«Financial Stability and Economic Performance* Jérôme Creel OFCE – Sciences Po & ESCP Europe Paul Hubert OFCE – Sciences Po Fabien Labondance ...»
Financial Stability and Economic Performance*
OFCE – Sciences Po & ESCP Europe
OFCE – Sciences Po
OFCE – Sciences Po
This paper aims at establishing the link between economic performance and financial
stability in the European Union from 1998 to 2011. We use the standard framework – both in
terms of variables and econometric method – of Beck and Levine (2004) to estimate this
causal relationship, independently from but controlling for the level of financial depth. We test how different measures of financial instability (an institutional index, microeconomic indicators, and our own statistical index derived from a Principal Component Analysis) affect economic performance (or components of aggregate dynamics like consumption, investment or disposable income) and find that financial instability has a negative effect on economic growth. Our results also suggest that the traditional result that financial depth positively influences economic performance is not confirmed for a subset of advanced economies like European countries.
Keywords: Financial depth; Aggregate dynamics; Financial stability, Banks, Nonperforming loans, CISS, Z-score, Principal Component Analysis.
JEL Classification: G10; G21; O40 We thank seminar participants at FESSUD Annual Conference 2013 and, in particular, Stephany Griffith-Jones, Piotr Banbula, * Thierry Philipponnat, Jézabel Couppey-Soubeyran, Emmanuel Carré, and Salim Dehmej for helpful conversations, suggestions and comments. This research project benefited of funding from the European Union Seventh Framework Program (FP7/2007under grant agreement n°266800 (FESSUD). Any remaining errors are ours.
Corresponding author: firstname.lastname@example.org. Address: 69, quai d’Orsay, 75007 Paris – France, Tel: +33(0)144185456, Fax:
+33(0)144185478. email@example.com, firstname.lastname@example.org.
1. Introduction This paper examines the relationship between macroeconomic performance and financial stability in the European Union (EU). Different views emerged from the literature on the links between finance and economic performance. On the one hand, credit is found to be determinant in the process of economic development. The literature often recalls the Schumpeterian view that entrepreneurs need credit to finance their innovations. Banks and financial markets are then viewed as facilitators. On the other hand, finance development appears to respond to economic growth. With economic expansion, firms and households are more likely to demand financial services. In both cases, the finance-growth relationship seems to be constrained by structural determinants such as the historical level of debt, the legal environment or the level of economic development. Beyond this finance-growth nexus which has already given rise to an abundant literature, we investigate whether financial instability affects macroeconomic performance.
Focusing on the financial stability issue is motivated from both an academic and a policy perspective. This topic has emerged in the academic debate since the crisis (Arcand et al., 2012 ; Cecchetti and Kharroubi, 2012 ; Beck et al., 2014). A major reason for addressing the question of financial stability is its public good’s nature (see e.g. Boyer et al., 2004): it is a non-rival good since its use does not prevent someone else from the same use, and it is nonexcludable since no one can be deprived from its use. After financial crises, new regulations are proposed to supervise and frame the financial system to preserve its properties as a public good (Cartapanis, 2011). In the case of a banking crisis at the micro level, financial stability has to be preserved to avoid that idiosyncratic shocks have a systemic impact through different contagion links: contractual, informational or psychological (Borio, 2003).
For instance, Lehman Brothers’ bankruptcy in September 2008 has affected the whole banking system through several channels. Contractually, its creditors were the first to be hit.
But very quickly, bankruptcy was analysed as a severe negative signal on financial markets and, in particular, on interbank markets. It induced uncertainty and suspicion among banking institutions that became suddenly reluctant to participate to the money market.
This informational or psychological link was transmitted all over the world and extreme tensions appeared on the European’s and US’ money markets, and consequently, affected the real economy. Moreover, payment systems are central to the smooth functioning of market economies and financial instability could potentially disrupt these.
European countries included financial stability in the European Treaty as an objective of the European Central Bank. According to Article 127(2) of the Treaty on the Functioning of the EU, the ECB has “to smooth the conduct of policies pursued by the competent authorities relating to the prudential supervision of credit institutions and the stability of the financial system”. But beyond this mandate, its policy formulation is difficult to achieve because of the difficulties to define, forecast and measure financial stability (Schinasi, 2004). In the absence of a consensus, it is acknowledged that these regulations should at least be implemented at the European level. The banking union is one step in that direction. The EU is thus an adequate level to investigate the financial stability-economic performance link, all the more so as European countries, thanks to financial integration and converging prudential regulations, are relatively homogenous compared to the rest of the world.
We have to stress that since we focus specifically on financially integrated and homogenous economies of the EU for which most countries share the same central bank and are part of the single financial market, the analysis of the global finance-growth nexus where advanced and developing countries are mixed, is beyond the scope of the paper. Our study documents the need for regulation in providing evidence about the effects of financial stability on aggregate dynamics.
We assess the finance stability-economic performance nexus following the dynamic panel estimations methodology of Beck and Levine (2004) dealing with endogeneity. This framework has been used by much of the related literature, hence producing results that are comparable across a wide array of previous contributions. On a sample of all European Union countries from 1998 to 2011, we test whether financial stability has a causal effect on economic performance and its subcomponents: consumption, investment and disposable income. The effect of financial stability is estimated independently of the level of financial depth, but we control for it to avoid capturing the specific effects of financial depth. We also control for a potential non-linear relationship between financial depth and financial instability with interaction terms, and show that financial instability affects economic performance independently of the degree of financial depth. We use different financial
instability indicators that measure the macro and the micro dimension of financial stability:
the Composite Indicator of Systemic Stress (CISS) provided by the ECB, aggregate prudential ratios for domestic banks for each country, stock market volatility and our own statistical index constructed with a Principal Component Analysis (PCA). We find that financial instability, captured by the CISS and non-performing loans, has a negative effect on economic performance. This result is robust to several panel specifications and estimators. It is interesting to note that introducing financial stability in this framework does not alter the financial depth effect. While the result may seem intuitive, this paper provides quantitative evidence on the negative effect of financial stability on economic performance. Financial instability, independently of financial deepening, represents a risk for the economy.
Moreover, we find that financial depth has no positive impact on economic performance in the EU. This result supports the view of Arcand et al. (2012) that the effect of financial depth on growth depends on the level of financial development. While these results do not undermine the seminal results that financial depth has a positive effect for developing countries, they suggest that the level of financial depth in the EU is such advanced that finance depth has no longer a positive effect on economic growth.
The main policy implication of this study is that the need for better micro and macro prudential regulations is made clear. In particular, the effect of banks’ non-performing loans has been shown to be damaging for economic performance, so as macroeconomic financial instability. Dealing with micro and macro financial stability, through the banking Union for instance or with the ECB’s OMT, would participate in improving the real economy.
The rest of this paper is organized as follows. Section 2 presents the related literature.
Section 3 describes the data and Section 4 the methodology and results. Section 5 concludes.
2. Related Literature This paper is related to the literature studying the link between finance and economic growth, and to the literature investigating the link between financial instability and growth.
2.1 Finance and Economic Performance Different perspectives on the relationship between finance and economic performance have been emphasized, and theoretical and empirical controversies on this subject exist since the beginning of the XXth century (Ang, 2008). The debate can be summed up as follows. Pros highlight that the development of finance induces a better allocation of resources, mobilizes savings, can reduce risks and facilitates transactions. The financial sector acts as a lubricant for the economy, ensuring a smoother allocation of resources and the emergence of innovative firms. Cons recall that stock markets have destabilizing effects and that finance liberalization leads to financial crises. These more sceptical authors believe that the link between finance and economic growth is exaggerated (Stiglitz, 2000; Rodrik and Subramanian, 2009). De Gregorio and Guidotti (1995) argue that the link is tenuous or even non-existent in the developed countries and suggest that once a certain level of economic wealth has been reached, the financial sector makes only a marginal contribution to the efficiency of investment. It abandons its role as a facilitator of economic growth in order to focus on its own growth. This generates banking and financial groups that are finally “too big to fail”, enabling these entities to take excessive risk since they know it will be mutualised via public authorities’ interventions. Their fragility rapidly transmits to other corporations and to the real economy. The subprime crisis is certainly a good example of the power and magnitude of the effects of correlation and contagion on financial markets.
Numerous empirical studies have investigated these questions. However, until recently, the literature highlighted a positive relationship between financial development and economic growth (Bumann et al., 2013). We can distinguish between cross-country, time series and panel studies.
Cross-country studies, mixing countries with different levels of development, generally found a positive effect of finance on economic performance with the notable exception of Ram (1999). King and Levine (1993) found that financial development indicators are positively associated with capital accumulation, total factor productivity growth and GDP growth. Focusing on the stock markets’ influence Demirgüç-Kunt and Maksimovic (1998) and Levine and Zervos (1998) concluded that liquid stock markets are positively related to GDP growth. Nevertheless, these cross-country studies suffer from severe limits. Most of them only intend to quantify how finance affects economic performance, neglecting the reverse causality. When they deal with this endogeneity bias, they include instrumental variables. But, as demonstrated by Ahmed (1998), this technique is not robust when data are averaged over decades, which is usually the case. Another limit of these cross-country analyses is the grouping of countries that are highly heterogenous. This problem is highligted by Ram (1999) who show that after defining subgroups into his sample, an important parametric heterogeneity is observed. This is due to the fact that the link beetween finance and economic performance is mainly determined by the financial structures, the legal environment, the preferences and the policies implemented in each country (Arestis and Demetriades, 1997; Demirgüç-Kunt and Maksimovic, 1998).
Time series studies have been developed in contrast to the above-mentioned limits. Arestis and Demetriades (1997) compare the finance-GDP growth link in Germany and in the United States. They find in Germany a relationship going from finance development to real GDP, whereas the reverse causal pattern runs for the United States. Xu (2000) also provides evidence of heterogeneity across countries. Arestis et al. (2001) compare the influence of banks and stock markets accross five developed countries. Their results show that both banks and stock markets promote GDP growth. But they also suggest that banks’ contribution is stronger than the stock markets’. Moreover, they point out that stock markets’ volatility has negative effects in Japan, France and the United Kingdom. This variety of results can be interpreted as a limit to time-series analysis. These studies also suffer from small sample constraints. To preserve degrees of freedom, variables included in the analysis are kept to a minimun and these studies are subject to the omitted variable bias.