Geoffrey Marsh

AI for Modern Investing

Blog Post created by Geoffrey Marsh Employee on Jul 26, 2018

I was at a leadership meeting in Barcelona and I got a call from our financial services CTO Daniel Knight letting me know he had just had a lunch meeting with the CIO of one of the largest wealth management firms in the world and that they had a major initiative that they were stuck on and if we could offer a solution they would buy it instantly.

That solution or problem was RoboAdvisor- or automating investment advice. This had quickly become one of the biggest things in wealth management.

I asked him a little about the use case and this is what he said. The millennials don’t want to spend time with a person managing their portfolio, but they need some kind of advice and also the firms don’t want to spend money for a person to manage a small portfolio that they are not going to make much money from. So, can we automate this using machine learning and AI to be able to consume data from inside an organization, outside an organization such as social and index feeds and then present something back?


This is our sweet spot so in a matter of weeks we built the Hitachi Vantara RoboAvisor solution. What this solution does is ingest data for a multitude of sources including index data that comes from a supplier, external financial data that is readily available, external social data, and internal data. With this we have a web front end that the user inputs their information, things like age, age they want to retire, dependents, and so on and then with the algorithms we have created it generates an investment profile and makes recommendations as to where to put your money.


I know I have overly simplified the technology, but I am more interested in what does this mean to both the organization that deploys it and the end user. What real tangible benefits will these stakeholders realize.


I am going to start with the firm. Currently the way most firms handle wealth management and it has been the same for decades is the customer meets with someone face to face to talk over things like personal details (when they want to retire, what are they saving for, what can they afford to put away) then they will invest the money into the funds, stocks, bonds etc that they think will get you the best return and help you reach your goals. This costs the firm money and if the portfolio is not making much from your average investor that just wants to save for retirement and then it may only be a break even. Think about it they have to pay that person something to sit in the office, they will pay commission on the funds they sell not to mention the basic overhead of a brick and mortar institution. And what about if they are long term investments and the person is not moving money around. That interaction with the customer could be complete for years. Do you really need a person to do it all every time? That could be a heavy OPEX handshake if you know what I mean.


By automating that piece you are essentially removing all of those costs. Yes, there will be the normal data center costs but no commission, no labor costs, greatly reduced overhead of a brick and mortar store if any.


Now let’s look at the end user of the application. I am around 40 years old, grew up at the same time as technology and work in technology so I am a perfect example. I cannot remember the last time I went into an institution to do anything with my money. I would be willing and have called a call center for something basic, but I would rather do it online or in an app. I am part of that generation while not as bad as my son, but I want to do my banking or investing and so forth when I want to do it not necessarily between 9 and 5 on weekdays. Also, with everything in the news I believe that trust factor of a human portfolio manager is not where it use to be, in the machine age I believe we trust what a computer is saying more than a human (I know a great AI topic for another blog). Bottom line is this, RoboAdvisor allows the investor to get financial advice, recommendations and could even move the money for them on their own terms and schedule. I know myself I would be more likely to use this than a human.


In talking with our financial services CTO, we know this is something every wealth management firm is either implementing or thinking about. Those that are not implementing it don’t want to be left behind but they don’t have an answer yet. We have an answer and while I cannot share the names of the organizations that have bought it I can assure you that you would recognize the names of them. In discussions with some of our other clients this solution could be used for insurance as well in the space of both quotes or new signup as well as claim adjudication.


This was the first in a series of blogs outlining some of the interesting solutions that we are and have built based on what our customers are asking for.