Post Keynesian endogenous money theory: A theoretical and empirical investigation of the credit demand schedule

ABSTRACT

The article investigates the relationship between interest rates and loan amounts provided by commercial banks from both a theoretical and an empirical perspective. Theoretically, some scholars belonging to the post Keynesian endogenous money tradition advocate that a decrease (increase) in interest rates leads to a positive (negative) effect on the amount of loans demanded by households and firms. On the other hand, some heterodox economists maintain that interest rates do not stimulate firms’ credit demand but that a certain degree of influence is allowed for loans provided to households. By applying a vector autoregression (VAR) and vector error-correction model (VECM) methodology to European Central Bank and Organisation for Economic Co-operation and Development data for the eurozone, this article proposes an empirical validation of such theoretical premises by analysing the relationship between the different types of credit provided by commercial banks and the corresponding interest rates. The main results show a negative relationship between the interest rates and the credit provided for the purchase of houses. Conversely, no significant relationship is found between loans granted to enterprises and loans for the purchase of consumption goods and the corresponding interest rates.

KEYWORDS:

• Post Keynesian endogenous money theory
• credit demand
• interest rates
• eurozone

JEL CLASSIFICATIONS:

• E11
• E12
• E40
• E50
• G21
• C32

Acknowledgments

I warmly thank all the participants of the Workshop held at Cassino University in April 2016 for comments and feedbacks received on a preliminary draft of this article. I am especially indebted to Antonella Stirati, Giuseppe Fontana, Mariana Mazzucato, the Graduate School in Economics of Roma Tre University, and two anonymous referees for guidance and stimulating debate on the matter. I am also grateful to Fabrizio Antenucci, Francesco Bianchi, Stefano Di Bucchianico, Riccardo Pariboni, Walter Paternesi Meloni, Luigi Salvati, Pascal Sotgiu, Francesca Tosi, and Simone Zardi for fruitful discussions.

Notes

¹In post Keynesian monetary theory, a strong debate related to the representation of the credit supply curve exists between the horizontalist and the structuralist approach. In this article we do not enter into this dispute, and for simplicity we assume a horizontal credit supply. For a thorough review of this debate, see Palley (1994, 1996, 2013), Lavoie (1996), Rochon (1999), Fontana (2003, 2004, 2009), and Deleidi (2016).

²Arestis (1987, p. 256), referring to disposable income, maintains that “this reflects the hypothesis that, as the level of economic activity expands, consumers go further into debt to finance purchases of durable goods.” It could be argued that an increase in households’ disposable income leads them to be more confident in their ability to repay their debts, thus inducing them to increase their level of indebtedness.

³As anticipated in Footnote 1, we endorse the horizontalist view (Eichner, 1987; Lavoie, 1996; Moore, 1988). As a consequence, the mark-up is exogenous—that is, independent of the volume of loans provided by commercial banks—and the supply of credit has to be represented horizontally (see Figure 1, Quadrant 1). For instance, this does not mean that the risk perceived by banks is not influenced by the economic activity but is independent of the quantity of loans provided by commercial banks, and there is no general functional relationship between the mark-up and the economic cycle (Lavoie, 1996). However, an increase in the risk perceived by the banking industry, for example due to a deep downturn in the economic activity, would lead to an upward shift in the supply of credit. In other words, banks, perceiving a greater insolvency risk and to preserve a higher rate of profit, increase the mark-up on loans. To be clear, the concept of the interest rate’s exogeneity can be summarized as follows: “Banks are price setters and quantity takers both in retail deposit and lending market” (Moore, 1989, p. 27) and the lending rate is “politically administered rather than a set of market-determined prices. Indeed, the short-term interest rate is simply another of the distributional variable” (Eichner, 1987, p. 858).

⁴According to Garegnani (1979), the alleged negative relationship between the rate of interest and the level of investment would ensure the return to the traditional theory, because “admitting an elastic investment demand schedule leads to maintaining, on the one hand, the existence of a full-employment level of the rate of interest and, on the other, the presence of inflation, or deflation and unemployment, when the actual rate of interest is not the full employment one” (Garegnani, 1979, p. 79). Moreover, the possibility of the investment demand being negatively affected by the rate of interest would allow a limit in a short-run analysis of the principle of effective demand, leaving in the long run the level of investment determined by full-employment savings. According to Garegnani (2015, p. 131), “a spontaneous tendency for investment to adjust to the saving capacity of the economy … is the outcome of a theory of interest derived from the idea that the overall demand for ‘capital’ is highly elastic with respect to the rate of interest.” In other words, to keep the distance from the traditional theory and to extend the role played by the effective demand in a long-run analysis, we have to untie the link between the interest rates on the one hand and the level of investments, the firms’ credit demand, and, eventually, the process of accumulation on the other.

⁵For a thorough literature review on the attempts to extend the principle of effective demand to a long-period analysis, see, among others, Vianello (1985), Ciccone (1986), Garegnani (1992), Serrano (1995), Trezzini (1995), Cesaratto, Serrano, and Stirati (2003) and Freitas and Serrano (2015).

⁶The GDP deflator is provided by OECD statistics.

⁷The nominal variables regarding loans provided by commercial banks to enterprises for the purchase of consumption goods and houses are deflated by the following equation:

$$LLR=\frac{LL}{DEF}\ast 100$$

where LLRis the real loans, LL is the nominal loans, and DEF is the GDP deflator. In the same manner, we adjust the several types of nominal interest rates using the following equation:

$$INTR=INT-\pi$$

where INTR is the real interest rate, INT is the nominal interest rate, and π is the inflation rate.

⁸BLS indicators are measured by the weighted diffusion indexes that assume values that move from −100 to +100. Regarding the credit supply conditions, if the index moves towards +100, a tight approval criterion occurs. On the contrary, regarding the credit demand conditions, if the several indexes move towards +100, an increase in the credit demand takes place. All the BLS indicators used in this analysis cover the previous three-month period.

⁹The Schwarz Bayesian information criterion is more precise and accurate than the Akaike information criterion, as the former takes into account the sample size of the series.

¹⁰The short- and long-run causality are estimated by the methodology developed by Engle and Granger (1987), Granger (1988), Sims, Stock, and Watson (1990), Mosconi and Giannini (1992), Toda and Phillips (1993, 1994), Granger and Lin (1995), Toda and Yamamoto (1995), and Rambaldi and Doran (1996). In this article we make use of the long-run causality.

¹¹If the trace and the maximum eigenvalue are greater than the critical values and therefore significant, we reject the null hypothesis of no cointegration.

¹²In the present article, we show the results concerning the autoregressive process (please see in Table 4 LLRC5(−1), LLRE15(−1), LLRE5(−1), LLRH1(−1), LLRH15(−1), and LLRH5(−1) and in Table 5 LLRC1(−1, −2, −3), LLRC15(−1), and LLRE1(−1, −2, −3)) without interpreting the results, because they are not relevant to our analysis.

Additional information

Funding

This research is supported by EU Horizon 2020 grants: DOLFINS Nr. 640772.

Notes on contributors

Matteo Deleidi

Matteo Deleidi, University College London, London, UK.