Time for take-off?
29 October 2019
Perceptions of data and market ‘transparency’ in the securities finance industry has shifted dramatically over the years from a nice-to-have to essential rocket fuel for space age artificial intelligence and automation technologies
Image: Shutterstock
In years gone by, the collection of data in the securities finance industry was treated by some with suspicion and as a potential threat to revenue by bringing unwanted price transparency to a market that’s value could be linked to its opacity. Nowadays, however, mostly thanks to the rise of new technologies that require huge quantities of data to operate, those previously sceptical of the need for more data have changed their tune.
Growing buy-side demands quantify the their lending/borrowing programmes has in turn brought the sell-side round to the idea that the creation and sharing of data is here to stay. As well as this, people are seeing the art of creating, capturing and utilising data as a top priority as it can open up all kinds of revenue opportunities and empower decision making. The opportunities afforded by increasing the utilisation of relevant data, according to Broadridge’s marketing director, securities finance and collateral management, Martin Seagroatt, include trade pricing and electronic market aggregation, global inventory management and collateral optimisation through to post trade analytics. “Clear, centralised data can help traders take advantage of short-lived trading opportunities and allows risk managers to gain a more holistic view of risk,” he explains.
Additionally, Kabin George, global head of product management for securities finance at IHS Markit, one of the securities finance industry’s major data providers, says that data can be used to measure securities lending performance, comparing a lender or borrower against the market average. It is this comparison against industry peers that many in the industry once wanted to avoid but now the benefits of wider market transparency are being realised. For the individual trader, data transparency leads to more effective control of a lending programme through features such as post-trade benchmarking, as well as the ability to find additional alpha through accurate pricing of securities.
In terms of technology powered by data, Tim Smith, managing director of business and development at Hazeltree, says that data combined with technology solutions can empower market participants to get full transparency into their operations and get better informed decision to get the best-execution financing rates. “As the amounts of and analytics surrounding data increase so too does the requirement to find ways to use that data efficiently and effectively,’’ Smith explains.
Today, increased use of data is being embraced across the entire spectrum of market participants. For example, beneficial owners are looking to benchmark and make effective use of data. Despite securities lending only accounting for a small section of most beneficial owners’ investment strategy, a recent beneficial owner survey by DataLend conducted in conjunction with Funds Europe, found that 81 percent of beneficial owners are using securities lending data.
Harvesting the juicy fruit
A significant driver for the increasing demand for data is due to the securities finance desire to embrace artificial intelligence (AI) and machine learning (ML) to automate as much of their day-to-day processes as possible in order to focus on the more lucrative, and complex, trade opportunities; and these tools are insatiably hungry for data in order to operate as required.
Data is the juicy fruit that people want to sink their teeth into but harvesting the whole data field is simply not enough. Firms must utilise data correctly and pick up upon the ‘right’ data in order to reap the rewards. And while technology can be the key to unlocking success, AI and ML are only as good as the data fed into it.
Beyond the practical restrictions of cost and build-time, the biggest barrier to industry players leveraging these space-age technologies is that market data still suffers from a chronic problem of fragmentation, and a lack of standardisation; which is essential for AIs and ML solutions to process and correctly read the data they are fueled by. Simply, tackling the problem of fragmentation is crucial to harnessing the advantages of data - the data must be the same and it must be correct.
Alexandra Foster, director insurance, wealth management, financial services at BT, emphasises the importance of getting the data right and outlines that first and foremost, all of the different pieces of data must come together to form a single view. “Once that has been achieved, you can then start to use robotic process automation/AI to look for actionable insights from that data. When the data is fragmented, it is very hard to overlay anything on top of that to be able to get insight,” she says.
Foster explains how BT is focused on internally mapping out where organisations are taking data from and helping them achieve a consolidated view of this data through BT’s Radianz cloud, to try and help firms overcome data fragmentation.
Shiny bells and whistles
Removing data silos across organisations and standardising data across the entire securities lending industry will be a key focus for many market participants going forward. To this end, the International Securities Lending Association has gathered the market’s biggest data providers into a working group to tackle the issue. Moreover, incoming regulations such as the Securities Financing Transactions Regulation, which will demand huge quantities of transaction data to be reported from April next year, is also serving to keep people’s minds on the task at hand. But, regulatory requirements aside, why should firms on a tight budget, commit itself to an emerging technology that is yet to truly prove its worth? What can AI offer the securities finance market today or in the years ahead?
According to Seagroatt, AI can improve matching algorithms for trading in more illiquid markets and there are areas of front office decision-making where AI could process large volumes of data much quicker than is currently possible.
“These solutions would then enable robot-assisted traders to take advantage of short-lived trading opportunities before the wider market can react,” he explains. “A future AI-driven system could also suggest sophisticated integrated transaction combinations to meet trading goals that are more complex than a basic single trade approach. AI can be used in a wide variety of areas and we’ve seen some announcements around use cases in areas such as trade pricing. There are also applications for predictive analytics to help with settlement fails or reconciliations,” he adds.
Over at EquiLend, Nancy Allen, global product owner, DataLend, comments: “ML and AI are very real for us. We are actively working on using ML and AI across the lifecycle of the loan and have recently undergone a proof of concept with clients in our Post-Trade Suite.”
From a purely data perspective, Allen explains that EquiLend are working on developing more predictive analytics and have recently brought on board a dedicated data science team.
The pie in the sky
Indeed, the opportunities that can sprout from these technologies are plentiful, but is technology still just the pie in the sky in this respect, and is the broader industry ready for AI and ML?
Some industry experts say that there are simpler solutions have out there for accessing these data opportunities for the short term. George highlights that firms are always looking for solutions to identify unrealised revenue and new opportunities, and while AI and ML are notable buzzwords, they have not yet delivered concrete results.
Mark Steadman European head of product development and change management at DTCC, reinforces this. “We are still in an early adoption phase with a lot of new technologies,” he argues.
“Nevertheless, while technologies such as AI and ML hold a lot of promise, proven technology exists today that can enable more efficient and high-quality data processing in the short term.”
Broadridge’s Seagroatt also notes that a lot of things can be analysed using simple statistical analysis like regression, such as looking into the probability that a trade will fail based on certain characteristics. “AI does not always need to be employed,” he explains.
Seagroatt recommends firms to create the fundamental building blocks of their data strategy before trying to launch AI solutions, and explains that without an accurate data model, it is difficult to implement AI solutions.
Seemingly, the ideas of what the technology can do are there but, for securities lending at least, it’s the rocket that hasn’t really taken off … yet
Growing buy-side demands quantify the their lending/borrowing programmes has in turn brought the sell-side round to the idea that the creation and sharing of data is here to stay. As well as this, people are seeing the art of creating, capturing and utilising data as a top priority as it can open up all kinds of revenue opportunities and empower decision making. The opportunities afforded by increasing the utilisation of relevant data, according to Broadridge’s marketing director, securities finance and collateral management, Martin Seagroatt, include trade pricing and electronic market aggregation, global inventory management and collateral optimisation through to post trade analytics. “Clear, centralised data can help traders take advantage of short-lived trading opportunities and allows risk managers to gain a more holistic view of risk,” he explains.
Additionally, Kabin George, global head of product management for securities finance at IHS Markit, one of the securities finance industry’s major data providers, says that data can be used to measure securities lending performance, comparing a lender or borrower against the market average. It is this comparison against industry peers that many in the industry once wanted to avoid but now the benefits of wider market transparency are being realised. For the individual trader, data transparency leads to more effective control of a lending programme through features such as post-trade benchmarking, as well as the ability to find additional alpha through accurate pricing of securities.
In terms of technology powered by data, Tim Smith, managing director of business and development at Hazeltree, says that data combined with technology solutions can empower market participants to get full transparency into their operations and get better informed decision to get the best-execution financing rates. “As the amounts of and analytics surrounding data increase so too does the requirement to find ways to use that data efficiently and effectively,’’ Smith explains.
Today, increased use of data is being embraced across the entire spectrum of market participants. For example, beneficial owners are looking to benchmark and make effective use of data. Despite securities lending only accounting for a small section of most beneficial owners’ investment strategy, a recent beneficial owner survey by DataLend conducted in conjunction with Funds Europe, found that 81 percent of beneficial owners are using securities lending data.
Harvesting the juicy fruit
A significant driver for the increasing demand for data is due to the securities finance desire to embrace artificial intelligence (AI) and machine learning (ML) to automate as much of their day-to-day processes as possible in order to focus on the more lucrative, and complex, trade opportunities; and these tools are insatiably hungry for data in order to operate as required.
Data is the juicy fruit that people want to sink their teeth into but harvesting the whole data field is simply not enough. Firms must utilise data correctly and pick up upon the ‘right’ data in order to reap the rewards. And while technology can be the key to unlocking success, AI and ML are only as good as the data fed into it.
Beyond the practical restrictions of cost and build-time, the biggest barrier to industry players leveraging these space-age technologies is that market data still suffers from a chronic problem of fragmentation, and a lack of standardisation; which is essential for AIs and ML solutions to process and correctly read the data they are fueled by. Simply, tackling the problem of fragmentation is crucial to harnessing the advantages of data - the data must be the same and it must be correct.
Alexandra Foster, director insurance, wealth management, financial services at BT, emphasises the importance of getting the data right and outlines that first and foremost, all of the different pieces of data must come together to form a single view. “Once that has been achieved, you can then start to use robotic process automation/AI to look for actionable insights from that data. When the data is fragmented, it is very hard to overlay anything on top of that to be able to get insight,” she says.
Foster explains how BT is focused on internally mapping out where organisations are taking data from and helping them achieve a consolidated view of this data through BT’s Radianz cloud, to try and help firms overcome data fragmentation.
Shiny bells and whistles
Removing data silos across organisations and standardising data across the entire securities lending industry will be a key focus for many market participants going forward. To this end, the International Securities Lending Association has gathered the market’s biggest data providers into a working group to tackle the issue. Moreover, incoming regulations such as the Securities Financing Transactions Regulation, which will demand huge quantities of transaction data to be reported from April next year, is also serving to keep people’s minds on the task at hand. But, regulatory requirements aside, why should firms on a tight budget, commit itself to an emerging technology that is yet to truly prove its worth? What can AI offer the securities finance market today or in the years ahead?
According to Seagroatt, AI can improve matching algorithms for trading in more illiquid markets and there are areas of front office decision-making where AI could process large volumes of data much quicker than is currently possible.
“These solutions would then enable robot-assisted traders to take advantage of short-lived trading opportunities before the wider market can react,” he explains. “A future AI-driven system could also suggest sophisticated integrated transaction combinations to meet trading goals that are more complex than a basic single trade approach. AI can be used in a wide variety of areas and we’ve seen some announcements around use cases in areas such as trade pricing. There are also applications for predictive analytics to help with settlement fails or reconciliations,” he adds.
Over at EquiLend, Nancy Allen, global product owner, DataLend, comments: “ML and AI are very real for us. We are actively working on using ML and AI across the lifecycle of the loan and have recently undergone a proof of concept with clients in our Post-Trade Suite.”
From a purely data perspective, Allen explains that EquiLend are working on developing more predictive analytics and have recently brought on board a dedicated data science team.
The pie in the sky
Indeed, the opportunities that can sprout from these technologies are plentiful, but is technology still just the pie in the sky in this respect, and is the broader industry ready for AI and ML?
Some industry experts say that there are simpler solutions have out there for accessing these data opportunities for the short term. George highlights that firms are always looking for solutions to identify unrealised revenue and new opportunities, and while AI and ML are notable buzzwords, they have not yet delivered concrete results.
Mark Steadman European head of product development and change management at DTCC, reinforces this. “We are still in an early adoption phase with a lot of new technologies,” he argues.
“Nevertheless, while technologies such as AI and ML hold a lot of promise, proven technology exists today that can enable more efficient and high-quality data processing in the short term.”
Broadridge’s Seagroatt also notes that a lot of things can be analysed using simple statistical analysis like regression, such as looking into the probability that a trade will fail based on certain characteristics. “AI does not always need to be employed,” he explains.
Seagroatt recommends firms to create the fundamental building blocks of their data strategy before trying to launch AI solutions, and explains that without an accurate data model, it is difficult to implement AI solutions.
Seemingly, the ideas of what the technology can do are there but, for securities lending at least, it’s the rocket that hasn’t really taken off … yet
NO FEE, NO RISK
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities Finance Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities Finance Times
