One potential explanation for low formal credit use is that poor entrepreneurs generate returns to capital below borrowing costs and cannot afford the loans. I test this using a new, nationally representative data from Ecuador, focusing on entrepreneurs that say credit constraints are a major problem. I estimate returns to capital and find monthly returns between 3.5% and 21%, well above prevailing interest rates. Despite this, one third of the finance constrained sample expresses no demand for

more »

... a hypothetical loan. I estimate the determinants of demand for this loan, focusing on the role profitability may play. I find that measures of profitability are positively and significantly associated with demand, and that perceptions of profitability are among the strongest determinants. Meanwhile, assets, employees, duration, formality and past credit use have no predictive power. This suggests that some microentrepreneurs cannot afford prevailing interest rates and rationally eschew formal credit as a result. JEL Classifications: O12, O16, O54 presence of nonconvexities poor entrepreneurs can be shut out of credit markets as they lack the capital to meet collateral requirements and the returns to cover borrowing costs. Although there is some evidence that financial constraints prevent entry into entrepreneurship and enterprises from reaching an efficient scale (Paulson and Townsend, 2005) , recent work finds little evidence of production nonconvexities at low levels of capital. For example, McKenzie and Woodruff (2006) , using survey data from Mexico, find returns to capital around 15% per month for capital levels below $200. Using experimental data from Mexico the same authors find monthly returns that range from 22% to 30% (McKenzie and Woodruff, 2008). Meanwhile, De Mel, McKenzie and Woodruff (2008) using experimental data from Sri Lanka, find real monthly returns in the order of 4.6% to 5.3%. These results suggest many poor entrepreneurs generate returns to capital well above borrowing costs and should exhibit higher demand for formal credit. This paper contributes to the debate over the returns to capital for poor microentrepreneurs and their potential role in formal credit use using new, nationally representative data from Ecuador. The data set can contribute uniquely in several ways. First, it is one of the few large scale samples of urban microentrepreneurs, and provides further evidence of the low use of formal credit. As shown in Table 2 , among poor, urban microentrepreneurs (those with less than $1000 in assets), only 3.2% used formal credit to start their enterprises and only 1.1% use formal credit for on-going operations. Second, few previous papers have directly examined the role returns to capital might play in explaining formal credit use. I can directly address this question because, unlike other surveys, the Ecuadorian one asks detailed questions on the use of and demand for formal credit. This allows for more general statements about credit behavior as well as inferences about demand. Finally, through questions on the most pressing problems facing the firm I can identify firms that view financing constraints as a major problem. By narrowing the analysis of demand to this group I can eliminate a lack of need for credit as an explanatory factor and focus on the role that affordability might play. I first estimate returns to capital for entrepreneurs with $1000 of capital or less and two sub-samples; entrepreneurs that list a lack of funding or an inability to obtain credit as a major problem facing the firm, and entrepreneurs that do not. For the full sample I find monthly returns between 6.4% and 13.0%, which translate into uncompounded annual returns between 76% and 157%. This compares to median interest rates charged by Ecuadorian microfinance institutions of close to 20%, suggesting that many poor entrepreneurs likely can afford available formal credit. For the sub-sample of entrepreneurs that list financing constraints as a problem I find higher returns for capital levels of $500 or less, which is in line with credit constraints being more binding for this group, but lower returns for capital of $500 or more. Overall, however, the estimated returns lie above the threshold interest rate for all capital values, suggesting that affordability is not a concern for most finance constrained entrepreneurs. Next I examine demand for formal credit, gauged by a question which asks entrepreneurs if they are interested in a formal loan for any amount at a 20% interest rate. Surprisingly, one third of entrepreneurs who say financing constraints are paramount say they are not interested in the loan, with the majority citing interest rates that are too high as the main reason. To understand these responses, I estimate demand for the hypothetical loan for the self-identified finance constrained group as a function of observable characteristics, including actual and subjective measures of profitability. significantly more likely to be women, married and have a college education. They are more likely to be informal but, puzzlingly, more likely to keep accounts. They are more likely to operate in retail, an industry with high working capital needs, but less likely to operate in construction, repair, and transportation, industries with high fixed capital needs. Their enterprises are younger, but the average tenure, at 8 years, is still high. Average profits are slightly lower than for those who don't list financing constraints, but the difference, at $6.40 a month, is small. Meanwhile, firm size, as measured by assets and employees, is not significantly smaller, suggesting the group is not dominated by new or less successful enterprises. Overall the finance constrained sample is not markedly different, suggesting they are not necessarily shut out of credit markets because they are less desirable borrowers.