What makes firms decide whether to buy their inputs from another firm or to vertically integrate and become their own supplier? Contracting problems between buyers and suppliers motivate a vast literature on the boundaries of the firm (e.g., Coase 1937, Williamson 1975, Grossman and Hart 1986). In particular, firms may choose to enter into procurement contracts with suppliers or source goods internally depending on the degree to which future contingencies can be specified between the two parties (i.e., what economists refer to as “contract incompleteness”). For example, if there is no contractual way to mitigate the uncertainty about the value of an intermediate good, a buyer firm may prefer to vertically integrate and avoid having to renegotiate with the good’s supplier in the future. In this context, trade credit, or delayed payment, may play a key role in allowing suppliers to guarantee the quality of their goods, enabling market-based procurement relationships. In a new paper, Emily Breza (Columbia) and I test this claim empirically.
Trade credit is one of the most prevalent features in procurement relationships (Petersen and Rajan 1997, Cuñat and Garcia Appendini 2011). Even large corporations, which have a relative financing advantage over their suppliers, rely heavily on trade credit. For example, Walmart uses four times more supplier financing than short term external financing. Since Walmart is much larger and has better access to financing than its typical supplier, trade credit must have intrinsic value to corporations beyond traditional access-to-finance-based explanations.
We exploit a natural experiment that restricted the terms of trade credit that relatively smaller suppliers could offer to their common large client to study the client's decision to integrate or procure from a third party. The experiment occurred as follows. In January 2007 one of the two large supermarket chains that operate in Chile (the “Supermarket”) agreed with the Chilean government to reduce the number of days in which it paid its small suppliers from approximately 90 to no more than 30 days (the “Agreement”). Based on reports from the press, the Agreement was put in place because of the government's concern that the Supermarket exerted monopsonistic power over its small suppliers. As per the Agreement, suppliers were categorized as small by an arbitrary yearly revenues cutoff (roughly US$4.0 million). We identify the causal effect that the restriction on the ability to extend trade credit had on these procurement relationships by comparing the same product sold by two very similar suppliers, one exposed to the lower payment days and one not, before and after the Agreement was put in place.
Our analysis suggests that the restriction on the set of feasible contracts had important effects on the organizational form of the Supermarket and its supply chain. First, we show that after the Agreement, the probability that an affected supplier sells the same product to the Supermarket falls by 9% relative to an unaffected supplier. Further, for those products that are still sold after the Agreement, the reduction in days payable results in a 4.6% average reduction in prices. The reduction in prices implies a yearly interest rate of 32% given the 60 day change in days payable, significantly higher than the contemporary bank rates of 7%-11% that the Supermarket could obtain in formal credit markets. This is consistent with the notion that trade credit is very important to the large Supermarket, because absent the Agreement they could have always chosen to pay suppliers a lower price upon delivery of the good.
Second, we document how the Supermarket responds to the restriction in the ability of its suppliers to extend trade credit. For products that were mostly procured by affected firms, the Supermarket shifts purchases both to unaffected suppliers and to fully owned subsidiaries. Thus, for some products, the Supermarket chooses to modify its boundaries and vertically integrate to supply its inputs. We thus verify that the ability of suppliers to extend trade credit makes it more likely that the Supermarket chooses to procure externally.
Third, consistent with previous work (McMillan and Woodruff 1999), we find that the effects of the Agreement are significantly reduced for older firms, which have been able to establish a reputation. Also, the effects are reduced for products whose quality can be easily ascertained (“search” goods”). These results suggest that trade credit allows suppliers to post a bond to its client in the presence of information asymmetries about the quality of the good or of the firm. Older firms, or firms selling products whose quality is easily ascertained do not need to extend trade credit to guarantee the quality of their products.
Our results inform the debate on the desirability of policies that restrict the terms at which parties agree to transact. Restricting trade credit may help suppliers bear the cost of extending trade credit, a claim we may not test with our data. However, by not considering how the parties respond to this restriction, the Agreement may have ended up harming small and young suppliers. This is, presumably, exactly the opposite of the policy’s intended effect.
Coase, Ronald H, 1937, The nature of the firm, economica 4, 386–405.
Cuñat, Vicente, and Emilia Garcia-Appendini, 2011, Trade credit and its role in entrepreneurial finance, Handbook of Entrepreneurial Finance, edited by Douglas Cumming, Oxford University Press, forthcoming.
Grossman, S.J., and O.D. Hart, 1986, The costs and benefits of ownership: A theory of vertical and lateral integration, The Journal of Political Economy pp. 691–719.
McMillan, J., and C. Woodruff, 1999, Interfirm relationships and informal credit in vietnam, The Quarterly Journal of Economics 114, 1285–1320.
Petersen, Mitchell A, and Raghuram G Rajan, 1997, Trade credit: theories and evidence, Review of Financial Studies 10, 661–691.
Williamson, Oliver E, 1975, Markets and hierarchies, New York pp. 26–30.