Calculating correlations – how technology is sharpening ALM approaches

The right mix

Carola Lawton-Brown, RLAM

It is of little surprise that those charged with managing the assets of insurance companies and pension funds are now more focused on risk. The financial crisis of 2007-08, quickly followed by the sovereign debt crisis, has shaken the foundations of traditional equity and bond portfolios. Meanwhile, the crises have exposed correlations between assets and liabilities that were previously either hidden or of such low apparent probability that they were all but ignored. Now the heightened awareness of risk has led insurance companies and pension funds to look to alternative asset classes and investment strategies that are more closely aligned with their liabilities.

Alternative assets, particularly derivatives, are more complex than traditional equity and bond portfolios, while liability-driven investment (LDI) requires modelling of both liabilities and assets and the matching of cash flows. This in turn requires technology systems that can support a wider range of asset classes, with more sophisticated modelling and risk management capabilities. For many insurers and pension funds this means upgrading their asset portfolio and management technology. However, the individual nature of each business requires a careful choice of system that most closely matches their particular requirements.

Market fluctuations
Toronto-based Healthcare of Ontario Pension Plan (Hoopp) is among the pension funds moving towards LDI. "We are trying to make our assets act as much like our liabilities as possible so that we are not as impacted by market fluctuations," says Jim Keohane, chief information officer and senior vice-president of investments at Hoopp, which has $31 billion (£19.2 billion) of assets under management (AUM). "To this end, we use derivatives to cover our exposures and get a more effective control of risk than we could through a traditional equity and bond portfolio structure."

However, in taking this course of action, Hoopp ran into problems with the data management of its traditional portfolio management system. "For example, to capture an equity swap in our old system we had to break it into two securities - an equity and a fixed income. The difficulty was trying to get data out at the other end in a format that made sense. There were a lot of manual overrides in the system, which meant the information sometimes wasn't very timely, and there was a lot of opportunity for error," says Keohane.

Hoopp looked at nearly 20 portfolio management systems before deciding that the Dimension system from Copenhagen-based SimCorp best met its needs in terms of support for derivatives and the flexibility to cater for different types of portfolio structures. Other important factors in its choice of Dimension were that, although complex, the system did not require onerous in-house resources to operate, and it could handle collateral. "When you are using derivatives, there is a lot of exchange of collateral to manage credit risk. The fact that SimCorp offered a sophisticated collateral management module was important to us," says Keohane.

London-based British Telecommunications Pension Scheme Management (BTPSM) is the executive arm of the BT Pension Scheme, with nearly £34 billion of AUM. It too takes a liability-driven approach to investing, and its growing preoccupation with risk led it to look for a new risk platform.

"We must be aware of the risk we incur if we're unable to pay pensions when they fall due, so we need to consider the combined risk across assets and liabilities," says Wyn Francis, head of investment risk for BTPSM.

However, BTPSM was finding this challenging with the mix of third-party applications and proprietary tools it had in place. "When you're using different systems or approaches, it |invariably means risk is being estimated on |an inconsistent basis. None of the approaches we were using consistently took into account the correlation between assets and liabilities when it came to risk exposure."

In addition, the firm was finding it difficult to aggregate outputs from various systems and had to manipulate data in a large number of spreadsheets in order to construct appropriate overlay strategies to execute tactical and scheme-wide objectives. So BTPSM concluded it needed a risk management system that would allow it to manage its overall risk and overlay strategies, as well as enable it to understand the performance of external asset managers, and monitor investments at various levels. "Risk is intrinsic to the way we make decisions," says Francis.

BTPSM has a policy of keeping its organisational infrastructure light and of outsourcing services where possible, so it began to look for a third-party risk platform. BTPSM was surprised to find that few vendors understood its requirements - the ability to analyse and attribute risk across an aggregated and diverse set of assets and liabilities, all in the context of a pension fund.

One vendor that did was Toronto-based Algorithmics. The vendor also offered its risk platform Algo Risk on a hosted basis, meeting BTPSM's requirement for an outsourced solution, and enabling it to avoid "the big data oversight function that many in-house developments require", says Francis. Other factors in the organisation's decision were Algorithmics' track record of evolving its technology to keep up with the markets, and its strong corporate position. "[Choosing our risk platform] is a long-term decision," says Francis.

Another organisation that has been putting more focus on risk is London-based Royal London Asset Management (RLAM), which manages the assets for Royal London Insurance, as well as pension funds, charities and multi-managers, and which has £40 billion of AUM. Its bond portfolios include a range of government and government index-linked bonds, credit instruments and derivatives. The firm appointed a dedicated portfolio risk manager, Carola Lawton-Browne, in 2008, and also began looking for a robust risk model that would allow it to look at the risk of its portfolios from a number of angles, such as tracking error, value-at-risk, duration and performance.

Risk model
"The challenge was to find a risk model that was easy to use and where the output could be intuitively understood by fund managers, and where the model was recognised by the industry and consultants because that would help us with marketing the funds and reporting on risk," says Lawton-Browne. RLAM looked at whether it might be able to upgrade its existing system as well as investigating other systems on offer in the market, and ended up running two potential replacements in parallel with its existing system for eight months.

A deciding factor in choosing the UBS Delta portfolio analysis and risk management system from Swiss bank UBS was the fact that it offered flexibility in the way the firm could look at its risks, says Shalin Shah, chair of the bond risk committee at RLAM.
"We have our own sector definitions and assign bonds to them, and then look at the risk in terms of these sector definitions," he says. Compared with the other systems that had more rigidly defined sectors, UBS Delta made it easy for Royal London to create its own definitions.

Selecting an appropriate portfolio and risk management system is one thing; implementing it successfully often presents its own challenges. RLAM had a head start in that it ran a prototype UBS Delta implementation in parallel with its existing system for eight months. Converting to a production version of UBS Delta took only a matter of weeks, says Shah. Hoopp on the other hand, found the implementation of Dimension "a very challenging undertaking", Keohane says.

"We have a complex portfolio, while being a small organisation - around 300 employees," says Keohane. "It was challenging to devote enough resources to the project." Hoopp ended up bringing in around 50 contractors, including a number from Europe where there was more expertise in the Dimension system. In all, the implementation took around 18 months, with the organisation going live on the system in July 2010. "But the outcome is very good," says Keohane.

A major benefit of the new system for Hoopp is the automated handling of transactions, including derivatives, replacing spreadsheets and manual interventions. "Dimension takes in a swap transaction in its entirety and it manages the back-end cash flows that come out of it," says Keohane. The system also automatically handles the resets associated with the swaps. "So our data management is much improved and we get much more timely reporting out of the system," he says.

The system also allows Hoopp to add new instruments quickly and in an automated way. "This gives us better internal controls and reduces our error rate, so we have a lot more confidence in the financial information that we are producing with the system," says Keohane.

Major benefit
Francis at BTPSM highlights the importance of resolving data issues before implementing a system such as Algo Risk. BTPSM invests in around 20 fund managers with about 70 underlying funds. Arranging to feed the data on the funds through to Algo Risk proved problematic in some cases, with managers not always able to immediately supply the data in the required format. Also market data had to be carefully managed or the costs would have escalated quickly, says Francis.
Like Hoopp, BTPSM says the implementation effort has been worth it. Algo Risk has given the organisation a foundation for an investment risk management framework, says Francis. "We are now able to analyse risk on a number of levels - from scheme level down to individual holding level. We can aggregate risk across not just the multi-asset class portfolios, but across a data-intensive liability portfolio as well. We can analyse our risk factors, we can stress and scenario test and we can monitor concentrations," he says.

Furthermore, "we are in a position to give far more insightful information to the trustee investment committee", with the system, says Francis. And the new system proved itself during the recent sovereign debt crisis, where BTPSM was able to quickly pull together information on its risks and what it might be able to do about them.

RLAM is also in the position to provide better information to those ultimately responsible for taking the investment and risk decisions for the assets it manages now that it has upgraded its risk analysis system. Lawton-Browne reports regularly to the board and the market and risk committee of Royal London Group.
"We use the analysis from UBS Delta at the highest level in the company," she says. RLAM also uses the service to produce reports for client services. And it is able to use UBS Delta's facilities to produce model portfolios when pitching for new business.

"We are able to provide risk summary data for the model portfolios, so prospective clients are aware of the risks before they buy into RLAM," says Shah.
A factor that has become increasingly important to insurers and pension funds when choosing a portfolio and risk management system is the relationship with the vendor. The software isn't cheap (although a hosted service can reduce upfront costs) and, as we have seen, implementation projects can take more than a year and require considerable expert resources, both internal and external.

Furthermore, such systems can come to play a critical role in an organisation's operations, so need to be able to advance and grow with the organisation itself. Keohane says that unless the vendor has a continual development cycle and is responsive to requests for enhancements, a system can end up with a limited shelf life - which was Hoopp's experience with its previous system. It was critical to Hoopp that SimCorp offered a partnership approach to its future technology provision. BTPSM had a similar view.

"There was a willingness that Algorithmics would partner with us and start to address some of the challenges that meet the needs of a pension scheme manager," says Francis.

A look at some of the other technology announcements over the past year shows that insurers and pension funds and their investment arms share the same concerns as Hoopp, BTPSM and RLAM. For instance, Edinburgh-based Standard Life Investments, with around £143 billion of AUM, has a long-term relationship with its portfolio management system vendor, Boston-based Charles River Development. After seven years of using the Charles River Investment Management System, the firm recently upgraded to the latest version of the system. Meanwhile, Nationwide Investments, the investment arm of the Columbus, Ohio-based insurer Nationwide, which has around $61 billion of AUM, recently went live with the Findur trading and portfolio management system from New York-based OpenLink to help it manage its growing derivatives investments.

As insurers and pension funds faced with a volatile and uncertain economic environment look to a wider range of asset classes - many of which are far more complex than traditional equities and bonds - they require more sophisticated tools to help them manage their portfolios and risk. There is a choice of third-party systems available, although firms must take care to select technology appropriate to their needs, as well as look at whether the vendor has the ability to keep up with their requirements, both now and into the future.

Emerging markets share risk mangement requirements
Insurers and pension funds in emerging markets are faced with many of the same investment challenges as their counterparts in developed economies and are also looking to the latest generation of portfolio management and risk systems to support their business and risk management objectives.

Beijing-based China Re Asset Management Company (CRAMC) is the asset management arm of China Reinsurance Group, with 50 billion yuan under management. It wanted to improve the efficiency and effectiveness of its investment management and put in place a stringent compliance and risk assessment mechanism. It also wanted a system that would support further diversification of asset classes over time, in both local and international markets. The company, therefore, looked for a specialised system where user-friendliness and proven performance in mature markets were essential, says Chen Jun, general manager of the finance department at CRAMC.

After a careful review, CRAMC selected the Value system from Paris-based Sophis in July last year. Value most closely matched the company's requirements, including offering integrated trading, risk management and compliance, with straight-through processing of transactions. "Straight-through processing helps improve our investment efficiency and data-sharing," says Jun. In addition, the system supports a clear and flexible portfolio structure, as well as providing risk indicators, compliance management and performance attribution. "This will help us improve investment management capacity and expand our asset management business," says Jun.

• Tweet  
• Facebook  
• LinkedIn  
• Save this article
• Send to  
• Print this page  

• Topics
• Technology