+Offsite Detection of Insider Abuse and Bank Fraud Among U.S. Failed Banks 1989-2015- PDF
FDIC Center for Financial Research Working Paper No. 2017-06
John P. O'Keefe and Chiwon A. Yom

This Version: October 2017

ABSTRACT
We find evidence that material insider abuse and internal fraud were present in approximately 457 (37 percent) of the 1,237 U.S. failed commercial and mutual savings banks (hereafter, banks) between 1989 and 2015. Using a unique dataset of the incidence of insider abuse and internal fraud among U.S. failed banks we analyze the characteristics of failed banks with the ultimate goal of developing fraud detection models—parametric (logistic regression, Benford digit analysis) and non-parametric (neural networks). We obtain information on the incidence of insider abuse and internal fraud among failed banks from failing bank cases prepared for the FDIC Board of Directors, restitution orders (fines) supervisors assessed for bank employee fraud, and bond claims the FDIC made to recover fraud-related losses on failed banks. The supervisory data we use to quantify fraud among failed banks has not been used previously in published research and, we feel, provides more comprehensive information on fraud among failed banks than that available to academic researchers. Since fraudulent behavior lies outside the realm of rationale behavior modelled in economics, we develop a framework for internal bank fraud that provides rationale and support for our fraud detection models. This framework is based on previous studies of financial fraud and internal bank fraud in particular. We test this framework using regression analysis of the determinants of fraudulent behavior among failed banks between 1989 and 2015 and find that banks with insider abuse and fraud present overstated income and asset values, under-reported losses and consequently overstated net worth. The regression models of fraudulent failed banks provide information on the financial statement line items that can be used to identify fraud. We next use a recently developed second-order Benford digit test to identify those banks whose financial statements suggest tampering and purposeful misstatement. Our results suggest that material insider abuse and fraud at banks is detectable using Benford digit analysis of bank financial data for a period one-to-four years prior to failure. Unfortunately, we are unable to develop an accurate neural network model for fraud prediction.

JEL Codes: C45, G21, G28, M49
Keywords: Bank Failure, Benford’s Law, Fraud Detection

+Supervisory Discipline and Bank Capital Management: Evidence from Before, During and After the Crisis- PDF
FDIC Center for Financial Research Working Paper No. 2017-05
Gary Fissel, Stefan Jacewitz, Myron Kwast and Christof Stahel

This Version: June 2018

Previously circulated as "Does Depository Discipline Reduce Bank Risk? Evidence from Before, During, and After the Crisis"

ABSTRACT
This paper investigates the effectiveness of supervisory discipline on bank risk over the years immediately before, during and just after the recent crisis. It is the first study to consider the effects of informal supervisory enforcement actions in addition to formal actions. Informal enforcement actions are not only much more numerous than the formal enforcement actions used in previous studies, but they are also often confidential, whereas formal enforcement actions must be public. Access to this information allows a complete analysis of the effects of regulatory enforcement actions on bank capital. Pre-crisis, results strongly support the risk-reducing (capital enhancing) effects of informal actions and find that using only information on formal actions leads to substantial bias. During the crisis, formal actions became a much more effective tool for slowing declines in a bank’s capital ratios and informal actions were relatively less potent. Post-crisis, while it appears that the effects of enforcement actions are moving back toward the “normal” times of the pre-crisis period, the statistical relationship between supervisory discipline and target capital is less clear. In all three periods banks had strong incentives to achieve their capital targets while they were in the higher prompt corrective action capital zones. TARP capital helped quicken a bank’s adjustment speed to its capital target during the crisis, but appears to slow this speed post-crisis.

JEL Codes: G20, G21, G28
Keywords: Banks; Capital Regulation; Enforcement Actions

+Determining the Target Deposit Insurance Fund: Practical Approaches for Data-Poor Deposit Insurers- PDF
FDIC Center for Financial Research Working Paper No. 2017-04
John P. O'Keefe and Alexander B. Ufier

This Version: May 2017

Published as: O'Keefe, John and Alex Ufier. "Determining the Target Deposit Insurance Fund: Practical Approaches for Data-Poor Deposit Insurers." World Bank Group, FIRST Initiative, November 2017. Also available online.

ABSTRACT
In order for deposit insurers to be able to maintain public confidence following a bank failure, they must be able to act quickly to repay depositors and adequately fund the resolution of failed institutions. According to the International Association of Deposit Insurers, the best measure of deposit insurer funding adequacy is the Target Fund Ratio, the deposit insurance fund divided by total insured deposits, and each countries fund target will vary based on institutional needs. This paper presents a framework to assist countries, especially data-poor ones, in developing such a funding target. The paper employs a simulation approach that combines probability of default, loss given default, default correlation, and exposure at default to yield Target Fund Ratios required for the deposit insurer to remain solvent across a variety of different economic environments. This paper then uses the U.S. as an example country and compares results of the method to U.S. experiences and policy decisions.

JEL Codes: G21, G28
Keywords: Deposit Insurance, Bank Failure Prediction

+Lender-Borrower Relationships and Loan Origination Costs- PDF
FDIC Center for Financial Research Working Paper No. 2017-03
Philip Ostromogolsky

This Version: January 2017

ABSTRACT

Using a recently developed method of causal inference, this paper estimates the additional up-front loan origination costs that a small business can expect to pay when it first borrows from a new lender. I compare firms that borrow from a previously-unused financial institution with firms that borrow from a financial institution with which they have a preexisting financial relationship. I estimate that firms that borrow from a new financial institution can expect to pay $5,650 to $6,980 more in closing costs than firms that return to a previously-used financial institution. Based on these findings, I argue that a central function of origination fees is to pay for the production of detailed, firm-specific information that is valuable to the lender. I study a natural quasi-experiment wherein, for a small group of firms, selection into borrowing from a new lender is close to random. Returning to the wider population of small business borrowers, I use the method of Altonji, Elder, and Taber (2002, 2005) to account for endogeneity in firm's selection to borrow from a new lender. The method of Altoji, Elder, and Taber allows me to measure the degree to which a firm's selection to borrow from a new lender is driven by unobservables that also determine closing costs and to correct for any resulting bias. All analyses confirm that borrowing from a new financial institution causes firms to pay higher loan origination costs.

JEL Codes: G20, G21, G32, L26, C31
Keywords: small business finance, small business lending, small business loans, lending relationships, loan contracts, debt contracts, switching costs, information costs, causal inference, quasi-experiment.

+Does Deposit Insurance Promote Financial Depth? Evidence from the Postal Savings System During the 1920s - PDF
FDIC Center for Financial Research Working Paper No. 2017-02
Lee K. Davison and Carlos D. Ramirez

This Version: November 2016

ABSTRACT

This paper tests whether deposit insurance promotes financial depth by influencing depositor behavior. To do so, we rely on two schemes operating in the U.S. during the 1920s: the Postal Savings System and the deposit insurance schemes that some states had adopted. We exploit the discontinuity in deposit insurance across state borders and compute changes in postal savings deposits in cities located along the borders of states that did and did not have deposit insurance. We examine the relative growth of postal savings deposits in pairs of border cities when bank suspensions occurred within a short radius (10, 20, and 30 miles). Our results indicate that, following a bank suspension within a 10-mile radius, deposits in postal savings offices located in the non-deposit insurance state increased by 16 percent more than deposits in the neighboring postal savings office located in the deposit insurance state. The magnitude of the effect declines with bank suspension distance. It disappears when deposit insurance is not in effect. Using county-level data, we find that deposit insurance is associated with a 56 percent increase in local banking capacity.

JEL Codes: N12, N22, E44, G21, G23
Keywords: bank failures; Postal Savings deposits; deposit insurance; banking capacity

+Global Banks and Syndicated Loan Spreads: Evidence from U.S. Banks - PDF
FDIC Center for Financial Research Working Paper No. 2017-01
Edith X. Liu and Jonathan Pogach

This Version: November 2016

ABSTRACT

This paper explores the relationship between bank global exposure and their syndicated loan spreads. Linking syndicated loan information from Dealscan with confidential US bank foreign exposure data and borrower characteristics, we find that more bank global exposure is associated with a higher loan spread that is statistically significant. To analyze this relationship between global banks and loan spreads, we develop a theoretical framework where, in equilibrium, riskier borrowers are more likely to work with global banks. Using a Heckman selection model, we confirm that borrower risk characteristics indeed predict a higher likelihood of having a loan arranged by a global bank. However, after controlling for the bank-borrower selection effect, we continue to find that global banks charge on average an 12.7 bps higher spread on their syndicated loans, as compared with domestically focused banks. Finally, we explore a non-risk based alternative where firms with multinational operations are more likely to work with global banks for international banking services.