Debit Card Interchange Regulation and the Pricing of Deposit - PowerPoint PPT Presentation

Debit Card Interchange Regulation and the Pricing of Deposit Accounts: Evidence from the Durbin Amendment April 2016, Amsterdam Manuszak Wozniak Federal Reserve Board The views expressed here are those of the presenter and do not necessarily reflect the view of the Board of Governors or its staff

Background • U.S. debit card industry – Fastest growing and largest non ‐ cash retail payment method in the United States – Over 50 billion transactions in 2014 – Over $16 billion in interchange fees in 2014 • Section 1075 of the Dodd ‐ Frank Act (the Durbin Amendment) instructed the Board to issue regulations to establish “standards for assessing” whether debit card interchange fees are “reasonable and proportional” to certain costs of certain debit card issuers • Regulation II issued in June 2011, effective as of October 1, 2011 – Debit card interchange fees capped for covered issuers (more than $10 billion in assets) – Cap: $0.21 + 5 bps of the transaction value, $0.01 adjustment for fraud protection • Pre ‐ regulation per ‐ transaction average: $0.44 • Post ‐ regulation per ‐ transaction average: $0.24

Overview • Research questions – How did covered issuers alter their account pricing in response to lower interchange revenues following Reg II? (non ‐ linear pricing) – Did exempt issuers also alter their pricing, even though they were not directly affected by Reg II? (equilibrium effects) – What lessons should policymakers draw from the issuers’ response to Reg II? • Contributions – Prior to Reg II there was disagreement over how issuers would react to an interchange fee cap – this paper aims to provide a clear answer – Reg II provides a unique opportunity to look at competition/pricing within a two ‐ sided market – Insight into non ‐ linear pricing practices – Shows how to account for equilibrium pricing effects among banks

Theory I: How should we think about interchange fees? • Theoretical work ( Baxter 1984, Rochet and Tirole 2002) – Justifies the existence of interchange fees by the need to balance customers’ incentives to use cards and merchants’ incentives to accept them – Intuitively, interchange fees flow towards the side of the market with more elastic demand for card transactions • Theoretical prediction – Lower interchange fees = higher per ‐ transaction costs to issuers  higher fees for consumers

Theory II: How do issuers maximize profits? • Multiproduct firms – Deposit accounts/services – Loans – Credit cards • Bundled products – Deposit account – Payment services (check, debit card, ACH, ATM access) • Complex, nonlinear pricing – No per ‐ transaction fee – Free accounts / monthly account fee – Min balance to avoid fee / fee cannot be avoided

Data • RateWatch: Periodic survey of deposit/loan rates, terms, and fees at a sample of bank branches based on customer requests – We focus on terms and fees for three types of deposit accounts: (1) non ‐ interest checking, (2) interest checking, and (3) savings • Data structure: Quarterly bank ‐ level panel dataset – Observation: Average of a variable across all of a bank’s observed branches in a given quarter – Time period: Q1 2009 – Q2 2014

Empirical approach • Expected monthly fee   0 if FreeAccount 1     E  Fee MonthlyFee if FreeAccount 0 and AvoidFee 0        MonthlyFee P Balance MinBalance if FreeAccount 0 and AvoidFee 1  • Empirical approach: analyze each variable separately – FreeAccount – MonthlyFee – AvoidFee – MinBalance

Specification and identification challenges ��� �� �� �� � ��� ��� where w ait = 1 for checking account a at covered bank i when Reg II is in effect Conditions other than Reg II changed: θ at • – Cannot simply use prices before and after Reg II – Solution: Use (exogenous) variation associated with regulatory exemption to account for changing market conditions Big banks generally differ from small banks: γ ai • – Cannot simply compare (checking) accounts at covered and exempt banks after Reg II – Solution: Use observations over time to account for systematic pricing heterogeneity across banks Big banks experience different shocks over time relative to small banks: λ it • – Cannot only use checking accounts at covered and exempt banks before and after Reg II – Solution: Use observations for “exempt products” (savings accounts) at covered banks over time to account for bank/time ‐ specific shocks

Result: “Triple difference” panel data approach to identify changes in pricing • Compare fees/terms 1. Pre vs. post Reg II (time) 2. Covered vs. exempt (banks) 3. Checking vs. savings (products) • Difference ‐ in ‐ difference alternatives – Pre vs. post, covered vs. exempt, checking only  Fails to account for big bank ‐ specific shocks over time – Pre vs. post, checking vs. savings, covered only  Fails to account for product ‐ specific shocks over time

Results: Base regressions Account type FreeAccount MonthlyFee AvoidFee MinBalance Non ‐ interest checking ‐ 25.2%*** $0.94*** ‐ 1.9% $234.7 β NI 30.0% $7.68 86.7% $1,098.9 Post ‐ Reg II mean Interest checking 0.6% $1.55*** ‐ 1.3% $1,057.3*** β I 8.3% $13.18 97.1% $4,570.3 Post ‐ Reg II mean R ‐ squared 0.91 0.94 0.88 0.87 Observations 51,491 37,809 37,788 36,381 Banks 4,437 4,284 4,286 4,263 NOTE: Coefficients on Reg II*covered*account type reported, all dummy variables are suppressed. Includes two period leads and lags for effect of regulation. Post ‐ Reg II mean applies to covered banks. Robust standard errors in parentheses clustered at the bank level: *** p<0.01

Results: Base regressions Account type FreeAccount MonthlyFee AvoidFee MinBalance Non ‐ interest checking ‐ 25.2%*** $0.94*** ‐ 1.9% $234.7 β NI 30.0% $7.68 86.7% $1,098.9 Post ‐ Reg II mean Interest checking 0.6% $1.55*** ‐ 1.3% $1,057.3*** β I 8.3% $13.18 97.1% $4,570.3 Post ‐ Reg II mean R ‐ squared 0.91 0.94 0.88 0.87 Observations 51,491 37,809 37,788 36,381 Banks 4,437 4,284 4,286 4,263 NOTE: Coefficients on Reg II*covered*account type reported, all dummy variables are suppressed. Includes two period leads and lags for effect of regulation. Post ‐ Reg II mean applies to covered banks. Robust standard errors in parentheses clustered at the bank level: *** p<0.01

Results: Base regressions Account type FreeAccount MonthlyFee AvoidFee MinBalance Non ‐ interest checking ‐ 25.2%*** $0.94*** ‐ 1.9% $234.7 β NI 30.0% $7.68 86.7% $1,098.9 Post ‐ Reg II mean Interest checking 0.6% $1.55*** ‐ 1.3% $1,057.3*** β I 8.3% $13.18 97.1% $4,570.3 Post ‐ Reg II mean R ‐ squared 0.91 0.94 0.88 0.87 Observations 51,491 37,809 37,788 36,381 Banks 4,437 4,284 4,286 4,263 NOTE: Coefficients on Reg II*covered*account type reported, all dummy variables are suppressed. Includes two period leads and lags for effect of regulation. Post ‐ Reg II mean applies to covered banks. Robust standard errors in parentheses clustered at the bank level: *** p<0.01

Results: Base regressions Account type FreeAccount MonthlyFee AvoidFee MinBalance Non ‐ interest checking ‐ 25.2%*** $0.94*** ‐ 1.9% $234.7 β NI 30.0% $7.68 86.7% $1,098.9 Post ‐ Reg II mean Interest checking 0.6% $1.55*** ‐ 1.3% $1,057.3*** β I 8.3% $13.18 97.1% $4,570.3 Post ‐ Reg II mean R ‐ squared 0.91 0.94 0.88 0.87 Observations 51,491 37,809 37,788 36,381 Banks 4,437 4,284 4,286 4,263 NOTE: Coefficients on Reg II*covered*account type reported, all dummy variables are suppressed. Includes two period leads and lags for effect of regulation. Post ‐ Reg II mean applies to covered banks. Robust standard errors in parentheses clustered at the bank level: *** p<0.01

Additional challenge: Equilibrium effects ��� �� �� �� � ��� � ��� ��� where c ait is the competitive exposure of exempt bank i to covered banks when Reg II is in effect Exempt banks compete with covered banks and may respond to price • changes (induced by cost shocks) at covered banks Solution: Include variable accounting for exposure of exempt banks to • competition from covered banks – Share of market ‐ level deposits held by covered banks averaged across all local banking markets in which an exempt bank operates – Ranges from 0 (not exposed) to 1 (fully exposed) Introduces a “fourth difference” • 1. Pre vs. post Reg II (time) 2. Covered vs. exempt (banks) 3. Checking vs. savings (products) 4. “High vs. low” competition with covered banks (exempt banks)

Results: Equilibrium effects Account type FreeAccount MonthlyFee AvoidFee MinBalance Non ‐ interest checking  NI ‐ 27.4%*** $1.05*** ‐ 2.0% $311.8  NI ‐ 13.5%*** $0.54 ‐ 4.2% $549.1 Interest checking  I 0.0% $1.70*** ‐ 1.4% $1,240.9***  I ‐ 7.7%** $0.90** ‐ 1.3% $1,125.0*** 0.910 0.941 0.882 0.874 R ‐ squared 50,318 37,270 37,249 35,883 Observations 4,358 4,219 4,221 4,198 Banks NOTE: Coefficients on Reg II*covered*account type reported, all dummy variables are suppressed. Includes two period leads and lags for effect of regulation. Robust standard errors in parentheses clustered at the bank level: *** p<0.01, ** p<0.05, * p<0.10