The following interview has been edited for matters of length and clarity.

Edited by: Editor in Chief, Kyra Ward

Conducted by: Economics Graduate 2018 Adam Stromme

Interviewee: Grant Fuller, Irithmics co-founder.

Irithmics is a highly innovative and exciting company in the field of data analytics. Specifically, Irithmics endeavors to use, “… artificial intelligence technologies to expose new insights into the dynamics and behaviour of listed companies and their institutional investors.”

Background

Adam Stromme: How did you first get involved in the world of fintech (financial technology)?

Grant Fuller: I have worked for some wonderful companies and amazing people. After all the academics (St Andrews and Cambridge) I worked in risk management for the Bank of Tokyo Mitsubishi. I had a fantastic manager Oki-san, who completely transformed my view of finance and banking. When the time came, I moved to RiskMetrics who, to this day, remain the most inspirational firm I have ever worked for. If RiskMetrics were to start today it would be the fintech of fintechs and I am lucky to be in very regular contact with many people I worked with there. The culture and philosophy, which strongly resonated with me, has profoundly shaped the way I view business, clients and the people who work with us. Ethan Berman, the then CEO of RiskMetrics, summarized the philosophy of the firm as, “Change the world, have fun and make money – in that order” – we share that philosophy even today. People from RiskMetrics still believe that, and many have carried that motto into their subsequent firms.

Of course, you can’t think of finance, technology and innovation without thinking of Bloomberg, which is where I worked after RiskMetrics. The team and project was fantastic, truly entrepreneurial and completely focused on clients. Actually, the product we started turned out to be wonderfully successful and a true credit to the innovation of Bloomberg and the people within in. If you get an opportunity to work for Bloomberg, take it.

My path into fintech has been a long time in the making and I have had the huge privilege of being nurtured though both RiskMetrics and Bloomberg – two titans. Both of these firms taught me that great people make great products which clients obtain huge value from.

AS: How did your background in chemistry prepare you to work in this rapidly developing field?

GF: There are very few truly vocational degrees. The most powerful thing about education is not the knowledge you acquire, but the skills you develop to learn and apply what you know. For me, chemistry was a great springboard for what I am doing now.

Some incidental lessons chemistry provided which have been incredibly useful are:

* Read the instructions before you begin – it sounds basic, but you’d be amazed how easy it is to rush into something without understanding what’s expected. It can be as simple as a Mary Berry cake recipe, an employment contract, documentation of some code or an agenda item for a meeting.

* Prepare – in chemistry it’s important you have everything prepared, clean, weighed, measured before you begin. It doesn’t end well if you don’t. Very few chemists can “wing-it”. It’s similar in business: being prepared for what you’re going to do, projects and meetings helps enormously. It’s a total waste of everyone’s time when someone hasn’t prepared for a meeting – you really don’t want to be that person.

* Patience and persistence – there’s a lot of waiting around in a chemistry lab. It can be very tedious counting drips, scratching the slide of a round-bottomed flask or trying to spot the exact moment the solution turns pink. This is certainly true in business too: sometimes things move slowly, painfully slowly. You often find yourself wondering if you’ve done the right thing.

* Failure – chemistry, science in general, is very good at teaching ways things don’t work, ways not to do something. You only know, or more correctly you only believe, something is the right way because you haven’t failed at it yet.

Your Work

AS: What do you think is the biggest misconception people have about AI?

GF: Goodness, there are so many. Some are:

* AI is new: this isn’t the case. Many of the machine learning algorithms have been around for over 50 years: logistic regression in the 1950’s and artificial neural networks in the 70’s. These have been used in universities and industry for decades. Advances in processing capabilities (the speed of computers, and in particular GPUs) and the quantity of data now make the application of these so exciting. Deep learning and the fascinating advances being made by large companies like Google, Apple, Facebook and IBM as well as small companies like Irithmics show just how accessible machine learning and AI is becoming to a wide range of industries.

* If AI is the answer, then what is the question? AI, and more accurately, the machine learning algorithms that make it possible, are just software: combining maths, code and data. They’re not mystical or magical. Like all things, the correct tools, technologies and data need to be used. For some things, machine learning isn’t the best tool for the problem.

* AI will take jobs away from people: I know what people mean by this, but I don’t believe it. Humans have always used technology to transform the way we do things and compensate for our physical and biological inadequacies. Now we’re simply using technology to help us understand the data we’ve collected about the world, about ourselves and society. People have a remarkable ability to adapt and learn new things. For those individuals and organisations who learn to use or exploit this new technology, and learn to capitalise on the insights being unlocked, there is a very bright future.

AS: Tell us about some significant trends you see in fintech.

GF: That’s a hard question! For years, people have assumed that in order to improve you just had to do things better, faster, or cheaper. If you did it substantially better or substantially faster, you were called an innovator or a disruptor. But to genuinely disrupt you need to do something in a completely different way. Rear Admiral Grace Hopper said the most dangerous words in the English language are, “We’ve always done it this way.” To be a disruptor you need to do things differently.

Fintech is helping with this and covers a huge spectrum – from algorithmic and robotic trading, risk management and compliance, valuation and settlement to insurance, payments, transactions, mergers and acquisitions. There are even companies processing satellite images of car parks, the number of swimming pools, or rice fields to measure the productivity of regions and countries. Insurance companies now have access to data which allows them to underwrite new products and services (e.g. you can now get insurance on whether or not your solar farm will have enough sunshine to generate the required amount of electricity).

But it’s not just what we typically think of as finance: JPMorgan’s COIN is credited with “doing in seconds” what had previously taken human credit officers and lawyers 360,000 hours to do. That’s around 200 human working years they have achieved in a matter of seconds.

To return to your question – the most significant trend I see fintech providing is disruption. The ability to do things differently and help answer questions which we didn’t even know could be asked before – this is what fintech is providing.

AS: For many, financial data, analytics and research is associated with investors understanding a company, sector or portfolio. Your company Iritimics, however, is changing this viewpoint. Can you explain this further?

GF: Several different players make up capital markets: two most notable are the issuer (the organisation issuing the equity, debt or any other security) and the investors in those securities. You’re right, much data, analytics and research focus on the quantification and description relating to the security and for very good reason – investors need to understand their investment, its potential, its risks, its suitability etc. However, a lot of information can be obtained analysing the investors themselves: their behaviour and their dynamics.

Just as a biologist can arrive at an understanding of which food sources contain more sugar simply by watching which one a colony of ants prefers, so too can we arrive at understandings about individual listed companies and investors.

This is the basis of Irithmics: we analyse the behaviour of [institutional] investors to understand aspects of individual companies, portfolios and sectors. Of course, there’s quite a bit of clever technology and a great deal of data which makes this all possible, but the most important thing is that Irithmics is able to help our clients better understand and anticipate the forces affecting investors.

If you’re a publicly traded company it’s more important than ever to understand how investors are likely to react: it helps to anticipate and plan, it helps you manage and develop your shareholder register, it helps you explain to your board what factors are likely to be influencing investors.

Carillion, the UK construction firm, which went into administration earlier this year, serves as an example of the impact Irithmics’ technology can have analysing publicly traded companies.

Carillion collapsed on the 15th of Jan 2018 after a series of lengthy and protracted discussions with its bankers. Six months prior to that, the group’s CEO resigned on 10 Jul 2017, followed by a 30% drop in share price the following day as analysts questioned whether measures to conserve cash would be enough to stave off a rights issue. It’s natural to expect, and indeed see, much turbulence from investors between Jul 2017 and Jan 2018. However, what is much more insightful are certain key shifts in investor behaviour and dynamics. This suggests as far back as Sept 2016 (15 months before the collapse), investors had begun to treat Carillion with caution, by Jan 2017 data indicates investor behaviour had changed significantly. Vulnerability to selling by systematic strategies and shorting rose throughout April and May. In July 2017, City A.M. reported that Carillion was the most shorted stock on the London Stock Exchange, with over 27% of its stock being shorted. There are many more examples of investor behaviour changes relating to Carillion, in fact almost every listed company has examples of investor behaviour changing, of shareholder register ecosystem changing or of significant concern from markets.

What is fascinating however, is this isn’t the result of human insight or research: a machine has generated this, a computer has monitored and discovered these changes in behaviour towards a company. Very often, it identifies and highlights these before they become apparent to everyone else.

The World Today

AS: I recently saw a video from the London Stock Exchange describing how Irithmics is now helping LSE corporates better understand their investors and markets. How has this development affected your company’s future plans?

GF: The team behind Irithmics is fantastic, so it’s tremendously flattering for us all to have worked with the LSE on their exciting new platform and see the energy, passion and values of one of the world’s largest exchanges [ed. over £4 trillion is invested in companies listed on the

LSE]. It’s great to be part of the project, and the opportunity to provide our innovative technology directly to issuers, helping them estimate, measure, monitor and anticipate the behaviour of institutional investors.

You’re right of course, a company’s plans change – it’d be foolish if they didn’t. We have had a great deal of interest in our technology and data from a wide range of organisations: listed companies, hedge funds, public relations advisors, corporate brokers and regulators. These have generated work for us. We’ve accelerated our research and development plans, we’re interviewing graduates for a number of roles with our team outside Bath and we’re speaking with companies across every sector and on six continents. They’re some incredibly fascinating companies and not only do we get to learn about them, but we get to learn how institutional investors, so-called “Smart-Money”, are behaving towards them too.

It really is exciting to be part of something which is generating so much interest and which has the ability to enhance understanding.

AS: What suggestions would you give to someone interested in working in fintech?

GF: If you get the opportunity to do something different, take it. Learn as much as possible and take advantage of what people are willing to teach you. For me, RiskMetrics and Bloomberg were seminal periods in my career. I learnt so much from those organisations and the people which made them possible.

Thank You Grant for speaking with us, and we are all looking forward to seeing the wonderful work Irithmics will continue to produce in the future!