The Performance Effects of Regulatory Oversight Kristin Wilson and - - PowerPoint PPT Presentation

Kristin Wilson and Stan Veuger Harvard Business School & Harvard University Presentation for SBE / ARCS Conference May 10, 2011

The Performance Effects of Regulatory Oversight

Firms have a range of options in engaging with their regulators…

Arms-Length Collaborative

Increasing engagement

Managers report that they feel ineffective building “strong trust-based relationships” with key government stakeholders Managers disagree on whether enforcement relationships contribute positively or negatively to firm value

27% Effective 73% Not Effective 41% Positive 44% Negative

How does regulatory engagement affect firm performance?

Arms-Length Collaborative Evade resource-constrained regulator Increase costs of rent-seeking for managers and bureaucrats Collude with regulator Capitalize on regulator’s lower monitoring costs Reduce uncertainty about enforcement environment Knowledge transfer

How does regulatory engagement affect firm performance?

Cost of compliance Q1: Does increasing regulatory engagement increase the administrative burden of compliance? No. Compliance levels Q2: Does increasing regulatory engagement increase compliance? No. Knowledge transfer Q3: Does increasing regulatory engagement increase firms’

• perational efficiency?

Yes.

Application to US Commercial Banking

Regulation

 Capital controls and reporting requirements

Oversight

 Periodic on-site exams for “Safety and Soundness”

 Loan-level analysis  Management evaluation  Examiner discretion  Highly confidential

 Continuous monitoring between exams

Measuring engagement

We need a proxy for engagement with regulators that is not endogenous to financial performance … We use physical distance as an exogenous measure

• f cost of engagement between firms and regulators

Regulator Field Office (FO) Bank Headquarters (HQ) Travel time (minutes)

Support for use of distance proxy

• 1. Economic geography literature
• 2. Evidence from examination frequency
• 3. Evidence that supervisors perceive proximity as lowering

barriers to engagement

Support for use of distance proxy

• 1. Economic geography literature

 Physical proximity facilitates information exchange, knowledge

transfer, relational contracting

(Rajan & Petersen, 2002; Coval & Moskowitz, 2002; Malloy, 2005)

 Physical distance increases monitoring costs

(Kedia & Rajgopal, 2011; DeFond, Francis & Hu, 2011)

• 2. Evidence from examination frequency
• 3. Evidence that supervisors perceive proximity as lowering

barriers to engagement

Support for use of distance proxy

• 1. Economic geography literature
• 2. Evidence on examination frequency

 Period between exams 3 months longer for banks 2 hours more

distant, controlling for bank characteristics

 Discretionary variation in between-exam periods is 12 months

• 3. Evidence that supervisors perceive proximity as lowering

barriers to engagement

Support for use of distance proxy

• 1. Economic geography literature
• 2. Evidence on examination frequency
• 3. Evidence that supervisors perceive proximity as lowering

barriers to engagement

“Because state banks in Tennessee have closer geographical proximity to their primary regulator, communication is more direct and more effective…any institution may call staff members or the commissioner with questions or concerns and get a personal audience quickly. We encourage…close contact with us; no problem is deemed unimportant.” Tennessee Department of Financial Institutions

Problems with distance proxy

1.

Omitted variables

2.

Endogeneity of location choice

… we address both of these concerns through our empirical approach

Empirical Approach: Overview

Question DV

Q1 Is increasing supervisor distance associated with higher

administrative burden? Admin Costs

Q2 Are differences in performance due to risk-taking at

distant banks? Leverage, NPL, NIM What is the net effect on performance? ROE … followed by discussion of evidence on efficiency gains at co-located banks (Q3) and additional evidence to support conclusions

Empirical Approach: Data

Sample: US Commercial Banks, 2001-2009

 Unbalanced panel, around 3,500 banks per year  Bank size between $100 Million and $10 Billion  Predominantly community / regional banks  Excludes rural banks  Excludes AK, HI, RI, DC

Sources

 Financial data - Call Reports  Local area data - BEA, BLS, Census  Distance measured between banks and regulators using ArcMap

Multiple regulators

 Division of power between federal and state authorities  Banks select into one of three supervision regimes

Institution type Supervision National Banks (NB) OCC State Member Banks (SB-M) State Authorities (SBA) Federal Reserve Bank State Non-Member Banks (SB-NM) State Authorities (SBA) FDIC

Map of Regulatory Agency Offices

(FRS)

Map of Regulatory Agency Offices

(FRS + FDIC)

Map of Regulatory Agency Offices

(FRS + FDIC + OCC)

Map of Regulatory Agency Offices

(FRS + FDIC + OCC + SBA)

FDIC OCC SBA FRS

60 minutes 120 minutes 45 minutes 20 minutes

Empirical Approach: Identification

National Bank State Bank MSA

Econometric Specification

Dependent Variables:

 Admin Costs % Capital  ROE  Leverage  NPL ratio  Net Interest Margin

X:

 Firm Characteristics  Portfolio Composition  Economic environment  MSA fixed effects  Field office fixed effects

Empirical Approach: Controls

Firm Controls

 Age  BHC indicator  “Main enterprise”  Acquisition history  SEC registration  Foreign Indicator

Portfolio Composition

 Asset size  % Cash, Earning assets,

Loans

 Loan portfolio weights  % Deposits  Leverage, RWA ratio  NPL ratios, Capital

adequacy ratio

Empirical Approach: Controls

Economic Environment

 Unemployment rate  Unemployment growth  Labor force growth  Average local market share  Average local HHI  State exposure

All indicators deposit weighted at MSA/county-level and aggregated to institution level Fixed Effects

 Year  MSA (headquarters)  Field office

Endogeneity concerns

Agency choice

 Instruments: indicator for geographically closest field offices;

state land area (sq mi)

Distance to Agency

 Large multi-state banks are excluded from sample, banks to regional

deposit base

 Highly stable location (bank and FO), so MSA fixed effects should

minimize this concern

Key Variable = Agency Choicei,t * Distance to Agencyi,t

Administrative burden

…is higher at distant banks

Admin Costs (b) Distance to OCC x National Bank 0.0131*** Distance to SBA x State Bank 0.0067 Distance to FDIC x State Non-Member 0.0100** Distance to FRS x State-Non Member 0.0069 Total distance, firm-level, portfolio, and market controls

• included. MSA, State, FO Fixed Effects included.

Observations: 23,020

Costs as a percentage of capital (includes legal, auditing, telecommunications, data, consulting and insurance fees)

Evasion? Evidence of Risk-taking

 Leverage does not vary within regime  Non-performing loans (NPL) do not vary

within regime

 Net interest margin (NIM) same or lower at

more distant banks … implying that more distant banks experience lower returns without any offsetting benefits

High costs to low ROE

Administrative burden

… results in reduced profitability

Cost Ratio ROE (b) (b) Distance to OCC x National Bank 0.0131***

• 0.0116*

Distance to SBA x State Bank 0.0067 0.0112 Distance to FDIC x State Non-Member 0.0100**

• 0.0132**

Distance to FRS x State-Non Member 0.0069

• 0.0058

Total distance, firm-level, portfolio, and market controls included. MSA, State, FO Fixed Effects included. Observations: 23,020

More evidence

 Knowledge transfer

Are co-located banks more efficient?

 Industry collusion

Is this just evidence of “captured” regulators?

 Banking Crisis (2008-2009)

How do industry shocks affect regulatory engagement?

 Technology trend

Does technology adoption reduce the importance of “distance” as a measure of engagement?

Wrap-up

 In this setting, regulatory engagement has two measurable

benefits for firms

 Lower costs as closer engagement facilitates information

exchange

 Improved use of resources as firms draw on regulatory

expertise

 Evidence of heterogeneous agency behavior

 Contribution to our understanding of the role of ongoing

regulatory relationships in firm performance

“We Should Do This More Often”