SFC modeling and the liquidity preference theory of interest

Abstract

According to Lavoie and Reissl, stock-flow consistent (SFC) modeling with “fully specified” financial sector allows for a better understanding of the dynamics of monetary economies than less detailed models. Although detailed models can help to identify mechanisms that cannot be captured by simpler models, they also have limitations that are worth bearing in mind when it comes to assess the merits of both kinds of modeling. This note pinpoints some difficulties regarding the conceptual framework and formal treatment of the “fully specified” SFC model, leading to a more balanced assessment regarding economic models.

Keywords:

• Complex models
• endogenous money
• interest rate
• liquidity preference
• stock-flow consistency

JEL-Classification:

• E5
• E12
• E40
• E43
• E44

Notes

1 The authors actually focus on steady state deviations produced over the long run by a permanent shock (such as a permanent increase in public spending or a permanent change in monetary policy), assuming that the parameter values remain unchanged during the whole adjustment process. By contrast, in the Keynes–Asensio conceptual apparatus, uncertainty implies that parameter values (e.g., the importance of liquidity preference) may change in response to the shocks and that the system does not tend to a predetermined position in the long run.

2 Both models have an aggregate demand function—hence aggregate output—that decreases with the rate of interest at any point in time (which depicts an IS curve), and a money demand that decreases with the interest rate and increases with output (which, given the stock of money at a point in time, depicts an LM curve), although they differ strongly from each other—and from the standard IS/LM model—in terms of their conceptual foundations and economic properties. The importance of uncertainty in this regard is discussed in Asensio (2018).

3 Recall that the appendix aims at providing analytical support for the simulation results, which is obviously biased by the fact that the analysis itself is based in part on the simulation results. The second part of the appendix, which aims to support the second simulation exercise, is also affected by this problem.

4 The debt repayments issue also appear in other SFC models, as transpires from the standard transactions flows matrices presented in Godley and Lavoie (2007, pp. 285, 382), Caverzasi and Godin (2015, appendix, pp. 30-31) and Nikiforos and Zezza (2017, p. 1209), to only mention the references cited by Lavoie and Reissl (pp. 503–504). See also the appendix to this note.

5 Simulation results of course depend on the calibration. Because “it is easy to construct a case in which the opposite result obtains” (p. 523), it would have been helpful to have detailed information on what the authors consider to be “reasonable” parameter and initial values (pp. 514, 523, 526), all the more because, in the absence of analytically tractable solutions, this information is essential to address issues related to the sensitivity of the simulation results to parameter and initial values.

Additional information

Notes on contributors

Angel Asensio

Angel Asensio is an Associate Professor at Université Paris 13–Sorbonne Paris Cité, Villetaneuse, France.