«Research Division Federal Reserve Bank of St. Louis Working Paper Series Understanding the Accumulation of Bank and Thrift Reserves during the U.S. ...»
Figure 8: Cross-sectional Distribution of Bad, Nonperforming, and Nonaccruing Loans of Banks and Thrifts (2007:Q2–2010:Q2) Note: These ﬁgures represent the frequency of three ratios of bad loans to assets in the population of banks (top) and thrifts (bottom). Black bars identity the ﬁrst quarter used in our analysis (2007:Q1) and white bars identify the last quarter (2010:Q2). From left to right in each panel: (i) nonaccruing loans as a share of total assets, (ii) nonaccruing loans and nonperforming loans with payments due for 90 days or more as a share of total assets, and (iii) nonaccruing loans and nonperforming loans with payments due for 30 days or more as a share of total assets. We right-censored the histograms at 15 for banks and at 10 for thrifts. Source: Authors’ calculations based on Reports of Income and Condition and Thrift Financial Reports.
Figure 9: Cross-sectional Distribution of Excess-to-Required Reserves Ratio for Large and Small Commercial Banks (2008:Q2–2010:Q2)
Note: This ﬁgure represents the U.S. Treasury disbursement of CPP funds in billions of U.S. dollars, the disbursement net of repayments, and the number of beneﬁciary institutions. Source: Authors’ calculations based on U.S. Treasury Transaction Reports data.
Table 1: U.S. Reserves Requirements
Notes: Clustered (at the thrift level) standard errors and bootstrapped (for CLAD estimation) standard errors are reported in parenthesis. We deﬁne excess reserves and cash as funds over 110 percent of reserves requirements. Approximately 1.9 percent of the sample is left-censored at 0. r1yr-IOR is the diﬀerence between the 1-year return on U.S.
Treasury bills and the interest paid on reserves. Bad loans 1, 2, and 3 are nested, with bad loans 1 being the narrowest deﬁnition and bad loans 3 the broadest. Tier 1 cap ratio X Loan loss is the interaction eﬀect between the ratio of Tier 1 capital to risk-adjusted assets and the loan loss provision as a ratio to assets. The remaining variables are deﬁned in the text. *** p0.01, ** p0.05, * p 0.10.
Table 7: Elasticities of Excess Reserve: All Banks and Thrifts
Note: These coeﬃcients measure the elasticity of ER to changes in each of the covariates. Since the ﬁrst two covariates are logs and the dependent variable is a log, the log covariate elasticities are evaluated as dLog(y)/dLog(x). The second group of covariates are levels or ratios so they are evaluated as (dLog(y)/dx)X x-covariates. All variables are deﬁned in the text. *** p0.01, ** p0.05, * p 0.10.
Table 8: Role of Uncertainty: Tobit Regressions of Excess Reserves (log)—Banks by Size
Notes: This is the ﬁnal probit result from running a forward and backward stepwise procedure to determine the vector of signiﬁcant covariates used in calculating the propensity score. We estimate this probit at the level of BHC. TA refers to total assets; TL refers to total loans; ints55y5 is the interaction of real estate loan shares with real estate prices; top 40 indicates the BHC is among the largest 40 BHCs by assets. Federal Reserve Board is a dummy variable for whether any of the managers of the BHC have a current position on a regional Federal Reserve Board in the fall of 2008. y-y designates year over year. All covariates are signiﬁcant at the 5 percent conﬁdence level or better.
Table 12: Diﬀerence-in-Diﬀerences Analysis: Cash-to-Asset Ratios
AT T 1 is the diﬀerence between the cash-to-assets ratio in the target quarter and that in the previous quarter;
AT T 2 is the diﬀerence between the cash-to-assets ratio in a given quarter and the previous quarter. ** p0.05.
Table 13: Variables List and Correspondence from the CRs and TFRs
rcfd1406 Total Loans and Lease ﬁnancing receivables: svgl3936 Past Due 30-89 Days and still accruing, total ( ) Past Due 30-89 Days and Still Accruing rcfd1407 Total Loans and Lease ﬁnancing receivables: svgl3942 Past Due 90 Days or more and still
rcfd a223 Risk-adjusted assets (Net Risk-weighted Assets) svcc2375 Net Risk-Weighted Assets (CCR78) rcfd 7205 Risk-based Capital Ratio svcc7205 Risk-based Capital Ratio We thank audience members at the EEA Oslo 2011; AMES Seoul 2011; Midwest Macro 2011; CEA Calgary 2012; Santa Clara University; University of San Francisco;
UC Santa Cruz brown bag; the St. Louis Federal Reserve brown bag, and conference participants at the joint ECB-DBC Conference 2013. We would also like to thank George Bulman, Carlos Dobkin, Carl Walsh, Riccardo DiCecio, and David Wheelock for helpful comments. Hoda El-Ghazaly provided excellent assistance with data management and Judith Ahlers provided excellent editorial support. We are very grateful
to Ran Duchin for providing the underlying data used to proxy management quality. The working paper version of this article was previously circulated with this title: