Futures and Equities Research Benefits from S-PLUS

The multi-billion dollar business of trading of stocks, bonds, futures and equities requires quick decisions, extensive research and complex data analysis. For every dollar made or lost hundreds of hours have gone into researching and analyzing the numbers behind the trade. All of this research and analysis requires timely and complex analytical computing. Renaissance Technologies Corp., a New York based financial research and trading organization, has found S-PLUS to be the statistical package it turns to for in-depth, flexible data analysis.

Renaissance has enjoyed considerable success in the financial industry in recent years managing over $400,000,000 in computer based equities and futures trading. Renaissance supports its trading activities with large scale statistical modeling and a great deal of statistical research.

The computer software programs it uses for this research fall into three categories: forecasting, applied strategy and accounting/information management. S-PLUS functions as a research tool for its equities research and in some critical phases of futures research. Andrew Mullhaupt, senior research analyst, points to the sophisticated robust statistical models of futures trading costs created with S-PLUS as an example of how beneficial S-PLUS is for the work that Renaissance does.

"By using S-PLUS, we were able to implement a complex statistical model for futures trading," says Mullhaupt, "A significant part of our success in futures trading is attributed to the model we built in S-PLUS, one we could not have built without it. In this respect, S-PLUS has paid for itself many times over."

For Renaissance, S-PLUS has applications in other aspects of its business as well. For example, equities tick data is also analyzed with S-PLUS. Using dynamic loading, many years of tick data, surpassing 30 Gb of information, can be managed and analyzed. By utilizing the flexible programming environment of S-PLUS, Renaissance is able to prototype systems for screening and removing errors in this data. Additionally, S-PLUS supports the research, mathematical development and testing of investment strategies and significantly cuts down on their development time.

Furthermore, Renaissance has increased productivity using S-PLUS. Its unique capabilities enable Renaissance to develop a wide range of specialized facilities in a coherent, well documented and easy to learn environment. In one case, a research consultant was able to make contributions within three days of arrival at Renaissance where previously several weeks of acculturation were required before a researcher could do useful work.

For Mullhaupt, one of the advantages to S-PLUS is its ability to handle extremely large amounts of data. The datasets at Renaissance are 30 Gb in size, nearly 1000 times as large as an "average" dataset. With dynamic loading, S-PLUS is able to handle these datasets with efficiency.

"S-PLUS has competition, but they don't have anything close to what we do with S-PLUS," says Mullhaupt, "What we do here is original and involves enormously large amounts of data. Dynamic loading allows high performance, custom models to be designed and then the object oriented language allows us to interface with existing systems. We couldn't live without it, we can hook S-PLUS into our existing systems and you just don't find that in any other program."

Mullhaupt feels that it is very easy to integrate S-PLUS with other systems and that this feature is often overlooked in the financial community. Financial datasets are so large, they often need to be processed before any analysis can begin. One advantage to S-PLUS is that any front end processing device can be added to its existing functions.

"If we were a larger company, we might have been tempted to write our own language but then we would have had to support it," says Mullhaupt, "S-PLUS gets such wonderful support from StatSci's Technical Support Department. Also, S is a language with a large constituency in the statistical field. S-News, the e-mail group of S users, facilitates a constant dialog between S users for support and problem solving."

Statistical analysis software