Be the best
04 November 2014
CloudMargin’s Andy Davies and Stuart McHardy share collateral management
oversight best practices to keep beneficial owners informed in a fragmented world
Image: Shutterstock
The past few years have seen unprecedented change in the world of collateral management: the rise of central clearing for securities lending; interoperability between security depositories enabling greater mobility of collateral; regulation driven revolution with OTC derivatives; and the list goes on.
For most beneficial owners, the majority of collateral management has been outsourced to lending agents, service providers and triparty agents. Unless the indemnity offered is watertight in every case, the beneficial owner still needs to maintain visibility over the process for good governance.
Whether it’s management information systems reporting requirements from senior management, questions needing immediate answers from colleagues in risk management, audit, compliance or credit, or regulatory reporting concerns, having accurate, real-time visibility of every position and analysis across every business line is essential.
For the largest banks, brokers and central counterparties (CCPs), throwing money at the problem to build or buy technology solutions costing millions is a viable solution. For beneficial owners such as asset managers, insurance firms and pension schemes, as well as the traditional buy side, this approach is just not an option.
Adopting best practice for beneficial owners wanting oversight of collateral management consists of asking three key questions:
Collation: can you collate data from various sources into one place, giving a complete picture of collateral activity across all products?
Validation: can you validate the data you have? Is the collateral eligible and correctly priced, are your positions correctly valued?
Visualisation: can you ask questions of the data to see how your business is running? Can you make sense of the numbers to make the right decisions?
In an increasingly fragmented world, this goal of best practice can be a lot harder to achieve than it seems.
Collation of data
Lending agents, CCPs, brokers and service providers typically have great client reporting portals and provide real-time data to their clients but there is no consistency.
In a commonly seen scenario, an asset manager uses its global custodian as agent lender for its securities lending programme, manages cash via triparty repos directly with one of the main agents, maintains margin accounts with numerous brokers for futures and options trading, and trades directly with the large banks to hedge via OTC derivatives. The number of different external portals that the firm has to deal with, combined with reporting or extracts from internal systems used for directly collateralised products, becomes large.
Even assuming that all of the reporting is in the same currency, a big assumption, the formats between different portals (even between different products with the same counterparty) are rarely similar. It’s not a quick process to combine this data so the process to get a holistic view of all collateral in a single place is usually given to a junior member of staff armed with spreadsheets.
While spreadsheets offer amazing flexibility and are quick to implement, re-keying or the cut-and-paste of data is inefficient and error prone. Even in the best-run firms, this is unlikely to be performed more than once a day, usually looking only at yesterday’s closing balances.
This approach precludes real-time analysis and in a time of stress, such as a looming counterparty default, trying to do it quickly only increases the risks of error.
Many firms we speak to have tried automating their spreadsheets via macros and while this has some benefits, it introduces new risks of supporting this often-unstable new process. Frequently, the developers of these solutions are not always around to maintain them, having changed roles or moved firms leaving a reliance on unsupported technology to run a critical process. Even if the developer is still available, the internal costs of maintenance and keeping up to date with report and regulatory changes can be prohibitive.
In a best practice environment, this process requires dedicated technology to ensure completeness and accuracy of data received, the ability to process real-time updates and deliver operational efficiency.
Fortunately in 2014 this technology is widely available off the shelf from vendors and is very affordable. It shouldn’t cost more than the headcount that would otherwise be required to cut and paste the data.
Any reputable vendor should be able to receive this information directly from the source, without the beneficial owner having to be involved in downloading or reformatting the data themselves.
The best practice for collation of data should be a fully automated, intra-day, behind the scenes process with little (if any) human involvement.
Validation
Once the data has been collated, any firm seeking best practice should be looking to validate the information received. Has collateral been priced correctly? Is the collateral pledged to me eligible? Are my positions correctly valued? This is all key to the decisions that the business has to make in the next phase of the process.
At a very basic level, every external party will have separate market-data sources for pricing non-cash collateral and be using different FX rates within calculations. A bad price or erroneous FX rate will lead to over- or under-collateralisation, so it’s prudent to revalue all collateral using a single, internally approved pricing source with a single set of FX rates to identify issues with the reporting received.
A similar set of controls should be applied to testing eligibility of collateral, whether received directly, via an agent lender or allocated by a triparty agent. Any errors with codifying eligibility and concentration rules, or with the market data such as credit ratings and asset class, can lead to ineligible collateral being pledged or the wrong haircut being applied. Re-performing the eligibility tests using independent market data should validate this.
A third source of errors is in the valuation of positions, especially with more exotic products such as OTC derivatives. If a broker is acting as valuation agent for swaps and at the same time pledging collateral to cover the exposure, there’s a conflict of interest that should be checked.
It’s clearly essential for the validation of the data to check every price, every collateral position and every valuation, but to do this without automation is operationally impossible. The ideal solution should only require human intervention where exceptions have been identified.
Again, technology to do this is widely and cheaply available and should be deployed in any best practice environment. Any solution should be able to source market and price data from third parties (or consume the beneficial owner’s internal data where available), test eligibility across all products, from securities lending, triparty and bilateral repo, to listed and OTC derivatives, and value exotic positions.
For best practice, automated validation of collateral data should be intra-day, exception-based, and it should question every price, every allocation and every valuation.
Visibility
Getting the data together in one place and validating it is only part of the issue, being able to query it ‘on the fly’ to extract the real meaning and make sense of the numbers is just as critical.
Firstly, the beneficial owner needs to proactively see any situation in which it is under-collateralised, ineligible assets have been received or valuations are incorrect, and it needs to have full supporting data to go with it.
This will let them challenge their agents with minimal effort and ensure full coverage for their firms.
Just as importantly, beneficial owners need to have visibility over instances where they have over-pledged collateral and can recall it.
Sophisticated analytics can suggest ways to optimise the collateral pool, whereas trending and what-if analysis can predict the impact of market changes. Relative and absolute performance of different counterparties can be tested, making the possibilities almost endless.
Showing visually how collateral is split by counterparty, product type, asset class, geography and so on gives businesses far greater control and insight into their activity than can be achieved by yet another spreadsheet report.
As before, technology enables firms to get this visibility and bring best practice into play. No longer should decisions have to be made after trawling through endless spreadsheets, combining reports into new reports that someone manipulates by hand to give a final sea of numbers. On-the-fly report building and data queries are essential.
Technology has improved massively in the past few years and cutting-edge business intelligence features should be the norm in any collateral management process trying to adopt best practices.
To summarise, best practice for beneficial owners means that:
Data should be automatically collated from all sources via a fully automated, intra-day, behind the scenes process into a single view.
Every price, every allocation and every valuation should be validated at multiple times per day.
There should be real-time, exception-based visibility into the collateral books. Errors should be automatically identified and quickly resolved, and ad-hoc reporting should be easily produced.
Many technology firms in the vendor space claim to offer this as a turnkey or out-of-the-box solution for beneficial owners. The hard part for some beneficial owners is identifying technology that can actually deliver on the above while costing less than giving the problem to a junior with a spreadsheet.
For most beneficial owners, the majority of collateral management has been outsourced to lending agents, service providers and triparty agents. Unless the indemnity offered is watertight in every case, the beneficial owner still needs to maintain visibility over the process for good governance.
Whether it’s management information systems reporting requirements from senior management, questions needing immediate answers from colleagues in risk management, audit, compliance or credit, or regulatory reporting concerns, having accurate, real-time visibility of every position and analysis across every business line is essential.
For the largest banks, brokers and central counterparties (CCPs), throwing money at the problem to build or buy technology solutions costing millions is a viable solution. For beneficial owners such as asset managers, insurance firms and pension schemes, as well as the traditional buy side, this approach is just not an option.
Adopting best practice for beneficial owners wanting oversight of collateral management consists of asking three key questions:
Collation: can you collate data from various sources into one place, giving a complete picture of collateral activity across all products?
Validation: can you validate the data you have? Is the collateral eligible and correctly priced, are your positions correctly valued?
Visualisation: can you ask questions of the data to see how your business is running? Can you make sense of the numbers to make the right decisions?
In an increasingly fragmented world, this goal of best practice can be a lot harder to achieve than it seems.
Collation of data
Lending agents, CCPs, brokers and service providers typically have great client reporting portals and provide real-time data to their clients but there is no consistency.
In a commonly seen scenario, an asset manager uses its global custodian as agent lender for its securities lending programme, manages cash via triparty repos directly with one of the main agents, maintains margin accounts with numerous brokers for futures and options trading, and trades directly with the large banks to hedge via OTC derivatives. The number of different external portals that the firm has to deal with, combined with reporting or extracts from internal systems used for directly collateralised products, becomes large.
Even assuming that all of the reporting is in the same currency, a big assumption, the formats between different portals (even between different products with the same counterparty) are rarely similar. It’s not a quick process to combine this data so the process to get a holistic view of all collateral in a single place is usually given to a junior member of staff armed with spreadsheets.
While spreadsheets offer amazing flexibility and are quick to implement, re-keying or the cut-and-paste of data is inefficient and error prone. Even in the best-run firms, this is unlikely to be performed more than once a day, usually looking only at yesterday’s closing balances.
This approach precludes real-time analysis and in a time of stress, such as a looming counterparty default, trying to do it quickly only increases the risks of error.
Many firms we speak to have tried automating their spreadsheets via macros and while this has some benefits, it introduces new risks of supporting this often-unstable new process. Frequently, the developers of these solutions are not always around to maintain them, having changed roles or moved firms leaving a reliance on unsupported technology to run a critical process. Even if the developer is still available, the internal costs of maintenance and keeping up to date with report and regulatory changes can be prohibitive.
In a best practice environment, this process requires dedicated technology to ensure completeness and accuracy of data received, the ability to process real-time updates and deliver operational efficiency.
Fortunately in 2014 this technology is widely available off the shelf from vendors and is very affordable. It shouldn’t cost more than the headcount that would otherwise be required to cut and paste the data.
Any reputable vendor should be able to receive this information directly from the source, without the beneficial owner having to be involved in downloading or reformatting the data themselves.
The best practice for collation of data should be a fully automated, intra-day, behind the scenes process with little (if any) human involvement.
Validation
Once the data has been collated, any firm seeking best practice should be looking to validate the information received. Has collateral been priced correctly? Is the collateral pledged to me eligible? Are my positions correctly valued? This is all key to the decisions that the business has to make in the next phase of the process.
At a very basic level, every external party will have separate market-data sources for pricing non-cash collateral and be using different FX rates within calculations. A bad price or erroneous FX rate will lead to over- or under-collateralisation, so it’s prudent to revalue all collateral using a single, internally approved pricing source with a single set of FX rates to identify issues with the reporting received.
A similar set of controls should be applied to testing eligibility of collateral, whether received directly, via an agent lender or allocated by a triparty agent. Any errors with codifying eligibility and concentration rules, or with the market data such as credit ratings and asset class, can lead to ineligible collateral being pledged or the wrong haircut being applied. Re-performing the eligibility tests using independent market data should validate this.
A third source of errors is in the valuation of positions, especially with more exotic products such as OTC derivatives. If a broker is acting as valuation agent for swaps and at the same time pledging collateral to cover the exposure, there’s a conflict of interest that should be checked.
It’s clearly essential for the validation of the data to check every price, every collateral position and every valuation, but to do this without automation is operationally impossible. The ideal solution should only require human intervention where exceptions have been identified.
Again, technology to do this is widely and cheaply available and should be deployed in any best practice environment. Any solution should be able to source market and price data from third parties (or consume the beneficial owner’s internal data where available), test eligibility across all products, from securities lending, triparty and bilateral repo, to listed and OTC derivatives, and value exotic positions.
For best practice, automated validation of collateral data should be intra-day, exception-based, and it should question every price, every allocation and every valuation.
Visibility
Getting the data together in one place and validating it is only part of the issue, being able to query it ‘on the fly’ to extract the real meaning and make sense of the numbers is just as critical.
Firstly, the beneficial owner needs to proactively see any situation in which it is under-collateralised, ineligible assets have been received or valuations are incorrect, and it needs to have full supporting data to go with it.
This will let them challenge their agents with minimal effort and ensure full coverage for their firms.
Just as importantly, beneficial owners need to have visibility over instances where they have over-pledged collateral and can recall it.
Sophisticated analytics can suggest ways to optimise the collateral pool, whereas trending and what-if analysis can predict the impact of market changes. Relative and absolute performance of different counterparties can be tested, making the possibilities almost endless.
Showing visually how collateral is split by counterparty, product type, asset class, geography and so on gives businesses far greater control and insight into their activity than can be achieved by yet another spreadsheet report.
As before, technology enables firms to get this visibility and bring best practice into play. No longer should decisions have to be made after trawling through endless spreadsheets, combining reports into new reports that someone manipulates by hand to give a final sea of numbers. On-the-fly report building and data queries are essential.
Technology has improved massively in the past few years and cutting-edge business intelligence features should be the norm in any collateral management process trying to adopt best practices.
To summarise, best practice for beneficial owners means that:
Data should be automatically collated from all sources via a fully automated, intra-day, behind the scenes process into a single view.
Every price, every allocation and every valuation should be validated at multiple times per day.
There should be real-time, exception-based visibility into the collateral books. Errors should be automatically identified and quickly resolved, and ad-hoc reporting should be easily produced.
Many technology firms in the vendor space claim to offer this as a turnkey or out-of-the-box solution for beneficial owners. The hard part for some beneficial owners is identifying technology that can actually deliver on the above while costing less than giving the problem to a junior with a spreadsheet.
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
