Examining Bank Competition in Florida. Using Critical Mass as an Improved Competition Measure

Research Paper (undergraduate), 2017
6 Pages, Grade: 8.0



This paper analyses bank competition in Florida by examining and comparing several market concentration measures. Previous research showed that using the HHI in market power models can lead to erroneous results, a new measure was introduced: the critical mass. The results for the HHI and the C3 ratio’s lead to the conclusion that there is a low concentration in the market, while the results for the critical mass show that all banks in the market have power and that the market is highly concentrated. These conflicting results cannot be completely explained by changes of key variables over time. An extension of the data set with more recent data is needed to determine which concentration measure is right.

1. Introduction

The Hirschman-Herfindahl Index (HHI) is a commonly accepted measure of market con- centration. An HHI of zero points implies that the firms within the market have an equal size, the HHI increases when concentration increases and reaches its maximum of 10,000 points when a single firm controls the entire market. Agencies, like the US Department of Justice, use this measure as an input for many of their decisions. For example, in the Department of Justice Horizontal Merger Guidelines, it is indicated that mergers that result in an increase in the HHI of 200 points or more will significantly increase market power and thus warrant scrutiny (The US Department of Justice, 2010). Hence, their guidelines rest on the assumption that the HHI is an accurate measure of market concentration. Roberts (2014) discusses several critiques on the HHI. Firstly, measurement errors in determining market shares can have a large impact on the value of the HHI. Secondly, it does not consider dynamic aspects of competition, for example the entry of new competitors. Lastly, it is not industry-specific, so it does not take scale economies, entry barriers, or other factors into account. Agencies using the HHI are aware of these flaws and correct for it by including these other factors in their analyses.

The HHI is also often used in empirical estimations of market power. Bos et al. (Forthcoming) show that the derivation of the HHI rests on assumptions that can result in an omitted variable bias and an aggregation bias in these market power models.1 They derive an alternative competition measure, critical mass, that solves for these biases.

This paper examines banking competition in Florida by measuring and comparing the HHI, critical mass, and several other competition measures. First, the model and resulting competition measure calculations are explained. Second, the model is applied to banks in Florida and the results are given. Third, the results are discussed and the paper is concluded.

2. The model

Calculating the HHI is simple; it is the sum of all squared market shares in a certain market. Measuring the critical mass is however less obvious and requires the following specification which is taken from Bos et al. (Forthcoming):

illustration not visible in this excerpt

The Lerner index of a bank is thus related to its market share (θ i,t), its conjectural variation (λ i,t), and the interaction between the market share and the conjectural variation. Setting the partial derivative of (1) with respect to λ i,t equal to zero results in the point at which the market turns into an oligopoly. This point is defined as the critical mass: the market share at or beyond which firms collude. It is specified by:

illustration not visible in this excerpt

Estimating the conjectural variation (λ i,t) is critical but problematic, it is often treated as an omitted variable. Bos et al. (Forthcoming) found a solution to the problem and measure it using dynamic reallocation, they arrive at the following end-result:

illustration not visible in this excerpt


[illustration not visible in this excerpt]: Lerner index/markup

[illustration not visible in this excerpt]: the percentage change in firm i ’s market share at time t

[illustration not visible in this excerpt]: price elasticity of demand, which is estimated to be -0.15 (Bos et al., Forthcoming)

To account for endogeneity, equation (1) is estimated using first differences. The controls include measures of risk, earning assets, costs, and instruments for the third and fourth lags of [illustration not visible in this excerpt], λ i,t, and [illustration not visible in this excerpt]:

illustration not visible in this excerpt

3. Results

The data set was constructed by using FDIC call reports and consists of 26,782 firm- year observations over the time period 1984-2004 for the state of Florida. After removing duplicate observations and creating the lag variables 3,389 firm-year observations remained. In this section the common competition measures are discussed first, followed by an analysis of the critical mass and a comparison between the critical mass and the other measures.

3.1. Descriptives and the HHI

Table 3 shows the mean and standard deviations for the main variables of interest. The average market share in Florida is 0.03 percent, the Lerner index is close to 10 percent on average. The mean of the conjectural variation is close to -0.97 indicating that the banking market in Florida is close to perfect competition.

Table 1: Descriptive Statistics

illustration not visible in this excerpt

Considering the estimated conjectural variation, a low market concentration is expected. This concentration can be determined by calculating the HHI and the C3 ratio; the outcomes are depicted in Figure 1. The HHI is scaled in such a way that a value of 0.1 indicates a 1,000 points out of a possible 10,000. The C3 ratio is higher than expected; perfect competition would result in C3 ratio’s close to zero. However, a C3 ratio between 0 and 50 percent is still considered as an indicator of low concentration. This is confirmed by the HHI; it never reaches the value of 1500 which is classified by the US Department of Justice as the point at which a market becomes moderately concentrated. Thus, the conjectural variation, C3 ratio, and the HHI all show that there is a low concentration in the banking market in Florida.


1 They base their analysis on the model developed by Cowling and Waterson (1976) and Cowling (1976).

Excerpt out of 6 pages


Examining Bank Competition in Florida. Using Critical Mass as an Improved Competition Measure
Maastricht University
Financial Markets
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
442 KB
The author of this text is not a native English speaker. Please excuse any grammatical errors and other inconsistencies.
examining, bank, competition, florida, using, critical, mass, improved, measure
Quote paper
Colin Tissen (Author), 2017, Examining Bank Competition in Florida. Using Critical Mass as an Improved Competition Measure, Munich, GRIN Verlag, https://www.grin.com/document/376746


  • No comments yet.
Read the ebook
Title: Examining Bank Competition in Florida. Using Critical Mass as an Improved Competition Measure

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free