Panel data can be used to measure directly or infer indirectly the presence and role of informal financial networks. In the Townsend Thai data  (a collection of over 12 years of annual panel data for 900 households in 64 Thai) villages, networks are shown to play a beneficial role in smoothing consumption and investment against income and cash flow fluctuations. Villagers who lack formal financial access but are indirectly connected through networks receive the benefits of the formal financial system. Surveys of financial access that ignore these networks can understate the reach of financial access while hiding the needs of the truly vulnerable (e.g., poor households without any kin in their village). Complementarities between the formal financial system and informal networks show up in bridge loans for repayment and the transactions demand for cash, revealing highly active informal money markets.
The same logic and data make labor supply and hours data conform with those of a sophisticated risk syndicate and make the rate of returns on investment/occupations conform with the theory of modern finance—in particular a capital asset pricing model applied to technologies/occupations with common market/village risk. We found that families engaged in occupations like rice farming require a higher expected return because this activity does well only when the village as a whole is doing well, and conversely occupations which are not covariate with market risk are recognized as particularly valuable. However, heterogeneous risk preferences creates a policy warning: outside insurance targeting village/market risk can actually make some in a village worse off, those had been providing insurance to others.
Networks vary in extent and reach. Variation in village cooperation attenuated the impact of a large Thai microfinance initiative, the million baht fund initiative, in some locations. Being rated by the households themselves as the best village in cooperation lowers the default rate on loans between 5 and 8 percent, and rated as best by institutions (again in a survey before the million bhat program was even imagined) from 18 to 24%. While these ratings are not statistically significant in urban communities, training and compulsory savings, though effective in both, are even more effective in urban communities. Yet, as economies develop, networks can morph into other forms. Viewing villages as countries in the larger national sea, through the lens of economic (NIPA, balance of payments, savings-investment) accounts, informal remittances and gifts are shown to be a large part of the financial system, smoothing consumption and facilitating the flow of funds to high yield investment projects. The latter is hard to detect at the local level. Family-related firms without apparent geographic boundaries act as risk syndicates at the larger, national level, thus policies to break up Chaibol-or-syndicate-like arrangements should be implemented with some care. Moral hazard, willingness to default, and other obstacles cause less damage for groups of family-related firms, even those without bank access, relative to those borrowing from single or even multiple banks.
However, this kind of interconnectedness appears to leave the financial system vulnerable to credit chains, cascades of default, and systemic risk. So, is the shadow banking associated with informal networks in developing countries a good or bad thing? Fortunately, careful logic with the same perspective on risk bearing and investment provides the needed policy guidance. On the down side, unlimited ex post side trade does undercut the effectiveness of ex ante high power incentives that deal with information problems. Similarly, active informal secondary markets, which mark to market the value of collateral, create pecuniary externalities and inefficient outcomes. But there are remedies. Indexation with aggregate shocks is helpful, i.e. indexed funds and easier access to these hedges. Another is appropriate liquidity provision to deal with flow of funds and payments imbalances, as when the Central Bank recognizes that private markets are incomplete and so its job is to try to equalize rates of return on assets (adjusting for risk). Another is a free ex ante market in "exchange islands" for unwinding state-contingent trades, though as yet this has not been implemented in practice and exchange and over-the-counter regulations are not guided but economic theory. Markets versus regulation creates a false dichotomy. Financial system design is a vision for the creation ex ante of constrained-optimal markets and institutions and a commitment to carry out these plans ex post, especially in the face of realized mis-matches and adverse shocks.
"Heterogeneity and Risk-Sharing in Thai Villages ." Pierre Andre Chiappori, Krislert Samphatharak, Sam Schulhofer-Wohl, and Robert M. Townsend. Working paper, 2011.
"Villages as Small Open Economies ." Archawa Paweenawat and Robert M. Townsend. Working paper, 2011.