Skip to main content

Further Literature Comments

Further Literature Comments​

In this section, I mention various academic works that explore the intricate relationships between networks, firm growth, and aggregate volatility. These studies provide insights into how different factors contribute to economic fluctuations and firm dynamics.

Eigenvector Centrality and Causal Inference​

One potential substitute for causal inference is eigenvector centrality. For more details, you can refer to the study available on Nature's website.

Cascades in Networks and Aggregate Volatility​

  • Acemoglu et al. (2011) discuss the impact of cascades in networks on aggregate volatility. For a comprehensive understanding, refer to their work on the MIT Economics website.

  • Acemoglu, Carvalho, et al. (2012) explore the network origins of aggregate fluctuations. Their findings are detailed in a paper available here.

Volatility in Granularity and Networks​

Another significant contribution is by Kelly (2013), who investigates volatility in granularity and networks. The study is accessible through this link.

Log Shocks and Distribution​

Evidence suggests that log shocks are neither small nor normally distributed, challenging the careless application of the 1/n1/\sqrt{n} rule. This is supported by works from Chesher (1979), Boeri (1989), and Bottazzi (2006).

Firm Growth Dynamics​

  • Holly, Petrella, & Santoro (2012) examine the cross-sectional dynamics of firm growth, highlighting aggregate fluctuations. Their research is published in the Journal of the Royal Statistical Society and can be found here.

  • Stanley (1996) provides early measurements of growth rates distributions, emphasizing the significance of firm size in global GDP representation. For more, see this article.

Growth Rates and Distribution​

The hypothesis of normally distributed and uncorrelated geometric growth rates suggests a lognormal distribution in the long run. If growth rates are correlated, firm size distribution will diffuse faster, and vice versa if negatively correlated. Chesher (1979) applies a method fitting firm size time series with autoregressive processes, as detailed here.

Gibrat Hypothesis and Firm Growth​

The unit root nature of firm growth, often referred to as the Gibrat hypothesis, has been explored by Stanley et al. (1996b) and Amaral et al. (1997). Bottazzi and Secchi (2003a, b) explain the tent-shape of firm growth rates density as an emerging feature due to positive-feedback effects. Bottazzi (2001) discusses the diversification effect in relation to firm size and submarkets.

Fat Tails and Autocorrelations​

Before concluding on aggregate volatility explanations, it is crucial to consider the effects introduced by 'fat-tailed' log growth rate distributions and their autocorrelations.

For a more detailed discussion on the antecedent of Gibrat and growth rates, refer to the appendix section on Gibrat and Pareto.

Additional References​

  • SIREN provides a comprehensive resource for firm characteristics, accessible here.

  • For extensive/intensive margin accounting using French firms, see this pertinent working paper.

  • An honorary mention goes to Wilson (1923) for the arith-log paper, available here.

This section provides a comprehensive overview of the literature, emphasizing the importance of understanding firm growth dynamics and network effects on economic volatility.