Data Cleansing Products to Save Industry Time and Money

DataMetrics and other external data products provided by vendors such as BARRA, GFINet and Reuters are presenting solutions to this data problem. While BARRAhas traditionally been an application provider, it has also been providing a data product since the introduction of its equity market models in the 70s when little if any data was available elsewhere for its users. DataMetrics came about after RiskMetrics was spun off from J.P. Morgan in 1998. ""At that point DataMetrics became a separate product for providing data to all of our software applications, maintaining the standard RiskMetrics data sets and finally providing clean, end-of-day data to a number of clients not even using our software,"" says Malz. DataMetrics provides clients with a daily file of error free, aggregated, properly formatted risk data for the client to put into their risk engine, continues Malz. This file is mainly delivered via the Internet, but other delivery mechanisms are also available. Reuters, which owns a portion of RiskMetrics and provides DataMetrics with some of its market information, has also created a data offering of its own and is a newer entrant in the risk data product arena. GFINet, an execution platform is also providing a data product related to its overall volumes, including pricing, instruments traded and spreads.

So who is the best candidate for these types of externally provided risk data products? Williams says both large and small financial firms can benefit from these types of data products. ""There are clearly big institutions that have figured out this is the way to go,"" says Williams, who notes in her brief that institutions such as Barclays and HSBC in the U.K. and AIG in the U.S. have already adopted these types of risk data products. And of course, smaller institutions that might lack in the manpower and resources to perform these tasks internally could also benefit from these data products. ""The size of our clients is across the spectrum, we have some extremely large clients and some smaller institutions,"" says Malz. ""A lot of our growth has been among the smaller firms.""

Erwin Martens, managing director and head of risk management at Putnam Investments, agrees that this type of data product is well suited for many smaller firms. ""Particularly the smaller places that don't have the infrastructure are doing this so they don't have to reproduce those capabilities which are expensive in the end,"" he says. Martens adds that while Putnam is a large research-oriented firm it gets most of its data internally, but as a user of BARRA's Total Risk for Asset Management (TRAM) product, they also take some risk data directly from BARRA. ""The data cleaning is pretty heavy lifting and you have to apply quite a bit of resources to it either through high start up costs or maintenance costs along with subscribing to the data and ensuring that it's clean,"" adds Martens.

He speculates that firms usually dedicated at least two to three people on the technology side of the data problem sustaining servers, ensuring the scripts run and integrating and maintaining that data. Additionally, Martens says firms also need one or two people to understand the context of the data with the business sense to ""see why the data is wrong."" Williams adds that while risk data products will never eliminate these data teams, it could ""at least dramatically cut them."" Incorporating these risk data products into the risk department will also increase the efficiency of those hired to perform risk duties. ""The issue today is that people spending a lot of time on data are not data people, they are analysts and because they need the data to do their analysis they spend the better part of each day working with data,"" says Williams, who adds that a risk data product will make those people more efficient.

In the long run, Williams expects that cost savings and improved efficiencies, as well as an overall trend toward outsourcing, will lead more firms to purchase these types of risk data products. In the brief, she estimates that market penetrations rates among the top 100 financial institutions could reach 20-25% in the next two to three years. The brief states, ""We would expect to see institutions cutting overall costs for maintaining reference data for risk purposes between 25 and 50% depending on the asset classes and geographic breadth of coverage."" A wider vendor marketplace making these types of data products more readily available will also contribute to its adoption over the next few years. It is still unclear which vendors will prove successful and how many more will enter the marketplace. Whether the applications providers or traditional data vendors come out ahead in the game also remains to be seen, but Williams says either way she expects more offerings to hit the market. ""It just seems obvious to me,"" concludes Williams. ""As with other data, people don't have their own tickers anymore, they take the data from people like Reuters. I see no reason why this won't catch on as well.""

2 of 2