Unshrouding: Evidence from Bank Overdrafts in Turkey

• DOI: http://doi.org/10.1111/jofi.12593
• Published date: 01 April 2018
• Date: 01 April 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 2 •APRIL 2018

Unshrouding: Evidence from Bank Overdrafts

in Turkey

SULE ALAN, MEHMET CEMALCILAR, DEAN KARLAN,

and JONATHAN ZINMAN∗

ABSTRACT

Lower prices produce higher demand . . . or do they? A bank’s direct marketing to

holders of “free” checking accounts shows that a large discount on 60% APR over-

drafts reduces overdraft usage, especially when bundled with a discount on debit card

or autodebit transactions. In contrast, messages mentioning overdraft availability

without mentioning price increase usage. Neither change persists long after the mes-

sages stop. These results do not square easily with classical models of consumer choice

and firm competition. Instead, they support behavioral models where consumers un-

derestimate and are inattentive to overdraft costs, and firms respond by shrouding

overdraft prices in equilibrium.

MANY BUSINESS MODELS IN FINANCIAL services and other industries rely on expen-

sive add-ons that are tied to base goods. Examples include expensive overdraft

credit (add-on) tied to a “free” checking account (base good), back-end man-

agement fees tied to “free” investment advice, printer cartridges and printers,

luggage fees and airline tickets, and dealer-supplied maintenance and auto-

mobiles. A closely related practice is overage/penalty pricing. Businesses with

such revenue models typically focus their marketing and competitive strategy

on the base good, even though add-ons/overages can be critical revenue sources.

Retail banking provides a striking example: banks often market checking

accounts as free, even though many consumers end up paying high fees for

overdraft credit.1Government audits find that banks rarely market overdraft

services at the customer acquisition stage and even actively discourage em-

ployees from providing information on overdraft terms (General Accounting

∗Sule Alan is at the University of Essex and Koc University. Mehmet Cemalcılar is at the Koc

University. Dean Karlan is at the Northwestern University, IPA, J-PAL, and NBER. Jonathan

Zinman is at Dartmouth College, IPA, J-PAL, and NBER. We thank Yapi Kredi staff for their

cooperation; Michael Grubb, Ben Keys, David Laibson, Eva Nagypal, Josh Schwartzstein, Andrei

Shleifer, and audiences at Boston College, Harvard/MIT, Kellogg (Finance), LSE, UCL, University

of Maryland, University of Virginia, the Federal Reserve Bank of Philadelphia, the CFPB Research

Conference, and the NBER Law & Economics group for comments; and Benni Savonitto, Glynis

Startz, and Zachary Groff for research management and analysis support. Authors did not have

direct conflicts of interest but are associated with other entities concerned with overdraft fees.

Disclosure information is available online.

1An overdraft occurs if the checking account holder initiates a transaction that makes her

balance negative or more negative.

DOI: 10.1111/jofi.12593

481

482 The Journal of Finance R

Office (2008), Competition and Markets Authority (2014), Consumer Financial

Protection Bureau v. TCF National Bank (2017)). After acquiring customers,

banks in some markets including Turkey, the site of our experiment, blur the

line between positive and negative balances by highlighting for customers

an available-to-withdraw figure that adds the available overdraft credit line

amount to the checking account balance, while making information on disag-

gregated balances and finance charges more difficult to find. Turkish banks

during our study period basically drew attention to overdrafts only when pro-

moting the feature to existing customers. Yet, even those promotions did not

explicitly mention the price of overdraft credit.

Why would a bank hide information on overdraft costs? After all, a classically

rational consumer would simply infer that shrouded prices are high prices. But

recent behavioral theories show that shrouded and high prices can persist if

consumers tend to underestimate their add-on costs and firms cannot profit

from debiasing consumers with more transparent pricing or information about

competitors’ high add-on prices (Gabaix and Laibson (2006), Grubb (2015),

Heidhues, K¨

oszegi, and Murooka (2017)).2

Bank regulators respond to such practices in various ways. For example,

in the United States regulators require upfront consumer opt-in for debit card

and ATM overdrafts, in both the United States and United Kingdom regulators

caution banks against relying too heavily on overdraft revenue, and in Turkey

regulators cap overdraft prices.

But empirical evidence is lacking on key questions raised by theory and

policy. Do consumers actually underestimate costs of add-ons such as over-

drafts?, do firms actually have incentives to shroud add-on prices instead of

competing to debias consumers?, how do consumers allocate attention to add-

ons?, and how quickly does consumer learning about add-ons break a shrouded

equilibrium? In short, empirical evidence on what drives overdraft pricing,

advertising, and usage is limited and largely descriptive.

These questions are central to overdraft markets. Beginning in the 1990s,

overdraft revenue replaced monthly subscription fees as banks’ major source

of explicit income from checking accounts, shifting the pricing equilibrium

for retail banking in much of the world to “free if nonnegative balance, very

expensive if in overdraft.” In the United States, banks collect more than

$10 billion in overdraft revenue annually.In the United Kingdom, banks derive

almost as much income from overdrafts as from reinvesting checking account

deposits (Competition and Markets Authority (2014)). In Turkey, the site of our

experiment, the announcement of a price ceiling on overdrafts was immedi-

ately followed by a 1.4% reduction in bank share prices, with a 2.1% drop for

the most overdraft-reliant bank (Toksabay (2013)).

We worked with Yapi Kredi (YK), one of the largest banks in Turkey, to de-

sign a randomized direct marketing experiment that distinguishes between

2For related evidence on consumer perceptions of overdraft costs, see Armstrong and Vick-

ers (2012) and Stango and Zinman (2014). For related models of limited and reactive consumer

attention, see, for example, Gabaix (2014) and Bordalo, Gennaioli, and Shleifer (2015).

Unshrouding: Evidence from Bank Overdrafts in Turkey 483

Figure 1. Experimental design. (Color figure can be viewed at wileyonlinelibrary.com)

classical and behavioral models of add-on pricing and advertising. YK sought

to learn more about its optimal strategy for pricing and advertising its over-

draft product. The bank was particularly interested in understanding whether

its past promotional pricing and advertising content tactics were effective in

increasing demand, and if not why not. YK’s interest rate (60% APR) and prod-

uct design was in line with standard practices and regulations. As is common

in overdraft markets, the product was priced expensively relative to seemingly

close substitutes (like credit cards), and disproportionally to credit risk (as

recently found by Turkish authorities).

YK sought to target marginal overdrafters among its existing client base, and

hence the experiment varied the promotions that YK sent via SMS (text mes-

sage) from September to December 2012 to 108,000 existing checking account

clients who had not overdrafted during the previous few months. These clients

are likely representative of a substantial population of marginal overdrafters

in Turkey, and they share key characteristics with “banked” populations in

both more- and less-developed countries.

The experimental design produces random variation, both across clients

and over time, in overdraft prices and in messaging content, frequency, and

duration (see Figure 1). Our tests rely on overdraft usage comparisons across

groups receiving different promotions, since Turkish banks frequently use