Steps Ahead

AI-driven data analytics: the foundation for exceptional multi-asset client service

By: Matthew Hodgson, CEO & Founder

Investment banking is changing. Driven by relentless regulation, cost pressures, electronification and automation, sales and trading desks have undergone a revolution in recent years. Whereas previously, sales and traders would have been experts in specific asset classes, the modern trading desk has evolved to become more customer focused and truly multi-asset in capability and offering.

AI-driven multi-asset insights dashboard

As a result, banks’ sales and trading staff are forced to do more with less. But how can a single person truly get under the skin of every market they cover and deliver the personalised level of service their customers expect?

The answer lies in cutting edge AI-driven data analytics – and adoption is picking up pace across the industry. A recent survey from The Economist Intelligence Unit found that 77% of bankers believe that the ability to unlock the value of AI will be the difference between the success or failure of banks, while in a 2021 McKinsey survey, 56% of respondents reported AI usage in at least one function of their organizations.

Achieving a holistic view

Today, the root cause of the problem is that the vast majority of salespeople and traders have to deal with an environment that has been built up with disparate systems and manual spreadsheets. Tasks, as a result, are spread not only across asset classes but also functions such as risk and liquidity analysis. As multi-asset trading becomes more and more prevalent, this is no longer an acceptable solution to wrestle with the enormous opportunities that all firms are pursuing.

Forward-thinking banks are now looking to AI-driven data analytics technology that can standardise all transactions from every asset class across the global organisation under one data model, and in so doing unlocking actionable insights from it.

All analytics programmes require fundamental data foundations to be effective – especially when these programmes are multi-asset in nature. The very fabric of a firm’s data must be integrated and used in a way that is frictionless, for it to be truly valuable. As such, data sets must be harmonised and standardised across asset classes into one consistent format and cover as wide a set of relevant transaction and market data as possible. Furthermore, each data entry should be as comprehensive as possible, with all relevant fields captured for every transaction.

With the right technology in place, a bank can achieve a single, holistic view of transaction data across all its global locations, across all trading channels and across all asset classes, combined with the relevant market data. The real benefits of ML & AI can then be fully realised.

Applying AI

With successful aggregation and standardisation in place, AI can then be applied to the multi-asset data set enabling actions such as:

Automation of reporting

  • For example, generating a market colour report that, based on transaction data, tells a bank which of its clients is buying or selling across the asset classes it covers at any given time.
  • The time saving benefits of this are enormous for the employees on the trading floor who are tasked with writing these reports daily, freeing them up to focus on their core roles of driving value generation.
  • Mosaic’s natural language generation (NLG) is applied and after trends have been identified, summarises them in natural and appropriate language in multi-asset reports.

NLG – Natural language narration of data sets coupled with the visualisation

AI-driven automated daily rates report

Anticipation of client & franchise flow

  • In the rates and credit markets, many asset managers and hedge funds tend to re-balance their portfolios at month end.
  • By leveraging seasonality signaling, salespeople and traders are empowered to act on this information and show their clients the most suitable inventory at the right time and work in tandem with their clients to provide a hyper-personalised service to them.

Seasonality using pattern recognition techniques to identify seasonality of any of the following:

  • Month end dates
  • Quarter end dates
  • IMM dates
  • Particular days of the week

I see a lot of value in being proactively alerted to client seasonality as it gets all my salespeople in front of the opportunity.

Head of FX Sales
Tier 3 Bank

Liquidity prediction and recommendations

I am a Salesperson, across all the clients I cover – who should I be calling for today’s Axes?

This probabilistic AI model gives targeted recommendations of which clients to call given a particular Axe with a better understanding of how to re-cycle risk building for better inventory management.

These models understand which clients are most likely to trade a specific instrument, based on years of learning. A salesperson or trader can gain this intelligence at the click of a button instead of having to pick up the phone or socialize this with their colleagues.

In the figure below, the salesperson simply selects the name of the client they cover and are then presented with ‘Matching Axes’ – a list of the firms axes with the highest probability of the client having interest.

Identifying unusual client behaviour

  • Providing a highly tailored experience for a client base necessitates understanding of micro changes in their behaviour in order to provide personalized advice.
  • This understanding revolves around unusually large or small trades, unusual patterns or significant changes in customer behaviour be it frequency or previously unseen activity around tenor behaviour or when clients trade products for the first time.

Examples of anomaly detection

Anticipating client defection

  • AI can act as an early warning system that a client is migrating their business to a competitor. For example, from the 500 clients it covers, it is beginning to see decreased business from 12 of them that contribute around 5% of revenue, so it needs to take preventative action to avoid defection.

Client defection scenario

The defection alert prompted a conversation that unearthed that the client was hedging certain deals in an unexpected way. We made internal changes to accommodate pricing for these deals. This resulted in us retaining the client and winning 25% of deals.

Head of FX Sales
Japanese Bank

Stepping up to the challenge

The old idea that single systems specialising in specific asset classes are the only way to get the depth of market access or access to the right tools and intelligence no longer stands true.

Leveraging innovative platforms like Mosaic Smart Data’s MSX , firms can now deliver exceptional customer service across asset classes and maintain their competitive edge. In its Building the AI Bank of The Future report, McKinsey states that banks that harness the power of AI today will increase profitability, drive personalisation at scale, enable an omnichannel experience and benefit from speed and innovation.

As the world of capital markets becomes increasingly multi-asset, the time to position your firm for the future is now.