«Valentina Hartarska Associate Professor Department of Agricultural Economics and Rural Sociology Auburn University, 210 Comer Hall Auburn, AL 36830 ...»
Financing Constraints and Access to Credit in Post Crisis Environment: Evidence from
New Farmers in Alabama
Department of Agricultural Economics and Rural Sociology
Auburn University, 210 Comer Hall Auburn, AL 36830
Phone: (334)844-5666, email: firstname.lastname@example.org
Department of Agricultural Economics and Rural Sociology
Auburn University 209 B Comer Hall Auburn, AL 36830
Phone: (334)844-5630, email: nadolda@ auburn.edu Selected Paper prepared for presentation at the Agricultural & Applied Economics Association’s 2012 AAEA Annual Meeting, Seattle, Washington, August 12-14, 2012 Copyright 2012 by [Hartarska and Nadolnyak]. All rights reserved. Readers may make verbatim copies of this document for non-commercial purposes by any means, provided that this copyright notice appears 1 Financing Constraints and Access to Credit in Post Crisis Environment: Evidence from New Farmers in Alabama Abstract We use survey data to study the degree to which new farming operations in Alabama were financially constrained in the post crisis period. Next, we control for farmers’ self-selection out of the credit market and identify which farmers were able to secure loans during the 2009-2010 period. The results show that new farmers, who started any part of their operation after 2005, were financially constrained, but there was no evidence that their financing constraints were affected by the 2008 financial crisis. As expected, we find that lending was collateral driven, although lenders also considered farmers’ profitability and cash flows.
Keywords: financing constraints, access to agricultural credit, new farmers JEL Classification: G31, Q12, Q14 2 Financing Constraints and Access to Credit in Post Crisis Environment: Evidence from New Farmers in Alabama Financial market crises translate into limited access to credit with negative consequences for all producers including those in agriculture. We study how the 2008 crisis affected agricultural producers’ access to credit. Agricultural banks were less affected because they are small compared to non-agricultural banks. Since previous financial crises have affected agricultural lenders significantly, this time they were in a better position to manage risks (Briggeman et al., 2009; Ellinger, 2009). Nationwide, agricultural sector profitability peaked in 2008 but has decreased since. Consequently, while the share of problem loans of agricultural lenders remains less than 50% of that of non-agricultural banks, delinquencies have been increasing (Briggemann, 2011, Ellinger, 2011). Increased delinquency rates typically lead to elevated collateral requirements with a potential to worsen access to credit for agricultural producers, especially among more vulnerable groups (Briggeman and Zakrzewicz, 2009).
This paper sets out to determine the impact of the financial crisis on access to credit for new farming operations and to determine which farmers got credit in the post crisis environment.
The existence and magnitude of credit constraints for agricultural producers are non-negligible.
Nationwide, Briggeman et al. (2009) estimate that the value of production is 3% lower in creditconstrained farm sole proprietorships compared to those that are not credit-constrained. Credit constraints have also been found in agricultural cooperatives and shown to affect land values (Chaddad et al, 2005; Mishra et al., 2008) If the financial crisis has affected farmers’ ability to borrow, then new operations should be most affected, since they typically lack capital, experience, or both. Even with the introduction of special support programs, beginning farmers continue to face production and
Previous studies have found that new operations are financially constrained and that, for younger (and high debt) farmers, the financial constraints are affected by the business cycle (Barry et al, 2000; Bierlen and Featherstone, 1998). This group is most vulnerable because banks elevate collateral requirement when delinquencies are on the rise and new farmers typically have less assets to offer as collateral. Moreover, even when lenders make lending decisions based not on collateral but on projected performance, younger farmers are still at a disadvantage because they have lower return on assets compared to more established operations (Mishra et al., 2009).
Even prior to the financial crisis, farmers in Alabama, especially small sole proprietors, were financially constrained and used off farm spousal income to invest on the farm (Hartarska and Mai, 2008). In this paper, we use survey data collected in the fall of 2010 from new operations in Alabama to study the degree to which new operations were financially constrained during the post crisis period and to identify the factors affecting access to credit in the 2009-2010 period.
The remainder of the paper is organized as follows. Section 2 presents the conceptual framework and empirical specifications. Section 3 briefly describes the data. Section 4 summarizes the results. Conclusions are offered in Section 5.
2. Analytical framework and empirical specifications The analysis consists of first establishing if new operators have financing (or liquidity) constraints and whether these constraints have become more severe in the post crisis period.
Next, we identify the factors affecting farmers’ ability to obtain credit, in order to gain insights into possible ways to alleviate existing financing constraints.
markets. According to this literature, in the presence of high transaction costs and asymmetric information, loans are either rationed or available at a premium (Jensen and Meckling, 1976;
Stiglitz and Weiss, 1981). In such circumstances, external and internal finance are no longer substitutes and investment in firms facing high information costs, such as most new farming operators, is constrained by the availability of internal funds (Myers and Majluf, 1984). Since financial constraints do not affect all farmers uniformly, the extent of effective financing constraints that different operators face provides information on the ability of the financial system to cater to their financial needs in that time period.
Financial constraints are important in farming because farming is capital intensive and, while farmers do not like debt, many especially newer operations have limited ability to undertake profitable investment with only own funds. The lack of equity markets and seasonality of cash flows makes access to loans crucial and the ability of credit markets to alleviate financing constraints very important. Moreover, limited diversification opportunity and supply shocks lead to large variations in farmers’ net worth and profitability further restricting their investment.
The financing constraints approach pioneered by Fazzari et al. (1988) tests for differences in sensitivity of investment to internal funds in firms with different levels of informational opacity by comparing sub-samples defined according to priors that characterize constrained and unconstrained firms (e.g., new and established farms). For each sub-sample, a reduced-form investment equation is estimated where investment is modeled as a function of internal funds and investment opportunities determined from a variety of theoretical perspectives (Hubbard, 1998).ii A statistically significant difference in investment sensitivity to internal funds between subsamples indicates that one group is more credit constrained. Recently, Carreira and Silva (2010)
argue that numerous studies find that younger firms are more financially constrained than established firms.
We first estimate a reduced-form (change in) investment equation for the 2008-2010 period for two groups of Alabama operators: new (started any part of their operation between 2000 and 2004) and newest (since 2005) and test for the difference in sensitivity of investment to cash flows. In this framework, we also test for differences in financing constraints before and after the crisis of 2008 for each group. Following Hartarska and Nadolnyak (2008), investment is modeled as a function of operators’ investment opportunity and internally generated funds (typically defined as revenues minus expenses) to which we add change in liquidity since 2008
and controls.iii The estimated model is of the form:
∆ Investmenti = β0 + β1Inv Opportunity i + β2 Cash Flowi + β3 Change in Liquidity since 2008i +
where ∆ Investment is the percentage change in the value of Fixed Assets, Inv Opportunity is a measure of investment opportunity proxied by the change in the Return-on-assets ratio (ROA), Cash Flow is the cash flow measure that proxies for available internal liquidity, Change in Liquidity since 2008 is a dummy that measures the impact of the 2008 crisis on liquidity and takes the value of one if, after 2008, operators kept larger proportion of cash and liquid assets, compared to before 2008.iv Investment in farms differs from that in firms because, for farmers who own their land, the land is the largest part of fixed investments. Some operators may not be landowners, and landowners may not be working on their farms. The dependent variable measuring change in fixed assets may contain possible measurement error since the survey did not collect data on land
Furthermore, when farmers cannot obtain a loan to invest in fixed assets, they could lease the land, and there will be no change in investment but we argue that even if this is true it will be a dependent variable measurement error which does not lead to biased coefficient estimates.v We also note that the majority of farm operators in Alabama (68 % of all farm sales) are in livestock production (cow and calf) or poultry for which land is a less important capital asset compared to land in row-crop producing regions. In our sample row-crop producers are only four percent. Thus, we include farm operation types as explanatory variables. We also capture variation in land assets size by including farm assets classes at the beginning of the operation.
We also control for spatial land value differences by including the average county-level price of land. During the study period, there were no recorded drops in the price of agricultural land values, so possible bias is likely one-sided – increase in the value. Since the possible measurement error is in the left hand side variable, it will be swept away in the error term.
In this class of models, proper measurement of investment opportunities and cash flow (liquidity) is important. Farm operators who do not have investment opportunities would not invest even if they had cash. These two effects must be clearly separated to ensure that the Cash Flow variable (liquidity, net worth) is capturing internally generated funds and not investment opportunity to avoid attributing investment’s sensitivity to cash flow. To measure available liquidity, we directly ask farmers what percentage of their revenue they keep in liquid assets.
In the literature, investment opportunity is a measure of the expected value of future profits or discounted value of income from 1 extra $ investment. In large firms, this is typically the average q which, under certain conditions (Hayashi, 1982), serves as a proxy for marginal q or the fundamental q which is measured in various ways (see Bierlen and Featherstone (1998) for
by (previous year) employment growth, sales growth, or indicators of current profitability such as return-on-assets or ROA (see Carreira and Silva 2010). We measure investment opportunity with three categories for change in ROA – increase, decrease and no-change as the base. vi,vii Empirical evidence shows that farmers’ off-farm investment is affected by entrepreneurial and operation characteristics (Mishra and Morehart, 2001). Since money is fungible within the household, these factors may also affect farm investment. We include controls for entrepreneurial experience and experience in farming prior to starting this operation, whether the operator or the spouse work off farm to capture possible access to external funds, the age of the operation to capture experience, and gender of the entrepreneur to capture differences in preferences for investment. We also include the proportion of income coming from farming to control for hobby farming as well as the proportion of sales coming from various types of farming e.g., livestock (largest group and serving as the base), poultry, specialty crops, government payments, and others.
While panel data would be preferable, such data are too costly to collect, especially given the relatively small population of new operators in Alabama and the difficulty of soliciting financial information, as well as because of large expected attrition due to high percentage of failure of new enterprises. Instead, farmers were asked to provide information for change in the key variables during the period 2008-2010, which compensates partially for the lack of panel data. Nevertheless, we interpret the results cautiously and argue they are valid for the state of Alabama and the study period.
We next determine which farmers were able to overcome their financing constraints and secure loans. To answer this question, we estimate a probit model where the dependent variable
the market if they believed they would not be approved even if they applied, we need to control for farmers’ self-selection. Thus, we use a Heckman probit model as described by Van de Ven and Van Pragg (1981). The unobserved relationship is
where yj* is the credit received by operators and x includes variables affecting banks’ decisions to lend. However, instead of y*j, we only observe a binary outcome (received or did not receive loans) which is captured by a probit equation
The dependent variable for operator j is observed only if we observe a loan application from that operator. Thus, the selection equation (applied or did not apply for a loan) is