How artificial intelligence is changing financial planning

Developments in technology are already driving the delivery of advice and will continue to do so, with significant innovation taking place in the automated advice sector.

Artificial intelligence (AI) is now enabling ground-breaking solutions to be developed that will further liberate the human adviser to focus on the client and enable more efficient, consistent and safe advice to be delivered in minutes rather than days or weeks.

We have developed our own AI capability, Turo, which learns the advice policy of an adviser by identifying patterns and by rapidly configuring the platform so that each brand has their own way of delivering advice.

It currently works best when handling binary decisions, such as whether a client should transfer their defined benefit (DB) pension, take equity release or go into drawdown. It will also flag up cases which fall into a grey area and need to be looked at by an adviser.  However, as the system learns more, review cases such as these become fewer.

What about open banking?

Alongside the progress with AI, open banking offers far-reaching opportunities for the advice sector in supporting both clients and advisers themselves.

Open banking represents an evolution from a ‘closed’ model, where each financial institution retains and controls the information it collects about its customers, to an ‘open’ model, where customer data is shared safely.

It has the potential to prompt a whole new ecosystem that shakes up the competitive landscape, with the creation of new products and services based on data. This will have a knock-on impact for the advice profession, which will need to find and define its place in this new environment.

Open banking also comes with its own set of jargon - application programming interfaces (APIs) and third party providers (TPPs).

New APIs created by approved TPPs will provide a single point of access to a range of relevant products, services and expert advice, based on analysis of their aggregated information.

Third party firms will be able to offer services such as lending platforms that use sophisticated technology to analyse customer data and make behaviour predictions, in turn offering more competitive terms.

This is especially relevant when applied to investments and pensions; a customer could give permission to a TPP to access their data.  The TPP would then make investment suggestions based on the rich data, risk profile and the retirement desires and ambitions indicated by the customer.

Traditional firms such as accountants, business consultants, pension advisers, and many others that obtain business through recommendations or tie-ups with banks must therefore seek out new ways of doing business. This may mean looking at firms they haven't considered working with before in order to access data and craft new ways of doing business.  

To drill down into this point a bit further, some key changes that we expect are:

• The adviser of the future will need to be increasingly tech and data savvy, and advice firms will need to engage and connect event more deeply with their clients.  Consumers now expect tech-driven solutions across their lives via the brands they trust and this expectation is underpinned by immediacy - the advice sector must learn from this and embrace it.

• We could see a seismic shift in the ability to better understand people by having access to ‘real’ and relevant data. At the same time, we are likely to see a reduction in the cost of servicing clients through technological efficiencies in the data gathering process. As a result, advisers will be able to serve more and more clients, helping them to gain control of their saving and retirement goals while contributing to increased margins for their firms.

• Advice interaction will change from 'across the desk' to 'along the road'. This 'along the road' advice can help people see in real time how much they are spending or saving, and where opportunities exist to do better, for example putting small amounts aside each month when they've paid the bills, or reducing discretionary spending.

Machine learning, client outcomes and suitability

For me machine learning is the only solution for a profession under pressure, and a profession that is facing a wave of baby-boomer retirees in the next decade on top of the extra demand created by pension freedoms.

For us and companies like us, our overall objective is to support advisers. By harnessing AI underpinned by machine learning, the more mathematical aspects of advice can be taken care of, leaving the more interesting and subtler edge cases to be handled by the human adviser. Together, machine learning alongside human advisers are a powerful combination, especially if the data continues to flow.

That said, the future of advice will be very different to what it is now, and I believe this will be driven by innovation in machine learning.  

Clients will spend most of their time talking about their dreams and goals, their needs, wants and fears with their human adviser.

While this is the case now for many financial planning firms, the difference is that in the future the ‘hard facts’ will be collected by an AI-powered 'robo-paraplanner', which is authorised by the client to collect all of the pertinent details from a data aggregator. This will be achievable in a large part due to open banking and its European equivalent, the second Payment Services Directive (PSD2).

AI-driven products will make it much simpler to ‘learn’ an advice firm's philosophy and replicate it as part of the onboarding process.  Therefore, delivering advice in more complex areas of financial planning such as retirement planning and optimising income (and tax) in retirement is set to become easier and faster.  

Machine learning will also be able to detect anomalies. By overlaying AI which has been trained with the benefit of past advice (both good and bad), machine learning should be able to reassure advisers that decisions are generated using best practice.

The beauty of AI is it keeps on learning; the more data it has, the more it will ‘learn’ and be intelligent enough to highlight areas of potential concern.

It is this data that will bring about the most change in the future.  If more organisations share data such as insurance, firms could then use this to significantly improve the speed at which machine learning can support the advice sector.

In the future, we would hope data is made public for advisers to evaluate both their firm's advice policy and individual cases. This will ensure the cumulative experience of the entire profession is applied to every individual who receives advice.

Something we are considering is how AI might be trained on large datasets of 'good' DB advice, as well as data that relates to some of the DB scheme scandals we have seen of late. This kind of model (perhaps something the regulator could support) would allow simple, bigger picture analysis of DB transfer business conducted by a particular firm.

Overall, machine learning can be used to map the advice process, potentially identifying cases which don’t follow an advice firm’s prescribed way of doing things. The FCA has signalled an interest in how this data could be used to improve supervision.

This article originally appeared in Illuminate.