VoxEU Column Development Financial Markets

Financial development and output growth in developing Asia and Latin America: A comparative sectoral analysis

The Global Crisis put to the fore the possibility that the relationship between financial development and output growth may be non-linear. This column presents new evidence on the issue using data on output growth of ten sectors from Latin America and East Asia. The authors find large differences between the two regions in terms of the impact of financial depth on sectoral growth, and validate the negative impact of financial deepening on output growth in several sectors. The results confirm that the impact of financial development on sectoral growth may indeed be non-linear – i.e. it may promote growth only up to a point. 

The financial-growth nexus differs in advanced and developing countries

A lingering challenge in applied economics is measuring and controlling the quality of services – e.g. health care, education, and finance. Since services are measured at costs, their GDP share is correlated with per capita GDP. But short of controlling for the quality of services, it is not clear if the growing share of services in the GDP adds to welfare. The Global Crisis put to the fore the possibility that the relationship between financial depth and output growth may be non-linear – the development of the financial sector may benefit the real sector, but only up to a point.1 Beyond that point, further financial development may have no effect or even a negative effect on growth.2

Concerns about too much finance are more relevant for advanced economies that already have mature, sophisticated financial sectors with a good mix of banks and capital markets than for developing economies with backward financial sectors. Diminishing marginal returns to financial development are more likely at higher levels of financial development. However, there are fundamental measurement issues that affect both advanced and developing economies. The crux of the problem is that the standard measures of financial development are quantitative measures, such as the ratio of private credit to GDP. Quantitative measures are highly imperfect measures of financial development, which refers to the quality of the financial system, or its ability to allocate resources to the most productive uses. It is conceivable that even as the financial system expands in size, its capacity to channel resources efficiently stagnates or even recedes. Perhaps no country epitomises such risks better than today’s China.

East Asian countries generally have large financial sectors relative to their income levels. Yet, they remain well inside the global finance frontier, as evident in their recycling of much of their abundant savings through the financial markets of the advanced economies. Even within the context of East Asia, China has an exceptionally large financial sector, yet few would mistake the large size as evidence of financial development. To the contrary, there are widespread concerns that an unsustainable expansion of credit – i.e. unsustainable expansion of financial sector – is jeopardising financial stability and eroding the quality of investments.[3] A specific major concern is that state-owned banks may be channelling credit to state-owned firms at the expense of credit to the dynamic private sector. Latin America and East Asia are at similar income levels but one key difference is the relative abundance of savings in the latter. As such, a comparative analysis of the finance-growth nexus in the two regions is worthwhile.

New evidence on finance and growth from Latin America and East Asia

We explore some of the above issues by delving into the relationship between financial depth and sectoral output growth (Aizenman et al. 2015). The sectoral data encompass 10 sectors: Agriculture, mining, manufacturing, construction, public utilities, retail and wholesale trade, transport and communication, finance and business services, other market services, and government services. Our analysis covers 41 economies, including 11 East Asian and 9 Latin American economies which we compare.  

The importance of the quality of financial intermediation has been well recognised in the literature, though identifying its effect remains a work in progress (see Levine 2005 for a comprehensive review, and the review in our paper). Taking stock of the literature, one may conjecture that credit boom and bust cycles associated with financial deepening would disproportionately affect activities and sectors that rely on stable external finance and are subject to larger sunk costs. The tenuous link between financial depth and growth may reflect a host of factors, including the damaging and uneven effects of credit cycles, as well the rent seeking associated with distorted incentives in the financial sector, where excessive risk taking and financial innovation may precipitate instability that penalizes credit dependent sectors. This is all  the more likely if the greater short and medium run profits associated with financial innovations divert credit from the real economy to further deepening of speculative financial intermediation. 

Figure 1 plots average value-added per worker for 1996-2011. Across the ten sectors the level of value-added per worker in East Asia and Pacific is higher than the level in Latin America and the Caribbean. Figure 2 summarises regional differences in the quality of financial intermediation, including countries in East Asia and Pacific, Latin America and the Caribbean, and a few other areas with sectoral output data from GGDC, while measures of quality of financial intermediation are missing altogether for some countries. In the whole sample and both sub-samples, we find that bank private credit to GDP (financial depth) is negatively associated with the growth of construction sector. Bank private credit growth is also negatively associated with the growth of manufacturing sector in East Asia, whereas it is positively associated with the growth of finance, insurance, and real estate sector in Latin America.

Figure 1. Value-added (constant 2005 prices) per worker in ten sectors

Notes: This figure provides average value added per worker in East Asia and Pacific, and Latin American & the Caribbean, from 1995-2011.
Sources: Authors' caluculation on ten-sector database.

Figure 2. Regional differences in quality of finance, 2005-2011

Notes: This figure provides an average for each region measures of financial depth – Bank Private Credit to GDP (%, BCRY); efficiency – Lending-Deposit Interest Spread (%, SPRD); governance – Getting Credit Index (GTCR) and Resolving Insolvency Index (REIN), and access – Depositors with Commercial Banks per 1,000 Adults (DCBA) and Outstanding SME Loans from Commercial Banks (% GDP, LSME).
Sources: Authors' calculation on World Bank Global Financial Developement Database (BCRY, SPRD), Doingbusiness Database (GTOR, REIN); and IMF Financial Access Survey (DCBA, LSME).

For the East Asian economies, we find that lending-deposit interest spread is positively associated with the growth of finance, insurance, and real estate sector. The growth of the construction sector is negatively associated with lending-deposit interest spread in East Asia, whereas it has a positive association in Latin America. The growth of wholesale and retail trade sector is positively associated with financial efficiency in East Asia, whereas the association is negative in Latin America.

Figure 3. Economic significance on sectoral output growth

Notes: This figure provides economic significance (annualised) based on estimation results of ten years of data (1996-2011) of Bank Private Credit (%, BCRY; including lagged, current, and squared terms) and Lending-Deposit Interest Spread (%, SPRD) on Sectoral Output Growth (%), based on estimation results. The economic significance is calculated as a product of coefficient estimate and sample standard deviation of variable.
Sources: Authors' calculation on Groningen Growth and Development Centre (GGDC) ten-sector database and World Bank Global Financial Development Database.

Our empirical analysis uses data on sectoral composition of the economy with controls for the quality of financial intermediation. The quality of financial intermediation is impacted by prices (i.e. financial intermediation spreads) and quantities (e.g. financial depth, SMEs’ ease of getting credit). Financial spreads are, in turn, affected by access to credit, and quality of institutions (e.g. rights of creditors and efficacy of the judicial system), and the risk level of loans. Figure 3 provides the economic significance of the controls. Intriguingly, for several sectors the control variables for our proxy of de facto quality of financial intermediation tend to have levels of economic significance that are larger in Latin America and the Caribbean than they are in East Asia and Pacific. The economic significance tends to be larger for non-tradable sectors, i.e. public utilities, wholesale and retail trade, community and social services, finance, insurance, and real estate, than for other sectors. In addition, bank private credit to GDP tends to have smaller economic significance than lending-deposit interest spread. Since credit may be scarcer in Latin America, the marginal importance of the quantity of finance is larger than in East Asia. The larger credit base of East Asia implies that region has reached the stage where quality may be at least as important as the quantity.

The evidence suggests that the level of service flow of financial sector is good only up to a point, after which it becomes a drag on sectoral growth in the sample countries.  To verify the possibility of ‘financial Dutch disease’ – i.e. booming financial service flows reduces the supply of long-term funding to manufacturing and other sectors that rely on stable external finance – we add a lagged growth of finance and business services, and control for the interest spread as well as its interaction with the growth of finance and business services. We find some support of this hypothesis in the whole sample since the coefficient estimate on the interaction term is negative and statistically significant for the growth of manufacturing sector.

Concluding remarks

Overall, our evidence is consistent with the hypotheses we set forth at the outset, in particular the non-linear effect of financial development on growth and its uneven effect across sectors. In addition, we find that financial depth has a negative effect on manufacturing in East Asia and a positive effect on finance, insurance, and real estate sector in Latin America. Financial efficiency, as measured by lending-deposit interest spread, is positively associated with the growth of finance, insurance, and real estate sector. Construction sector growth is negatively associated with the spread in East Asia, but positively in Latin America. Several of the differences between the regions may reflect the greater scarcity of finance in Latin America in comparison to East Asia. It may be that the expansion of East Asia’s financial sector has reached the stage where the quality of finance may be at least as important as its quantity. We also find some evidence of a financial Dutch disease – the faster the growth of financial services and the larger the lending-deposit interest spread, the slower the growth of the manufacturing sector. Future research would benefit greatly from better measurement of the quality of finance by controlling for, for example, the degree of financial repression and the role of directed credit and state-owned banks.

References

Aizenman J, B Pinto and V l Sushko (2013), “Financial sector ups and downs and the real sector in the open economy: Up by the stairs, down by the parachute”, Emerging Markets Review, 16 (2013) 1–30.

Aizenman J, Y Jinjarak and D Park (2015), “Financial Development and Output Growth in Developing Asia and Latin America: A Comparative Sectoral Analysis,” NBER Working Paper No. 20917.

Alfaro, L, S Kalemli-Ozcan and V Volosovych (2003), “Why doesn’t Capital Flow from Rich to Poor Countries? An Empirical Investigation” University of Houston, December.

Boyd, J H and B D Smith(1992), “Intermediation and the equilibrium allocation of investment capital: Implications for economic development”, Journal of Monetary Economics 30, 409–432.

Cecchetti and Kharroubi (2012), “Reassessing the impact of finance on growth”, in proceedings of the Second International  Research Conference of the Reserve Bank of India, forthcoming.

Eichengreen, B, and K J Mitchener (2003), “The Great Depression as a Credit Boom Gone Wrong”, BIS Working Paper No. 137, September.

Jordà, Ò, M Schularick, and A Taylor (2011), “When Credit Bites Back: Leverage, Business Cycles, and Crises,” NBER Working Paper No. 17621

Ju J and S-J Wei (2011), “When Is Quality of Financial System a Source of Comparative Advantage?” Journal of International Economics, 84(2), pages 178-187.

Minsky, H P (1974), "The modeling of financial instability: An introduction," in Modelling and Simulation, Vol. 5, Proceedings of the Fifth Annual Pittsburgh Conference, Instruments Society of America, pp. 267.72.

Philippon, T, and A Reshef. (2013). “An international look at the growth of modern finance”, The Journal of Economic Perspectives 27 (2): 73-96.

Rousseau, P, and P Wachtel (2011), “What is Happening to The Impact of Financial Deepening on Economic Growth?” Economic Inquiry, 49, 276-288.

Schularick, M, and A M Taylor (2009), “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles and Financial Crises, 1870–2008”, NBER Working Paper 15512.

Footnotes

1 Boyd and Smith (1992) show that the quality of financial intermediation has first-order effects on capital flows and economic growth, providing a nice interpretation to the Lucas paradox (1992) of capital flowing uphill, a topic that gained even more attention in the context of the global imbalances in the 2000s (see Laura et al. 2003, Ju and Wei 2011 and the references therein). Cecchetti and Kharroubi (2012) study how financial development affects growth at both the country and industry level. They find the level of financial development is good only up to a point, after which it becomes a drag on growth. These results are in line with Rousseau and Wachtel (2011)’s ‘vanishing effect’ – i.e. credit has no statistically significant impact on GDP growth over the 1965-2004 period. Looking at the more recent data, Philippon and Reshef (2013) concluded that at the very high end of financial development, rapidly diminishing social returns may have set in.  Aizenman et al. (2013) found that periods of accelerated growth of the financial sector are more likely to be followed by abrupt financial contractions than are periods of slower financial sector growth. Though these studies do not identify the mechanisms associated with the ‘vanishing effect’ of finance, they are consistent with Minsky (1974)’s  hypothesis over time that financial deepening may eventually divert financial resources from financing real activities into speculative and ultimately destabilising risky and bubbly yield-seeking financial investments.

2 For example, the Global Crisis was preceded by a wave of financial innovation which produced many complex high-tech financial products but entailed little obvious benefit for growth. Excessive financial innovation that overwhelms the regulatory capacity of regulatory authorities can culminate in financial crisis which sets back growth for some time.

3 The recent empirical literature validated the key role of credit cycles. Schularick and Taylor (2009) present long-run historical data showing that financial instability was often the result of "credit booms gone wrong". Their analysis lends support to the Minsky-Kindleberger view of financial crises (Eichengreen and Mitchener 2003). The credit system seems all too capable of creating its very own shocks, judging by how well past credit growth predicts future financial crises. The dynamic role of credit overhang is further validated by Òscar et al. (2011). Analysing over 200 recessions in 14 advanced countries between 1870 and 2008, they find that financial crisis recessions are more costly than normal recessions in terms of lost output. For both types of recessions, credit-intensive expansions tend to be followed by deeper recessions and slower recoveries. Credit growth also affects the behaviour of other key macroeconomic variables such as investment, lending, interest rates, and inflation. 

2,835 Reads