Bank Resilience and the Architecture of Market based Intermediation
Description
James Tarr, DBA, is a faculty leader at Colorado Technical University with expertise in economics, management, operations, and finance. He holds a BSBA in Economics from The Ohio State University, an MBA in Accounting, and a DBA in Operations Management from Cleveland State University. His research interests include financial stability, banking and market-based intermediation, online student success, and innovative approaches to teaching and learning. Dr. Tarr is particularly interested in translating research into practical strategies that strengthen organizational decision-making, financial-system understanding, and student outcomes. Patrice is a full-time instructor of Finance at American Intercontinental University where she has worked for 21 years. She currently teaches all finance courses at the Masters, Bachelors, and Associate level as well as has developed curriculum in all finance related courses. She also has developed courses at the Doctoral level for California Southern University. Patrice obtained her DBA in Finance in 2017 from North Central University, and her BS and MBA in Finance from St. John's University, New York. Patrice has also held various NASD/FINRA principal and representative licenses throughout her 20-year tenor on Wall Street at Citigroup and Prudential. Her roles included: the Controller of the Debt Capital Markets Division (Citigroup), the Chief Financial Officer of the Prudential Asset Backed Subsidiaries, and the Business Unit Manager of various bond trading desks (both firms). Additionally, while in various capacities throughout her Wall Street career, she facilitated the college internship program where she mentored high school and college students. Her research on higher-ed has been published by Cogent Education, and her finance research has been published by ProQuest. Various articles related to financial regulation are being prepared for publication. Patrice lives on the ‘Jersey Shore' with her husband, two daughters, and her poodle. She enjoys cooking, traveling, dancing, and volunteering in various community endeavors.
Files
Steps to reproduce
The dataset was constructed from publicly available U.S. financial and regulatory data, primarily the Federal Reserve’s Financial Accounts of the United States (Z.1) and related Federal Reserve/FRED series. Quarterly observations cover 2013Q1–2025Q4 for the primary analysis, with 2009Q4–2025Q4 used for robustness testing. Sector-level series for depository institutions, broker-dealers, money market funds (MMFs), mutual funds, and other intermediaries were used to construct measures of repo-market participation and relative intermediary scale. Bank resilience was measured using the aggregate Tier 1 capital-to-risk-weighted-assets ratio. Raw series were aligned quarterly and transformed into the ratios and institutional shares used in the analysis. Repo measures were constructed from applicable Financial Accounts federal-funds-and-repurchase-agreement series. Relative-scale measures include MMF assets relative to private depository-institution assets and GDP. An alternative MMF repo measure excluding Federal Reserve Overnight Reverse Repurchase Agreement (ON RRP) activity was also constructed. Variable definitions, source series, and transformations are documented in the data dictionary. Data-quality procedures included verification of source definitions, quarterly alignment, reconciliation of constructed measures to underlying series, and auditing of labels and transformations. A previously mislabeled broker-dealer repo-liability series was identified and corrected, with the correction documented in the replication workbook. Manuscript values, tables, and principal regression estimates were reconciled against the final dataset. The principal dynamic analysis uses quarterly changes and distributed-lag models relating financial-architecture measures to contemporaneous and lagged changes in Tier 1 capitalization, with lagged dependent variables and HAC/Newey-West inference. Robustness analyses include alternative HAC bandwidths, compositional transformations, an extended sample, ON RRP adjustment, structural-stability diagnostics, leave-one-quarter-out analysis, split samples, multiple-testing adjustments, and power calculations. The accompanying Master Research Dataset and replication workbook contains the quarterly data, derived variables, data dictionary, model results, robustness analyses, and documentation needed to reproduce the study from the identified public source series.