Ep. 44 The Long-Short | How artificial intelligence and machine learning are shaping the future of asset management?

Published: 02 November 2022

Brought to you by the Alternative Investment Management Association (AIMA). This podcast aims to provide a window into the world of alternative investments.

Each episode will examine topical areas of interest from across the alternative investment universe with news, views and analysis delivered by AIMA’s global team, as well as a host of industry experts.

This week, The Long-Short takes a step through the looking-glass to imagine what the world of asset management could look like with the help of artificial intelligence, machine learning and automation.

We’ve enlisted the help of our friends at SS&C, Dr Zeynep Hizir, EMEA Development Director, and Richard Atkinson, Managing Director, to help us understand this new innovative era of finance.

Listen to this episode and subscribe on Spotify

Listen to this episode and subscribe on Apple Podcasts

Listen to this episode and subscribe on Google Podcasts

Listen to this episode and subscribe on Amazon Music

Additional resources

Read the transcript 

Hosts: Tom Kehoe, AIMA; Drew Nicol, AIMA

Guests: Richard Atkinson, Managing Director, SS&C, Dr. Zeynep Hizir, EMEA Business Development, SS&C

Interlude: Bill Kelly, President and CEO of the CAIA Association


Tom Kehoe, AIMA  00:00

Before we start today's podcast, we would like to bring to your attention an issue very close to our hearts.

Drew Nicol, AIMA  00:06

One billion children globally suffer from some form of abuse each year, including half a million in the UK with one in fourteen being physically abused. These dreadful statistics have only gotten worse in recent years, and sadly, more children are expected to be at risk as the cost-of-living crisis worsens.

Tom Kehoe, AIMA  00:22

We urgently need to do more to prevent and treat child abuse.

Drew Nicol, AIMA  00:27

Help for Children is a charity primarily supported by the alternative investment industry that is dedicated to eradicating the silent epidemic. Its fundraising events, including its annual gala, are critical to raising much-needed funds to do so.

Tom Kehoe, AIMA  00:41

Next month, on the 17th of November sees the return of London's alternative investment community to host its annual benefit dinner and we'd love for you to join us there.

Drew Nicol, AIMA  00:49

This year the theme is James Bond as a nod to 60 years of 007 in 2022. The evening is hosted by comedian and impressionist Jon Culshaw, who BBC Radio Four listeners may know from Dead Ringers. There will also be a silent auction, a raffle, and a live auction.

Tom Kehoe, AIMA  01:05

The London annual benefit dinner is always a fun night, it's a great opportunity to catch up with friends from across the industry while raising much-needed funds in the fight against child abuse

Drew Nicol, AIMA  01:14

Tickets are still available. Please do contact Fern Gray at [email protected].

To read more about the great work being undertaken by the help for children charity, including how to participate in events in your area. Please go to hfc.org. Thank you.

Tom Kehoe, AIMA  01:36

Hello and welcome to The Long-Short. My name is Tom Kehoe.

Drew Nicol, AIMA  01:40

And I'm Drew Nicol.

Tom Kehoe, AIMA  01:41

Artificial intelligence, machine learning and automation. They're just some of the buzzwords that you're likely to have heard during conversations about the future of technological innovation and financial markets.

Drew Nicol, AIMA  01:51

Though often we gloss over these terms without dwelling on what they actually mean, and how these various tools can or can't help managers run their businesses.

Tom Kehoe, AIMA  01:59

So, we have therefore decided to fix this omission and get answers to the questions that you've always wanted to ask about these technologies.

Drew Nicol, AIMA  02:07

To help us on this we've called upon our friends at SS&C, who are market leaders in fund administration, and live and breathe innovation across the front and back offices. We are joined by Dr. Zeynep Hizir, a member of the EMEA business development team at SS&C fund services. And Richard Atkinson, a managing director in systems development at SS&C technologies, you're both very welcome to The Long-Short.

Richard Atkinson, Managing Director, SS&C  02:30

Thank you for having us.

Tom Kehoe, AIMA  02:32

Dr. Zeynep, and Richard, as we said, at the top of the program, we've spoken a lot about the increasingly prominent role that technology of various forms plays in the business of running a fund. And so, for those of our listeners outside of the sector, what does that mean, in practice? Who of you wants to take that question first?

Richard Atkinson, Managing Director, SS&C  02:53

Thanks, Tom. Technology trends do tend to move in waves, from the initial research phase through to early adoption and into widespread adoption and delivery of value. The most interesting waves are really big waves, that change our lives dramatically, and not just within the world of work. If I think back over my lifetime, the big waves that spring to mind were back in the 1980s, the development of a PC, every everyone in the office getting a computer, rather than just one computer that might have been in a server room somewhere. If you were lucky. It was really the democratisation of computing at that stage.

And then, later in the early 1990s, the widespread rollout of graphical user interfaces with the rise of Microsoft Windows. Then also in the 1990s, mobile phones first going mainstream. The internet was another huge technology wave that particularly came in at the end of the 1990s the advent of e-commerce, and really connecting the world and starting to migrate software back off your computer and into the cloud. The smartphone revolution was another one, you know, particularly the launch of the iPhone in 2006, and subsequent smartphones. They became, you know, fundamental to our lives at about the same time as the rise of social media, which moved more of our personal relationships online as well.

Since then, increases in network bandwidth in both our sort of fixed and mobile networks have enabled the migration of broadcast television to streaming services and the now ubiquitous use of video conferencing, you know. Fund managers will be using all of these things and consider them part of the routine operation of their businesses. If I had to pick a single new wave (development) that still has to reach its full potential and fund managers might be continuing to look at, it would be the rise of artificial intelligence and more specifically machine learning and intelligent automation, which I know we're going to be talking about it in more detail.

Drew Nicol, AIMA  05:19

So just looking at the funds industry or the alternative investment industry specifically, I imagine there is a huge discrepancy between what many of the smaller players might consider sort of must-have technologies of today, compared to the larger players. If we think about the type of technology they should be expected to be wielding today, or what might be on their wish list, how has that changed over the years? And where are we now?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  05:48

I can take that one because I like the topic of startups or the smaller organizations that are entering the marketplace. One thing that Richard just mentioned is the power of Intelligent Automation. Intelligent Automation allows organisations to reimagine the operating model, enable and augment the workforce and gain business agility. This business agility is what smaller firms or startups typically have compared to larger organisations. So, through Intelligent Automation, larger firms are gaining this agility and keeping their competitive advantage.

Drew Nicol, AIMA  06:26

So, we've mentioned automation there and AI (artificial intelligence). So, let's dwell on that for a second, because automation really is one of those terms that will come up again and again, when you read events, and then generally sort of read around market trends. So, just to focus on that one for a second before we go into the others. What does that actually mean in your world of fund administration? And why is it such an important development?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  06:48

So, the term automation refers simply to the technique of making a process operate automatically. In its simplest form, intuitively, automation refers to speed and efficiency. So, automation isn’t an important development, but what makes it an important development is that we can combine automation with other technological advances, that have automation technologies, such as RPA (robotic process automation), and AI. When we combine these technologies, RPA and AI, we have Intelligent Automation. And that's where the important development in the world of financial services, not just fund administration, comes into play.

So, if I can just expand on that, a common example given in the space is comparing robotic process automation to Excel macros in that RPA is Excel on steroids, particularly in the hedge fund world. Excel macros have been widely used in investment research, or in reconciliation or calculating hedgerows, where macros only operate in Excel documents. RPA can combine different inputs and different systems to achieve the desired outputs. What this means is that tasks and processes can be automated so that the front office or human workers are relieved of non-value-added tasks for better and more accurate decision-making.

Tom Kehoe, AIMA  08:04

Dr. Zeynep, you've brought a new term into my lexicon, which is our RPA or robotic process automation. For our listeners, like myself, who are not that minded in terms of these technological phrases, how can you boil down RPA to its most simple example, you've given an example of it in practice when it comes to the funds industry. But what would be a simpler example again, when describing RPA?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  08:39

RPA is robotic process automation. So, actually, it's mimicking the human swivel chair movements. For example, downloading information from an email, copying it, putting it in an Excel spreadsheet, running a macro, and then outputting it into a different report. RPA can do this automatically because, in this sequence of actions, there was no value added by a human worker. So, RPA can do this and RPA is not intelligent. If you have a task or process that is not good, or streamlined, RPA will do it just as badly, but it will do it faster and without making any errors. I hope that was clear.

Tom Kehoe, AIMA  09:22

Yeah, so as for the more labour-intensive type of work where you can get the robot to come in and do what a human would take much longer to do, right?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  09:34

Exactly. So, it's typically actions that are copy pasted that don't require any human thought process into it. The swivel chair movements, taking information from one place, putting it somewhere else and sending out an email. These are all actions that can be enhanced by RPA. But we combine RPA with AI which is still simulating human thinking. This is now what we call Intelligent Automation. And this is really where the value is added.

When I started doing my research in RPA. First, it was “oh, RPA is the best thing in financial services. It's excellent. It's magnificent, everyone should do RPA, you don't need to use the IT department, everybody can be an RPA person”. And then, as I continued my research, we discovered the loopholes of RPA, and where RPA falls short. And it was the worst thing, you can't do anything without it, and there is no ROI on the RPA. Then when we were able to combine RPA with other technologies, AI, or OCR (optical character recognition), which resulted in Intelligent Automation. And that's really where we can gain value. AI can process high volumes of unstructured data, so now we can process paper-based and or manual tests that would traditionally take lots of human decision-making and time and copy and paste. This is now automated. This is what we call digital workers.

Tom Kehoe, AIMA  11:13

So RPA then is the most basic form, and artificial intelligence is a much more advanced form. Right? How then would you define machine learning? Is that somewhere in the middle?

Richard Atkinson, Managing Director, SS&C  11:27

So, artificial intelligence is computers trying to mimic in some way human thought. And machine learning is a subset of artificial intelligence, which is using data to build models, and then make predictions given new data. So, the models essentially recognise patterns within the data and can operate at a very large scale. I studied neural networks in the early 1990s when I was at university, and at that point they were interesting, but no one really knew how to use them to solve real-world problems. And it really took another 20 years of research for deep learning to be introduced, and then to make significant strides forwards.

There were watershed moments in 2011, and 2012, with image recognition tasks. Since then, there have been further significant milestones with voice recognition, machine translation, natural language processing, and many other areas. For example, over the past 10 years, we've all been very accustomed to being able to talk to our digital devices and have them understand us, and machine learning has been the key technology that's enabled that shifting capability.

Tom Kehoe, AIMA  12:46

So, when we think about the funds business, we're looking at the business from the back office to the front office, where can artificial intelligence and automation then support the operations of the hedge fund business?

Richard Atkinson, Managing Director, SS&C  13:03

There are a few examples of where we’ve used machine learning within our fund administration business being within our cash wire system. We process a million plus payments per year that are mostly high-value and same-day settlements and payments. And there are extensive controls around that process, including maker checker controls on the payment and a callback process to give a segregation of duties and a two-factor verification on them. But we wanted to provide additional information to the person approving the payment to advise them if the payment looked unusual compared to previous payments that have been made. That's not a particularly easy problem for traditional software development approaches. So, we tackled the problem by training a machine learning model using key attributes of each payment that could then provide fast insight to the person approving the payment. Whether that particular wire looked unusual. It was not a hard stop for a payment, but an advisory to the person approving the payment, to take extra care and ensure that the payment is appropriate for their client. And since then, we've applied similar anomaly detection for profit and loss items on accounting reports within our news GoCentral application.

We also process a large number of loan agent notices, which is a challenging area to support because we receive a lot of documents that are very diverse in terms of their formats. Some we receive as PDFs attached to emails, some are just emails, some are received by fax and the volume is incredibly uneven, as it tends to be relatively low during the month and then has a significant spike with 60 to 70% of documents received in the last four business days of the month. That makes it very challenging to resource with a service team doing the processing because it leads to some very long hours being worked on month-end and even more so at quarter-end. Applying traditional software development techniques to extract information from those loan agent notices only achieved relatively low straight-through processing rates. And so over the past few years, we've been applying machine learning to that problem as well as using natural language processing in order to significantly improve those straight-through rates.

So, we convert the loan agent notices into text using OCR, and then we pass that text to a neural network that's been trained on previous loan agent notices to extract the key economic attributes that are needed in order to update the accounting system. So, for a rate-setting notice, it will be the loan facility identified, the effective date, the new base rate spread, and the all in rate, for example. And so, the neural network is taking that unstructured text data and then turning it into structured data in order to feed the accounting system. It's doing that pattern recognition in order to identify the correct fields in the loan agent notes and provide materially better results. And so, it's another example of where we've had good success with the technology.

Dr. Zeynep Hizir, EMEA Business Development, SS&C  16:28

Thank you, Richard. Those were really great in-depth examples, I want to just touch upon some of the new techniques that we are using in our world in the hedge fund space.

Most recently, in March 2022, SS&C acquired Blue Prism, which was a market leader in robotic process automation. So, this has enabled us to unlock more capabilities in our Intelligent Automation space. Some of the ways how we're enhancing hedge funds are facilitating the speed of new account opening and speed to market, improving sanctions checking for offshore hedge funds, institutional entity checking to improve compliance, improved productivity, and accuracy of reporting internally for decision-making or externally for clients and regulators. We're improving fund management risk, reducing costs through automated reconciliation services, improving costs and speed of bringing new products to market, and cybersecurity, which are all key areas that are challenging in the hedge fund space.

Bill Kelly, President and CEO of the CAIA Association 17:35

Hi, this is Bill Kelly, President and CEO of the CAIA Association, and you're listening to AIMA’s The Long-Short podcast. Join me in episode 14 where I discuss my vision for improving financial literacy and understanding of the alternative investment industry, as well as keeping CAIA’s curriculum up to date with the market, and this is a never-ending job. Enjoy and stay educated.

Drew Nicol, AIMA  18:02

Can I just pick you up on something that I think you've both mentioned so far, which is this need for data to be standardised. And that really is the essential component of applying automation and a lot of the AI really can't be stressed how important that aspect is. But there are areas of our industry at least where that is very difficult, and the one that comes to mind is ESG. So, could you just help our listeners understand in a little bit more depth, where data standardisation comes in? And maybe use ESG as an example of why it's so essential?

Richard Atkinson, Managing Director, SS&C  18:40

So, we certainly do take in data from a number of ESG vendors and integrate it into our security master to then provide aggregated reporting for both investment managers and for investors as well. There's a large number of data points from each of the ESG data vendors on each issuer. But there is still significant variability between the datasets. That variability can include scope divergence, where there are different pillars that are being evaluated within each environment, social governance, rating, weight divergence where you know how the scores for each pillar are then combined into an overall E S or G rating, and also measurement divergence, even at the kind of very bottom layer of that pyramid, in terms of what is being measured. For example, whether it's carbon dioxide emissions or carbon dioxide equivalent emissions.

There was some academic work done on the topic by the MIT Sloan School of Management and the University of Zurich, which showed that the correlation between ESG ratings ranges between 38% and 71%, whereas correlations between credit ratings are typically around 99%. So, there's definitely some way to go in terms of reaching an agreement in terms of methodology. Tesla was one key example where, in 2019, MSCI rated the company in the top 10% for the Environment Rating, FTSE rated them in the bottom 10% and Sustainalytics, which is now owned by Morningstar, rated them somewhere in the middle. And the reason for the divergence was that MSCI was evaluating them on the end product, zero-emission cars. FTSE was evaluating them on their factory emissions, and Sustainalytics was looking at both. So there still isn't a consensus in the evaluation methodology for an environment rating, let alone how to combine all three of those into an overall ESG rating.

There's definitely a lot of opinions still in the mix, and I think there'll be a range of opinions that continues to exist, you know, across data vendors, across investment managers, and also, across investors as well.

I think probably the key for investment managers is to provide transparency around the process that they use and how it integrates into the investment process. And then also for investors to dig beneath an ESG rubber stamp, as it were, to make sure that the investment process is actually meeting their objectives and priorities as well.

Drew Nicol, AIMA  21:34

So is this an example then, for a limitation, at least for now in AI, or Intelligent Automation, as you describe it, insofar that it's being held back by issues in the wider market, and the datasets you're getting. Are the other tools and services that you work with every day, sort of hamstrung until this issue is resolved?

Richard Atkinson, Managing Director, SS&C  21:57

I think in some ways, machine learning can help to try and solve some of these problems. So, for example, back in the loan agent notice example, the data that we get in from different loan agents is very diverse, and the formats are very different. They contain mostly the same information; they don't always contain everything that they need. But by training a machine learning model on that data, we're effectively using that machine learning model to standardise that data into a structured format, in order to feed an accounting system. So, in that case, you know, we're using machine learning to get us to the normalised example. With something like ESG data, there's still a lot of, you know, sort of taxonomy issues and opinions within the data, which machine learning model is not necessarily going to solve? It is a question of getting to the end of the processes and going, did that produce the outcome that the investors were looking for?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  23:06

Can I just add to that? Since ESG is an important topic, I also wanted to highlight that putting aside the data and taxonomy issues, we should highlight that Intelligent Automation can play a significant role in helping companies across industries achieve their sustainability goals and deliver financial benefits. This is done through evaluating the company's specific sustainability goals and identifying the key places where Intelligent Automation could make them more achievable, accelerate the benefits, and equip the organisation with the appropriate capabilities to measure and govern the progress, which is critically important for internal and external communication. So, while there are still issues that are being brought up to highlight, there's a lot that Intelligent Automation can add to the ESG space.

Drew Nicol, AIMA  24:05

I think it might be useful to actually just take a step back because I think a big component of this conversation that we've not addressed yet, is the drivers of adoption. Thank you for the examples you've given because they do answer that question somewhat in that, you know, greater efficiency is a self-evident benefit. But there are also external factors, as I understand it, at least, and one of those, I believe, is the demands of regulators on the alternative investment Industry, when it comes to reporting and you know, ESG comes up again in that field.

So just to put it to you, how big a factor is the increasing regulation? As we've covered many times in The Long-Short, there is much more to come in that arena. Are there any particular examples that stand out to you?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  24:56

As you highlighted, there are a lot of regulatory requirements that is putting a lot of pressure on organisations, because complying with the regulation is quite costly. It's allocating a lot of human resources and technology resources in being compliant with the regulations. A recent example, under UK EMIR, in September, phase four of EMIR saw uncleared margin rules come into play. This requires very complex calculations, which are very much facilitated by Intelligent Automation. So not only is it error-proof, it's also 24/7. It's auditable and compliant.

Drew Nicol, AIMA  25:42

And again, is that something where with you talking about EMIR, does the data need to be clean going in or can Intelligent Automation, by virtue of being intelligent, go some way to consolidating issues in the datasets?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  25:56

Intelligent Automation can consolidate the dataset, we've mentioned putting the intelligence into the automation, we can now also manipulate unstructured data, or we can recognise handwriting with a lot of accuracy. So, this has enormous benefits, in terms of the reporting to stakeholders and regulations, and keeping everything auditable and compliance

Richard Atkinson, Managing Director, SS&C  26:24

In terms of something like UMR, though, you know, generally speaking, its structured data that's going into that process, it's the clients trading portfolio, it's the other risk numbers that need to go into there. And so, you know, there's always a need to apply the right tool to the job, and putting a neural network in that process probably wouldn't be the right tool for that. Because, you know, there's a clear way you get to the right numbers and traditional software development and feeds is generally the right way to do that.

Dr. Zeynep Hizir, EMEA Business Development, SS&C  27:00

I think Richard raised a very important differentiator. All of these available techniques are not always appropriate for every situation. That's why it's very important to partner up with the right technology partner, and solution provider, so that we can identify the optimal levels of where to use RPA, where to use AI, where to just use a simple API, to make your organisation run efficiently. I think this is really a differentiator and a competitive advantage to understanding where to use which technology.

Tom Kehoe, AIMA  27:37

And we will have information in the episode notes, for those of you who want to read more about that, and indeed, the work of SSC technologies. But when it comes to operating all of this new technology, undoubtedly employees everywhere, are having to become increasingly competent when it comes to the use of IT and being able to stay on top of what seems like an increasing number of trends and ways of managing and working with technology.

So, what effect is this having on hiring and indeed the skills that you're looking for the types of roles across the industry and how they're likely to evolve over the next few years? And also then, should we all be learning to code? I mean, kids, the next generation, they're already coding. You know, are we in a situation where robots are going to take over our jobs? Or is it more a case of we're going to manage the robots?

Richard Atkinson, Managing Director, SS&C  28:42

I mean, I think employees do need to be competent with IT. The pandemic drove that home when we all started working from home and to some extent, we became our own IT support departments. I also do think it's helpful for everyone to understand the basics of coding, just to understand how technology can be deployed and how it can be productively used. But I don't think everyone needs to be building systems or writing games in their spare time. But certainly, an understanding of the technology is useful. I think if a machine learning system is going to take part of your job, it's generally going to be the least interesting part, it's going to be the part that is repetitive.

You know, if we get a loan agent notice through a machine learning pipeline, and it's the same format, the chances are it's going to flow straight through, and the machine will recognise it. If it's a new format that the machine learning pipeline has never seen before it won't get it right. You know, it's going to fall into the exception queue and the service team will have to have a look at it and then they may, if it's one they've never seen before, they may go oh, that's different, not seen one of these before. But then they'll proceed to work it out and go, alright, well, that's the principal amount and as the new base rate and spread, and they will find what they need to find, in order to interpret that new notice.

In the machine learning systems that we're building today, you train your machine learning model, but it doesn't have that cognitive flexibility to be able to learn new things on the fly. But people do, you know, so they won't display creativity, they don't display problem-solving. They can't learn without experience. But people can. They can't perform conceptual reasoning or abstract thinking. These are all areas that researchers are working on, but machine learning systems today can't do those things.

In terms of hiring and potential careers for our children, I think there's plenty of reason for hope. I recently started a book that I bought just before the pandemic, but not got around to reading and it was called ‘Range’ by David Epstein. It's about how generalists triumph in a specialised world. And, they had a fascinating fact that isn't new, but I was completely unaware of it. It's called the Flynn effect, and it's that IQ tests need to be periodically renormalised, in order to keep the average score at 100. So, it's been documented in over 30 countries, and the gains are significant, on average a three-point increase every 10 years. So, when you accumulate that, an adult who scores average on an IQ test today, if they were compared to an adult, 100 years ago, they will be in the 98th percentile. The gains don't tend to be related to directly observable objects, but in abstract words, or imperceptible concepts like law and pledge and citizen solving problems without a previously learned method. And these are precisely the areas where machine learning doesn't do so well. So, you know, we're smarter than our parents, and our children are smarter than us. I related that story to my daughter who's 16, and she was completely unsurprised, and she said she knew it already, but then she is 16.

Tom Kehoe, AIMA  32:22

So, our jobs are safe for now. As you say, it's more humans working with the machines rather than machines taking over entirely. That’s comforting to know, but, technology continues to advance and progress.

Richard Atkinson, Managing Director, SS&C  32:36

That's right. The gains that we get from working with machine learning systems are increases in productivity, you know, so that translates obviously to your business value and to rising living standards. So, there are benefits from adopting the technology. If firms, you know, don't necessarily have the skills and experience in-house to be able to deal with it, there are always firms to partner with to be able to take advantage of some of the capabilities, which are now available.

Drew Nicol, AIMA  33:07

You've teed me up perfectly for our final question, which is just to naturally ask you to look ahead a few years, and where we're going. And I think you've sort of teased that already, but if I'd say sort of five years and where the alternative investment industry will be in terms of technological adoption? I think we've mentioned one or two areas where it's a little bit aspirational to get there, maybe if I could put it to both of you, sort of five years from now, is there anything that we will have that we don't have now? And is there anything that will just be more common than it is now?

Dr. Zeynep Hizir, EMEA Business Development, SS&C  33:42

I'll kick that off, and Richard, maybe you can add to it. So today, Intelligent Automation provides companies with an alternative labour source. A digital workforce that executes business processes, 24/7, with incredible speed and productivity that never makes mistakes has 100% audibility, and can tap into one or more advanced AI technologies that we've mentioned. So that's what's happening today, maybe a decade ago, we would never be able to project that technology is advancing at an exponential speed that is. Whereas we humans can only project linearly. That's why it's very difficult to imagine what the next five years will be like. We expect to see a much wider adoption of technology as the alternative investment industry will get used to deploying these technologies as it sees the benefits of using this technology.

So, we will see the adoption rates go higher, return on investments go higher and advance linearly, whereas the technology has already surpassed human capabilities. With these facts we have discussed in this session, the alternative investment industry will need to quickly adopt these available technologies so that we can get to where we're going to be in five years, quicker.

Richard Atkinson, Managing Director, SS&C  34:56

Yeah, I'll just add to that. The technology trends that we talked about that have sort of come in large waves, you know, they take time. They seem, sometimes, to move slowly. But then when you look back, you realise that everything has changed. And actually, you've become accustomed to the fact that everything has changed, but it crept up on us somehow. And so, I think in five years’ time, you know, the state of everything will have advanced, all the waves that we already have seen will be further progressed, and I'm sure we will see a lot more artificial intelligence, machine learning, Intelligent Automation in the mix of people's businesses and personal lives as well. You never know what's going to come up, you know, sometimes you do get new waves arriving that have been out there for a while, but people weren't too aware of them. Sometimes you even find that all of a sudden, it's reached that critical point where you can get huge value from an aspect of technology, which wasn't there before. And those things aren't always readily apparent. But I think it will be interesting to see how technology is developing, particularly in the machine learning space, because it's got so much potential, and it's really still at the early stage.

Dr. Zeynep Hizir, EMEA Business Development, SS&C  36:23

I just want to quickly add to that, if you don't mind. So, Richard brought up a really good point about investing in machine learning models. And I want to highlight that at SS&C, we invest thousands and thousands of hours in modelling these machine-learning programs and models. So that if an organisation does not want to allocate resources to enhance these capabilities in-house, they can partner up with organisations where they can focus on their core competencies and outsource where they don't need to develop these competencies in-house.

Drew Nicol, AIMA  37:01

That's a good point, actually. And that whole aspect of the conversation around the role of service providers being specialists and providing the services, especially as it relates to the smaller fund managers that we spoke about before, then that's a very important part of this conversation. And maybe we'll have to get you back to focus on that because as you say, it's vital.

Personally, though, I'm just dwelling on the fact that we're still going to have a job in a few years. That's my one takeaway from this. I'm very reassured to hear that. And we mentioned at the top of the episode that this was meant to be focused on just sort of answering the simple questions that maybe get glossed over and we hinted that was for our listeners, but it was more for me, really, and I've certainly learned a lot. So, thank you both very much for your time, and your insights have been extremely valuable.

Richard Atkinson, Managing Director, SS&C  37:51

Thanks very much. Nice to talk to you today.


This podcast is the sole property of the Alternative Investment Management Association (AIMA). This audio production and content are intended as indicative guidance only and are not to be taken or treated as a substitute for specific advice, whether legal advice or otherwise. AIMA permits use or sharing of the content in media or as an educational resource, provided always that proper attribution is made. The rights in the content and production, including copyright and database rights, belong to AIMA.