Automation: past, present and future
31 December 2014
What does securities finance need to automate, and why? Experts discuss the issues
Image: Shutterstock
The move towards automation is often mooted but never completed, and securities finance is no different. Jonathan Lombardo of Pirum raises the issues with Stuart Skeel of BNY Mellon, Kit Newman of Barclays and Andy Krangel of Citi, and asks how far securities finance has progressed in the automation game, and which level it must tackle next.
Jonathan Lombardo: Before the financial crisis in 2008, what were the drivers that led you to identifying automation as a route forward?
Andy Krangel: The driver for automation has been predominantly process improvement/efficiency-based. Until 2008, securities finance was growing exponentially to the extent that participants were looking at doing more with the personnel and systems they had. We had many middle- and back-office people servicing large accounts. But, as the financial crisis showed, that was unsustainable in the long term.
Lots of thinking about automation has come post-2008, when revenues decreased dramatically, very quickly. Before the financial crisis, we were all mainly focused on growing the business. Now, we can’t afford to have numerous people working on manual operational tasks—it’s no longer feasible and we have to be more efficient. We were looking at it before the crisis, but it wasn’t on agendas as much as revenue growth.
Stuart Skeel: The need for automation had been created by the trading desk with a view to increasing volume and revenue. Operations required a level of automation in order to keep up, but it is now becoming fundamental to what the settlement and exposure teams do today.
Much of what we do around international equities and non-cash triparty is run and managed in London, with billing and supporting functions handled by other locations. We have close-knit teams, with only a relatively small number of people managing a large volume of trades and collateral. Our operations have historically been located close to the trading desk, as we feel that to be an efficient way to operate. For instance, it used to be that same-day pre-pay was only really supported by operations as a favour to the desk and for that traders wanted operations located close by.
Automation helped them to keep up with what traders were doing, so operations were looking for the best way to reconcile. Everything flows from reconciliation. Once that is perfected, exposure and billing follow on—it was for that reason that we looked towards automation.
Everything is a risk
Lombardo: With volumes and activity on the rise pre-crisis, scalability was becoming more of an increased burden on day-to-day processing. What steps were taken to identify processes that were lacking efficiency and required automation?
Kit Newman: There would be a squeeze on the function to perform better and then it would be analysed. Costs and benefits would be considered, as would the possible effects on the business of doing it externally or internally.
Krangel: The initial drive was from a process improvement perspective and not doing more with less. I have run operations in the past and there have been processes, such as billing, that have generated a lot of operational challenges. Billing is a tricky area, because delays in reconciling and agreeing bills can result in being months behind collection, which can cause serious problems. I’d say that is still a challenge in securities finance.
Newman: That’s right. Billing wasn’t seen as a risk in the past. Most were focused on the borrower the going bust. That’s changed to the point where all exposures to borrowers are considered risks, including bills outstanding and corporate actions. The longer they are left, the greater the chance that the loan will be returned, leaving the lender uncollaterised.
Lombardo: Realising in-house capacity was still stretched and your client base was continuing to expand. Was an in-house solution considered or was the reach of external vendors the only viable solution?
Skeel: That decision was reached by looking at the best way to access borrowers. Looking at it from that perspective meant that we were far less likely to tackle it in-house. Also, offerings were already out there that could connect to a majority of our counterparties, which made a lot more sense when connecting with so many.
Krangel: I totally agree. Technically, we have multiple borrowers that we can lend to, so there is no way we could ever build a reconciliation structure with all of them that is practical. There is not a single, industry-wide lending system out there.
There are certain things that an agent lender can do for its entire borrower network, such as send out its available inventory in a consistent format. But to actually create a reconciliation process with multiple counterparties, I don’t think we would’ve have cracked that internally given the market system variation.
Connectivity is king
Newman: I think counterparty connectivity is the most important area for the buy side, particularly how the most important information is accessed without having someone to ask for it over the telephone or via email.
Whichever vendor offered the most cost-effective solution and the biggest community of agent lenders was the one to go for. But broker-dealers also considered the fact that multiple vendors were out there, each with different networks using their offerings, and so many would often sign up to lots of services as they came online. Eventually, they would decide which one was most appropriate.
I also think it was about understanding what a vendor can provide, because there wasn’t a general awareness about what was out there. The decision was probably based on existing relationships with certain vendors.
Skeel: There are needs that require links to other participants in the market, so it makes sense to use a vendor because it removes the need to cultivate each link individually.
Newman: Of course, there is also internal infrastructure to consider when looking at vendors. The broker-dealer might have solutions in common with lenders in terms of how currency positions are financed, for example. These components had to be factored in when choosing a vendor.
Lender-led
Krangel: The decision to build in-house or use a vendor system tends to be driven at a firm level. With our lending system, we have an in-house build and have done it that way for some time. We have considered vendors but the lending system is quite specific to the operating model. It also has to link to internal systems, such as custody.
One of the challenges of using a vendor is that generally the system is being built for many organisations, not just for the particular agent lender. If a specific need exists, the agent lender has to justify why it needs what it needs over the requirements of other users, or alternatively pay for that development. In-house lending systems allow greater flexibility and customisation.
I think that a lending system is very different to an interface into the market.
Lombardo: When looking across vendor solutions, how did functionality come into the decision making process?
Skeel: We originally looked at connectivity, then considered functionality. Those were the two key priorities, in that order. Yes, one system may provide access to more counterparties, but as a lender we also thought that we could choose a system that worked better for us and then push for the borrowers to accept it. If they wanted to reconcile with us, and we know they did, then we could say that this system is the one that we want to use. Whether or not they were open to that is difficult to say because they hedged their bets on what they were going to use anyway, but connectivity and functionality certainly made that choice for us.
Lombardo: Is there scope for a global solution?
Krangel: I think that it is beneficial to have more than one provider, both from a pricing and risk perspective. Without competition, there is less incentive to improve or price competitively.
My problem with a global solution is that it is putting all of your eggs into one basket. I would certainly want to consider how long it would take to transition everything onto to the single system before making that decision.
But, ultimately, I think that today market participants are focused on cost control and will look at alternative solutions that are as efficient as long as potential transitions don’t trigger capital costs. It has to be done effectively and wisely.
It’s real, this time
Lombardo: Moving on from vendors and solutions, has the introduction of real-time become a critical component of intra-day risk management? Why or how does real-time become critical to your business?
rangel: Consider central counterparties (CCPs): real-time is critical for them. It’s also important for pre-trade, so that it can be viewed and differences identified before it becomes a real trade. Finally, the ability to calculate required value processing (RQV) in real-time is essential, because trading throughout the day needs to be collateralised on a real-time basis.
Skeel: The information throughout the trading day that enables us to look at how efficient we are makes real-time critical. Collateral optimisation is going to be more and more important going forward and real-time tools are needed to move towards that optimisation. I also think regulations coming out next year will demand more real-time pre-matching information.
Newman: The potential for a trade to fail at 4pm means that counterparties are going to want to be able to borrow replacement securities at the same time, and they will want to plan that in advance. The only way that is going to happen is if a counterparty can match up and collateralise in real-time. The market is moving in this direction.
Lombardo: Firmly believing that knowing where you stand at any point in time during the trading day is crucial. The ability to re-evaluate exposures in multiple intervals, giving firms the ability to reassess risk and reposition, is critical for moving the industry forward.
What next?
Lombardo: Finally, what is the main driver for you in 2015 to increase automation?
Skeel: We need to concentrate on better marks and returns functionality next. That’s a main focus for us.
Krangel: Among other things, we are going to make better use of the mark-to-market process. We are not using that enough and we should be.
Newman: We want to make pre-trade matching more efficient with counterparties across the business.
Jonathan Lombardo: Before the financial crisis in 2008, what were the drivers that led you to identifying automation as a route forward?
Andy Krangel: The driver for automation has been predominantly process improvement/efficiency-based. Until 2008, securities finance was growing exponentially to the extent that participants were looking at doing more with the personnel and systems they had. We had many middle- and back-office people servicing large accounts. But, as the financial crisis showed, that was unsustainable in the long term.
Lots of thinking about automation has come post-2008, when revenues decreased dramatically, very quickly. Before the financial crisis, we were all mainly focused on growing the business. Now, we can’t afford to have numerous people working on manual operational tasks—it’s no longer feasible and we have to be more efficient. We were looking at it before the crisis, but it wasn’t on agendas as much as revenue growth.
Stuart Skeel: The need for automation had been created by the trading desk with a view to increasing volume and revenue. Operations required a level of automation in order to keep up, but it is now becoming fundamental to what the settlement and exposure teams do today.
Much of what we do around international equities and non-cash triparty is run and managed in London, with billing and supporting functions handled by other locations. We have close-knit teams, with only a relatively small number of people managing a large volume of trades and collateral. Our operations have historically been located close to the trading desk, as we feel that to be an efficient way to operate. For instance, it used to be that same-day pre-pay was only really supported by operations as a favour to the desk and for that traders wanted operations located close by.
Automation helped them to keep up with what traders were doing, so operations were looking for the best way to reconcile. Everything flows from reconciliation. Once that is perfected, exposure and billing follow on—it was for that reason that we looked towards automation.
Everything is a risk
Lombardo: With volumes and activity on the rise pre-crisis, scalability was becoming more of an increased burden on day-to-day processing. What steps were taken to identify processes that were lacking efficiency and required automation?
Kit Newman: There would be a squeeze on the function to perform better and then it would be analysed. Costs and benefits would be considered, as would the possible effects on the business of doing it externally or internally.
Krangel: The initial drive was from a process improvement perspective and not doing more with less. I have run operations in the past and there have been processes, such as billing, that have generated a lot of operational challenges. Billing is a tricky area, because delays in reconciling and agreeing bills can result in being months behind collection, which can cause serious problems. I’d say that is still a challenge in securities finance.
Newman: That’s right. Billing wasn’t seen as a risk in the past. Most were focused on the borrower the going bust. That’s changed to the point where all exposures to borrowers are considered risks, including bills outstanding and corporate actions. The longer they are left, the greater the chance that the loan will be returned, leaving the lender uncollaterised.
Lombardo: Realising in-house capacity was still stretched and your client base was continuing to expand. Was an in-house solution considered or was the reach of external vendors the only viable solution?
Skeel: That decision was reached by looking at the best way to access borrowers. Looking at it from that perspective meant that we were far less likely to tackle it in-house. Also, offerings were already out there that could connect to a majority of our counterparties, which made a lot more sense when connecting with so many.
Krangel: I totally agree. Technically, we have multiple borrowers that we can lend to, so there is no way we could ever build a reconciliation structure with all of them that is practical. There is not a single, industry-wide lending system out there.
There are certain things that an agent lender can do for its entire borrower network, such as send out its available inventory in a consistent format. But to actually create a reconciliation process with multiple counterparties, I don’t think we would’ve have cracked that internally given the market system variation.
Connectivity is king
Newman: I think counterparty connectivity is the most important area for the buy side, particularly how the most important information is accessed without having someone to ask for it over the telephone or via email.
Whichever vendor offered the most cost-effective solution and the biggest community of agent lenders was the one to go for. But broker-dealers also considered the fact that multiple vendors were out there, each with different networks using their offerings, and so many would often sign up to lots of services as they came online. Eventually, they would decide which one was most appropriate.
I also think it was about understanding what a vendor can provide, because there wasn’t a general awareness about what was out there. The decision was probably based on existing relationships with certain vendors.
Skeel: There are needs that require links to other participants in the market, so it makes sense to use a vendor because it removes the need to cultivate each link individually.
Newman: Of course, there is also internal infrastructure to consider when looking at vendors. The broker-dealer might have solutions in common with lenders in terms of how currency positions are financed, for example. These components had to be factored in when choosing a vendor.
Lender-led
Krangel: The decision to build in-house or use a vendor system tends to be driven at a firm level. With our lending system, we have an in-house build and have done it that way for some time. We have considered vendors but the lending system is quite specific to the operating model. It also has to link to internal systems, such as custody.
One of the challenges of using a vendor is that generally the system is being built for many organisations, not just for the particular agent lender. If a specific need exists, the agent lender has to justify why it needs what it needs over the requirements of other users, or alternatively pay for that development. In-house lending systems allow greater flexibility and customisation.
I think that a lending system is very different to an interface into the market.
Lombardo: When looking across vendor solutions, how did functionality come into the decision making process?
Skeel: We originally looked at connectivity, then considered functionality. Those were the two key priorities, in that order. Yes, one system may provide access to more counterparties, but as a lender we also thought that we could choose a system that worked better for us and then push for the borrowers to accept it. If they wanted to reconcile with us, and we know they did, then we could say that this system is the one that we want to use. Whether or not they were open to that is difficult to say because they hedged their bets on what they were going to use anyway, but connectivity and functionality certainly made that choice for us.
Lombardo: Is there scope for a global solution?
Krangel: I think that it is beneficial to have more than one provider, both from a pricing and risk perspective. Without competition, there is less incentive to improve or price competitively.
My problem with a global solution is that it is putting all of your eggs into one basket. I would certainly want to consider how long it would take to transition everything onto to the single system before making that decision.
But, ultimately, I think that today market participants are focused on cost control and will look at alternative solutions that are as efficient as long as potential transitions don’t trigger capital costs. It has to be done effectively and wisely.
It’s real, this time
Lombardo: Moving on from vendors and solutions, has the introduction of real-time become a critical component of intra-day risk management? Why or how does real-time become critical to your business?
rangel: Consider central counterparties (CCPs): real-time is critical for them. It’s also important for pre-trade, so that it can be viewed and differences identified before it becomes a real trade. Finally, the ability to calculate required value processing (RQV) in real-time is essential, because trading throughout the day needs to be collateralised on a real-time basis.
Skeel: The information throughout the trading day that enables us to look at how efficient we are makes real-time critical. Collateral optimisation is going to be more and more important going forward and real-time tools are needed to move towards that optimisation. I also think regulations coming out next year will demand more real-time pre-matching information.
Newman: The potential for a trade to fail at 4pm means that counterparties are going to want to be able to borrow replacement securities at the same time, and they will want to plan that in advance. The only way that is going to happen is if a counterparty can match up and collateralise in real-time. The market is moving in this direction.
Lombardo: Firmly believing that knowing where you stand at any point in time during the trading day is crucial. The ability to re-evaluate exposures in multiple intervals, giving firms the ability to reassess risk and reposition, is critical for moving the industry forward.
What next?
Lombardo: Finally, what is the main driver for you in 2015 to increase automation?
Skeel: We need to concentrate on better marks and returns functionality next. That’s a main focus for us.
Krangel: Among other things, we are going to make better use of the mark-to-market process. We are not using that enough and we should be.
Newman: We want to make pre-trade matching more efficient with counterparties across the business.
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 Íø±¬³Ô¹Ï 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 Íø±¬³Ô¹Ï Finance Times
