Abstract
This article discusses the role that macroeconomic uncertainty plays in banks’ decisions on the optimal asset allocation. Following the portfolio model proposed by Baum et al . (2005), the article aims at disentangling how Italian banks choose between loans and risk-free assets when uncertainty on macroeconomic conditions increases. The econometric results confirm that macroeconomic uncertainty is a significant determinant of banks’ investment decisions, also after controlling for other factors. In periods of increasing turmoil, banks’ ability to accurately forecast future returns is hindered and herding behaviour tends to emerge, as witnessed by the reduction of the cross-sectional variance of the share of loans held in portfolio.
Acknowledgements
I am grateful to F. Cannata, A. De Vincenzo, C. Guerzoni, S. Iannotti, J. Marcucci for their comments. An anonymous referee provided very helpful suggestions for improving the article. My special thanks go to P. N. Smith for his continuous guidance. All remaining errors are my own. The opinions expressed herein are those of the author and do not necessarily reflect those of Banca d’Italia.
Notes
1 For a survey, see Quagliariello (Citationin press).
2 The subscript t is omitted since the model describes a one period problem.
3 For example, as far as monetary policy is concerned, in 1986q–1986q2 and 1987q4–1988q1 ceilings on the supply of credit that each bank may provide to the aggregate of its customers (so-called massimale sugli impieghi) were still in use. These constraints, intended for controlling liquidity and aggregate demand, limited banks’ ability to expand their shares in the loan market. As a result, even in the absence of specific obligations for banks to subscribe Treasury bonds, the banking system's demand for bonds increased (Cotula, Citation1989). Also supervisory rules were particularly strict. According to the banking law, intermediaries were classified as retail banks, providing only short-term credit, and medium and long-term credit institutions. Until 1990, when the Bank of Italy decided to liberalize banks’ branch networks, the opening of new branches was limited by the system of ‘branch plans’. These legal barriers between different categories of banks and the administrative constraints on the opening of branches were an obstacle to the enlargement of banks’ activity (Ciocca, Citation2004).
4 According to the classifications provided by the Bank of Italy, banks are grouped into five categories depending on the size of their total assets: major banks (total assets greater than 45 billion euros), large (total assets between 20 and 45 billion euros), medium-sized (7–20), small (1–7), minor (total assets up to 1 billion euros).
5 For Italy, a quarterly survey on firms’ expectations regarding future inflation is available only since December 1999.
6 Strictly speaking this measure is not the quarterly volatility, but the average of the monthly volatilities in a given quarter. As an alternative measure I also use the quarter-end conditional variance. Results are unchanged.
7 I am grateful to the referee for raising this point.
8 In principle, if bank's size is a proxy for the type of loans a bank grants, the evidence provided in this section might partly overcome the shortcomings related to loan heterogeneity mentioned above. Unfortunately, this is not necessarily the case, since Italian banks are not specialized according to their size.