Panel and geospatial data for U.S. FDIC insured banks fiduciary activities and annual performance analyses over the periods 2016 to 2018

Created in 1934, the Federal Deposit Insurance Corporation (FDIC) is the independent agency of the United States Government tasked to protect depositors of insured banks located in the U.S. against the loss of their deposits in case of bank failure. To achieve this objective, the FDIC collects data on banks condition and income through quarterly Financial Report calls, which are then publicly released as “Statistics on Depository Institutions (SDI)”. The present data article, as a follow up to Niankara and Ismail, 2019 that focuses on U.S. banks’ exposure to foreign counterparty risk, describes an extract from the quarterly SDIs that is compiled into a panel of 16209 observations on 5403 U.S. FDIC insured Banks, observed over the three-year periods of 2016, 2017 and 2018. Since our objective is to bring forth a useful data source for analyzing U.S. FDIC banks fiduciary activities and annual performance changes over the past recent years, our constructed sample contains all FDIC insured banks with end of year (4th quarter) financial reporting for each of the 3 fiscal years 2016-2018. We further supplement this data with U.S. county and state level geospatial data that allow analysts and business researchers to address questions with temporal significance, but also spatial relevance through appropriate modelling and mapping of the data. Finally, we demonstrate the usability of the data using R based descriptive analytics, with computer codes provided for prospective Analysts and business researchers.


Data
The Federal Deposit Insurance Corporation (FDIC) is the independent agency of the United States Government tasked to protect depositors of insured banks located in the United States against the loss of their deposits in case of insured bank's failure [2]. To facilitate its oversight and regulatory missions, the FDIC collects Quarterly data on banks condition and income through its Thrift Financial Report (TFR) calls, which are filed with the Office of the comptroller of the Currency (OCC), as of the close of business on the last calendar quarter, in accordance with the Dodd-Frank Act [2]. The collected data is subsequently released to the general public by the FDIC. The current data availability at the FDIC's website covers from as early as 4th quarter 1992 to 4th quarter 2018 [3]. Each quarterly report is provided as a Zipped file containing 61 comma separated variables (.csv) data Files, describing the various aspect of FDIC Bank performances. Combined together, these 61 "csv" data files provide the full quarterly data of all reporting FDIC insured banks during that quarter.
Given our interest in extending the analysis in [7], which focused exclusively on U.S. global banks' performances outside of the country, we now turn our attention to U.S. banks' fiduciary activities and annual performances overtime inside the country, by focusing our data query on the 4th Quarter reports for the most recent three years of reporting, namely (Q4-2016, Q4-2017, and Q4-2018). After downloading these three zipped files of 61 "csv" data files each, we extracted from each unzipped folder, the 10 "csv" data files describing the aspect of FDIC banks fiduciary activities and performances of interest to us. These include: All_Reports_XXXX1231_Assets and Liabilities All_Reports_ XXXX1231_Performance and Condition Ratios Specifications Value of the data This data provides Analysts and Business researchers with a unique opportunity to investigate patterns, and trends in United States FDIC insured financial institutions' performances using descriptive analytics. The data is useful for research in areas including Banks performance, efficiency, and risk management; It also allows investigations of banks' fiduciary activities at the extensive and intensive margins, and their relationship with banks' performance, efficiency and risk management strategies. Researchers could use the data to address questions of temporal significance, but also of spatial relevance to the US banking industry. Finally, the data could also inspire other works, which might consider expending the data coverage to periods before 2016, while adding state and county level macro-economic data, for much richer analyses that put the banking sector within the overall context of the US economy. Where the "XXXX" represents the year index of the file (2016,2017,2018). From each of the 10 files we selected our variables of interest, and merged these variables into a single cross-sectional report file under the file name "Z_combXXXX" for each of the three years. In this way we had "Z_Comb2016", "Z_Comb2017", and "Z_Comb2018" representing the end of year reports of all FDIC insured banks during the fiscal years 2016, 2017 and 2018 respectively. Using Excel, these three cross-sectional data sets were then combined into a single Panel data in long form, named "Z_CombPanel" and spanning the three fiscal years 2016e2018. This excel file was then imported into the R statistical Software, and saved into an R object Named "CombFDICBanPanel" [1].
This version of the data is contained in "folder 0" of the attached supplementary materials. After further treatments of the data as described in the "material and methods" section below, the resulting final data "CombFDICBankPanelC" is retrievable from "folder 1" of the attached supplementary materials. "Folder 2" of the supplementary materials contains the state level aggregated data without (statesOutcomeDat) and with (statesOutcomeDat1) geospatial information, which are used for the mappings in Figs. 3e7. Although not explicitly described in the article, the county level aggregated data with (countiesOutcomeDat2) and without (countiesOutcomeDat) geospatial information are retrievable from "folder 3". "Folder 4" on the other hand, contains the original U.S. state level "gadm36_U-SA_1_sf.rds" and county level "gadm36_USA_2_sf.rds" spatial meta data downloaded from the GADM library for merging with the treated bank performances data. Finally the last folder "Folder 5" in the attached supplementary materials, contains the aggregated data summaries across respectively bank charter class "bkclassOutcomeDat", FDIC supervisory regions "fdicdbsOutcomeDat", asset concentration hierarchy "specgrpOutcomeDat", and data collection period/year "yearOutcomeDat".

Experimental design, materials, and methods
The subset of data (variables) described in the present article is extracted from the R data object "CombFDICBankPanel" in [1]. Using this initial data file, and the R statistical Software [4], we proceeded to further treatments of the data into its final version "CombFDICBankPanelC" presented in more details here. The applied data treatments included: Sub-setting the initial data "CombFDICBankPanel" to keep only the few variables discussed in the present article; Creating three new financial ratios: (TPIMATOTr, depiR, and depdomR); where "TPIMATOTr" is the ratio of managed assets in fiduciary accounts to total assets; "depiR" is the share of interest bearing deposits in total deposits; and "depdomR" is the share of domestic deposits in total deposits; Defining the qualitative variables as factors and saving the resulting data into the R data object "CombFDICBankPanelB"; Adjusting the bank's data "CombFDICBankPanelB" for compatibility with U.S. state and county level geospatial data, which we extract from the GADM database of Global Administrative Areas [5]. The adjusted data is then saved into the R data object "CombFDICBankPanelB2"; Incorporating descriptive labels for the generically named variables in the adjusted Panel Data "CombFDICBankPanelB2", and saving the final result into the R data object "CombFDICBankPanelC".
"CombFDICBankPanelC" is the final sample with 16209 observations on 5403 U.S. FDIC insured banks observed over a period of three years. From a statistical sampling standpoint, this sample is the Categorical Asset Concentration Hierarchy-An indicator of an institution's primary specialization in terms of asset concentration: 1 International Specialization: Institutions with assets greater than $10 billion and more than 25% of total assets in foreign offices. 2 Agricultural Specialization: Banks with agricultural production loans plus real estate loans secured by farmland in excess of 25% of total loans and leases. 3 Credit-card Specialization: Institutions with credit-card loans plus securitized receivables in excess of 50% of total assets plus securitized receivables. 4 Commercial Lending Specialization: Institutions with commercial and industrial loans, plus real estate construction and development loans, plus loans secured by commercial real estate properties in excess of 25% of total assets. 5 Mortgage Lending Specialization: Institutions with residential mortgage loans, plus mortgage-backed securities, in excess of 50% of total assets. 6 Consumer Lending Specialization: Institutions with residential mortgage loans, plus credit-card loans, plus other loans to individuals, in excess of 50% of total assets. 7 Other Specialized < $1 Billion: Institutions with assets less than $1 billion and with loans and leases are less than 40% of total assets. 8 All Other < $1 Billion: Institutions with assets less than $1 billion that do not meet any of the definitions above, they have significant lending activity with no identified asset concentrations. 9 All Other > $1 Billion: Institutions with assets greater than $1 billion that do not meet any of the definitions above, they have significant lending activity with no identified asset concentrations. Ratio Core capital (leverage) ratio -Tier 1 (core) capital as a percent of average total assets minus ineligible intangibles. Tier 1 (core) capital includes: common equity plus noncumulative perpetual preferred stock plus minority interests in consolidated subsidiaries less goodwill and other ineligible intangible assets. The amount of eligible intangibles (including mortgage servicing rights) included in core capital is limited in accordance with supervisory capital regulations. Average total assets used in this computation are an average of daily or weekly figures for the quarter. rbc1rwaj Ratio Tier 1 risk-based capital ratio -Tier 1 (core) capital as a percent of risk-weighted assets as defined by federal regulators for prompt corrective action during a time period. rbcrwaj Ratio Total risk-based capital ratio -Total risk based capital as a percent of risk-weighted assets as defined by federal regulators for prompt corrective action during a time period. Nimy Ratio Net interest margin -Total interest income less total interest expense (annualized) as a percent of the average of all loans and other investments that earn interest or dividends (Average Earning Assets). Noijy Ratio Net operating income to assets -Net operating income (annualized) as a percent of the Year-to-date average of the total assets represented on the balance sheet (average total assets). Average total assets in year-end quarter 4 (December) is calculated as: (Previous December earning assets þ March earning assets þ June earning assets þ September earning assets þ December earning assets)/5 ntlnlsr Ratio Net charge-offs to loans -Gross loan and lease financing receivable charge-offs, less gross recoveries, (annualized) as a percent of average total loans and lease financing receivables. Average total loans in year-end quarter 4 (December) is calculated as: (Previous December total loans þ March total loans þ June total loans þ September total loans þ December total loans)/5 elnantr Ratio Credit loss provision to net charge-offs -Provision for possible credit and allocated transfer risk as a percent of net charge-offs. If the denominator is less than or equal to zero, then ratio is shown as "NA." ernastr Ratio Earning assets to total assets ratio -Income before income taxes and extraordinary items and other adjustments, plus provisions for loan and lease losses and allocated transfer risk reserve, plus gains (losses) on securities not held in trading accounts (annualized) divided by net loan and lease charge-offs (annualized). This is a number of times ratio (x) (continued on next page) Ratio Net loans and leases to total assets -Loan and lease financing receivables, net of unearned income, allowances, and reserves, as a percent of total assets. depdastr Ratio Total domestic deposits to total assets -Total domestic office deposits as a percent of total assets. Eqv Ratio Equity capital to assets -Total equity capital as a percent of total assets. asset5 Numeric Year-to-date Average total assets ($) -Year-to-date average of the total assets represented on the institution's balance sheet. ernast5 Numeric Year-to-date Average earning assets ($) -The average of all loans and other investments that earn interest or dividends. eq5 Numeric Year-to-date Average equity ($) -The average of total equity capital (includes preferred and common stock, surplus and undivided profits). lnlsgr5 Numeric Year-to-date Average total loans ($) -The average of total loans and lease financing receivables, net of unearned income.  Note: The information tabulated here is extracted from the R data frame "yearOutcomeDat" in the attached supplementary material, which was produced by splitting the shared data "CombFDICBankPanelC" into its cross-sectional planes using "dplyr" [6], and then computing the mean and standard deviations (shown in parenthesis) for a select key performance measures. For each performance measure, the name of the data object capturing its computed mean in "yearOutcomeDat" is preceded by "BkAvrg" followed by the name of the variable itself as in the R data file "CombFDICBankPanelC". Similarly, the name of the data object capturing its computed standard deviation in "yearOutcomeDat" is preceded by "BkSd" followed by the name of the variable itself as in the R data file "CombFDICBankPanelC". (For more details, see the attached R computer codes). Note: The information tabulated here is extracted from the R data frame "specgrpOutcomeDat" in the attached supplementary material, which was produced by splitting the shared data "CombFDICBankPanelC" into its planes formed by banks' primary area of specialization, using "dplyr" [6],  Note: The information tabulated here is extracted from the R data frame "bkclassOutcomeDat"in the attached supplementary material, which was produced by splitting the shared data "CombFDICBankPanelC" into its planes formed by banks' charter class, using "dplyr" [6], and then computing the mean and standard deviations (shown in parenthesis) for the select key performance measures. (See the attached R computer codes, for more details).  The key experimental factors in the data are described in Table 1 and include: (i) FDIC insured banks demographic information, further summarized through the histogram representations in Fig. 1 for banks' distribution across asset concentration hierarchy and Fig. 2 for banks' distribution across bank charter class. The remaining experimental factors in Table 1 include (ii) Performance and Condition measures, and (iii) Annual Fiduciary Activities.
In addition to the description of the variables in Table 1, using the R library "dplyr" [6], we provide descriptive analytics of FDIC banks performances across key banks' demographic characteristics including: Years of reporting [see Table 2]; Asset concentration Hierarchy [see Table 3]; Bank charter class [see Table 4].