ABSTRACT
This article examines factors contributing to the substantial change in the terms of payment, such as letter of credit (LC) and cash terms, in Korean export transactions during the period 1997–2015, using the Autoregressive Distributed Lag bounds test approach. We find that high competition in the world market, the expansion of exports of information technology products and intra-firm trade, and the improvement of information and communication technology (ICT) decrease the share of exports on LC terms in total exports in the long run. However, the increase in exports to developing countries raises the use of LC terms. On the other hand, the increase in intra-firm trade and ICT improvement raise the share of exports settled in cash, and higher trust between trading partners increases the use of cash terms.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 The data are seasonally adjusted and based on monthly average.
2 Although they are not quite equivalent, the terms of payment in international trade can be interchangeably referred to as trade finance instruments, as the former entails financing during cross-border shipment.
3 In this article, cash terms consist of simple remittance and cash on delivery (COD) and cash against document (CAD). Simple remittance includes both cash-in-advance and open account terms. COD and CAD are variants of post-shipment payment terms. Refer to International Trade Administration (Citation2007) for the complete description of various international payment terms.
4 The description of LC and documentary collection terms is drawn on Hwang and Hyejoon (Citation2013).
5 It includes all the IP addresses assigned to companies, as well as households.
6 We thank a referee for suggesting this idea.
7 Closely related is the extensive margin, where trade may increase when a trading relationship is newly established.
8 See Banerjee et al. (Citation1993), Pesaran and Shin (Citation1997), and Pesaran, Shin, and Smith (Citation2001) for details regarding the recovery process of the long-run coefficients from the estimated ARDL model.
9 crisis1 is set to one for 1997m10–1998m12 and zero otherwise, and crisis2 is set to one for 2007m8–2009m4 and zero otherwise.
10 The sample size is assumed to be 1,000 for Pesaran, Shin, and Smith (Citation2001)’s critical value. On the other hand, for the sample size of 80, Narayan (Citation2005)’s critical value at the 1% significance level is given by 4.59. Therefore, the null hypothesis is rejected for both critical values.
11 One explanation for the less precise estimates is that the decrease in the demand of LC terms during the sample period may be met by the increase in the demand of other payment terms like documentary collection terms rather than cash terms. In 2015, the export share of the remaining terms other than cash and LC terms accounts for approximately a quarter.
12 We find that the adjustment process to the long-term equilibrium occurs very quickly. For instance, the estimate of the error correction coefficient implies that once the exports share on cash terms deviates from the long-term equilibrium, more than 73% of the deviation is corrected within a month.