Spoilt for choice
20 August 2019
With enticing new technologies such as AI and electronic trading coming to the fore, industry experts discuss how technologies are changing the dynamic of the industry and what risks and opportunities they present
Image: Shutterstock
鈥楾echnology鈥� is often touted as the ambiguous answer to the many challenges and ambitions of the securities lending industry, but consensus on which new innovation should be prioritised for which problem is hard to come by.
Artificial intelligence (AI), blockchain, transaction automation, machine learning and real-time stock streaming capabilities are just some of the new tools on offer that come with promises to increase efficiency like never before and lower costs in the long-term. But, is a hefty investment in emerging technology really the antidote to all the industry鈥檚 woes or just a glitzy and expensive distraction?
Today, many see fintech and regtech in one form or another to be the ultimate solution to the challenges posed by the incoming 麻豆传媒 Financing Transactions Regulation (SFTR) and other regulatory frameworks soon to shake-up the industry.
At the same time, discussions around the rise of electronic trading, the leap to blockchain technology and the integration of AI into decision making are taking centre stage at industry events and boardroom discussions all over the world.
However, while some are hailing the rise of the machines, others are warning to keep the new risks and challenges associated with technology in sharp focus moving forward.
What鈥檚 on offer?
Industry experts suggest that electronic trading and the increasing usage of data is enabling better pricing decisions and empowering the industry in the deployment of proprietary trading algorithms.
It has also been predicted that there will be a migration towards higher proportions of trading occurring over electronic venues to allow for a better allocation of resources and to increase standardisation.
As well as benefiting from electronic trading and AI, the industry can also benefit from automation. Paul Lynch, global head of product at EquiLend, a securities lending platform, says that automation can reduce the risk of error-prone manual processes and can allow firms to focus intellectual capital on client engagement, performance drivers and risk mitigation.
Similarly, Martin Seagroatt, global marketing director for securities finance and collateral management at Broadridge, says automation can also facilitate growth in trading volumes with a reduction in operational risk.
Seagroatt sees new tech bringing lower costs across the trade lifecycle as the main benefit of automation, along with freeing up time for staff to focus on decision-making and reallocation of personnel to more productive tasks.
鈥淟arger buy-side institutions are showing more interest in bringing some of their lending activities in house,鈥� he explains. 鈥淎utomation could reduce the middle- and back-office costs of running in-house desks and make insourcing less expensive. Finally, greater automation can facilitate innovation and the ability to launch new products and business lines more rapidly.鈥�
Need versus want
Considering the digital smorgasbord of tools on offer, it鈥檚 unsuprising that the various demographics of the market are looking to focus on different applications for their IT investments.
Discussing the technologies that are expected to take centre stage for banks in the securities finance industry over the next 12 to 18 months, Guido Stroemer, founding partner at HQLA?, says that their solutions should centre around improving balance sheet management.
鈥淏ank treasurers need real-time balance sheet reporting to feed optimisation engines which in turn feed into execution platforms. I believe blockchain and AI technologies can help the modern-day bank treasurer get the right financial resource to the right place at the right time,鈥� he explains.
Elsewhere, Lynch says that his clients have expressed a keen interest in screen-based collateral trading and management, screen-based swaps optimisation, additional post-trade automation tools, including AI, and expanded their usage of central counterparty trading facilities for securities lending.
However, although industry participants aspire to be the first to apply HAL 9000 to securities lending, the practical demands of the day are forcing many to settle for more modest additions to their systems.
Seagroatt says that the uptake of these new technologies such as AI, blockchain, cloud and digital or data in securities finance has been gradual. He suggests this is mainly because many IT budgets are currently focused on adapting to regulatory change resulting from SFTR, Central 麻豆传媒 Depositories Regulation (CSDR), Brexit and the uncleared margin reform rules.
Seagroatt predicts that the adoption of cloud technologies will continue to grow due to greater understanding of the security and data protection, as well as the benefits around flexibility and scalability that the cloud can offer. He also expects to see more firms apply robotic process automation to the securities finance lifecycle due to its relative simplicity to deploy versus more advanced AI techniques.
The need to address the reality of the market鈥檚 current challenges before getting lost in space-age technology was also raised at the 麻豆传媒 Finance Technology Symposium in May by David Lewis, senior director at FIS. Lewis said at the time that, as a result of SFTR, lots of people will have to look at the nuts and bolts of their trading systems and behaviours and potentially need to change the way they鈥檙e booking things to make sure they鈥檙e getting it right.
鈥淚t can be tedious to some, and we鈥檇 perhaps rather talk about AI, machine learning and of the other shiny bells and whistles, but you need to get the base technology and data right first,鈥� he said. 鈥淲e are not going to get the most from automation and the efficiency it can bring until we get the basics right. There is lots of interesting data out there, which gets our attention, brings innovation and drives competition. We talk about AI and machine learning but until we strengthen the foundations of our basic data, we won鈥檛 see transformational change.鈥�
Making lemonade
For the optimists of the industry, the incoming regulations also present an opportunity to improve. 鈥淲hile SFTR and CSDR are currently taking up a great deal of resources, they are also enablers of technological change in the future. SFTR in particular is driving the standardisation of processes. This lays the foundation for greater automation and could facilitate the use of smart contracts over a distributed ledger,鈥� Seagroatt says.
Moreover, Seagroatt believes firms could start to leverage these datasets, along with other alternative data, to guide decision making.
鈥淪FTR and the common domain model being discussed by the International 麻豆传媒 Lending Association will enable greater standardisation. This will lead to higher volumes of electronic trading - which in turn provides a foundation for greater automation, algorithmic trading and the use of AI.鈥�
On the topic of AI, Seagroatt predicts that there are areas of front-office decision-making where AI could process large volumes of data in a very short space of time.
He describes how data crunching could feed into PC-based cognitive assistants and decision support tools. 鈥淭hese solutions would then enable robot-assisted traders to take advantage of short-lived trading opportunities before the wider market can react. 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,鈥� he says.
Meanwhile, Lynch believes that the usage of data, and the sophistication of how that data is used, will only increase with time. He says: 鈥淭he industry has become 鈥榮uper users鈥� of data. Many of our clients are buying data at a record level and consuming it daily.鈥�
鈥淚t is being used for trading, client and firm performance benchmarking and daily quantitative analysis. In addition, different types of firms and different departments within existing client firms are consuming securities finance data鈥攆or example, beneficial owners and portfolio managers.鈥�
The honey is sweet but the bee has a sting
Although there is much industry fanfare around the possibilities offered by emerging technologies, new roads bring new dangers.
Seagroatt warns that many industry experts and quants are also deeply sceptical about the ability of AI to ever predict financial market movements. 鈥淎I can help to increase the speed and accuracy of predicting the probability of events occurring by processing larger volumes of data faster than a human ever could,鈥� he says. 鈥淗owever, it would be a mistake to rely too heavily on AI for things like predicting risk. AI is also not good at predicting black swan events, tail risk, and the impact of events with no historic precedent such as Brexit.鈥�
Seagroatt also highlights that AI and quantum computing could allow hackers to crack encryption keys, although these technologies can also be used to create more complex forms of encryption. Firms therefore need a robust cybersecurity framework to monitor and mitigate threats as they evolve, he says.
Beyond the risk of quantum hackers, more familiar concerns and limitations also remain. HQLA?鈥檚 Stroemer notes that it鈥檚 important for technology solutions to be able to interoperate with one another, otherwise it will create silos which may cause inefficiencies and impede market adoption. Elsewhere, model risk is a major issue, where a machine learning solution can be difficult to govern due to a lack of transparency, according to Seagroatt. 鈥淕reater complexity means fewer people understand what the AI is actually doing,鈥� he says. 鈥淭here are difficulties bridging the gap between the work data scientists are doing and staff who have deep domain knowledge of industry dynamics or the trade lifecycle.鈥�
Risk management
Where there鈥檚 risk, regulators are sure to be close by and Seagroatt believes that they are increasingly concerned about the shifting of risk from the banking sector to the fintech sector, which is less regulated.
鈥淚t is critical that market participants perform sufficient due diligence to ensure fintech providers have mature security procedures in place. There is also an increasing reliance on a few big cloud computing or AI providers, who could be classed as systemically important/too big to fail. There are other concentration risks where firms could all use similar AI algorithms or datasets to guide trading, resulting in herd effects,鈥� Seagroatt says.
He concludes: 鈥淭hese are some of the known knowns. However, there are likely to be many unknown unknowns and technology-driven risks we haven鈥檛 even thought about yet. Greater interconnectivity in the market could increase the severity and magnitude of problems with contagion effects occurring very rapidly and unexpectedly. A common global regulatory framework around areas such as AI would greatly help with identifying and mitigating some of these risks.鈥�
Artificial intelligence (AI), blockchain, transaction automation, machine learning and real-time stock streaming capabilities are just some of the new tools on offer that come with promises to increase efficiency like never before and lower costs in the long-term. But, is a hefty investment in emerging technology really the antidote to all the industry鈥檚 woes or just a glitzy and expensive distraction?
Today, many see fintech and regtech in one form or another to be the ultimate solution to the challenges posed by the incoming 麻豆传媒 Financing Transactions Regulation (SFTR) and other regulatory frameworks soon to shake-up the industry.
At the same time, discussions around the rise of electronic trading, the leap to blockchain technology and the integration of AI into decision making are taking centre stage at industry events and boardroom discussions all over the world.
However, while some are hailing the rise of the machines, others are warning to keep the new risks and challenges associated with technology in sharp focus moving forward.
What鈥檚 on offer?
Industry experts suggest that electronic trading and the increasing usage of data is enabling better pricing decisions and empowering the industry in the deployment of proprietary trading algorithms.
It has also been predicted that there will be a migration towards higher proportions of trading occurring over electronic venues to allow for a better allocation of resources and to increase standardisation.
As well as benefiting from electronic trading and AI, the industry can also benefit from automation. Paul Lynch, global head of product at EquiLend, a securities lending platform, says that automation can reduce the risk of error-prone manual processes and can allow firms to focus intellectual capital on client engagement, performance drivers and risk mitigation.
Similarly, Martin Seagroatt, global marketing director for securities finance and collateral management at Broadridge, says automation can also facilitate growth in trading volumes with a reduction in operational risk.
Seagroatt sees new tech bringing lower costs across the trade lifecycle as the main benefit of automation, along with freeing up time for staff to focus on decision-making and reallocation of personnel to more productive tasks.
鈥淟arger buy-side institutions are showing more interest in bringing some of their lending activities in house,鈥� he explains. 鈥淎utomation could reduce the middle- and back-office costs of running in-house desks and make insourcing less expensive. Finally, greater automation can facilitate innovation and the ability to launch new products and business lines more rapidly.鈥�
Need versus want
Considering the digital smorgasbord of tools on offer, it鈥檚 unsuprising that the various demographics of the market are looking to focus on different applications for their IT investments.
Discussing the technologies that are expected to take centre stage for banks in the securities finance industry over the next 12 to 18 months, Guido Stroemer, founding partner at HQLA?, says that their solutions should centre around improving balance sheet management.
鈥淏ank treasurers need real-time balance sheet reporting to feed optimisation engines which in turn feed into execution platforms. I believe blockchain and AI technologies can help the modern-day bank treasurer get the right financial resource to the right place at the right time,鈥� he explains.
Elsewhere, Lynch says that his clients have expressed a keen interest in screen-based collateral trading and management, screen-based swaps optimisation, additional post-trade automation tools, including AI, and expanded their usage of central counterparty trading facilities for securities lending.
However, although industry participants aspire to be the first to apply HAL 9000 to securities lending, the practical demands of the day are forcing many to settle for more modest additions to their systems.
Seagroatt says that the uptake of these new technologies such as AI, blockchain, cloud and digital or data in securities finance has been gradual. He suggests this is mainly because many IT budgets are currently focused on adapting to regulatory change resulting from SFTR, Central 麻豆传媒 Depositories Regulation (CSDR), Brexit and the uncleared margin reform rules.
Seagroatt predicts that the adoption of cloud technologies will continue to grow due to greater understanding of the security and data protection, as well as the benefits around flexibility and scalability that the cloud can offer. He also expects to see more firms apply robotic process automation to the securities finance lifecycle due to its relative simplicity to deploy versus more advanced AI techniques.
The need to address the reality of the market鈥檚 current challenges before getting lost in space-age technology was also raised at the 麻豆传媒 Finance Technology Symposium in May by David Lewis, senior director at FIS. Lewis said at the time that, as a result of SFTR, lots of people will have to look at the nuts and bolts of their trading systems and behaviours and potentially need to change the way they鈥檙e booking things to make sure they鈥檙e getting it right.
鈥淚t can be tedious to some, and we鈥檇 perhaps rather talk about AI, machine learning and of the other shiny bells and whistles, but you need to get the base technology and data right first,鈥� he said. 鈥淲e are not going to get the most from automation and the efficiency it can bring until we get the basics right. There is lots of interesting data out there, which gets our attention, brings innovation and drives competition. We talk about AI and machine learning but until we strengthen the foundations of our basic data, we won鈥檛 see transformational change.鈥�
Making lemonade
For the optimists of the industry, the incoming regulations also present an opportunity to improve. 鈥淲hile SFTR and CSDR are currently taking up a great deal of resources, they are also enablers of technological change in the future. SFTR in particular is driving the standardisation of processes. This lays the foundation for greater automation and could facilitate the use of smart contracts over a distributed ledger,鈥� Seagroatt says.
Moreover, Seagroatt believes firms could start to leverage these datasets, along with other alternative data, to guide decision making.
鈥淪FTR and the common domain model being discussed by the International 麻豆传媒 Lending Association will enable greater standardisation. This will lead to higher volumes of electronic trading - which in turn provides a foundation for greater automation, algorithmic trading and the use of AI.鈥�
On the topic of AI, Seagroatt predicts that there are areas of front-office decision-making where AI could process large volumes of data in a very short space of time.
He describes how data crunching could feed into PC-based cognitive assistants and decision support tools. 鈥淭hese solutions would then enable robot-assisted traders to take advantage of short-lived trading opportunities before the wider market can react. 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,鈥� he says.
Meanwhile, Lynch believes that the usage of data, and the sophistication of how that data is used, will only increase with time. He says: 鈥淭he industry has become 鈥榮uper users鈥� of data. Many of our clients are buying data at a record level and consuming it daily.鈥�
鈥淚t is being used for trading, client and firm performance benchmarking and daily quantitative analysis. In addition, different types of firms and different departments within existing client firms are consuming securities finance data鈥攆or example, beneficial owners and portfolio managers.鈥�
The honey is sweet but the bee has a sting
Although there is much industry fanfare around the possibilities offered by emerging technologies, new roads bring new dangers.
Seagroatt warns that many industry experts and quants are also deeply sceptical about the ability of AI to ever predict financial market movements. 鈥淎I can help to increase the speed and accuracy of predicting the probability of events occurring by processing larger volumes of data faster than a human ever could,鈥� he says. 鈥淗owever, it would be a mistake to rely too heavily on AI for things like predicting risk. AI is also not good at predicting black swan events, tail risk, and the impact of events with no historic precedent such as Brexit.鈥�
Seagroatt also highlights that AI and quantum computing could allow hackers to crack encryption keys, although these technologies can also be used to create more complex forms of encryption. Firms therefore need a robust cybersecurity framework to monitor and mitigate threats as they evolve, he says.
Beyond the risk of quantum hackers, more familiar concerns and limitations also remain. HQLA?鈥檚 Stroemer notes that it鈥檚 important for technology solutions to be able to interoperate with one another, otherwise it will create silos which may cause inefficiencies and impede market adoption. Elsewhere, model risk is a major issue, where a machine learning solution can be difficult to govern due to a lack of transparency, according to Seagroatt. 鈥淕reater complexity means fewer people understand what the AI is actually doing,鈥� he says. 鈥淭here are difficulties bridging the gap between the work data scientists are doing and staff who have deep domain knowledge of industry dynamics or the trade lifecycle.鈥�
Risk management
Where there鈥檚 risk, regulators are sure to be close by and Seagroatt believes that they are increasingly concerned about the shifting of risk from the banking sector to the fintech sector, which is less regulated.
鈥淚t is critical that market participants perform sufficient due diligence to ensure fintech providers have mature security procedures in place. There is also an increasing reliance on a few big cloud computing or AI providers, who could be classed as systemically important/too big to fail. There are other concentration risks where firms could all use similar AI algorithms or datasets to guide trading, resulting in herd effects,鈥� Seagroatt says.
He concludes: 鈥淭hese are some of the known knowns. However, there are likely to be many unknown unknowns and technology-driven risks we haven鈥檛 even thought about yet. Greater interconnectivity in the market could increase the severity and magnitude of problems with contagion effects occurring very rapidly and unexpectedly. A common global regulatory framework around areas such as AI would greatly help with identifying and mitigating some of these risks.鈥�
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 麻豆传媒 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 麻豆传媒 Finance Times
