Fimat moves to Yolus risk management platform

Source: Yolus

Fimat, the brokerage subsidiary of French bank Société Générale, is using a risk management framework from Yolus to help consolidate risk information across a number of the broker's legacy systems.

Fimat will use the Yolus platform to support risk management functions for its brokerage operations across a range of asset classes, including listed futures and options, foreign exchange (FX) spot, forward and options, cash equities and bonds, bond repos, and contracts for difference (CFDs), officials say. The platform will also help Fimat to start clearing over-the-counter (OTC) equity options and index options for its hedge fund clients.

Following a pilot implementation, which was completed in February of this year, the bank underwent a six-month implementation to install the Yolus technology, which ended in July. The next phase of the implementation is focused on generating client statements-predicted to go live next month-and making the system real-time, due to commence in 2005.

Yolus is a software spin-off of Dresdner Bank. The vendor's technologies act as a data consolidation layer, communicating with numerous systems and housing all relevant risk data in one place.

Rob Devayya, managing director for Yolus, says that the vendor supports standard components for connecting with databases, reading files and linking to enterprise middleware. "We have connected with over 100 different third-party and in-house systems using the APIs (application program interfaces) provided, or via files and reports delivered by these systems," adds Devayya.

At Fimat, Yolus is running on Linux-Intel platforms, but Devayya says that the system can also be deployed in a Microsoft Windows NT environment or in a Unix environment. "Our system is written in Java, which gives us platform independence," he says.

Claire Bridel, head of credit and risk for Fimat, says that there were several drivers for the new project. Regulatory reporting requirements from the Bank of England had been an early motivator.

However, by using Yolus as a data consolidator, Fimat will now be able to feed its downstream applications, such as client statements, accounting systems, and risk systems, through a single platform, says Bridel.

Another advantage of maintaining data centrally is that it promotes better communications within the company, says Bridel. "Instead of having to find someone who deals with a particular issue, you just have the data available," she says.

The Yolus application is also helping Fimat to launch a new service: clearing OTC equity options and index options for hedge fund clients, says Bridel. "It's through the prime brokerage that we got to develop cross margining-we got more sophisticated in terms of risk-based margining rather than product by product," says Bridel.

However, until Yolus was installed, Bridel says that calculating margin requirements across asset classes tended to be a manually intensive process for her department. Automating that process makes the business much more scalable, she says.

The better a firm is able to monitor risk the easier it is to take on new clients regardless of their creditworthiness, says Bridel. "Intraday monitoring allows you to disregard the creditworthiness of clients because you can rely purely on what's in their account," she says.

Prior to Yolus, Fimat didn't have a consolidation tool, says Bridel. "We'd take all the futures positions from the GMI System, [a back-office clearing and accounting software solution from SunGard], all the foreign exchange (FX) and options positions from Reuters Kondor, and custody data from our custody system. We'd do some consolidation manually in my department, then calculate the VaR (Value-at-Risk) out of the risk engine," she adds.

Fimat's risk engine is a proprietary technology that uses Monte Carlo simulation to calculate three-day, 99 percent VaR figures.

For the selection process, Fimat reviewed offerings from Reuters and Aleri. both vendors decline to comment on their participation in the deal.

"Reuters didn't have a product that was tailored to our specific needs," Bridel says. "The reason we didn't take Aleri was because the language was more proprietary. It would have been harder for us to change the code ourselves. With Yolus, once you've gained knowledge of the system you can program it yourself because it's all Java."

Devayya agrees, saying, "Our technology is designed to be extended by others. For functionality that is not available out of the box, we provide a set of APIs that enable other IT professionals to write and deploy their own JMX components into a Yolus system."

Now that Fimat has been able to consolidate client risk data across its various systems, the next step is to feed that data into client reporting systems, says Bridel. This should be completed in December, so that clients, both from the buy side and other brokers, begin receiving consolidated statements as of the beginning of 2005.

"We should have our clients' end-of-day positions and statements across all products by the end of December. Then we'll concentrate on the intra-day feeds, because the intra-day feeds might not be the same as the close-of-day feeds, so it's really a separate development," says Bridel.

Devayya explains that no changes will have to be made to the Yolus platform itself to accommodate real-time updates. "Our technology is real-time; the distinction is only down to the frequency with which the underlying deal-capture systems deliver their data," he adds.

Comments: (0)

Featured job
All Jobs »