Artificial Intelligence: A Glimpse into the Financial Industry

In 2012, when a team of the University of Toronto perfectioned the ImageNet project by getting computers to associate words with images just as accurately as humans, little did we know that eight years later the AI and Robotics combined market value would be expected to reach $153 billion.

Right before cutting to the chase, it is important to note that when talking about AI solutions, the financial world should be considered as a stand-alone system in which companies may leverage machine learning technologies and make money in the dark. For instance, while we might know what an e-commerce giant such Amazon does to profile its users and suggest products, financial companies generally tend to take a more protective approach to ensure their competitive edge by recruiting from big tech companies such Google, Microsoft and IBM in order to build huge in-house AI clusters and leverage these kinds of technologies for trading and investing.

The increasing trend of AI is a complex and transversal phenomenon to the financial industry, with a wide variety of application and relevant implications.

Firstly, within an industry pervaded by unstructured data sets, AI solutions are taking the lead in translating and squeezing out lots of insights through natural language processing. An example of this application come from Alex Lu, co-founder of Kavout who said: “We are building something called sentiment analysis…our algorithms leverage all the data relative to the investors, traders, news, blogs and come up with a score that represent what people think about specific stocks”. In 2014, Goldman Sachs invested $15 million in Kensho, a leading American provider of next-generation analytics natural-based processing that last year was sold for $550 million to S&P Global, representing the largest price on an AI company to date.

Moreover, virtual financial assistants can be developed by integrating sentiment analysis with predictive analysis, based on learning which variables could trigger specific economic events, such as the 2008 financial crisis: depending on customer risk appetite, the machine can provide advice on whether to buy, hold or sell stocks.

At a corporate level, Robot-Advisory represents a further enhancement of virtual assistants. By filtering human activity through detailed online questionnaires regarding risk acceptance, financial goals and customer investment time ranges, the Bionic-Advisor comes up with a specific asset allocation. Once a financial portfolio has been created, the machine automatically monitors and manages investments, rebalancing weights, executing trades, calculating tax-loss harvesting and so on. However, Bionic Advisory has still to come up with a more adaptive asset allocator system by taking into consideration people’s behaviour. According to a recent study, while most Robot-Advisors assumes you stick to the strategy for 30 or 35 years, most people actually change their financial planning every 3 to 5 years.

Focusing on transaction services, AI has shown to be particularly effective in supporting Fraud Detection by analyzing behavioral spending, tracking down historical transactions and eventually highlighting any strange behavior of both online and in-person banking. Through Citi Ventures for instance, Citibank has recently renewed its partnership agreement with Feedzai, a leading global data enterprise powered by advanced real-time machine learning which allows the company to offer an adaptive model and better minimize transaction risk in the financial industry. The company manages $5 billion worth of transactions daily and protects 10 of the largest 25 global banks against transaction fraud. In addition, it has raised $82 million in funding from investors including Data Collective DCVC, Sapphire Ventures and Oak HC/FT.

Over the last few years, the expansion of AI applied to finance has been particularly noticeable in Personal Financial Management. A growing number of start-ups, such as Wallet, use algorithms that collect web footprint data and analyses consumer spending behaviour to deliver a graph of consumer’s behaviour and well as spending advice. However, this poses several concerns, as consumers should share private data on consumer habits with service providers. This explains the growing public concern and opposition in various jurisdictions, due to consumer privacy and data protection laws.

Overall, big banks and asset managers seem to be mainly focused on improving adaptive processes and developing their own Robot-platforms. For instance, FidelityGo, Merrill Edge or Vanguard Personal Adviser Services are few of the most prominent own Robot-Advisors while JP Morgan in 2016 spent a total amount of $9,5 billion in technology, of which $600 million were dedicated for fintech solutions and $3 billion “dedicated to new initiatives”. Bank of America itself invested around $3 billion in 2016, a strategy founded on technology and AI that brought the bank to reach the second most profitable year in the company history.

Among the pros that could be listed, undoubtedly the reduction of cost, customer experience and rapidity of performing qualitative tasks are the main reasons why big institutions have embraced digitalization and are collaborating mainly with fintech companies, which, by the look of things, is set to play an increasing role in the financial industry.

Gianluca Biancardi and Giorgio Guarrella

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s