Skip to main content

Agent-based Computational Economics for Policy Analysis

Periodic Reporting for period 1 - ACEPOL (Agent-based Computational Economics for Policy Analysis)

Reporting period: 2019-03-01 to 2021-02-28

Credit markets, financial markets and, relatedly, the accumulation of debt, significantly impacts upon competition and affects both industry and macroeconomic fundamentals. How can economists better capture and understand the mechanisms trough which the financial side affects the real economic side?

With the aim of grasping the nature of these mechanisms, the ACEPOL project proposed the adoption of agent-based computational economic models in combination with up-to-date calibration and validation econometric techniques. In addition, and after the outbreak of the Covid-19, also the necessary lockdown implemented by many governments strongly impacted competition, market selection and aggregate economic activity. The ACEPOL action has therefore broadened its scope to also provide prompt policy suggestions to policy makers.

The main objective of the ACEPOL project is therefore twofold. On the technical aspect, the project aimed to improve the adoption and the efficiency of calibration and validation techniques. On the applied aspect, the project aimed to propose policy suggestions over the competition, innovation and growth domains.
From the beginning to the end of the project the fellow worked both on technical and applied aspects of the research. In this summary only the two most policy-relevant results are summarized:

First, after taking stock of the evidence about the growing influence of institutional investors in publicly owned companies -- their participation in the U.S. stock market has been increasing from 10% in 1952 to over 70% in 2006 -- the fellow developed an agent-based model to explain this phenomenon and calibrated it on aggregate EU data. The fellow has shown that a strong presence of impatient (short-term) investors among the shareholders of a firm has a negative effect on firm R&D activities and a positive impact on shares buybacks. In addition, also an increase in the stock-based remuneration of the firm’s executives might exert a positive impact on the amount of share buybacks committed by the firm. Overall, these two forces impact upon competition and growth. From a policy perspective, the analysis carried out suggests that the combination of high-growth and high-competition is feasible, but only under specific conditions: if a sufficient proportion of the shares must be controlled by investors with a long-term horizon and if corporations do provide the right incentives for the management to prefer real investments over buybacks.

Second, the author started from the evidence that the Covid-19 lockdown exerts a strong effect on the firm finances, also impacting on firm churning and market selection. How many more bankruptcies there will be in 2020, with respect to the number of bankruptcies that are observed in normal year? And, more importantly, will the most efficient firm be able to survive? Will the least efficient firms go bankrupt? By employing an agent-based model calibrated using the balance sheet data of one million French firms, the fellow shown that, with respect to normal times, the number of firms in solvency distress is almost double during the pandemic year. Furthermore the overall productivity in France is expected to decline by about 10% due to the lockdown and the impact this exerts on firm bankruptcy. In particular, this is due to the fact that during the lockdown phase, the market selection process has worsened substantially. Firm-level efficiency can only marginally explain the probability of survival of the single firm. Size and available liquid resources matter more. Thus, government intervention is needed in order to limit business failures and to avoid the generation of long-term hysteresis effects. Which firms to bailout? According to the analysis carried out, the government shall focus on helping those firms which are more efficient and that can guarantee higher growth rates in the future, rather than providing generic financial support to all firms.
"On the technical aspect, the fellow created of a set of new agent-based models that go beyond the state of the art. In particular, the fellow developed three agent-based models tackling the main research questions of the project. Two of these models are mostly focused on the effects of credit and financial markets on firm and industrial dynamics. The third is instead focused on their effects on macroeconomic aggregates and combines a previously existing macroeconomic model with an interbank network. All the models developed have been calibrated and validated with empirical data using either the method of simulated moments over a high-dimensional parameter space, via an indirect calibration approach or with the help of an information theoretic criteria. Additionally, after the outbreak of the Covid-19 pandemics, the fellow developed a new agent-based model to study the effect of different lockdown waves on liquidity, solvency and bankruptcy of French business firms; and he developed, using Italian data, an input-output model to study the regional exposures to the lockdown. In addition the fellow put forward a companion paper for the replication of the results of the most complex of the above mentioned models. This requires a strong effort but should become the state of the art in the future, since it eases the replication of simulation papers and a faster transmission of knowledge between researchers.

At the societal level, the impact of the project has been brought to the fore trough two policy round tables to which the fellow actively participated. The first round table was more technical, covering topics related to the calibration and validation of agent-based models and it has been held at the Bank of England premises in June 2019. The second, online due to the Covid-19 pandemics and held in December 2020, has been organized by the ""Institut de la gestion publique et du développement économique”. The fellow presented at this round table his work on the effects of the lockdown on market selection and had the possibility to confron and discuss the ideas stemming from the ACEPOL action, together with researchers from OECD, Banque de France and the French ministry of economics and finance."