The Supermultiplier-Cum-Finance. An Application to the Credit-Led Boom before the 2008 Crash

ABSTRACT

The supermultiplier model is gaining momentum. To become a cornerstone of modern macroeconomics, it should explain the dynamics of advanced market economies in a coherent, complete, and simple way. In particular, it is supposed to clarify the possibilities and limits of credit-driven economies, like the one observed in most advanced countries after 1995. This paper contributes to these goals by integrating the Sraffian supermultiplier with the post-Keynesian hypothesis of credit money endogeneity. This hypothesis is somehow modified when autonomous banks are able to accelerate credit above output growth, as happened after 2002. The gap between credit and output growth implies that a part of the loans is financing non-output transactions (land, old houses, shares), usually with a speculative bias. The consequences of a persistent gap are demand depression and asset inflation, as we observed after 2008.

KEYWORDS:

• Supermultiplier
• endogenous credit money
• financial instability
• post-keynesian economics
• sraffian economics

JEL CODES:

• E11
• E12
• E32

Acknowledgments

During the long preparation of this paper, we received helpful comments from John S. McCombie (University of Cambridge), Richard Werner (University. of Southampton), Gumersindo Ruiz (University of Malaga), and Eladio Febrero (University of Castilla – La Mancha). Our gratitude to all of them Also, to the people attending the seminar in Centro Sraffa (Università Roma Tre) on the 30th May 2018, where a first version of the paper was presented. The detailed and constructive comments of the referees deserve explicit recognition.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 Sraffa (1960). Among the Sraffians that look at Keynes in order to rebuild the ‘surplus approach’ see Pasinetti (1973), Pasinetti (1979), Garegnani (1978-79), and Eatwell (1983).

2 See Keynes (1936) (books I to V), and Kalecki (1937). Lavoie (2014) offers a comprehensive study of the post-Keynesian approach. See, in particular, chapters 4, 5, and 6.

3 Some recent examples: Skott, Santos, and Oreiro (2021) insists on the need of ‘fine -tuning’ to accelerate the adjustment. Hein & Woodgate (2020) study the formal stability conditions in an economy driven by government expenditure financed with debt. O Dejuán and Dejuán (2018) shows the interaction between financial and real forces by means a predator-prey model. Ó. Dejuán and McCombie (2019) focuses on the role played by the securitization strategy and mortgage-backed derivatives from 1995 till 2008. Dejuán, Portella and Mateo (2020) analyze the impacts of decarbonization using a disaggregate multiplier-accelerator model.

4 Pasinetti (1973) remarks two traits of economic analysis based on VIS: (1) intermediate consumption and capital consumption are ‘transformed’ into indirect labor, and indirect fixed capital; (2) The value-added in each VIS (the sum of wages and profits) coincides with the value of final output: Y_c= VA_c; Y_k= VA_k; Y_z= VA_z. A proper multisector economy would require the use of input-output techniques. Here we offer the typical macroeconomic analysis where goods are produced with the same technology but differ in their function.

5 In this sentence, the words in italics signal the Sraffian contribution. Demand expectations are linked to ‘normal prices’, i.e.. the prices of production forged by competition.

6 The reader should distinguish between (a) the macroeconomic equilibrium conditions: savings = residential investment + productive investment. After dividing by income, σ = z+h, and (b) the sectoral conditions derived from our particular expenditure assumptions: residential investment is eventually paid with savings out of wages; productive investment, with savings out of profits (or business savings): σ_w= z, σ_r= h.

7 Two advantages of this definition of investment: (1) It is immune to Garegnani’s (1978-79) criticism to the marginalist remnants of Keynes’ investment function. (2) The separation between two types of investment dilutes Harrod’s fear of the extreme instability of the multiplier-accelerator model (Harrod, 1939), (Dejuán, 2017).

8 In chapter 3 of The General Theory, Keynes refers to the ‘long-term state of expectations.’ Keynes focused on the psychological factors prevailing in an environment of fundamental uncertainty. Following Schumpeter [2008] 2012, we focus on an objective aspect: the potential market of the industries driving the economy. A complete picture should add to building companies and banks, the institutions and policies that favor residential investment.

9 The assumption that productive investment is financed with business savings was used by Robinson (1962), p. 86, and Shaikh (1991). Of course, such equality does not mean that saving determines investment. To prevent such a conclusion, we could differentiate between ex-ante investment (production of machines under the expectation of an increase in aggregate demand), and ex-post investment (purchase and installation of the new machines assuming that demand expectations are confirmed).

10 The first presentations of the supermultiplier, after Serrano (1995), have a Sraffian flavor. See, Bortis (1997), Cesaratto, Serrano, and Stirati (2003), O. Dejuán (2005), White (2009), Aspromourgos (2013), Freitas and Serrano (2015), Girardi and Pariboni (2016), Dejuán (2017), Pérez-Montiel & Manera (2020). Post-Keynesian and neo-Kaleckian contributions to the supermultiplier appear in Allain (2015), Lavoie (2016), and Palley (2018). Critics of the supermultiplier model include Sraffians, post-Keynesian and Marxian economists: Palumbo and Trezzini (2003), Lavoie (2010), Skott et al. (2021). Following Ciccone (1986), the critics argue that normal capacity is not an attractor. The debate appeared in vol. 2 of Political Economy. The Surplus Approach.

11 In this paper, the expression ‘persistent demand’ refers to the level or rate of growth of demand that has ruled in the recent past and is expected to last long enoung to justify an increase in capacity. In the expression ‘transient demand’ the adjective might be replaced by transitory ,temporary or seasonal.

12 From the very formulation of the supermultiplier (SM = 1/(1-c-kγ)) we can derive a similar limit. By construction, (c+kγ) has to be lower than 1, leaving room for proper autonomous demand, the driver of the system. The last statement implies that the maximum rate of growth is related to the minimum propensity to consume in the following way: $\hat{\gamma }<{\displaystyle \frac{\left(1-\underset{\_}{c}\right)}{k}}$.

13 The hypothesis of money endogeneity has been developed by economists in the post-Keynesian and the circuit traditions. Among post-Keynesians, see Kaldor (1981), Moore (1988), Wray (1990), and Rochon and Rossi (2017). Among the ‘circuitists’ see Graziani (1989), Parguez and Seccareccia (2000), and Lavoie (2016).

14 The stock-flows consistent model was formally introduced by Godley and Lavoie (2007). See also Nikiforos and Zezza (2017). It is not always easy to find the above identities in the financial accounts and balances. Investment banks transform the nature of the assets and liabilities. An example: the securitization process initiated in 1995 converted mortgage loans into mortgage-backed securities.

15 Alternatively interest payments can be computed as i’ times the outstanding debt: INT_t= i’·[DEBT]_t. Consistency requires that i·CR_t = i’·[DEBT]_t. The second formulation will be used in section 4 to simplify the definition of debt service: DS_t= (i’+a)·[DEBT]_t. For the time being we will use the first formulation that locates the financial transfer in the year when each loan is granted.

16 Banks also charge fees for credit card management, direct debiting, currency exchange, and the like. To simplify our presentation, we assume that these revenues roughly match banks’ interest payments to depositors and bondholders.

17 For the sake of simplicity, in this paper we assume a constant interest rate. In the numerical exercise of the Appendix there would be no problem to introduce new scenarios with different rates altering the transfer of income. The impact on credit demand it is beyond our reach.

18 An alternative way to measure liquidity includes in the denominator the value of non-liquid assets: λ’ = ${\displaystyle \frac{{\left[DEP\right]}_t}{\overline{X}\cdot P_{xt}}}$

19 The array of financial indicators is not exhaustive. For business students, the most relevant indicator is the leverage coefficient: ‘borrowed funds over equity’. Table 2 does not mention it because, in our model, firms finance productive investment entirely with equity. 

20 The endogenous deterioration of the structures of debt (Minsky, 1964), is another way to present the issue at stake. Credit for non-output transactions appears directly in Werner (2005, 2014) and indirectly in Schularick and Taylor (2012), and Borio and Lowe (2002).

21 Notice the differences between output inflation (P_y) and asset inflation (P_x). During the last boom (1995-2008) credit doubled for both non-produced assets and produced goods. The residential and stock exchange markets experienced important bubbles: their price indexes almost trebled. On the contrary, the adjustment in the industrial market came through quantities: cars’ factories simply doubled the output of luxury cars.