers. It can offer pre-approved personal loan to customers too. Customers can get loan approval in a few minutes and start accessing the loan money! Financial institutions (FIs) in Russia and Poland have started offering instant personal loans to customers through ATMs.

3.ATMs with video conferencing capabilities: Customers can have a video conference with teller through these ATMs. The teller can control the machine and its functions, helping customers to complete the banking transaction.

4.ATMs with aerial imaging plate: This is a contactless ATM with holographic floating interface. Customer can see the ATM user interface floating in midair. He can perform hand gestures to make a selection like pressing a button without touching the actual screen.

5.Apple's virtual ATM: Apple has recently filed a patent for iTunes-based ATM network. The apple phone user can launch the application and borrow instant cash from nearby iPhone users. The borrowed money will later be settled through iTunes accounts.

6.Gold-dispensing ATMs: One of top Swiss gold refiners in the world has launched Smart Gold ATM in Singapore. This gold vending ATM machine allows users to convert cash into 24 karat gold coins / bars. Likewise, ATMs can be used to dispense other goods, including lottery tickets, gift cards, movie tickets, postal stamps, and other consumer goods as well.

7.Bitcoin ATMs - Unlike traditional ATMs, these act as bitcoin currency exchange machines. They are connected to bitcoin exchange, allowing users to exchange cash for bitcoins or move money to public key on the blockchain. In future, when the regulations on bitcoin currency are formalized, these would be easily accepted by users worldwide.

ATM 2.0, the future of self-service banking, can be envisioned as the convergence of ATMs and smartphones with expanded interoperability.

ATMIA has rolled out the red carpet to celebrate this memorable milestone and urged billions of ATM users to join this mega event to unveil the true power of ATM!

May 3, 2017

Rethinking loyalty programs in cards and payments

- By Siddhartha Chanda and Souna Uthappa

Customer loyalty programs have always been important for any business and financial services (FS) industry is no exception. To understand why loyalty is so important, we can recollect Pareto's principle which states that 80 percent of the business comes only from 20 percent of customers. In other words, loyal and returning customers make a business workable and profitable.

Coming to the FS industry and in particular the card business, when we talk about loyalty programs, the first thing that comes to the mind is the point per spend model or more commonly called cash back program. This can be substantiated by data where in the US, about 91% of card companies offers the cash back program to the consumers at the rate of one point per dollar spent. Historically, cashback programs have been quite successful in customer retention, as they directly offer monetary benefits. CEB suggests that point-per-spend program directly enhances card usage by two times and increases account retention with half as likely to attrite - thereby increasing the length of the customer relationship.

However, the future of cashback programs looks bleak, as tighter margins could make them outdated by the end of the decade. In addition, regulators are incessantly scrutinizing interchange fees, which govern most funding projects. As a result, interchange fees have already started seeing a downward trajectory -- it has been reduced to 50 basis points (bps) or lesser from 175 bps in major markets.. On the other hand, in markets where interchange fees have been cut, issuers have started using higher fees and annual percentage rates to fund rewards, which is driving down customer satisfaction levels. In the debit card segment, where interchange fees were drastically cut in the wake of Dodd-Frank act, leading issuers have stopped offering reward programs.

At this juncture, FIs are moving away from traditional loyalty programs to a model which focusses more on services and features. In terms of services, some of the banks and card companies are looking beyond rewards and are focusing more on pricing features like trimming cash advance and annual purchase rates for best card holders, offering fee waivers for some of the services, and more. Another American financial services giant is putting its money in a coalition loyalty program, where it has collaborated with some of the top-rated retailers in the region and customers can earn and redeem at any of the retailer's locations.

Digital innovations using emerging technologies like big data, analytics, and machine learning are being used to come up with innovative features for customers. For instance, in one of the banks, big data is being used to get insights about a particular growing community, which can be used to create a customized loyalty program. Analytics helped the bank to get data about education levels, home ownership, affluence, and other factors. Based on the bank's geographic footprint, customers were categorized into segments and then their financial needs were identified. By getting these segmentation insights, banks could strengthen customer relationships and increase customer base within this community by 10 to 15 percent.

In addition, innovative digital apps with high-levels of personalization are helping banks earn loyalty of customers in a digital world. For instance, new apps are being built which automatically calculate the average monthly income of customers, helping them plan their expenses and savings better. During months of high-income, the extra amount gets moved to savings automatically. In another mobile app, unimportant transactions and budgets are not shown to customers, instead customers focus on must know information like the amount that can be spent, upcoming expenses, and automatic transfer of money to savings at set intervals, etc. Such intuitive features and much more will be definitely liked by customers.
Reward programs have become a key differentiator amongst financial institutions to foster a more customer-centric brand image, thereby retaining valuable customers.

The traditional model of cashback has become expensive, which comes from the interchange fees is under scrutiny by regulators. Leading institutions should overcome these challenges by discarding reward points in favour of non-monetary rewards and targeted campaigns through digital innovations.