Empirical exercises in estimating the effects of different types of financial institutions’ functioning on economic growth

Abstract

While recent studies of the finance-growth nexus have focused on the use of proxies which more accurately capture the theorized functioning of the financial sector, they have tended to focus either on the functioning of the financial sector as a whole, or on the dominant institutions within the sector. Little attention has been paid to a comparison of the relative effects of different types of financial institutions on economic growth. This article attempts to get a deeper understanding of the finance-growth process by disaggregating the total financial sector impact and examining the individual and relative effects of each type of institution in the financial sector. We explore the empirical properties of alternative specifications of models of the impact of financial institutions’ functioning on economic growth, by conducting a number of exercises. These exercises experiment with various model specifications to represent the long- and short-run impacts of the financial institutions’ functioning on economic growth, using cointegration and error correction methodologies.

Notes

¹ See Gupta (1985, p. 341), King and Levine (1993), Odedokun (1996), Rajan and Zingales (1998), Arestis et al. (2001), Fisman and Love (2002) and Berger et al. (2003).

² For example, Tang (2006), in examining the effect of financial development on economic growth in Asia-Pacific Economic Cooperation (APEC) countries, focused primarily on the operations of the stock market and banking sector, and on the effect of capital flows. Although an attempt was made to account for the size of nonbank financial institutions by measuring their total liquid liabilities, this does not suffice as a proxy of their functioning.

³ See Schumpeter (1934), Goldsmith (1969), McKinnon (1973), Shaw (1973), Bencivenga and Smith (1991), Greenwood and Jovanovic (1990), Stiglitz and Weiss (1981), Bernanke (1983), Diamond (1984) and Fama (1985).

⁴ For example, Mazur and Alexander (2001) use the ratio of liquid liabilities of the financial system to Gross Domestic Product (GDP) and the ratio of deposit money bank domestic assets to deposit money bank domestic assets plus central bank domestic assets, as broad proxies of financial depth. Thieβen (2005) uses several financial indicators (including broad money/GDP, importance of banks relative to the central bank, share of credit to the nonfinancial private sector in total credit and stock market capitalization/GDP), none of which adequately measure the theorized functioning of the financial sector.

⁵ Loans are subtracted from total assets so as to ensure that when deposits increase, assets will not necessarily increase by the same proportion.

⁶ This assumption leads to an unavoidable conceptual limitation of DRISK, as there is an obvious absence of threshold limits beyond which any further diversification would represent an allocation of resources away from the most productive usage. Nonetheless, in the absence of more data, this measure suffices as a proxy for the degree of diversification of idiosyncratic risk, as it is reasonable to expect that as the share of loans becomes more concentrated in a few sectors, the diversifiable risk associated with the loans in the portfolio becomes greater, thus indicating a need for more diversification.

⁷ If there are 13 sectors in the economy, then the maximum SD is 26.65. That is, √[12*(0–7.7)² + 1*(100–7.7)²]/13 = 26.65

⁸ For building societies, the sectoral classification of loans is different, and so mortgages which finance Building Lots and Land, Commercial and Semi-Residential Properties, Agricultural and other types of projects are productive, while mortgages for Owner Occupied Properties, Housing Schemes and Tenanted Properties are noted to be for consumption.

⁹ For commercial banks and Financial Institution Act (FIA) Institutions, figures for connected party loans and investments are sourced from the Bank of Jamaica prudential reports. No such data are available for building societies, and as such CORPC is not calculated for this type of institution. This is, however, not a major concern, as building societies, by virtue of their area of specialization are not expected to be heavily involved in this function, because they tend to lend heavily to individuals rather than firms.

¹⁰ Levine (1996) explains by noting that, ‘because they make it easy for dissatisfied investors to sell quickly, the liquid markets may weaken investors’ commitment and reduce investors’ incentives to exert corporate control.’

¹¹ This measure is similar to that used by Vergil (2002) to examine the effects of exchange rate volatility on trade.