Skip to main content

CSBS 2022 Annual Data Analytics Competition
The Role of Community Banks During the Pandemic 


Final Presentations & Winners


A three-member team of students from the William & Mary won the CSBS 2022 Data Analytics Competition. This year’s competition challenged students to develop a data analytics model that demonstrates the role community banks played during the pandemic using data from the Paycheck Protection Program (PPP).

“We started the CSBS Data Analytics Competition in 2021 to give college students an opportunity to explore data analytics solutions to real-world banking questions, and to provide CSBS and its members with creative ideas and valuable solutions to interesting and important bank-related data analytics questions,” said Emil Phillips, CSBS Senior Vice President over Research and Analytics.

CSBS selected William & Mary students Gio DeFrank, Junghee Mun, and Kristina Posner as winners of the CSBS 2022 Data Analytics Competition. Led by faculty advisor Dr. Joseph Wilck, the William & Mary team developed and tested the following hypothesis: Were community banks who partnered with financial technology (fintech) firms more successful in distributing PPP loans to small businesses with racial minorities and underbanked populations? 

To complement PPP loan data from the Small Business Administration’s database that CSBS matched to most U.S. commercial banks, the William & Mary team added the FDIC’s community bank financial data and information from the FDIC’s annual “How America Banks” survey. But because of many missing observations, the team created a classification neural network model to predict the probability of a business being minority owned.

Once the team finished scrubbing and cleaning the data, they used logistic regression models and random forest classification models to test their hypothesis. The team found that fintech partnerships with community banks was not a factor in predicting whether a PPP borrower was a minority or lived in a low or moderate income (LMI) area and did not significantly contribute to boosting minority or LMI outreach. The team concluded that fintech partnerships could help better diversify the banking ecosystem and encouraged further research in this area and for better data collection to help determine outcomes.

“We have all been extremely impressed with the creativity and rigorous technical skills from all the students involved in the CSBS 2022 Data Analytics Competition,” said Tom Siems, CSBS chief economist and creator of the competition. He added, “Indeed, the Data Analytics Competition is one of the ways we engage with the academic community as it effectively combines both research on banking with data analytics, and helps us be part of building the workforce of tomorrow. We hope all participants are inspired to continue their academic studies in economics, finance, data science, accounting, or other fields that offer a valuable perspective on community banking and financial sector supervision.”

The other finalists in the CSBS 2022 Data Analytics Competition include teams of students from the University of California at Irvine (second place), Carnegie Mellon University (third place), and Southern Methodist University (fourth place).

The University of California at Irvine team was led by faculty advisor Dr. Gary Richardson. The five-member team included Abhishek Karimpuzha Devarajan, Qi Wang, Qingqing Yu, Wanjing Xu, and Yixuan Cai.  The team created several multiple regression models and concluded that PPP lending via community banks appears to have reduced county-level unemployment during the pandemic more than PPP lending via other institutions, particularly larger banks. They found that $1 billion in PPP lending via community banks reduced unemployment by about 0.6%, whereas $1 billion in PPP lending via other institutions reduced unemployment by about 0.1%.

The three-student Carnegie Mellon University team was led by Dr. Gabriela Gongora-Svartzman and comprised of students Thomas Yu Chow Tam, Shun Tomita, and Jamie (Jeong Yeon) Lim. Using multivariate regression models, the team concluded that community banks seem to have helped local communities and local employees mainly in counties in the Midwest and Great Plains states, and their results were inconclusive regarding their hypothesis that communities with higher proportions of community bank-originated PPP loans have lower bankruptcy rates.

The three-student SMU team was led by faculty advisor Dr. Eli Olinick. Melissa Serrano, Annalise Sumpon, and Madi Tedrow used regression and statistical models that found statistically significant differences in the business type and business ownership (e.g., racial minority or gender) for businesses that received PPP loans from community banks and/or non-community banks.

As the first-place winners of the 2022 CSBS Data Analytics Competition, the William & Mary team will collect $5,000. The second-place team from the University of California at Irvine will collect $2,500, while the third-place team from Carnegie Mellon got $1,500 and the fourth finalist team from SMU got $1,000.
CSBS’ Data Analytics Task Force provides oversight and direction for the annual CSBS Data Analytics Competitions. 

Archive - 2022 Challenge Details


Develop a data analytics model that demonstrates the role community banks played during the pandemic using data from the Paycheck Protection Program (PPP).

This competition is open to teams of 3-5 undergraduate and/or graduate students working with a faculty advisor.


The number of community banks in the United States has fallen from more than 13,000 in the mid-1980s to less than 5,000 today. These community-focused banks have consolidated mainly as a result of competitive pressures. Research shows that community banks are essential to maintaining economically fruitful communities, and losing these banks could be a significant blow to local infrastructure. 

One example of the importance of community banks was their role in distributing Paycheck Protection Program (PPP)1 loans during the Covid-19 pandemic. The PPP was designed to help small businesses keep their workers employed during the pandemic by providing funds through a short-term loan backed by the Small Business Administration (SBA). Preliminary research by CSBS shows that state-chartered banks were the primary distributor of PPP loans, and that community banks played an outsized role in the distribution of PPP funds. 



CSBS invites small teams of three to five undergraduate and graduate students at U.S. universities (working with a faculty advisor) help scholars, policymakers, and the public better understand the role of community banks during the COVID-19 pandemic. Using data provided by the SB, CSBS and elsewhere, teams should develop a hypothesis about the role community banks played during the pandemic, create a data analytics model to test their hypothesis and draw data-driven insights from their research.  

For selected proposals (see competition phases below), the final research project should consist of a written report that includes at least one of the following supporting data analytics components: 

  1. Statistical model 
  2. Optimization model 
  3. Business Intelligence tool/dashboard 
  4. Other data analytics component 



The winning teams will be selected by a panel of judges, with an emphasis on the following (see Appendix 1 for the scoring rubric): 

  1. Use of data and analytical methods/models 
  2. Creativity and innovative thinking 
  3. Persuasion through data-driven arguments 


Competition Phases 

  • Phase I – Proposal Submission: Teams should submit a detailed project proposal by Jan. 28, 2022, for eligibility to advance to Phase II. A proposal template is included for your reference in Appendix 2. Limit your proposal to two pages. Proposals must have a faculty sponsor. Send proposals to: [email protected]  
  • Phase II - Proposals Reviewed by CSBS: Up to eight teams will be selected to advance to Phase III. Selected teams will be notified by mid-February to submit final reports by mid-April. Proposals will be evaluated on your research idea (hypotheses), expected methodology and model(s), and presumed insights from your analyses. 
  • Phase III - Eight Teams Work on Competition: The eight selected teams will work on their reports and must submit their written research reports and data analytics components by April 22, 2022. Each team should email their reports/components to [email protected] 
  • Phase IV – Four Finalist Teams Make Oral Presentations: Four (finalist) teams from Phase III will make oral presentations (20 minutes, plus 10 minutes Q&A) to a panel of CSBS judges on May 5, 2022, to determine first through fourth place winners. (The winning team may also be invited to present their findings at the annual CSBS State and Federal Supervisory Forum in late May.) 


Eligible Participants and Prizes

The CSBS Data Analytics Competition is open to all undergraduate and graduate students at U.S. universities. Teams of three to five college students are required to be sponsored by a research faculty advisor at their university. Prize money will be awarded to the top four finalists as follows: $5,000 to the first-place team; $2,500 to the second-place team; $1,500 to the third-place Team; $1,000 to the fourth-place Team.  

Appendix (Scoring, Template, About)

Top Category

1129 20th Street, N.W., 9th Floor, Washington, DC 20036 | Tel. 202.296.2840 | Fax. 202.296.1928