The Post Keynesian theory of the banking firm originally

The Post Keynesian theory of the banking firm, originally developed by Oreiro (2004) and Silva and Oreiro (2007), advances in the analysis of the determinants of bank spread in Brazil as it demonstrates that “a permanent pilocarpine hydrochloride of banking spreads can be obtained through a policy of lower interest rates and/or through a change in social conventions regarding the “safe” or “normal” value of the interest rate” (Silva and Oreiro, 2007, p. 43).
Silva et al. (2007) make the macroeconomic aspects more relevant because they consider the history of macroeconomic instability of the Brazilian economy and its high levels of bank spreads in international terms. An extremely used variable to measure the stability of the Brazilian economy is the interest rate, which is believed to be largely responsible for the high levels of bank spread. Below this comparison is made, where it is found that, in fact, the trajectories of these two variables have similar behavior (see Graph 1).
Empirically, an important part of the papers on the determinants of bank spread in the Brazilian economy uses the two-stage method of Ho and Saunders (1981) to separate the macroeconomic determinants of the microeconomic. An example of a study using this methodology is developed by Afanasieff et al. (2002), which makes use of panel data to uncover the main determinants of bank spread in Brazil. In this paper, the authors take into account the variables that address the characteristics of banks, macroeconomic conditions, the deposit insurance regulation, overall financial structure and the legal and institutional indicators. The results show that “the pure spreads are sensitive to both, market structure and volatility effects, and also that the effects are quite heterogeneous across countries” (Afanasieff et al., 2002, p. 10). Note that despite the downward trend observed in the banking spread in Brazil, these rates remain extremely high when compared to international standards and that this difference cannot be explained by inflation, as shown in Chart 2.
The work developed by Nakane (2001) and Belaisch (2003) closely follow the microeconomic approach. Both sought, although using different methodologies to analyze the determinants of bank spread in Brazil. Their conclusions are similar: the market structure in the Brazilian banking sector is not in any of the two extremes, i.e., corresponds neither to a monopoly/cartel nor to a model of perfect competition. Thus, the structure turns out to be not perfectly delimited, being acknowledged as a structure of oligopoly, since studies show evidence of some market power in this industry. To Belaisch (2003), the Brazilian banking system has a non-competitive market structure, which may be the explanatory factor for the low rate of banking intermediation and for the relative inefficiency of the Brazilian banking sector. Belaisch\’s (2003) hypothesis is that a non-competitive market structure may justify high rates of bank spread, discouraging deposit volumes and higher loans. The results show that Brazilian banks have oligopolistic behavior, given that the hypotheses of monopolistic behavior and a perfectly competitive demeanor in the banking system are rejected.
Nakane (2001) also conducts an empirical test that allows him to analyze the degree of competition in the Brazilian banking sector. The result found was that the market structure of this industry is highly competitive, although not perfectly competitive, but it is also not a monopoly or cartel as advocated by the general thought. The methodology used to identify and measure the degree of competition in an industry was an adaptation of the methodology developed by Bresnahan–Lau which seeks to test the significance of market power in banking intermediation in Brazil. The demand function for long-term bank loans is represented by the following expression:where L is the aggregate amount of bank loans in real terms, r is the interest rate of the loan market in real terms, Y is an indicator of economic activity, and α1, α2 and α3 are coefficients to be estimated. The interaction term between r and L rotates the demand curve for bank loans, which allows the identification of the parameter of market power.