NIBCapital is an independent merchant bank with a global distribution network. Headquartered in The Hague in The Netherlands, the highly rated bank focuses on northwestern Europe and global clients in selected industry sectors, providing corporate finance, risk management and structured investment solutions.
When the bank decided that it needed to upgrade its market risk management, it knew that although its priority was to tackle interest rate risk in its banking book, it wanted a solution that would improve risk management across its operations. NIBCapital chose Algorithmics' Algo Market as the system best able to meet not only its immediate requirements for interest rate risk management, but also for its trading and mortgage books, enabling it to take a portfolio approach across all of its risks. The firm has since gone on to implement Algorithmics' collateral and credit limit management solutions as well, with the opportunity to achieve integrated risk management at an enterprise level.
NIBCapital's project began when in 2001 it realized that its risk management practice needed a major overhaul. At the time, it was managing the interest rate risk on its banking book using functions in a front-office system. "We knew we had to move away from this towards a more advanced approach to risk management," says Ad Keijsers, responsible for risk systems at NIBCapital.
Although it initially made a business case for a new system just to handle interest rate risk, the bank also wanted to tackle credit spread risk in its trading and investment books, as well the risks in its mortgage and other portfolios, and to calculate value-at-risk across all its portfolios. "But we decided to do this all step by step," says Keijsers.
NIBCapital reviewed the risk management systems that were currently on the market. The bank was familiar with Algorithmics through its asset management operation, which was already in the process of implementing Algo Risk for asset managers. The asset management division had selected Algo Risk largely because the solution covered the broadest spectrum of financial products of any system on the market, says Keijsers, and further investigation suggested that it was also the most appropriate system for the bank's needs, especially for taking an integrated approach to risk management.
NIBCapital began implementing Algo Market in late 2001, and went live with interest rate risk management for the banking book in April 2002. The bank then moved on to incorporate credit spread risk in the trading book, and later in the mortgage portfolio.
Integrating all portfolios into a single risk system has a number of advantages, says Keijsers. "If you only look at one portfolio at a time you can't get the tradeoffs with risk you have in other portfolios. For example, all the interest rate risk that we run in the bank has a diversifying effect on our credit spread risk, but to measure the diversification it is necessary to have all the portfolios in one place, and to be able to measure all the different risk factors that we need," he says.
The other major advantage of integrating the various risks is that the bank is able to cover all its requirements in a single system. "Being a small bank, we cannot afford to have a large number of applications, each of which covers only a portion of our portfolios," says Keijsers.
NIBCapital fully supports Algorithmics' move toward pre-defined configuration choices, combining an ongoing commitment to flexibility with a drive to reduce total cost of ownership. "When we went further and started to implement the credit spread risk management, we really started to see the advantages of the flexibility in the system," says Keijsers.
The bank used a proprietary data mapping application from Algorithmics to help it transform the data into the format required by the system. Identifying where the data for risk analysis would come from, and checking it for consistency, completeness and accuracy was a major task, as it always is with risk management implementations.
The bank also installed another proprietary application, this time an advanced scenario generator for creating risk and stress scenarios. The scenario generator is one of the building blocks of Algorithmics' award-winning Mark-to-Future methodology, whereby risk is broken down into units by instrument for computational efficiency, allowing near real-time valuation of even the largest portfolios. NIBCapital
intends to investigate introducing Mark-to-Future, partly because of the performance improvements it would provide, but also because the bank is interested in some of the new developments that Algorithmics is making with its systems that are based on the methodology, such as web-based reporting of risk figures.
Mark-to-Future is a basic concept for Algorithmics, and it is something we want to buy into," says Keijsers. "We have some specific reasons for wanting to be part of it-for example, we think it will make our applications go faster -but we also think we will lose out if we don't start implementing Mark-to-Future if Algorithmics' [major developments] are built on this. I also expect there are some hidden advantages which we will only see if we have a good look at it."
At the moment, NIBCapital runs a daily risk analysis batch process. All positions in its loan, trading, investment and mortgage portfolios are fed into Algo Market, which produces a set of standard reports. These include basis point values of positions, greeks, value-at-risk, and some stress scenarios defined by the bank. The numbers calculated by Algo Market are also fed into a web-reporting tool, developed by NIBCapital, for further analysis. The tool enables the bank to do things like look at credit spread risk in terms of certain industry sectors and rating categories. The tool allows the bank to drill down through the risk data to the individual deals if required. "We can look at, say, the interest rate basis point value for our derivatives trading, and we might see a sensitivity of 100,000 euros there, and we can drill down to it and see all the deals that make up this number," says Keijsers.
The bank also uses Algo Market for ad hoc reports and stress scenarios as required. For example, the bank's asset and liability committee recently asked for an analysis of the full diversification effect between interest rate and credit spread risk across portfolios.
With more sophisticated facilities for managing market risk now in place, NIBCapital moved on to the next part of its project, and looked towards automating the management of collateral in its financial markets deals. Although its activity in this area is still limited-it has only 35 counterparty collateral agreements so far-the bank's use of collateral is growing, and the traditional method of using spreadsheets for handling confirmations was proving too slow and unreliable. A review of systems on the market showed that Algorithmics had one of the few systems that could meet NIBCapital's requirements, and the bank implemented Algo Collateral in 2003.
Counterparty risk was next on the bank's list, and NIBCapital recently successfully implemented Algo Credit's limit manager functionality. "Previously, our back-office system was handling the counterparty risk for our financial markets group, but we weren't happy about the module," says Keijsers. The bank considered how best to tackle the problem, and reviewed its approach to counterparty risk. "We decided that we wanted to upgrade how we looked at counterparty risk. The idea, which Algorithmics provided, was that we should work with potential future exposure, and link [the limit system] to our existing implementation of Algo Market," says Keijsers. The fact that the bank had already implemented Algo Market was a big advantage, and was critical in the bank choosing Algo Credit over other systems available on the market. "[The two integrated solutions] give us the scope to develop a more sophisticated way to approach counterparty risk in the future," says Keijsers.
So far, the bank has only implemented the system for its financial markets group, where the counterparties are 95 percent financial institutions. But the bank is keeping open the option of incorporating its loan portfolio into the system. It is aware that a number of other Algorithmics' customers use Algo Credit in this way. "This gives us the confidence that we can easily include loans in the future," says Keijsers.
NIBCapital began implementation of Algo Credit for financial markets at the end of 2003, and completed the project in August 2004. One of the solution's major advantages is that it allows the bank to set limits by time buckets. This enables it to differentiate the risk between, say, a 10-year illiquid bond and an instrument with a maturity of a week.
For the moment, NIBCapital is not attempting real-time incremental risk calculations for new deals in the limit system. "We have taken a mark-to-market plus add-on approach, which is basically the same way it was done in the back-office system," says Keijsers. "That is the way we approach our projects-we do something simple in the first instance, where we learn the system, then we build further on that."
NIBCapital used Algorithmics' consultancy services as a catalyst in the early stages of each project to accelerate knowledge transfer and to make rapid progress with early milestones. Once knowledge transfer was complete, however, the firm believed that it was critical that the skills and responsibility for final implementation and ongoing operations be transferred to the NIBCapital team.
Today, NIBCapital has a team of knowledgeable users of Algorithmics' solutions. In turn, the team has been able to apply their experience and knowledge to new Algorithmics solutions, thereby ensuring that the firm realizes the greatest possible return on its investment.
NIBCapital has taken the same approach with Algo Credit. Having learnt the system with the help of Algorithmics' consultancy on the first phase, the bank plans to implement the use of potential future exposure, and extend coverage to its loan portfolio using its own resources as much as possible. "We have felt very comfortable working with Algorithmics," says Keijsers. "Whenever we do projects with them, there is always a sense that we are in control, and that we are not being pushed in any way."
Once the loan portfolio is incorporated into Algo Credit, the bank will have an integrated system for all its counterparty risk. This will allow it to produce consolidated risk figures across all counterparties of the bank. "There will be one, and only one, system that can cover everything. And we want to use it more in the sense of a portfolio management system, to look at all our exposures in a portfolio way,"
Although NIBCapital has done calculations that suggest that it will be profitable for the bank to move to value-at-risk based reporting for regulatory capital reporting, this is not primarily why it implemented its new risk systems, and it has not yet moved to this way of reporting. Nor has it attempted any other return-on-capital calculations for the Algorithmics applications. "These have been very much quality-based projects. Our main concern was that our old risk management tools weren't up to speed," says Keijsers.
NIBCapital regards Basel II in a similar fashion. The bank has not yet analyzed precisely how Algo Market, Algo Collateral and Algo Credit might help it with its Basel II reporting, but it believes that it is in a better position to meet the new regulatory requirements than it was with its old systems.
I'm confident that [the Algorithmics systems] add significantly to our whole way of looking at risk management, and the Basel II process," says Keijsers. "In fact, I can't even imagine the situation we would be in if we hadn't done these implementations. I simply can't imagine doing without the numbers we are producing in the daily process now.
We are using the value-at-risk figures for economic capital calculations, for example. It is unthinkable that we wouldn't have these figures."
NIBCapital is using Algo Market for the new International Financial Reporting Standards for accounting. The bank uses the system for the valuation and hedge accounting of derivatives. "For hedge accounting, you have to prove before the period that you are accounting for that your hedges will be effective, and afterwards, you have to prove that your hedges were effective. Algo Market isideal for establishing which derivatives to use in a hedge relationship," says Keijsers.
NIBCapital is cautious about claiming that its new risk management facilities have helped its competitiveness. "There is not always a direct link to be made from risk management to competitive advantage," says Keijsers. However, the risk systems give the bank comfort that its positions aren't too risky for its' own good. "If we didn't have that insight into our positions and our risk then we wouldn't really be able to do our business confidently," he says.
The NIBCapital Case Study is one in a series of Algorithmics' case studies. For information about Algorithmics and to view other client case studies, visit www.algorithmics.comAlgorithmics
More on Structured Products
Regulator set to focus on backtesting and replicability of index products
The watchdog’s priorities for 2015 include drawing up new powers of product intervention
Contineo initiative set to transform structured product sector
Trade body says issuers will face unnecessary legal and compliance bills under Esma plans
Sign up for Risk.net email alerts
Oxford professor David Vines argues that the carrot is as important as the stick
Sponsored webinar: IBM
Watch highlights of this year's London conference
Operational risk and the challenges of defining and dealing with conduct risk
There are no comments submitted yet. Do you have an interesting opinion? Then be the first to post a comment.