This summer Rachel Gosch took on some challenging endeavors: mastering baking to an exact science, competing against friends in Extreme Uno and wrangling real-time data while conducting research in industrial math and statistics.
As a BHP math and business double major, Rachel chose research over the beach this summer in preparation for the research thesis she will be required to complete later in her academic career. In an effort to gain experience in the research realm, she applied for a REU (Research Experience for Undergraduates) with the Worcester Polytechnic Institute in Worcester, Massachusetts. To apply she submitted a transcript, letters of recommendation, an essay describing her interests and topics she would be interested in working on. Rachel and only 11 other undergraduate students were chosen nationwide.
Rachel’s REU was focused on Industrial Mathematics and Statistics. Once the group arrived, the students split up into 3 teams of four students each. Each team was then assigned an industrial sponsor that presented a problem for the research team to solve by the end of the summer. Rachel’s group was paired with Wellington Management, an investment advisor to more than 2,100 institutions located in over 50 countries.
Wellington had observed work published by three authors working with the Philadelphia Federal Reserve Board. The authors had created a process for filtering real-time data through the Kalman filter to find the Business Condition Index (economic status) of the economy. The process began by collecting four numbers from their U.S. market: GDP, M1, import sales and retail sales. As these four numbers come in and constantly change, they were then funneled into the Kalman filter, which essentially “cleaned up” the data making the numbers manageable and easier to use in determining the current Business Condition Index of the U.S.
Wellington requested help in implementing this process in their operations in other large countries, including: Australia, Canada, Germany, Japan and the UK. The idea was to collect the Business Condition Indexes from each country so they could then be compared to illustrate where business is looking good and where business isn’t doing so hot. If Rachel’s team succeeded, Wellington would be able to use the data in the portfolio planning.
When Rachel first learned she would be working with a portfolio management firm, she assumed she should refresh on finance concepts. However, for this project, her team actually did a lot of research on engineering concepts like signal processing and the Kalman filter, which was a surprise to see those concepts applied to financial math. She quickly realized that working with real-time data would be her biggest challenge, “I’ve heard professors say this before, but working with real-time data is extremely difficult! It’s messy, and when you’re dealing with varying frequencies and information availability, it’s challenging to write programs that can work with whatever data you give them.” She also learned how valuable programming and coding skills can be when working in some types of financial math analyst positions, “I took an intro programming class in the spring pass/fail just so I’d have a basic exposure to coding and I’m so glad I did!”
Outside of researching financial math, Rachel spent a lot of time with her fellow team members. During their free time, the group would play extreme UNO. Rachel described the game as normal UNO with additional, intense rules. She also took to baking with another teammate. They tried recipes over and over again until they perfected them, just as mathematicians do. Being in Massachusetts, the team made frequent trips to Boston to enjoy the city and attended a Boston Red Sox game. Even as Rachel enjoyed her down time, she often noticed her mind wandering back to the research project.
Rachel did not anticipate how invested she would become in the outcome of her team’s research. It didn’t take long before she was putting in long hours after the workday had finished. “Getting results that work…is really exciting,” Rachel explained, with a smile on her face. Rachel was not the only one infected with excitement, her team members rallied with her. Specifically, she recalled an evening the team had finally compiled the data into a useable format. Team members were calling and texting one another to rejoice the great news and review the numbers, which were stored in a file, named “Pretty Data.”
At the end of the summer each group made a presentation to their industrial sponsor to share their findings and hopefully, solutions. Rachel’s team pleased Wellington with the presentation of their hard work, which paid off in an answer to the company’s request. However, the work did not cease here because Rachel continues to correspond with her team via email to further tweak the process.