Like security prices, retail deposit interest rates cluster around integers and "even" fractions. However, explanations for security price clustering are incompatible with deposit rate clustering. A theory based on the limited recall of retail depositors is proposed. It predicts that banks tend to set rates at integers and that rates are "sticky" at these levels. The propensity for integer rates increases with the level of wholesale interest rates and deposit market concentration. When banks set noninteger rates, rates are more likely to be just above, rather than just below, integers. The paper finds substantial empirical support for the theory's implications.RECENT EMPIRICAL RESEARCH HAS DOCUMENTED a widespread tendency for security prices to cluster around integers and "even" fractions of currency units. An example is the decreasing relative frequencies of stock price quotes and transactions made at integer-dollar, half-dollar, quarter-dollar, and eighthdollar price levels. Standard price theory has difficulty rationalizing such clustering because it depends on the arbitrary choice of the pricing system's unit of account. This has led to new theories which explain even-fraction clustering as an implicit convention that reduces traders' costs of negotiating security prices or as a collusive mechanism for maintaining anticompetitive bid-ask price spreads among security dealers. This paper reports a similar integer and even-fraction clustering of interest rates on retail bank deposits. Because theories of security price clustering are not applicable to the clustering of deposit interest rates, a new theory based on consumers' limited recall is proposed. In addition to explaining integer clustering, this theory has a number of other consequences. It predicts that noninteger deposit rates should be more frequently observed at rates just above integers rather than just below integers. It also predicts that retail deposit rates are stickier~less likely to change! when rates are quoted as even, rather than odd, fractions. Additional implications are that * Kahn and Pennacchi are from the University of Illinois, Sopranzetti is from Rutgers University. We are very grateful to

No abstract

The recent wave of bank mergers has raised concern over its effect on competition. This paper examines the influence of concentration and merger activity on consumer loan interest rates. It uses Bankrate, Inc. survey data on loan rates quoted weekly by large commercial banks in ten major U.S. cities during the period 1989 to 1997. The pricing behavior of banks is analyzed for two types of loans: new automobile loans and unsecured personal loans. Market concentration has a significant positive impact on the level of personal loan rates, but not on auto loan rates. Auto loan rates show little change around the time of significant bank mergers, suggesting that their relevant market is nationwide. However, consistent with mergers changing the market power and size structure of personal loan markets, the personal loan rates of banks involved in mergers show a significant decline relative to their market rivals during the period prior to the merger's completion. Merger participants' decision to lower personal loan rates could also reflect their desire to gain regulatory approval of the merger. The paper also tests for the existence of leader-follower relationships in loan pricing and finds that it is more widespread in markets for automobile loans. Interest rates on both types of loans respond asymmetrically to a change in equivalent maturity Treasury security rates, being more sensitive to a rise than a fall. In addition, personal loan rates are less responsive in more concentrated markets.

No abstract

Like security prices, retail deposit interest rates cluster around integers and "even" fractions. However, explanations for security price clustering are incompatible with deposit rate clustering. A theory based on the limited recall of retail depositors is proposed. It predicts that banks tend to set rates at integers and that rates are "sticky" at these levels. The propensity for integer rates increases with the level of wholesale interest rates and deposit market concentration. When banks set noninteger rates, rates are more likely to be just above, rather than just below, integers. The paper finds substantial empirical support for the theory's implications.RECENT EMPIRICAL RESEARCH HAS DOCUMENTED a widespread tendency for security prices to cluster around integers and "even" fractions of currency units. An example is the decreasing relative frequencies of stock price quotes and transactions made at integer-dollar, half-dollar, quarter-dollar, and eighthdollar price levels. Standard price theory has difficulty rationalizing such clustering because it depends on the arbitrary choice of the pricing system's unit of account. This has led to new theories which explain even-fraction clustering as an implicit convention that reduces traders' costs of negotiating security prices or as a collusive mechanism for maintaining anticompetitive bid-ask price spreads among security dealers. This paper reports a similar integer and even-fraction clustering of interest rates on retail bank deposits. Because theories of security price clustering are not applicable to the clustering of deposit interest rates, a new theory based on consumers' limited recall is proposed. In addition to explaining integer clustering, this theory has a number of other consequences. It predicts that noninteger deposit rates should be more frequently observed at rates just above integers rather than just below integers. It also predicts that retail deposit rates are stickier~less likely to change! when rates are quoted as even, rather than odd, fractions. Additional implications are that * Kahn and Pennacchi are from the University of Illinois, Sopranzetti is from Rutgers University. We are very grateful to