Syndicate content

Optimal Financial Structures for Development? Some New Results

Asli Demirgüç-Kunt's picture

One of the interesting debates in the finance and development literature is on financial structures: does the mix of institutions and markets that make up the financial system have any impact on the development process? Last week we hosted an interesting conference on the topic at the World Bank (click here for the agenda and papers). Those of you who have been following this literature will know this is not the first time this topic has been discussed – we held a conference on financial structures over ten years ago.

What do financial structures look like? How do they evolve with economic development? What are the determinants and impact of financial structures? Years ago Ross Levine and I, along with many others, tried to answer these questions and saw clear patterns in the data. One stylized fact: Financial systems become more complex as countries become richer with both banks and markets getting larger, more active, and more efficient. But comparatively speaking, the structure becomes more market-based in higher-income countries. We also saw that countries did not get to B from A in a single, identical path. You didn’t have any market-based financial structures in the lowest-income countries, but as soon as you got to lower-middle income, financial structures became very diverse: Costa Rica was bank-based, whereas Jamaica was much more market-based; Jordan was bank-based, Turkey was market-based etc. etc. So countries were all over the place and the correlation between GDP per capita and financial structure was less than 30 percent.

So we looked at what other factors are correlated with financial structure and saw that even after controlling for gdp/cap there were significant and sizable correlations with a country’s legal origin, legal codes and their enforcement, quality of accounting standards and information availability, historical and cultural differences, political systems and so on. One potential conclusion from all this is that the development process is multi-faceted, and since every country does not follow the same path in development, their financial structures can show variations too. And if you look at it that way, this also shifts the emphasis away from the financial structure itself and more towards the services that the financial system provides. So the crucial issue is what the financial system does; whether it provides the necessary services, and not really how or through which institutions and markets they are delivered. If the underlying legal, institutional, and political systems are different – even at the same level of gdp/cap – the financial structure will reflect that; but as long as the system functions well, this will not be all that relevant.

Well, that was our conclusion then. At that time, we were less concerned about the concept of an “optimal” financial structure – the idea that at each stage of development, the financing needs of countries are likely to be different. Since the industrial structure and the enterprises tend to differ in terms of size, risk, and financing needs at each stage of development, the demands of the economy for financial services will also vary, resulting in different financial structures that are better able to perform these functions as the economies develop. As Justin Lin articulated in his opening presentation at the conference, this framework suggests there is an optimal financial structure for each country and deviations from this optimum may curtail economic activity.

The conference included many interesting papers, some of which I will blog about in future posts. Franklin Allen and his colleagues provide a comprehensive historical review of financial structures in the US, UK, Germany and Japan and how they contributed to the development process. Despite the fact that financial structures clearly evolved in response to demand, their work illustrated how difficult it is to identify a unique optimal path. Clearly finance is an integral part of an organically evolving political-economic system. To the extent that the economy innovates and grows, the financial system will both reflect those dynamics and shape the process of technological innovation and economic development. Papers by Steve Haber, Charlie Calomiris, Fenghua Song and Anjan Thakor emphasize that politics is a central and inextricable component of this synergistic evolution.

Among the empirical papers, Erik Feyen, Ross Levine and I take a fresh look at our earlier cross-country results and this time explore whether deviations from an optimal financial structure is associated with development. Bob Cull and Colin Xu use firm-level data across many countries, looking at how labor growth rates of firms vary with their country’s financial structure. Thorsten Beck, Dorothe Singer and I focus instead on banking structure and try to see if bank size and banks’ relative importance among all other financial institutions affect financial access by firms.

What do we conclude? What are the policy implications? This was the focus of a panel chaired by Mahmoud Mohieldin. Overall, compared to our earlier work, new evidence suggests that indeed different financial structures may be better at promoting economic activity at different stages of a country’s economic development. We always knew both banks and stock markets are good for development, but we were also very uneasy when we saw our earlier findings used as justification to construct elaborate stock markets in some of the poorest countries of Sub-Saharan Africa. Hence, these new results clearly suggest those types of interventions are not a good idea.

But do we know enough to target a given financial structure for particular countries? Should we actively intervene and “fix” financial structures? No. However, if we do see that market or bank development is too skewed compared to what we would expect, these findings give us a reason to dig much deeper. Among the questions we can ask are:

  • Are taxes, regulations, or legal impediments inducing nonfinancial sector firms to rely excessively on markets, rather than the services provided by banks – or vice versa?
  • Are barriers to competition providing protection to large banks and hindering the entry of small banks?
  • Are there taxes that limit the development of different industries in the economy?

Hence much of the new research can justify a more focused, in-depth investigation of particular characteristics of the financial system that might be hindering economic progress. Stay tuned for more.


Dear Dr. Demirguc-Kunt, Thank you for sharing this very important discussion that is in progress with us. This is indeed an important area and I wonder if the discussion should not be broadened beyond banks and financial markets to include insurance, mutual funds and pensions as well. There are, in my view, two "mega-trends" that are nearing their conclusion: 1. In a post-biometric / UID / fibre-optic / 3G world cash, credit cards, ATMs, etc. are on their way out and perhaps even the traditional branches with large teller counters and big vaults. It may perhaps be a blessing in disguise (as was the case for mobile and fixed line phones) that developing countries failed to establish the very expensive traditional branch infrastructure despite an enormous policy push to do so. 2. The risk management technologies in all of these different domains are converging rapidly and it is now hard to separate the thinking behind capital rules for insurance from those for banks and securities markets. The product-markets have launched credit insurance and credit derivatives in a domain such as credit once thought of an exclusive preserve of banks. However the older challenges of managing moral hazard, adverse selection, irrational exuberance, fraud remain. Does this imply a fundamentally different design for financial systems that are not quite as product based as they have been historically? Perhaps go closer to the direction that Australia has gone? We are holding a conference in India at IFMR ( in August on precisely this issue so that we can launch a very carefully structured research agenda that will we hope inform policy making in this very broad area. Your inputs and guidance for us as we plan this would be much appreciated. Sincerely, Nachiket Mor

Submitted by asli on
Dear Nachiket, I completely agree with your points. Actually in one of the papers in the conference, we looked at the institutional structure and tried to see if relative importance of banks vs. other financial institutions made a differencce for access to financial services by firms. I will blog on that paper at a later date. Your conference sounds very interesting and timely. Best regards, Asli

Add new comment