The beginning of an era
18 Mar 2020
The AI era is still in its infancy in the asset servicing world, but how far can it go?
Image: Shutterstock
The world we live on, according to scientists is around 4.543 billion years old. Our world history teaches us about the stone age, bronze age, ancient Egypt, early Islamic civilisation, and more. Each era is interesting in its own way but given the age of the Earth, the pace of change in development from the Victorian era until today is truly remarkable.
Today, a vast majority of industries are heavily reliant on technology and it is crucial to the way we conduct business, particularly in the financial services industry. Technology has soared since the start of the 20th century and it is involved in our everyday lives today in the 21st century, with technology only expected to accelerate and evolve at lightning pace.
Looking back at how quickly technology has advanced over the last 200 years, you can only wonder what the world will look like within the next 200 years. Technology is vital in the financial services world, especially within the asset servicing industry, where enhanced technology is key. How do firms prepare for the future when the change is occurring so rapidly?
Within the asset servicing industry, Andreas Burner, chief innovation officer of SmartStream Technologies, observes that there is a large number of people who understand the potential of artificial intelligence (AI) and have started utilising it within their firm. However, Burner says that the industry is still very much at the beginning of the AI era.
A new way to work
Discussing whether AI poses a threat to jobs, Andreas Burner, chief innovation officer of SmartStream Technologies, emphasises that people have always been afraid of redundancy.
Burner says the world is constantly developing, using the example of newspapers moving to digital content, he notes that people were worried that automation would mean loss of jobs. However, he explains “all these things have in fact created more jobs. Certain roles will need to adapt and develop to meet the demands of new technology”.
Echoing this, Alastair McGill, general manager data control solutions at Broadridge, comments that AI is not something we should be scared of but something we should embrace. McGill highlights that we should not assume right away that intelligent automation will lead to massive job losses.
According to McGill, Gartner research indicates that more jobs will be created by AI rather than will be lost to it.
He says: “One of the factors is the adoption and use of AI to augment human tasks rather than to replace them. New types of jobs will grow as new technology is adopted. The analogy used often is the Industrial Revolution – even though agricultural jobs went away, new jobs (and new types of jobs) were created.”
Additionally, McGill stipulates that in a heavily regulated industry like financial services, human oversight of intelligent automation will “remain critical” well into the future. However, McGill suggests we are still many years away from creating the kind of artificial general intelligence that outperforms humans when it comes to critical thinking about relationships between certain types of information.
Reinforcing this, Davide Zilli of Mind Foundry, explains that humans will always have a role to play, “algorithms need to be fully transparent in their decisions, easily validated and monitored by a human expert”.
A treasure trove of technology
Although concerns will remain, the consensus suggests that technology is more abundant in its opportunities than its downfalls, and is seen to be more useful as opposed to threatening. With AI providing the ability to remove manual, mundane processes, the human workforce will have more time to take on creative tasks, enabling greater efficiency in the workflow.
Richard Street, global head of client coverage for RBC Investor & Treasury Services (RBC I&TS), indicates that “AI should accelerate the process of automating manual activities, which will enable the industry to deliver end-to-end solutions more quickly”.
Street comments: “If you look at the private capital and real assets space, deploying capital is a complex activity involving the engagement of lawyers, investment bankers and advisers, which is very different from the automation of execution and settlement in the traditional asset world.”
He explains that there is a role for AI to play in evolving and perfecting this process, more so than focusing on perfecting the mature real money investment management processes.
McGill also weighs in discussing Broadridge’s experience with implementing the latest technology. He says: “Replacing a traditional rules-based matching engine with a self-learning machine learning model can provide significant efficiency benefits by giving the system the ability to automatically build and improve its matching schemes based on new activity.”
McGill adds: “IT projects to onboard or maintain matching schemes are no longer required as the solution analyses historical matched data and suggests a matching scheme that will process new data with very high levels of accuracy. As the training dataset grows, the model improves, keeping straight-through-processing rates high without the need for constant development and testing.”
“Overall performance also improves significantly as machine learning-based solutions take advantage of in-memory distributed computing platforms designed to process huge amounts of data efficiently.”
Reflecting on the opportunities of AI, Street identifies that it enables it to increase the quality of the services that our industry provides, and thereby drive client demand for faster service and better insights based on the high-quality data which can be derived from those automated tools.
Technical issues
Although AI does have plenty of advantages, AI projects can suffer from a “cold start”, which is one of the key challenges. McGill explains that this is where predictions are initially low quality as sufficient training data is not available to generate an accurate model.
McGill explains: “A solution with the ability to generate models from historic training data can mitigate the cold start problem, leading to shorter project timelines and a faster return on investment by providing high-quality predictions from the start.”
According to McGill, self-service and transparency are critically important to the industry as firms are cost sensitive in regard to change processes.
He also states that auditing requirements are of paramount importance. “Many AI-based solutions available to the market operate on a ‘black box’ basis, requiring vendor data science staff to handle the generation of new models or optimisation of existing ones at significant cost”, he adds.
Weighing in on this, Zilli from Mind Foundry, says: “Machine learning tools must introduce this full accountability to evolve beyond unexplainable ‘black box’ solutions and eliminate the easy excuse of ‘the algorithm made me do it’.”
RBC I&TS’ Street also warns that it is important to recognise that AI is not a ‘fix-all’ solution. He stresses that anyone who is building an operation on AI alone has got it wrong.
He adds: “You have to put a process in place and automate it as best you can. AI can evolve that process faster and better because it’s learning and remembering things more efficiently.”
To infinity and beyond
Although technology is evolving at a rapid pace, industry experts suggest that it is moving more slowly in the asset servicing industry. But arguably, never before has technology been so crucial amidst increased velocity of regulation.
McGill suggests that the asset servicing industry “has been under consistent financial pressure for the last five years”. He notes: “Rising asset prices have largely masked underlying declines in assets under custody/assets under administration fees.”
“Restoring profitable growth amidst these pressures will require more innovative solutions to improve operational efficiency. There are a number of areas in which the industry can benefit from the adoption of AI.”
“Leading firms will rapidly innovate using AI which will form the basis of leaner, more efficient operational functions,” he adds.
Also looking to the future of the asset servicing sector, Burner notes that tomorrows’ applications will continuously support the users in their needs and will deliver answers to questions before a user is asking.
Burner explains that AI applications “will provide excellent user interfaces that will deliver condensed and relevant information to the user in the right moment”.
He believes that in a few years, “the financial services industry should operate in a similar function to weather forecasts, which are easy to access and they will be something you can look at daily and is automatically delivered to you in terms of information.”
“Although we are a long way off from this at the moment it is something that could be possible in the future”, he concludes.
Today, a vast majority of industries are heavily reliant on technology and it is crucial to the way we conduct business, particularly in the financial services industry. Technology has soared since the start of the 20th century and it is involved in our everyday lives today in the 21st century, with technology only expected to accelerate and evolve at lightning pace.
Looking back at how quickly technology has advanced over the last 200 years, you can only wonder what the world will look like within the next 200 years. Technology is vital in the financial services world, especially within the asset servicing industry, where enhanced technology is key. How do firms prepare for the future when the change is occurring so rapidly?
Within the asset servicing industry, Andreas Burner, chief innovation officer of SmartStream Technologies, observes that there is a large number of people who understand the potential of artificial intelligence (AI) and have started utilising it within their firm. However, Burner says that the industry is still very much at the beginning of the AI era.
A new way to work
Discussing whether AI poses a threat to jobs, Andreas Burner, chief innovation officer of SmartStream Technologies, emphasises that people have always been afraid of redundancy.
Burner says the world is constantly developing, using the example of newspapers moving to digital content, he notes that people were worried that automation would mean loss of jobs. However, he explains “all these things have in fact created more jobs. Certain roles will need to adapt and develop to meet the demands of new technology”.
Echoing this, Alastair McGill, general manager data control solutions at Broadridge, comments that AI is not something we should be scared of but something we should embrace. McGill highlights that we should not assume right away that intelligent automation will lead to massive job losses.
According to McGill, Gartner research indicates that more jobs will be created by AI rather than will be lost to it.
He says: “One of the factors is the adoption and use of AI to augment human tasks rather than to replace them. New types of jobs will grow as new technology is adopted. The analogy used often is the Industrial Revolution – even though agricultural jobs went away, new jobs (and new types of jobs) were created.”
Additionally, McGill stipulates that in a heavily regulated industry like financial services, human oversight of intelligent automation will “remain critical” well into the future. However, McGill suggests we are still many years away from creating the kind of artificial general intelligence that outperforms humans when it comes to critical thinking about relationships between certain types of information.
Reinforcing this, Davide Zilli of Mind Foundry, explains that humans will always have a role to play, “algorithms need to be fully transparent in their decisions, easily validated and monitored by a human expert”.
A treasure trove of technology
Although concerns will remain, the consensus suggests that technology is more abundant in its opportunities than its downfalls, and is seen to be more useful as opposed to threatening. With AI providing the ability to remove manual, mundane processes, the human workforce will have more time to take on creative tasks, enabling greater efficiency in the workflow.
Richard Street, global head of client coverage for RBC Investor & Treasury Services (RBC I&TS), indicates that “AI should accelerate the process of automating manual activities, which will enable the industry to deliver end-to-end solutions more quickly”.
Street comments: “If you look at the private capital and real assets space, deploying capital is a complex activity involving the engagement of lawyers, investment bankers and advisers, which is very different from the automation of execution and settlement in the traditional asset world.”
He explains that there is a role for AI to play in evolving and perfecting this process, more so than focusing on perfecting the mature real money investment management processes.
McGill also weighs in discussing Broadridge’s experience with implementing the latest technology. He says: “Replacing a traditional rules-based matching engine with a self-learning machine learning model can provide significant efficiency benefits by giving the system the ability to automatically build and improve its matching schemes based on new activity.”
McGill adds: “IT projects to onboard or maintain matching schemes are no longer required as the solution analyses historical matched data and suggests a matching scheme that will process new data with very high levels of accuracy. As the training dataset grows, the model improves, keeping straight-through-processing rates high without the need for constant development and testing.”
“Overall performance also improves significantly as machine learning-based solutions take advantage of in-memory distributed computing platforms designed to process huge amounts of data efficiently.”
Reflecting on the opportunities of AI, Street identifies that it enables it to increase the quality of the services that our industry provides, and thereby drive client demand for faster service and better insights based on the high-quality data which can be derived from those automated tools.
Technical issues
Although AI does have plenty of advantages, AI projects can suffer from a “cold start”, which is one of the key challenges. McGill explains that this is where predictions are initially low quality as sufficient training data is not available to generate an accurate model.
McGill explains: “A solution with the ability to generate models from historic training data can mitigate the cold start problem, leading to shorter project timelines and a faster return on investment by providing high-quality predictions from the start.”
According to McGill, self-service and transparency are critically important to the industry as firms are cost sensitive in regard to change processes.
He also states that auditing requirements are of paramount importance. “Many AI-based solutions available to the market operate on a ‘black box’ basis, requiring vendor data science staff to handle the generation of new models or optimisation of existing ones at significant cost”, he adds.
Weighing in on this, Zilli from Mind Foundry, says: “Machine learning tools must introduce this full accountability to evolve beyond unexplainable ‘black box’ solutions and eliminate the easy excuse of ‘the algorithm made me do it’.”
RBC I&TS’ Street also warns that it is important to recognise that AI is not a ‘fix-all’ solution. He stresses that anyone who is building an operation on AI alone has got it wrong.
He adds: “You have to put a process in place and automate it as best you can. AI can evolve that process faster and better because it’s learning and remembering things more efficiently.”
To infinity and beyond
Although technology is evolving at a rapid pace, industry experts suggest that it is moving more slowly in the asset servicing industry. But arguably, never before has technology been so crucial amidst increased velocity of regulation.
McGill suggests that the asset servicing industry “has been under consistent financial pressure for the last five years”. He notes: “Rising asset prices have largely masked underlying declines in assets under custody/assets under administration fees.”
“Restoring profitable growth amidst these pressures will require more innovative solutions to improve operational efficiency. There are a number of areas in which the industry can benefit from the adoption of AI.”
“Leading firms will rapidly innovate using AI which will form the basis of leaner, more efficient operational functions,” he adds.
Also looking to the future of the asset servicing sector, Burner notes that tomorrows’ applications will continuously support the users in their needs and will deliver answers to questions before a user is asking.
Burner explains that AI applications “will provide excellent user interfaces that will deliver condensed and relevant information to the user in the right moment”.
He believes that in a few years, “the financial services industry should operate in a similar function to weather forecasts, which are easy to access and they will be something you can look at daily and is automatically delivered to you in terms of information.”
“Although we are a long way off from this at the moment it is something that could be possible in the future”, he concludes.
NO FEE, NO RISK
100% ON RETURNS If you invest in only one asset servicing news source this year, make sure it is your free subscription to Asset Servicing Times
100% ON RETURNS If you invest in only one asset servicing news source this year, make sure it is your free subscription to Asset Servicing Times
