International Trade, Risk and the Role of Banks* Friederike Niepmann and Tim Schmidt-Eisenlohr† July 26, 2016 Abstract International trade exposes exporters and importers to substantial risks. To mitigate these risks, firms can buy special trade finance products from banks. Based on unique data with global coverage, this paper explores under which conditions and to what extent firms use these products. 15 percent or $2.5 trillion of world exports are settled with letters of credit and documentary collections. Letters of credit are employed the most for exports to countries with intermediate contract enforcement, and they are used for riskier destinations than documentary collections. The 2007/2008 financial crisis affected firms’ payment choices, pushing them to use more letters of credit. These patterns follow naturally from a model of payment contracts in international trade.
Keywords: trade finance, multinational banks, risk, letter of credit JEL-Codes: F21, F23, F34, G21
*
The authors are grateful to JaeBin Ahn, Andrew Bernard, Giancarlo Corsetti, Galina Hale, Charles Kahn, Rod Ludema, Morten Olsen, Philipp Schnabl, Valerie Smeets, Catherine Thomas and an anonymous referees for helpful comments, and also thank workshop participants at the New York FED, NYU and UAB, the 2014 EITI Conference, the Ifo Institute conference on State Export Credit Guarantees in a Globalized World and the 2014 conference of the CEPR working group on the Macroeconomics of Global Interdependence. Special thanks go to Peter Ware and the SWIFT Institute for providing data on SWIFT traffic and for being very responsive and attentive to the authors’ questions. The authors also thank Geoffrey Barnes for excellent research assistance, and Banu Demir and JaeBin Ahn for providing some of the data used in this research. † The authors are staff economists in the Division of International Finance, Board of Governors of the Federal Reserve System, Constitution Avenue NW, Washington, D.C. 20551, USA. Emails:
[email protected] and
[email protected]. The views in this paper are solely the responsibility of the authors and should not be interpreted as reflecting the views of the Board of Governors of the Federal Reserve System or of any other person associated with the Federal Reserve System.
1
Introduction
International trade involves risks. Firms manage some of these risks by choosing among a variety of payment contracts, often employing letters of credit (LCs) and documentary collections (DCs) offered by banks to limit their exposures.1 Exploiting unique data derived from electronic bank communications worldwide, this paper presents first reliable evidence on the prevalence of LCs and DCs in world trade and sheds new light on the factors that determine their use. Contrary to the common view, LCs are most popular in exports to intermediate-risk countries as opposed to high-risk countries, a finding that follows naturally from the key tradeoffs exporters face. Why study firms’ payment choices? Because they are central for international economic activity. Consider that when exporters settle transactions with LCs and DCs, the availability of these bank products and their costs directly impact the profitability of the trade and, therefore, firms’ export behavior. Also, LCs and DCs connect economies. Because banks in the importing and exporting country need to collaborate to provide them, the availability and costs of these instruments depend on both domestic and foreign conditions. The importance of banks’ trade finance services was highlighted by the 2007/2008 financial crisis and policy makers subsequent responses. For example, while many development banks had trade finance programs before the crisis, the support was ramped up substantially after 2008, especially the public confirmation of LCs.2 Moreover, the risk weights on LCs originally proposed by the Basel Committee on Banking Supervision in 2010 were brought down over concerns that the new rules on capital and leverage would make trade finance too costly and harm trade. Despite the relevance of understanding firms’ payment choices for trade costs and the public provision of trade finance, knowledge of the topic is still very limited. How important trade finance instruments are for international trade and when they are used has been difficult to assess due to a lack of data. This paper makes significant progress in answering these questions by exploiting unique information from the Society for Worldwide Interbank Financial Telecommunications (SWIFT), the provider of the single most important communications platform for banks worldwide. SWIFT estimates that about 90 percent of worldwide LC bank-to-bank flows go through its network. Our data set contains the number and value of LC and DC messages sent through SWIFT by location of the advising and 1
In a letter of credit, the importer’s bank guarantees payment to an exporter upon proof that the goods were delivered to the importer. A DC involves ownership documents that are forwarded by the exporter’s bank to the importer’s bank; the importer receives the DC only upon payment. The two payment methods are illustrated in figures 1 and 2. For a more detailed discussion, see section 2. 2 In the midst of the crisis, the G20 committed to extending the public support for trade finance by $250 billion, worried that firms would stop exporting without bank guarantees. The International Finance Organization, which is a part of the World Bank Group, now runs a $5 billion program that mostly confirms LCs; see IFC (2012).
1
the issuing bank and allow us to study these trade finance instruments around the world since 2003. Trade finance products are very relevant for exporting although less prevalent than what previous studies have found.3 By value of exports, LCs were used in 2011 for 13 percent of world exports ($2.3 trillion), while DCs were employed for 1.8 percent ($319 billion). These data points imply that the often-cited 47 percent of bank-intermediated trade found in the IMF-BAFT survey may have substantially overstated the size of trade finance.4 They also differ from the numbers reported in Antr`as and Foley (2015) who study the payment choices of one large U.S. food exporter. These findings highlight the value of the SWIFT data with their broader coverage and representativeness. The use of LCs and DCs varies considerably across importing and exporting countries. While 27.3 percent of exports to low-income countries are settled with LCs, only 4.9 percent of exports to high income (non-OECD) countries rely on this instrument. Similarly, 4.7 percent of exports to the United States employ DCs but this instrument is hardly ever used for shipments to Russia. This paper documents three other key facts. First, many country pairs do not use LCs at all. While 19,563 country pairs in the data traded with each other in 2010, only 7,446 of or 38.1 percent of these exchanged at least one LC message. Zeros are especially high for trading pairs with Sub-Saharan African countries as importers and exporters, indicating that firms in these countries might not have the option to employ LCs due to a lack of supply. Second, the 2007/2008 financial crisis had an impact on how international trade transactions were settled. Around the time of the Lehman bankruptcy in September 2008, the use of both LCs and DCs rose significantly. The increase in LCs was considerably larger than that of DCs, suggesting that exporters prefer LCs in an uncertain global economic environment. Third, the average size of trade transactions differs by type of payment contract. For U.S. exports, LC transactions are by far the largest with an average of $669.7 thousand, followed by DCs ($120.4 thousand). Transactions that do without bank intermediation (cash-in-advance and open account) are on average much smaller ($37.4 thousand). This ordering is consistent with substantial fixed costs in the provision and use of trade finance products. The formal regression analysis explores the role of risk for firms’ payment choices. The 3
In this paper, trade finance refers to services provided by banks that support firms’ international trade transactions. Trade credit refers instead to supplier credit provided by the exporter when sales are settled on an open account. 4 Part of the discrepancy may be due to a wider interpretation of bank-intermediated trade by survey respondents. See IMF-BAFT (2011) and also Asmundson et al. (2011).
2
risk that companies face when trading internationally is related to the degree to which contracts can be enforced abroad. Accordingly, the use of LCs and DCs in international trade is closely linked to the quality of legal institutions in both the exporting and the importing country. We find that (i) LCs are used the most for exports to countries with intermediate contract enforcement. (ii) The use of DCs largely increases with contract enforcement in the destination country. (iii) Compared to DCs, LCs are used for riskier destinations. (iv) The use of LCs and DCs also first rises and then falls with contract enforcement in the exporting country with no significant differences between LCs and DCs. These findings can be rationalized by theory. Consider the first finding. Although the main objective of an LC is to reduce financial risk for the exporter, it tends not to be used for either the least risky or most risky destinations. The basic intuition is that the value of risk mitigation through bank intermediation is offset to a degree by the cost of the intermediation. Because banks can reduce but cannot eliminate the risk of a trade transaction, the fees they charge rise with the remaining risk they take on. For the riskiest destination countries, bank fees are so high that exporters prefer cash-in-advance. In that case, the importer pays before the exporter produces, and payment risk is eliminated. Similarly, LCs are not used for lowrisk destinations; for those transactions, the exporter can save on bank fees by bearing the risk itself. The model in section 6 formalizes these arguments, showing that the presented empirical findings follow from the key tradeoffs that firms face in their choice of payment contracts. Literature Previous empirical work on payment contracts focuses mostly on the trade-off between cash-in-advance and open account.5 A few papers provide evidence on the use of LCs, but none of them has comprehensive data or explicitly analyzes the use of DCs. Antr`as and Foley (2015), who focus on one large U.S. food exporter, show that the firm’s use of LCs decreases with the degree of contract enforcement in the destination country and that the firm is more likely to use an LC when interacting with a new customer. Hoefele et al. (forthcoming) exploit firm-level data from World Bank surveys to study how country characteristics affect firms’ payment choices. They find that product complexity affects the choice between pre- and post-delivery payments. Glady and Potin (2011) employ information from SWIFT for 2006, reporting that the number of LC transactions increases with country risk. Demir and Javorcik (2014) use data from Turkey to investigate the role of industrylevel competition for the optimal payment terms. While the authors observe the share of exports that use LCs, they pool bank-intermediated trade with cash-in-advance transactions 5
For cash-in-advance transactions, the importer pays before the exporter delivers. For open-account transactions, the exporter delivers before the importer pays.
3
for their analysis.6 Ahn (2014) has detailed data on the payment choice of Colombian firms, focusing on the prevalence of open account. Two papers about the effects of shocks to the supply of trade finance and credit also contain information on payment forms. Del Prete and Federico (2012) study the effect of financial shocks on trade based on detailed information on loans and guarantees extended by Italian banks. Using matched importer-exporter data from Colombia, Ahn (2013) investigates whether financial shocks in 2008–2009 affected the use of LCs.7 Several papers propose payment choice models based on the theory in Schmidt-Eisenlohr (2013). Antr`as and Foley (2015) extend the framework to a dynamic setting. Glady and Potin (2011) introduce heterogeneous firms and asymmetric information into the model. Hoefele et al. (forthcoming) derive predictions on the role of industry complexity. Ahn (2014) develops an alternative model that explains the popularity of open account transactions by their self-liquidating and recourse nature.8 None of the aforementioned papers studies DCs or derives predictions that match the empirical patterns documented here. More broadly, the analysis in this paper relates to the large literature on trade credit9 and to the literature on financial conditions and comparative advantage.10 The remainder of the paper is structured as follows. Section 2 provides background information on payment terms in international trade. Section 3 introduces the SWIFT data. Section 4 documents key facts. Section 5 presents the regression analysis with a focus on the role of risk. Section 6 develops a model of payment choice that explains the empirical patterns and reports results from calibration and simulation exercises. Section 7 discusses implications of our analysis.
2
A Brief Primer on Trade Finance
Trade is risky and takes time. For this reason, it matters how the payment of a transaction is organized. The four most important payment forms in international trade are open account, 6
Turkcan and Avsar (2014) employ the same data to test key predictions of the model in SchmidtEisenlohr (2013) on the use of cash-in-advance and open account. 7 For more papers on the effect of financial shocks on trade, see also Amiti and Weinstein (2011), Paravisini et al. (2015) and Niepmann and Schmidt-Eisenlohr (2013b). 8 In an earlier version, Ahn (2011), the focus was on the effect of changes in aggregate default risk on the ratio of exports to domestic sales; in that model, LCs become relatively less attractive in a crisis because risks increase both for importers and banks. 9 Granting trade credit corresponds to settling a transaction with an open account, one of the payment forms we consider. Several papers analyze the substitutability between inter-firm finance and bank finance, although none of them exploit data on bank guarantees as we do here. Among the theoretical contributions are Ferris (1981), Petersen and Rajan (1997), Biais and Gollier (1997), Wilner (2000), Burkart and Ellingsen (2004) and Cunat (2007). Empirical studies include Ng et al. (1999), Love et al. (2007), Giannetti et al. (2011) and Klapper et al. (2012). See also Eck et al. (2014). 10 See, in particular, Beck (2003) and Manova (2013) on how financial development affects trade patterns.
4
cash-in-advance, LCs and DCs. Each type differs in the allocation of risk, the financial cost and the extent of bank involvement. Table 1 summarizes key differences between the four choices. Table 1: Tradeoffs between different payment forms
exporter importer
risk ++ -
OA financial costs + -
risk ++
CIA financial costs +
risk + -
DC financial costs ++ -
risk -
LC financial costs + +
Note: The table illustrates how the four payment contracts (open account, cash-in-advance, letters of credit, documentary collections) differ in terms of their implied allocations of risk and financial costs. The notation has the following interpretation. Risk: ++ all risk, + some risk, - no risk. Financial costs: ++ interest expenses and fees, + interest expenses or fees, - zero costs.
Cash-in-Advance Under cash-in-advance (CIA), the importer pays first for the good and the exporter produces the good after receiving payment. Since sales revenues only realize later, the importer has to pre-finance the payment to the exporter. At the same time, the importer faces the risk that the exporter does not deliver or supplies products of low quality. The importer, thus, bears both the financial cost and the risk of the transaction. Open account
With an open account (OA), the exporter delivers the good and the
importer pays upon receipt. Now the exporter bears the financial cost and the risk of the transaction. On the one hand, the exporter has to pre-finance working capital since it produces before being paid. On the other hand, the importer may delay payment or may not pay at all. Exporters can cover all or some of this risk through trade credit insurance. In addition, they may sell some or all of their claims to factoring companies to raise working capital. Banks can be involved in both cash-in-advance and open account transactions. They typically help with financing the firms’ pre-payment or working capital and may assist exporters in finding insurance providers or factors. Letter of credit LCs are the most common trade finance product provided by banks. The importer initiates the LC transaction (figure 1) by having its bank issue the instrument to the exporter. The LC guarantees that the issuing bank will pay the agreed contract amount when the exporter proves that it delivered the goods, for example, by providing shipping documents confirming the arrival of the goods in the destination country. To cover the risk that the issuing bank will not pay, an exporter may have a bank in its own country confirm the LC, in which case the confirming bank agrees to pay the exporter if the issuing bank defaults. Documentary collection
Another way for banks to assist firms in international trade,
although less common, are DCs. In contrast to an LC, a DC does not involve payment guarantees. Instead, the exporter’s bank forwards shipping/ownership documents to the 5
Figure 1: How a letter of credit works
Contract
Execution
1. Contract
5. Shipment
6. Submit documents.
2. Apply for letter of credit.
4. Authenticate letter of credit.
9. Payment
10. Payment 11. Release documents.
7. Send documents.
8. Payment
3. Send letter of credit .
importer’s bank; the documents, which transfer the legal ownership of the traded goods to the importer, are handed to the importer only upon payment for the goods (figure 2). Figure 2: How a documentary collection works
2. Export goods
Exporter
Importer 1. Sales Contract
3. Submit Documents
5. Payment
8. Payment
Advising Bank (Exporter’s)
7. Payment
6. Deliver Documents
Issuing Bank (Importer’s)
4. Send Documents
A DC provides less security to the exporter than an LC. In the latter case, an exporter is paid by the issuing bank/confirming bank upon proof of delivery regardless of whether the importer paid. With a DC, the exporter is remunerated only if the bank receives the payment from the importer. As mentioned, trade credit insurance is another way for exporters to handle risk in international trade. An important difference is that credit insurance shifts the risk from the exporter to the insurer, whereas LCs and DCs reduce the risk inherent in the transaction since they increase the importer’s incentives to pay. LCs and DCs are shortterm trade finance instruments. According to the International Chamber of Commerce Trade Register, the average maturity of a confirmed LC is 70 days, while the average maturity of 6
an importer LC is about 80 days (see ICC (2013)).
3
A New Data Set from SWIFT
3.1
Other data sources
Data on trade finance are scarce, especially those with world-wide coverage, and are presented only in a few papers. International organizations were first to collect information through surveys.11 In the IMF-BAFT survey, for example, bankers were asked to guess how world trade was financed (IMF-BAFT (2011)). In 2014, the Committee on the Global Financial System under the umbrella of the Bank for International Settlements (BIS) gathered various trade finance data collected by BIS member countries to provide cross-country estimates (see BIS (2014)). Antr`as and Foley (2015) explore data from one large U.S. food exporter who disclosed how it settles its international shipments. More recently, comprehensive customs data for individual countries has become available. Ahn (2014) documents the use of trade finance in Columbian and Chilean imports, while Demir et al. (2014) study Turkish exports. Since it is hard to know how representative and accurate surveys are, it remains an open question how much of trade is intermediated by banks worldwide, with figures differing substantially across sources. There is also only limited knowledge of cross-country differences in the use of trade finance instruments and of the factors that drive them. This paper contributes to filling these gaps by exploiting a new data set derived from SWIFT messages sent by banks around the world.
3.2
The SWIFT data
The Society for Worldwide Interbank Financial Telecommunications (SWIFT) network provides a communications platform through which financial institutions exchange standardized financial messages. SWIFT reports that more than 10,500 corporations and financial institutions in 215 countries use this platform and estimates that about 90 percent of worldwide LC bank-to-bank flows go through the SWIFT network. While there is no data that allows us to directly verify this global number, we show below that for those countries where detailed data is available, SWIFT messages basically capture all trade paid for with LCs. In addition, there is ample anecdotal evidence that there is no substitute for the SWIFT system.12 11
See, for example, ICC (2009) and IMF (2009). For example in the context of recent discussions by Russia about setting up a BRICS alternative to SWIFT, in mid-2015, the Deputy Chairman of the Central Bank of the Russian Federation stated: ’Seriously speaking, there is no (alternative) to SWIFT in the world right now.’ See Forbes (2015) for details. 12
7
According to the SWIFT Institute, even messages between different parts of multinational banks are often on SWIFT.13 This allows banks to use SWIFT’s traffic analysis tools and the marginal costs of sending messages are relatively low. When a bank in the importer’s country issues an LC, it sends a so-called MT700 message to the advising bank in the exporter’s country; when it releases a payment related to a DC, it sends an MT400 message. The message specifies the terms of the LC or the payment, including the names of the trading parties, the banks involved and the goods traded. It can happen that, at some point, a bank wants to cancel or amend the terms of an LC or a DC. For this purpose, it can send message types MT707 and MT407. According to the SWIFT Institute, this is a relatively rare event. Banks also issue so-called standby LCs using message type MT760. Standby LCs provide a general backstop against non-performing counterparties and work more like guarantees or insurance. They seem to be used more for domestic transactions and services as well as long-term international investments than for standard international trade. Unfortunately, SWIFT does not disclose information on any of these additional message types. Due to confidentiality, the SWIFT Institute provides the messages data aggregated up to the country-pair level. Our data set contains the number of MT700 and MT400 by sender and recipient country at a monthly frequency from 2003 to 2012. From the fourth quarter of 2010 onward, we also know the total value of the messages in U.S. dollars. Because the value data is available for only a relatively short period, and message counts and amounts are highly correlated, we mainly use the count data for the analysis. Unfortunately, we do not observe messages by industry. Since there is evidence that the use of trade finance instruments differs across industries (see, for example, Hoefele et al. (forthcoming)), some of the differences in their use across countries could be driven by the industry composition of trade. Due to the nature of the data, we cannot investigate this issue and have to accept this limitation.
3.3
Accuracy of the SWIFT data
The reader might wonder how accurate the SWIFT data are. Fortunately, we have a way to check this. The three data sets derived from customs records presented in Ahn (2014) and Demir et al. (2014) deliver arguably the most reliable and representative numbers since exporters and importers have to declare at the transaction level which payment form they employ. For comparison, we therefore calculate the same moments as reported in the aforementioned studies for Chile, Columbia and Turkey based on the SWIFT data. Results are 13
The SWIFT Institute is an entity set up by SWIFT that, among others, funds research and provides access to data derived from the SWIFT messaging system.
8
shown in table 2. For all three countries, the SWIFT data deliver numbers almost identical to the customs data. So the SWIFT data seem to provide an accurate measure of aggregate trade finance activity. Table 2: Data on LC and DC use Country World US2)
Colombia Chile Turkey
Ex/Im LC DC Source Survey data / Individual firm data 471) IMF-BAFT (2011) 13.0 1.8 SWIFT 5.5 10.7 Antr`as and Foley (2015) Ex 8.5 1.1 SWIFT Customs data 5 Ahn (2014) Im 5.0 0.01 SWIFT 10 Ahn (2014) Im 10.1 0.3 SWIFT 15 Demir et al. (2014) Ex 15.7 2.7 SWIFT
Note: The table summarizes results from previous studies and surveys on the shares of LCs and DCs in international trade and compares them to values calculated from SWIFT data. All numbers indicate the percentage share in the value of total trade (either from customs data, individual firm data or from guesses of survey participants). The Ex / Im column specifies if a data point refers to exports or imports. All numbers are for 2011 except those from Antr` as and Foley (2015), who report averages from 1996 to 2009. 1) The IMF-BAFT survey does not ask about LCs or DCs but asks about bank-intermediated trade. Respondents may subsume additional payment forms under this category (IMF-BAFT (2011)). 2) Antr` as and Foley (2015) use information from one U.S. food exporter while SWIFT covers total U.S. exports.
Since we are also interested in the country-pair dimension, we go one step further and compare numbers also at the bilateral level. To this end, we regress SWIFT LC shares on the LC shares from customs data. We do not allow for a constant in these regressions. Columns (1), (3) and (5) report results for all country pairs for which information is available both in the customs data and in the SWIFT data.14 Table 3: SWIFT data and Customs data
LC share Customs Observations R-squared
Chile (1) (2) 0.600*** 0.755*** (0.0601) (0.0902) 43 15 0.703 0.833
Colombia (3) (4) 0.538*** 0.905*** (0.111) (0.184) 26 15 0.486 0.634
Turkey (5) 0.715*** (0.0536) 69 0.724
Note: The table shows how the SWIFT data compares with customs data from Columbia, Chile and Turkey. First, we compute the share of import transactions that use LCs by source country for Chile, Columbia and Turkey based on the SWIFT data. Then, we obtain the equivalent shares from country studies that use customs data. Finally, we regress the SWIFT shares on the shares obtained from customs data for each importing country separately. Columns (1) and (3) present the results based on the full matched samples. Columns (2) and (4) report results on imports from the 15 most important source countries in each case. Column (5) reports results for the full sample excluding Iran. Standard errors are in parentheses. *, ** and *** denote significance at the 10%, 5% and 1% level.
There are several reasons why the estimated coefficients are not equal to one. First, 14
We exclude Iran in the regression for Turkey because sanctions forced SWIFT to disconnect Iran from its network in April 2012.
9
the customs data likely suffer from some form of measurement error.15 Columns (2) and (4) of table 3 support this. In these columns, regressions are run for the top 15 source countries of Chile and Colombia, respectively. The fit improves when only large suppliers are included and the coefficients get closer to 1. Another reason for a discrepancy is that sometimes banks involved in the transaction are located in a third country, i.e. in a country different from the source or destination country of the shipped goods. In this case, SWIFT message flows at the bilateral level might not perfectly match up with the respective bilateral trade flows. These third countries appear to be primarily tax havens and offshore financial centers. The LC flows observed from and to these countries are large relative to the trade flows. To account for this issue, we exclude offshore financial centers and tax havens from the regression analysis that uses bilateral message flows. All results are robust to the inclusion of these countries.16 A third reason for a coefficient other than 1 is that trade occurs in different currencies and that the method to convert trade flows into U.S. dollar values may differ across data sets. To summarize, while the match is not perfect, the results show that SWIFT data is highly informative about the use of trade finance instruments not only in the aggregate but also at the bilateral level.
4 4.1
The Facts The prevalence of LCs and DCs in world trade
With the external validity of the SWIFT data established, this section provides new facts on the use of payment forms in international trade. In particular, we provide the first reliable estimates on the prevalence of LCs and DCs in world trade. The statistics reported below are based on observed trade values as well as LC and DC message values in 2011 if not mentioned otherwise. The underlying trade data are from the IMF Directions of Trade Statistics. LCs and DCs in world trade According to the SWIFT data, LCs cover about 13 percent or $2.27 trillion of world trade. DCs are employed less frequently, for only 1.8 percent or $319 billion of world exports and imports. These numbers are far off those resulting from the IMF-BAFT survey, which suggested that 47 percent of world trade are bank-intermediated 15
For example, in the Columbian data, firms choose between 11 different payment types. Ahn (2014) drops 4 of them when computing his statistics. 16 A list of tax havens and offshore financial centers can be found in the data appendix. 24.7% (36.2%) of LC flows in terms of value have an offshore center as importer (exporter). 55.6% of LC flows have an offshore center either as exporter or importer. If Switzerland, Hongkong and Singapore are not designated as offshore centers, these shares fall to 5.4%, 7.6% and 12.9%, respectively. Results based on a sample with tax havens and offshore financial centers are reported in tables B.12 and B.13 in the online appendix.
10
(IMF-BAFT (2011)). Part of this discrepancy may be driven by the fuzzy definition of trade finance in the survey. Survey respondents might have included other trade finance products but LCs and DCs in their estimates. In any case, this often-cited number has led to an overstatement of the importance of LCs and DCs in world trade. Figure 3 shows the use of LCs and DCs by the top 10 world importers and exporters in 2011, revealing substantial heterogeneity across countries. For example, 36.4 percent of Chinese imports are based on LCs while only 4.9 percent of Japanese imports rely on this instrument; DCs are employed for 4.7 percent of Korean but only for 0.004 percent of Russian exports.17 Figure 3: The use of LCs and DCs by the top 10 world exporters and importers
Note: The chart has the top 10 world importers on the left and the top 10 world exporters on the right. For each of these countries, it displays the share of international shipments that is settled with LCs or DCs in 2011, based on SWIFT message values and export values in that year.
A closer look at the figures for the United States uncovers some key differences to Antr`as and Foley (2015). LCs are used more frequently and DCs less frequently than what the data from the food producer studied in the aforementioned paper would suggest. The firm relies on LCs for 5.5 percent and on DCs for 11 percent of its shipments. In contrast, the SWIFT data show that 8.5 percent of U.S. exports are settled with LCs but only 1.1 percent with DCs. Our data set also provides evidence on the use of LCs and DCs in U.S. imports, information that was not available previously. The share of LCs in U.S. imports is 1.9 percent. The share of DCs is 5.2 percent. LCs and DCs by income regions
Many institutions provide trade finance support to
low income countries, based on the view that trade finance is particularly useful for these countries but often difficult to obtain. To shed some light on this issue, we consider the use of LCs and DCs by region. 17
In the online appendix, we provide a full list of countries and their LC and DC intensities in 2013.
11
We split countries into five income regions following the World Bank classification and create a separate group for offshore financial centers and tax havens. Figure 4 displays the use of LCs and DCs in imports on the left and exports on the right by these six groups of countries. Not considering tax havens and offshore financial centers, LCs are most prevalent in imports to low-income countries (27.3 percent), while they are used only rarely for imports to high-income countries (4.9 percent). DCs are hardly ever employed in imports to lowincome countries (0.1 percent), while they are most common in exports by lower-middle income countries (around 2.9 percent). As mentioned earlier, LC and DC traffic in offshore financial centers and tax havens appears disproportionate relative to these countries’ total exports and imports, probably stemming from the fact that banks in these countries provide trade finance services for parties outside of the country. Figure 4: The use of LCs and DCs by income region
Note: The chart shows the share of international trade transactions that use LCs and DCs by importer and exporter region in 2011. Countries are grouped according to their income and status as offshore financial center or tax haven. The six groups are: HIO - high-income OECD countries; HINO - high-income non-OECD countries; UMI - upper middle income countries; LMI - lower middle income countries; LI - lower income countries; OC/TH: tax havens or offshore financial centers.
As work by Helpman et al. (2008) shows, there is a large number of country pairs in the world that exhibit zero trade flows. Similarly, there is a large number of country pairs between which no LC and DC messages are sent. In 2010, there were 19,563 country pairs with positive trade flows, of which 7,446 (5,191) pairs exchanged LC (DC) messages.18 This means that only 38.1 (26.5) percent of country-pairs with positive trade flows used LCs (DCs) in that year. At the same time, however, country pairs with positive LC (DC) message flows accounted for 96.7 (90.7) percent of total trade in the data. Figure 5 illustrates that there is substantial variation in the number of unidirectional country pairs that trade with each other but have zero LC traffic. For example, the first bar 18
Country pairs are defined unidirectionally, so exports from the United States to China and exports from China to the United States are counted separately.
12
in the chart displays the fraction of country pairs with South Asian countries as importers for which SWIFT recorded at least one LC message in 2010. While roughly 60 percent of country pairs with South Asian countries as importers or exporters send or receive LC messages, more than 70 percent of country pairs with Sub-Saharan African countries as importers or exporters have zero LC traffic. Figure 5: Positive SWIFT flows by region
Note: The chart shows the fraction of unidirectional country pairs for which we observe positive LC flows by geographic region in 2010 conditional on positive trade flows. For example, the first (second) bar shows the share of country pairs with South Asian countries as importers (exporters) that had positive LC traffic in that year. The acronyms stand for: SA - South Asia; MENA - Middle East and North Africa; HINO - high-income non-OECD; high-income OECD; EAP - East Asia and Pacific; ECA - Eastern Europe and Central Asia; LAC - Latin America and Caribbean; SSA - Sub-Saharan Africa.
This finding supports the view of international institutions like the World Bank and the Asian Development Bank that for many country pairs trade finance products are not available, probably due to the presence of substantial fixed costs of offering trade finance services for banks. The provision of LCs and DCs requires a well-functioning banking network between the source and the destination country, which takes time and is costly to establish. To the extent that access to trade finance is beneficial for countries, international development banks might have a role in supporting the formation of additional bank links.
4.2
Changes in the use of LCs and DCs over time
Figure 6 shows the use of LCs and DCs in world trade over time.19 Chart (a) on the left plots the number of LC and DC messages relative to real world exports from 2003 to 2012. There is a secular decline in the use of both trade finance instruments over this period. The vertical dotted lines in the charts indicate 2008q3, the quarter when Lehman collapsed and the financial crisis deepened. Both LC and DC messages saw a strong pick-up at that 19
We thank an anonymous referee for the suggestion to explore changes in the use of LCs during the financial crisis.
13
time. The increase in LCs was much larger compared to DCs. Chart (b) on the right illustrates this clearly. It plots the moving average of the share of LC messages in total trade finance messages. Until 2008q3, the fraction of LCs was relatively stable at around 52 percent. From 2008q3 onward, it moved quickly to reach a share of 61 percent. While some of the rise might be explained by the policy initiatives of the G20, the pattern still suggests that firms shifted towards LCs to limit their risk exposures during the crisis. The finding contrasts with survey evidence reported in Malouche (2009) who concludes that DC and LC shares did not increase in the wake of the Lehman collapse. Figure 6: The use of LCs and DCs over time (a) LCs and DCs / real exports
2003q1
2005q3
2008q1 Date LCs
2010q3
2013q1
.5
.2
.52
.4
.6
LCs / (LCs + DCs) .54 .56
.8
.58
1
.6
(b) LCs / (LCs + DCs)
2003q1
DCs
2005q3
2008q1 Date
2010q3
2013q1
Note: The figure shows the use of LCs and DCs over time. Panel (a) depicts the ratio of LCs and DCs over real exports (nominal exports deflated with the U.S. GDP deflator). The ratio is indexed to 1 in 2003q1. Panel (b) shows the ratio of LC messages over the sum of LC and DC messages. To filter out short-term fluctuations, we smooth the ratio in panel (b), showing a 5-quarter moving average with weights 1/2/3/2/1. The vertical line in both panels indicates 2008q3.
4.3
Average transaction sizes
Importers and exporters have to pay a fee for LCs and DCs. Because a part of the fee is fixed, covering document handling, screening and monitoring costs, the attractiveness of these payment forms should increase with the size of an export transaction. While we do not observe the transaction value of single LCs and DCs, we can compute the average size of LC and DC transactions based on the total value of LCs and DCs and the corresponding number of SWIFT messages by importing and exporting country. To compare the average size of these transactions to the average trade transaction, additional information is needed. While not easily available for all countries, we obtain data on the number of shipments (so-called cards) for U.S. trade from the United States Census Bureau. In 2012, the average value of a U.S. export transaction that employed an LC was $669.7
14
thousand.20 The average value of those using DCs was $120.4 thousand. The average value of a U.S. export transaction –regardless of the underlying payment form– was $40.3 thousand. Thus, an average LC transaction is more than sixteen times as large as an average trade transaction, while an average DC transaction is roughly three times as large. The ranking survives even when average transaction sizes are analyzed for each importing country separately.21
5
Regression Analysis
As discussed earlier, LCs and DCs reduce the risk of transactions for exporters and importers. In this section, we explore the role of both exporting and importing country risk for the use of LCs and DCs formally, based on regression analysis. The risk that firms face when trading internationally decreases with the quality of legal institutions in the trading partner’s country so we proxy the inverse of country risk by contract enforcement. The analysis delivers two main results. First, the use of LCs and DCs is hump-shaped in contract enforcement both in the importing and the exporting country. Second, LCs are used for riskier destinations than DCs. These patterns can be explained by payment choice theory as we show in section 6.
5.1
The role of importer and exporter risk
Importer risk
We start by studying the link between the use of LCs and DCs and
importing country risk, estimating the following equation: log(𝑌𝑖𝑗𝑡 ) = 𝛽1 log(exp𝑖𝑗𝑡 ) + 𝛽2 log(distance𝑖𝑗 ) + 𝛽3 law𝑖𝑡 + 𝛽4 (law𝑖𝑡 )2 + 𝛽5 log(GDP per capita𝑖𝑡 ) + 𝛽6 (GDP per capita𝑖𝑡 )2 + 𝛽7 log(fin. developm.𝑖𝑡 ) + 𝛽8 LC requirement dummy𝑖𝑡 + 𝛽9 bilateral controls𝑖𝑗 + 𝛼𝑗𝑡 + 𝜖𝑖𝑗𝑡 .
(1)
𝑌𝑖𝑗𝑡 stands for the number either of LC or of DC messages sent to banks from importing (destination) country 𝑖 to exporting (source) country 𝑗. The dependent variable is regressed on the log of exports, on the log of distance, and on contract enforcement in country 𝑖, the destination of the shipped goods. We use as a proxy for contract enforcement the rule 20
To compare transaction sizes across payment types, we restrict the sample to countries for which each payment type is observed and exclude tax havens and offshore financial centers. This reduces the number of countries to 60. Results are very similar when we calculate the numbers for the unbalanced data set and include offshore financial centers and tax havens. 21 For 95 percent of U.S. export destinations, LCs have a higher average value than DCs. For 100 (80) percent of the countries, LCs (DCs) have a higher average value than transactions that use neither LCs nor DCs.
15
of law index from the World Bank’s World Governance Indicators (law𝑖𝑡 ). Since the index takes negative values for some countries, it is mapped on the interval between 0 and 1, where 1 corresponds to the highest degree of the rule of law observed. The regression also includes rule of law squared to allow for a non-linear relationship.22 It further controls for the destination country’s stage of financial development. In line with the extensive literature on the topic, we use as a measure of financial development the variable private credit by financial institutions over GDP from the World Bank’s Financial Structure Database (see Beck (2003) and Manova (2013)). Because some countries require firms to use LCs when trading internationally, the equation includes a dummy variable that takes the value of 1 when such a requirement is in place in country 𝑖 at time 𝑡.23 Log of GDP per capita and log of GDP per capita squared are also added as regressors, to ensure that the effect of contract enforcement is not due to omitted factors correlated with a country’s overall development. Moreover, all regressions control for a set of country-pair specific variables (contiguity, common colonizer, common language, time difference) and exporter-time fixed effects. Standard errors are clustered by importing country. Note that we do not include destination country fixed effects, so identification comes primarily from cross-country variation in contract enforcement. This is the correct specification for testing the aggregate non-linear relationship that we are interested in.24 Table 4 presents evidence on the relationship between SWIFT traffic and the degree of contract enforcement in the importing country. Column (1) shows the results for LCs. Column (2) is for DCs. In both columns, the coefficient of the linear rule of law term is positive and significant, while the coefficient of the squared term is negative and significant. Thus, the relationship between a destination’s rule of law and the use of the respective trade finance instrument in world trade is non-linear. The coefficients of financial development and distance are also highly significant and positive, indicating that LCs and DCs are used more for exports to more financially developed countries as well as for long-distance trade. To explore differences in the relationship between the use of DCs and LCs and a country’s 22
In the WP version Niepmann and Schmidt-Eisenlohr (2013a), we show that semi-parametric estimation that allows for an entirely flexible form delivers very similar results. 23 Information on countries’ documentation requirements is from the IMF’s AREAER Database. 24 Including destination country fixed effects would instead deliver an estimate of a local non-linear effect, that is, the model would be misspecified for estimating the relationship we want to study. As McIntosh and Schlenker (2006) put it: ’The marginal effects of a relationship which is globally quadratic in X must depend on the non-demeaned value of X, and so inherently cannot be identified in deviations from a group mean.’ See their paper for technical details.
16
Table 4: LCs, DCs and Destination Country Rule of law: intensive margin
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it
(1) log(# LCijt ) 0.560*** (0.0267) 0.306*** (0.0974) 7.299*** (2.145) -6.592*** (1.595)
(2) log(# DCijt ) 0.610*** (0.0368) 0.429*** (0.0976) 8.439*** (3.177) -4.917** (2.314)
-1.068* (0.577) 0.0544 (0.0357) 0.624*** (0.134) 0.553*** (0.209) 0.0984 (0.109) 0.164 (0.148) 0.479*** (0.119) 0.0905 (0.0626) -0.0604*** (0.0203) 49,404 0.626
-0.531 (0.666) 0.0350 (0.0419) 0.383*** (0.124) 0.296 (0.194) 0.571*** (0.176) -0.00874 (0.174) 0.690*** (0.115) 0.238*** (0.0790) -0.0653*** (0.0226) 35,437 0.599
DC dummymijt rule of lawit * DC dummy typemijt rule of law2 it * DC dummy typemijt log(GDP per capit ) log(GDP per capit )2 fin. developmentit LC requirementit contiguityij common colonizerij common languageij common legal originij time differenceij Observations R-squared
(3) log(# messagesmijt ) 0.567*** (0.0228) 0.345*** (0.0789) 7.259*** (2.135) -6.778*** (1.448) -1.985* (1.146) 1.389 (3.537) 1.981 (2.564) -0.963** (0.424) 0.0532** (0.0257) 0.525*** (0.104) 0.456*** (0.171) 0.300*** (0.103) 0.0933 (0.125) 0.584*** (0.0890) 0.159*** (0.0488) -0.0609*** (0.0144) 84,841 0.593
Note: This table analyzes the relationship between the use of letters of credit and documentary collections and destination country rule of law at the intensive margin. log (# LC𝑖𝑗𝑡 ) and log (# DC𝑖𝑗𝑡 ) are the logs of the number of MT700 and MT400 messages sent from country 𝑗 to country 𝑖, respectively, in year 𝑡. log (# SWIFT𝑚𝑖𝑗𝑡 ) is the log of the number of SWIFT messages of type 𝑚 ∈ {𝐿𝐶, 𝐷𝐶}. DC dummy𝑚𝑖𝑗𝑡 takes a value of 1 if the message refers to a DC and a value of 0 if the message is related to an LC. All regression include exporter × year fixed effects. Standard errors are in parentheses and are clustered by importing country. *, ** and *** denote significance at the 10%, 5% and 1% level.
rule of law, we estimate the following pooled regression: log(𝑌𝑚𝑖𝑗𝑡 ) = 𝛽1 log(exp𝑖𝑗𝑡 ) + 𝛽2 log(distance𝑖𝑗 ) + 𝛽3 law𝑖𝑡 + 𝛽4 (law𝑖𝑡 )2 + 𝛽5 DC message dummy𝑚𝑖𝑗𝑡 + 𝛽6 law𝑖𝑡 × DC message dummy𝑚𝑖𝑗𝑡 + 𝛽7 log(𝑋𝑖(𝑗)𝑡 ) + 𝛼𝑗𝑡 + 𝜖𝑖𝑗𝑡 .
17
(2)
Now log(𝑌𝑚𝑖𝑗𝑡 ) stands for the number of messages of type 𝑚 sent from importing country 𝑖 to exporting country 𝑗 in year 𝑡. The regression includes a dummy that takes the value of 1 if the message refers to a DC and zero if the message is related to an LC, as well as interactions between the dummy and the rule of law variables. 𝑋𝑖(𝑗)𝑡 collects the control variables specified earlier. Column (3) of table 4 shows that the coefficient of the linear term differ significantly for DCs and LCs, being larger for DCs.25 Panel (a) of figure 7 illustrates what this means. It plots the relationship between the use of LCs and DCs, respectively, and the rule of law index based on the results in column (3). LCs are employed the most for exports to countries with intermediate contract enforcement. Their use peaks at a rule of law index of 0.52, equivalent to Panama’s in 2012. In contrast, the use of DCs largely increases with the importer’s rule of law. The relationship is concave, implying that an improvement in the degree to which a contract can be enforced in the destination has a larger positive effect on the use of DCs when contract enforcement is low. The share of exporters that use DCs is highest for a rule of law index of 0.86, which corresponds to Chile’s in 2012. Thus, the use of DCs in exports reaches its maximum at higher contract enforcement than the use of LCs. Figure 7: Estimated relationship between the use of LCs and DCs and the rule of law (a) Importer rule of law
-1
1.5
-.5
2
log(messages) 0 .5
log(messages) 2.5 3
1
3.5
1.5
(b) Exporter rule of law
.2
.4
.6 importer rule of law
DC messages
.8
1
.2
LC messages
.4
.6 exporter rule of law
DC messages
.8
1
LC messages
Note: The figure illustrates how the use of letters of credit and documentary collections varies with the rule of law of the importer on the left and of the exporter on the right, based on the estimated coefficients displayed in column (3) of tables 4 and 5, respectively. The solid lines show the log number of LC messages. The dashed lines depict the log number of DC messages.
To see that rule of law is an economically important determinant of firms’ payment choices, consider the following example based on the estimated coefficients in column (3). If Guatemala (normalized rule of law of 0.34) had the same rule of law as Brazil (normalized rule of law of 0.55 in 2012), the number of LC messages sent from Guatemala would increase by 19.8 percent. In contrast, if Brazil had the same rule of law as Israel (normalized rule of 25
When an additional interaction between rule of law squared and the message type is included, we find that the coefficient on the quadratic term is the same for both message types.
18
law of 0.77 in 2012), LC messages sent from Brazil would decline by 36.3 percent. The moves from Guatemala to Brazil and from Brazil to Israel correspond to roughly one standard deviation of the rule of law index. Exporter risk
We conduct a parallel regression analysis to study the role of contract
enforcement in the exporting country. Regressions now include exporter variables and importer-year-fixed effects. Standard errors are clustered by exporting country. Table 5 shows the results, which resemble those for destination country risk. SWIFT traffic also displays a non-linear relationship with source country rule of law. Moreover, greater financial development of the exporting country promotes the use of trade finance instruments. Unlike with destination country rule of law, source country rule of law does not appear to have a differential effect on the use of LCs versus DCs. This is indicated by the insignificant coefficient of the interaction term between the message type dummy and rule of law in column (3). The result is further illustrated in panel (b) of figure 7, which plots the estimated relationship between LC/DC traffic and rule of law in the exporting country. Table 5: LCs, DCs and Source Country Rule of law: intensive margin
log(exportsijt ) log(distanceij ) rule of lawjt rule of law2 jt
(1) log(# LCijt ) 0.711*** (0.0329) 0.385*** (0.0847) 9.395*** (2.463) -6.336*** (1.848)
(2) log(# DCijt ) 0.633*** (0.0352) 0.417*** (0.106) 9.114*** (2.838) -5.232** (2.063)
50,181 0.624
35,615 0.593
DC dummymijt rule of lawjt * DC dummy typemijt rule of law2 jt * DC dummy typemijt Observations R-squared
(3) log(# messagesmijt ) 0.643*** (0.0309) 0.358*** (0.0835) 10.47*** (2.578) -6.690*** (1.946) 1.223* (0.653) -3.600* (2.073) 2.336 (1.535) 85,796 0.540
Note: This table analyzes the relationship between the use of letters of credit and documentary collections and source country rule of law at the intensive margin. log (# LC𝑖𝑗𝑡 ) and log (# DC𝑖𝑗𝑡 ) are the logs of the number of MT700 and MT400 messages sent from country 𝑗 to country 𝑖, respectively, in year 𝑡. log (# SWIFT𝑚𝑖𝑗𝑡 ) is the log of the number of SWIFT messages of type 𝑚 ∈ {𝐿𝐶, 𝐷𝐶}. DC dummy𝑚𝑖𝑗𝑡 takes a value of 1 if the message refers to a DC and a value of 0 if the message is related to an LC. All regressions control for the log of GDP per capita, the log of GDP per capita squared, LC requirements and the log of financial development in the source country, as well as bilateral controls. All regression include importer × year fixed effects. Standard errors are in parentheses and are clustered by exporting country in columns. *, ** and *** denote significance at the 10%, 5% and 1% level.
The share of LCs in total SWIFT traffic
To test more directly whether LCs are used
more for exports to riskier destinations than DCs, we compute the share of LC messages in total SWIFT messages (LC+DC) to destination 𝑐 in year 𝑡 and regress this variable on the 19
importer’s rule of law index and other controls. Regression results are presented in table 6. Column (1) includes exporter-year-fixed effects only. In column (2), the importer’s rule of law is added as an explanatory variable. Its addition increases the 𝑅2 of the regression by 11 percentage points. At the same time, the associated coefficient is highly significant and negative, indicating that the share of LCs in total SWIFT traffic decreases as the ability to enforce a contract in the importing country increases. Column (3) shows that this result is robust to the inclusion of a battery of country variables, which together add only little explanatory power.26 Table 6: Share of letters of credit in bank-intermediated trade (1)
(2)
log(exportsijt ) log(distanceij ) rule of lawit
-0.357*** (0.0356)
(3)
(4) LC shareijt -0.00689* (0.00361) -0.0228** (0.0103) -0.308*** (0.0966)
rule of lawjt Exporter × Year FE Importer × Year FE Observations R-squared
Yes No 30,439 0.151
Yes No 30,439 0.264
Yes No 30,439 0.277
No Yes 30,659 0.315
(5)
(6) 0.00530*** (0.00192) -0.0211*** (0.00615)
-0.0843*** (0.0231) No Yes 30,659 0.322
-0.128** (0.0503) No Yes 30,659 0.335
Note: This table shows how the importance of letters of credit relative to documentary collections changes with rule of law. The dependent variable LC shareijt is the share of letter of credit messages (MT700) in total trade finance messages sent (MT700+MT400) in year 𝑡. Columns (1) to (3) explore the role of importing country characteristics. Columns (4) to (6) study the role of exporting country characteristics. Columns (3) and (6) control for the log of GDP per capita, the log of GDP per capita squared, LC requirements and the log of financial development in the destination and source country, respectively as well as bilateral controls. None of the coefficients of these controls are significantly different from zero. Standard errors are in parentheses. They are clustered by importing country in columns (1) to (3) and by exporting country in columns (4) to (6). *, ** and *** denote significance at the 10%, 5% and 1% level.
Columns (4) to (6) repeat the exercise for source country risk. A stronger rule of law in the exporting country also leads to a smaller fraction of LC messages in total trade finance messages. Note, however, that the effect of higher source country risk is considerably smaller than that of destination country risk. An increase in the importer’s rule of law index of 0.1 decreases the share of LCs in total traffic by around 30 percentage points. In contrast, the same increase in the exporter’s rule of law lowers the share by only 15 percentage points. Furthermore, the inclusion of source country risk only increases the 𝑅2 of the regression by 0.7 percentage point and so explains a much smaller fraction of the variation in LC shares 26
Variables included in the regression but not displayed in columns (3) and (6) of table 6 are: GDP per capita, GDP per capita squared, financial development, a dummy for a documentation requirement, the number of time zones between country pairs, dummies for contiguity, a common official language and a common colonizer. Note also that when rule of law squared is added as a regressor, its coefficient is insignificant, which suggests that the share of LC messages indeed decreases linearly with the rule of law index.
20
across countries than destination country risk. Extensive margin
Source and destination country rule of law also explain the extensive
margin of firms’ use of LCs and DCs in international trade. Table 7 presents the results from logit regressions, in which the dependent variable takes the value 1 if there was LC or DC traffic between a country pair in year 𝑡 and zero otherwise.27 Controls and clustering are as in the baseline regressions. The probability that a country pair uses LCs and DCs is also hump-shaped in the exporter’s and importer’s rule of law and increases in the financial development of both countries. Table 7: LCs, DCs and Rule of law: extensive margin
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it fin. developmentit
(1) 𝐿𝐶 𝐷𝑖𝑗𝑡 0.635*** (0.0282) -0.174 (0.123) 7.484*** (2.080) -6.068*** (1.630) 0.896*** (0.150)
(2) 𝐷𝐶 𝐷𝑖𝑗𝑡 0.650*** (0.0339) 0.585*** (0.131) 12.65*** (3.446) -6.640** (2.760) 0.389** (0.160)
rule of lawjt rule of law2 jt fin. developmentjt Exporter × Year FE Importer × Year FE Observations Pseudo R-squared
Yes No 100,144 0.528
Yes No 100,958 0.529
(3) 𝐿𝐶 𝐷𝑖𝑗𝑡 0.733*** (0.0286) -0.123 (0.118)
(4) 𝐷𝐶 𝐷𝑖𝑗𝑡 0.665*** (0.0306) 0.401*** (0.138)
8.650*** (2.160) -5.405*** (1.744) 0.808*** (0.149) No Yes 101,266 0.538
9.482*** (2.843) -3.985* (2.254) 0.549*** (0.157) No Yes 92,070 0.537
Note: This table analyzes the relationship between the use of letters of credit and documentary collections and rule of law 𝐿𝐶 and 𝐷 𝐷𝐶 are dummy variables that take the value of 1 if the importer-exporter pair 𝑖𝑗 had at the extensive margin. 𝐷𝑖𝑗𝑡 𝑖𝑗𝑡 positive LC and DC traffic, respectively, in year 𝑡. Columns (1) and (2) explore the role of importing country characteristics. Columns (3) and (4) study the role of exporting country characteristics. Columns (1) and (2) (3 and 4) control for the log of GDP per capita, the log of GDP per capita squared, LC requirements and the log of financial development in the destination (source) country. Standard errors are in parentheses. They are clustered by importing country in columns (1) and (2) and by exporting country in columns (3) to (4). *, ** and *** denote significance at the 10%, 5% and 1% level.
5.2
Additional results and robustness
Trade in services
LCs and DCs are instruments that were originally created to facilitate
the shipment of physical goods. Accordingly, these instruments are largely used for inter27
Note that country pairs are defined unidirectionally, so U.S. exports to China are counted separately from Chinese exports to the United States.
21
national trade in goods. However, LCs and DCs also seem to be used occasionally for trade in services. When we control for bilateral trade in services from the World Banks’ Trade in Services Database, the associated coefficients are positive and highly significant for both LC and DC messages, indicating that LCs and DCs are employed for services trade, although to a much smaller extent than for trade in goods (see online appendix table B.1). Related party trade
A significant fraction of international trade is intra-firm, that is,
between two affiliated companies (see Bernard et al. (2009)). Since the incentives between related parties are typically more aligned, an intra-firm transaction should be less risky than an inter-firm transaction. Accordingly, firms should rely less on banks’ trade finance services in the former case. Information on the share of intra-firm trade is not available worldwide, but the United States Census Bureau provides such data for U.S. transactions. Including the share of U.S. intra-firm exports in our baseline regressions, we find that U.S. exports to destinations that receive a higher share of intra-firm exports employ considerably fewer LCs (see online appendix table B.2). The effect of intra-firm trade on DCs is not significant.28 More robustness
We conduct additional exercises to show that our results are robust.
We control for the cost of importing and exporting as well as financial development squared. We also drop China, India and South Korea from the sample, countries that exhibit high LC shares for historical reasons. Results also hold when regressions are run on cross-sections of different years. The hump-shaped relationship between the use of LCs and destination country risk even survives when different trade finance data from U.S. banks’ regulatory filings available at the Federal Reserve are used. Details on these robustness checks can be found in the online appendix. We also address the worry that exports could be correlated with unobserved factors that also affect firms’ choices of payment contracts, generating an endogeneity problem. We would ideally want to run regressions on the share of LC trade over total trade. But this is not feasible for the bilateral data set with many exporting and importing countries because we observe message counts and not values for most of the sample period. At the same time, we only know export values and not the number of export transactions. Fortunately, more detailed information is available for the United States. We normalize the left-hand-side variable by dividing the number of LC (DC) messages by the number of shipments from the United States to the respective importing country; the new dependent variables reflect the shares of U.S. shipments that use LCs and DCs. These are regressed on the various explanatory variables as before (see columns (3) and (4) of table B.2). In an additional exercise, we compute the share of U.S. exports that use LCs and DCs based on the value data from SWIFT for 2011 and 2012 and estimate equation (1) (see 28
We also ran regressions on U.S. imports including the share of intra-firm imports as explanatory variables. The associated coefficient was not significant.
22
columns (5) to (6) of table B.2). While the normalizations naturally change the magnitudes of the coefficients, and standard errors become larger when regressions are based on only two cross-sections resulting in much fewer observations, the results imply the same qualitative relationship as identified in the baseline regressions.
6
A Model of Payment Choice
In this section, we lay out a model that rationalizes the documented empirical patterns. We derive results on transaction sizes, the role of source and destination country risk and the extensive margin. We also calibrate and simulate the model to show that it can account for the hump-shaped relationship between the use of LCs (DCs) and destination country risk. The model can also generate the observed increase in LC transactions after the Lehman collapse.
6.1
The different payment contracts
In the model, one exporter is matched with one importer. They play a one-shot game and have a choice between four payment contracts: cash-in-advance (CIA), open account (OA), documentary collection (DC) and letter of credit (LC). Both firms are risk-neutral. The exporter has all bargaining power and makes a take-it-or-leave-it offer to the importer specifying the payment contract, the price and the quantity of the goods to sell. 𝑅 denotes the sales value of the goods in the destination country and 𝐾 the production costs in the source country. Firms are either good or bad. A good firm always fulfills a contract. A bad firm breaks it whenever it is profitable. The share of good firms in the source country is given by 𝜂 and in the destination country by 𝜂 * . If a firm does not voluntarily fulfill a contract, its trading partner can attempt to enforce it in court. This is successful in the source country with probability 𝜆 and in the destination country with probability 𝜆* . Cash-in-advance
Under cash-in-advance terms, the importer first pays for the goods
and then the exporter delivers them. The exporter can decide to keep the money without delivering the goods. Exporters that are of the good type (share 𝜂) always fulfill the contract, whereas bad exporters (share 1 − 𝜂) always try to get away without producing. If an exporter defaults on the contract, the importer goes to court, which enforces the contract with probability 𝜆. The exporter sets the price of the goods 𝐶 to maximize its expected profit.29 It also respects the participation constraint of the importer. The expected profit 29 Two bad-exporter strategies need to be distinguished: pooling and separating. Under pooling, a bad exporter imitates the good exporter. Under separating, it chooses a different strategy that reveals its type.
23
of a good exporter can be derived as: [︀ ]︀ E Π𝐶𝐼𝐴 = (𝜂 + (1 − 𝜂)𝜆)𝑅 − 𝐾. 𝐸
(3)
The profitability of cash-in-advance terms increases in the degree of contract enforcement 𝜆 and the share of good firms 𝜂 in the source country. Open account
With an open account, the exporter first sends the goods and then the
importer pays for them. Good importers (share 𝜂 * ) pay the agreed price in any case, while bad importers (share 1 − 𝜂 * ) try to get away without paying. When faced with a bad importer, the exporter goes to court, which enforces the contract with probability 𝜆* . The exporter maximizes expected profit respecting the participation constraint of the importer, which results in: [︀ ]︀ E Π𝑂𝐴 = (𝜂 * + (1 − 𝜂 * )𝜆* )𝑅 − 𝐾. 𝐸
(4)
Because the commitment problem is on the importer’s side, better contract enforcement and a higher share of good firms in the destination country (higher 𝜆* and 𝜂 * ) increase the profitability of an open account. Documentary collection In a documentary collection transaction (figure 2), banks handle documents that transfer the ownership rights from the seller to the buyer. A DC ensures that the importer receives the documents only after paying for the goods. Because the importer typically needs the documents to fully employ the delivered goods, a DC improves the reliability of payment compared with an open account.30 We make two assumptions to capture these features of a documentary collection. First, we assume that with a documentary collection fewer firms try to cheat. More specifically, the (︁ share )︁ of firms that voluntarily fulfill their contracts increases by a factor of 1 + 𝜑𝐷𝐶 ∈
1, 𝜂1* . Payment then happens
with probability 𝜂 * (1 + 𝜑𝐷𝐶 ) + (1 − 𝜂 * (1 + 𝜑𝐷𝐶 ))𝜆* . More generally, define the probability of payment as a function of 𝜑 as: Λ(𝜑) = 𝜂 * (1 + 𝜑) + (1 − 𝜂 * (1 + 𝜑))𝜆* .
(5)
Second, the exporter needs to pay a fixed transaction fee 𝐹 𝐷𝐶 to the bank for its handling Following Schmidt-Eisenlohr (2013), we assume that conditions are such that only the pooling case arises. 𝑅 This is the case if 𝐾 > 𝜂1 . See appendix D for details and a formal derivation of all results. 30 Even with a DC arrangement, however, the importer may still not pay. For example, when the importer can take possession of the goods and divert them without demonstrating legal ownership. Alternatively, the importer may not pay because it no longer wants the goods or because it simply does not have the funds (for example in bankruptcy). We abstract from the latter case in the analysis here.
24
of the documentary collection. Expected profits are then as follows: [︀ ]︀ E Π𝐷𝐶 = Λ(𝜑𝐷𝐶 )𝑅 − 𝐹 𝐷𝐶 − 𝐾. 𝐸
(6)
Relative to an open account, the exporter’s expected profit increases due to the higher probability of being paid. However, the profit is reduced by the fee 𝐹 𝐷𝐶 , which the exporter pays at the beginning of the transaction. Note that the fee is not proportional to the value of the trade so that a documentary collection features increasing returns to scale. Letter of credit
When a trade is settled with an LC (figure 1), banks do not only hand
over documents to the importer as in a DC but they may also advance the importer’s payment. Banks therefore have a strong incentive to screen and monitor importers. Accordingly we assume that the share of importers that try to get away without paying decreases by more with an LC than with a DC. Specifically, we assume that the share of importers that always pay increases by a factor of 𝜑𝐿𝐶 > 𝜑𝐷𝐶 . There remains a risk that the bank does not get paid, which happens with probability 1 − Λ(𝜑𝐿𝐶 ). At the same time, the fixed fee that the bank charges for an LC to cover screening, monitoring and document handling costs is higher than for a DC. It is given by 𝑚 > 𝐹 𝐷𝐶 and is independent of the value of the transaction. Under perfect competition, banks charge an importer the monitoring cost 𝑚 plus the expected loss from extending the guarantee. The letter of credit fee is thus given by: 𝐹 𝐿𝐶 = 𝑚 + (1 − Λ(𝜑𝐿𝐶 ))𝐶 𝐿𝐶 .
(7)
The exporter’s expected profits with an LC can be derived as: Π𝐿𝐶 =
1 − (𝑚/𝑅) 𝑅 − 𝐾. [2 − Λ(𝜑𝐿𝐶 )]
(8)
Because the risk that the exporter does not deliver is eliminated by an LC, expected profits are independent of the enforcement probability 𝜆. However, profits depend on the enforcement probability 𝜆* in the importing country through the fee 𝐹 𝐿𝐶 . Recall that the probability Λ(𝜑) that the importer pays increases in 𝜆* . Thus, under this payment form, the higher the risk that the importer does not pay, the lower the profits are. As with a DC, the LC case contains an element of increasing returns to scale since the monitoring cost 𝑚 is fixed. The average cost of an LC decreases in transaction size 𝑅.
25
6.2
Results
Transaction sizes As discussed above, both DCs and LCs imply fixed document handling, screening and monitoring costs that give rise to increasing returns to scale. Therefore, the higher the value of a contract is, the more attractive DCs and LCs become. Fixing the ration 𝑅/𝐾, it is easy to show that for small transaction sizes (low 𝑅), firms should rely on OA or CIA to save on fixed costs. Transactions with intermediate values should be settled with DCs. When transactions are very large, LCs are most attractive; they imply the highest fixed costs but at the same time reduce payment risk the most.31 These results perfectly match the empirical patterns uncovered in the data. Importer risk The payment forms differ in how risk is allocated between trading partners (see again table 1 for a summary). Figure 8 illustrates the relationship between the optimal payment contract and destination country risk. It plots the exporter’s expected profits under OA, CIA, LC and DC as a function of 𝜆* . Recall that under each payment form the exporter is differently exposed to the risk that the importer does not pay. The higher the exporter’s exposure, the larger is the effect of a change in the probability 𝜆* on expected profits, and the steeper is the profit line in the figure. Figure 8: Profits as a function of contract enforcement in the destination country 1
0.9
CIA OA LC DC
0.8
0.7
Profits
0.6
0.5
0.4
0.3
0.2
0.1
0 0.6
0.65
λ1
0.7
0.75
λ2
0.8
λ*
0.85
0.9
λ3
0.95
1
Note: The graph plots the expected exporter profits under the four payment forms (cash-in-advance, open account, documentary collection and letter of credit) as a function of contract enforcement 𝜆* in the destination country. Parameters are: 𝜆* ∈ [0.6, 1], 𝜆 = 0.5, 𝑅 = 2, 𝐾 = 1, 𝐹 𝐿𝐶 = 0.21, 𝐹 𝐷𝐶 = 0.04, 𝜑𝐿𝐶 = 0.6, 𝜑𝐷𝐶 = 0.2.
OA allocates all the risk to the exporter. Profits under this payment form therefore respond the most to changes in 𝜆* . Since a DC and an LC reduce the risk that the importer gets away without paying, expected profits with an LC or a DC are less responsive to changes in 𝜆* . When the trade is settled on CIA terms, the commitment problem on 31
Results on transaction sizes are proven formally in appendix D. See proposition 2.
26
the importer’s side is eliminated. Therefore, CIA profits are independent of the degree of contract enforcement in the destination country, resulting in a flat line in the figure. The ordering of the slopes is unique and implies that if each payment contract is optimal for some value of 𝜆* , then CIA is chosen for destination countries with the weakest contract enforcement.32 As the probability that a contract is enforced rises, first LCs dominate, then DCs. An open account is optimal when the risk that the importer does not pay is particularly low. In the figure, the optimal payment form switches at 𝜆1 , 𝜆2 and 𝜆3 . As we show below, the unique ordering leads to a hump-shaped relationship between the use of LCs and DCs and destination country risk when there is a random shock to the profitability of the different payment forms that varies across trading partners. Exporter risk The empirical analysis showed that the use of LCs and DCs is not only hump-shaped in destination but also in source country contract enforcement. It is possible to extend the model to generate this result. For an illustration, consider the choice between CIA, OA and LC. A hump-shaped relationship between the use of LCs and source country contract enforcement arises if the profitability of LCs decreases with source country risk. Why might this be the case? The reason could be related to the document handling of LCs. Sometimes, LCs get rejected by the issuing bank for formal reasons, for example, because paperwork is incomplete or contains errors. This problem is likely to be bigger when legal institutions in the source country are weak, making LCs more costly and risky in this case.33 To incorporate this consideration into the model, one could, for example, assume that LC profits increase with source country contract enforcement 𝜆: [︀ ]︀ (︀ * )︀ E Π𝐿𝐶 = 𝜆 + (1 − 𝜆* )𝜑𝐿𝐶 𝑅 − 𝐾 − 𝐹 𝐿𝐶 − 𝛿𝑅(1 − 𝜆), 𝐸
(9)
where 𝛿 < 1. Then, if all payment forms are used for some level of 𝜆, the use of LCs first increases and then decreases in 𝜆. The intuition is similar to before. Profits under OA are independent of 𝜆, while CIA profits decrease in this variable. The derivative of LC profits with respect to 𝜆 is smaller than the derivative of CIA profits with respect to 𝜆. As a result, LCs are employed the most for intermediate levels of 𝜆. When source country risk is low, firms can avoid the fees associated with LCs by choosing CIA. When source country risk is high, they settle transactions with OA. Extensive margin As detailed earlier, the probability that a country pair uses LCs is also hump-shaped in source and destination country risk, that is, country pairs that trade 32
This is proven formally in appendix D. See proposition 1. The same argument applies to DCs. This might explain why the relationship to exporter risk is very similar for DCs and LCs. 33
27
with each other but do not use LCs are concentrated in high-risk and low-risk countries. This makes sense. First, LCs are not very attractive to use for trade with these countries, leading to a low level of demand. As establishing and maintaining a network to process LCs is costly for banks, low demand may give rise to a situation where, for some country pairs, banks do not offer LCs at all.
6.3
Calibration and simulations
In this section, we present simulations of the model. We set several parameters exogenously. These are: 𝜂 = 𝜂 * = 0.6, 𝜑𝐷𝐶 = 0.1, 𝜑𝐿𝐶 = 0.5, 𝑅 = 2 and 𝐾 = 1. 𝜆 is chose to equal the average export-weighted rule of law of all source countries, which is 0.734. We then simulate the model and select the fixed costs of a DC, 𝐹 𝐷𝐶 , and the letter of credit monitoring cost, 𝑚, to match the value shares of LCs and DCs in total transactions. Further details of the calibration are discussed in appendix E. Importer risk Figure 9 shows the payment choices resulting from the calibrated model. In line with our empirical findings, the use of LCs exhibits a hump-shaped relationship with destination country enforcement. Similarly, DC use follows a hump shape that is, however, shifted to the right and at a much lower level. When weighted by the import shares as given in table E.1, LC use and DC use add up to their counterparts in the data (13 and 1.8 percent, respectively). Figure 9: Simulation: Shares of payment forms as a function of importer contract enforcement 1.0
CIA OA LC DC
Share
0.8 0.6 0.4 0.2 0.0 0.4
0.5
0.6
0.7
λ∗
0.8
0.9
1.0
Note: The graph shows the shares of the four payment contracts in total transactions for different levels of contract enforcement 𝜆* in the destination country in the calibrated model. Parameters are: 𝜆* ∈ [0.4, 1], 𝜆 = 0.734, 𝑅 = 2, 𝐾 = 1, 𝑚 = 0.0725, 𝐹 𝐷𝐶 = 0.03, 𝜑𝐿𝐶 = 0.5, 𝜑𝐷𝐶 = 0.1.
28
Simulating the crisis Following the Lehman collapse in 2008q3, the share of transactions that use LCs increased by 53 percent (see panel (a), figure 6). This rise in the importance of LCs can be generated in the model by a change in risk. Suppose there is a shift in the trade distribution from its starting point in table E.1, where an equal share of imports gets downgraded from the top two rule of law deciles to the next two deciles. We find that keeping the costs of DCs and the monitoring costs of LCs fixed, replicating the change in the crisis requires a shift of 6.2 percent from the top two deciles each to the next two deciles. This shift corresponds to a reduction of the trade-weighted average destination country enforcement from 0.795 to 0.768. As a result, LC use increases from 13.2 percent to 20.2 percent in the model, a 53% increase. The new distribution is shown in the last row of table E.1. The counterfactual is also able to generate the relative increase in LC use over DC use. However, our simulation over-predicts the extent to which LCs expand relative to DCs. In the data the share of LCs in total trade finance messages increases from 51 to 61 percent, or a 20 percent increase. In our counterfactual the corresponding ratio increases by 62 percent.
7
Summary and Implications
Using comprehensive data from SWIFT that covers trade finance transactions world-wide, this paper studies the payment choice of firms in international trade. It documents substantial heterogeneity in the use of LCs and DCs across countries and regions. It also shows that the 2007/2008 financial crisis affected firms’ trade finance choices. The focus of the analysis is on the role of contract enforcement in the importing and the exporting country for firms’ payment choices. In contrast to the common view, the use of LCs is highest for exports to and imports from countries with intermediate contract enforcement, a pattern that can be explained by a model of firms’ payment choices. The presented findings have several implications for ongoing discussions around trade finance. First, since LCs are used heavily for some destinations, increases in the cost of trade finance that may come from increased due diligence requirements and new rules on capital and leverage have the potential to impact real economic activity not only at home but also abroad, with heterogeneous effects across source and destination countries. Second, policymakers have interpreted the low usage of trade finance for shipments to less-developed economies as evidence of a gap in the provision of trade finance by commercial banks.34 This is consistent with the large number of zero LC flows that we observe for SubSaharan countries. Our theory shows that there is another reason why LCs may be used little for such destinations, which present a relatively high risk of non-payment by importers. 34
There are ongoing discussions at the WTO whether the private sector is able to meet the demand for trade finance, especially in poor countries (see Working Group on Trade, Debt and Finance (2014)).
29
The high risk means that LCs for these countries are expensive, and firms may optimally decide to use cash-in-advance terms instead of an LC. The relative importance of supply and demand factors cannot be disentangled in our data, but it is crucial to distinguish between them for policy interventions. A further exploration of which firms, industries and countries are especially constrained in their access to LCs and other trade finance instruments is a key task for future research.
A
Data Appendix
Data Sources - SWIFT MT400 and MT700 messages: SWIFT Institute. - Bilateral trade data: Quarterly trade data is from the IMF’s Directions of Trade Statistics; yearly data is obtained by summing over 4 quarters. - U.S. trade data: Information on the share of related party trade and trade cards (number of shipments) is from the United States Census Bureau. - Data on services trade: Trade in Services Database from the World Bank. - Rule of law: World Government Indicators provided by the World Bank, normalized as follows: 𝑙𝑎𝑤𝑛𝑒𝑤 =
𝑙𝑎𝑤𝑜𝑙𝑑 −min(𝑙𝑎𝑤𝑜𝑙𝑑 ) . max(𝑙𝑎𝑤𝑜𝑙𝑑 )−min(𝑙𝑎𝑤𝑜𝑙𝑑 )
- Nominal GDP per capita: World Development Indicators provided by the World Bank. - Distance and other gravity variables: CEPII. For a description, see Mayer and Zignago (2005). - Financial development: proxied by private credit by deposit money banks over GDP from Financial Structure Database provided by the World Bank. For a description, see Beck et al. (2009). - Letter of credit requirements: Information on whether a country requires that importers use of letters of credit is from the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (ARERER). - U.S. GDP deflator: FRED Economic Data provided by the Federal Reserve Bank of St. Louis. - Cost to import and cost to export: Doing Business Indicators provided by the World Bank, US$ per container.
Offshore Centers and Tax Havens (excluded) Antigua and Barbuda, Azerbaijan, Netherlands Antilles, United Arab Emirates, Bahrain, Bahamas, Belize, Bermuda, Barbados, Cayman Islands, Cyprus, Gibraltar, Grenada, Hong 30
Kong, Oman, Ireland, Jordan, Lebanon, Macao, Monaco, Maldives, Malta, Mauritius, Saint Lucia, Seychelles, Singapore, Switzerland, Taiwan, Vanuatu, Samoa
31
References Ahn, JaeBin, “A Theory of Domestic and International Trade Finance,” IMF Working Papers 11/262, International Monetary Fund November 2011. , “Estimating the Direct Impact of Bank Liquidity Shocks on the Real Economy: Evidence from Letter-of-Credit Import Transactions in Colombia,” 2013. mimeo. , “Understanding Trade Finance: Theory and Evidence from Transaction-level Data,” May 2014. International Monetary Fund, mimeo. Amiti, Mary and David E. Weinstein, “Exports and Financial Shocks,” The Quarterly Journal of Economics, 2011, 126 (4), 1841–1877. Antr` as, Pol and C. Fritz Foley, “Poultry in Motion: A Study of International Trade Finance Practices,” Journal of Political Economy, 2015, 123 (4). Asmundson, Irena, Thomas William Dorsey, Armine Khachatryan, Ioana Niculcea, and Mika Saito, “Trade and trade finance in the 2008-09 financial crisis,” Technical Report 2011. Beck, Thorsten, “Financial Dependence and International Trade,” Review of International Economics, May 2003, 11 (2), 296–316. , Asli Demirguc-Kunt, and Ross Levine, “Financial institutions and markets across countries and over time - data and analysis,” Policy Research Working Paper Series 4943, The World Bank May 2009. Bernard, Andrew B, J Bradford Jensen, and Peter K Schott, “Importers, exporters and multinationals: a portrait of firms in the US that trade goods,” in “Producer dynamics: New evidence from micro data,” University of Chicago Press, 2009, pp. 513–552. Biais, Bruno and Christian Gollier, “Trade Credit and Credit Rationing,” Review of Financial Studies, 1997, 10 (4), 903–37. BIS, “Trade finance: developments and issues,” Technical Report, Bank for International Settlements 2014. Burkart, Mike and Tore Ellingsen, “In-Kind Finance: A Theory of Trade Credit,” American Economic Review, June 2004, 94 (3), 569–590. Cunat, Vicente, “Trade Credit: Suppliers as Debt Collectors and Insurance Providers,” The Review of Financial Studies, 2007, 20 (2), 491–527.
32
Demir, Banu and Beata Javorcik, “Grin and Bear It: Producer-Financed Exports from an Emerging Market,” 2014. mimeo, University of Oxford and Bilkent University. , Tomasz Kamil Michalski, and Evren Ors, “Risk-Based Capital Requirements for Banks and International Trade,” Research Paper FIN-2014-1064, HEC Paris 2014. Eck, Katharina, Martina Engemann, and Monika Schnitzer, “How trade credits foster exporting,” Review of World Economics, 2014, pp. 1–29. Ferris, J Stephen, “A transactions theory of trade credit use,” The Quarterly Journal of Economics, 1981, pp. 243–270. Forbes, “Russia Wants To Convince BRIC Partners To Create Alternative Banking System,” June 2015. [Online; posted 1-June-2015]. Giannetti, Mariassunta, Mike Burkart, and Tore Ellingsen, “What You Sell Is What You Lend? Explaining Trade Credit Contracts,” Review of Financial Studies, 2011, 24 (4), 1261–1298. Glady, Nicolas and Jacques Potin, “Bank Intermediation and Default Risk in International Trade-Theory and Evidence,” 2011. ESSEC Business School, mimeo. Helpman, Elhanan, Marc Melitz, and Yona Rubinstein, “Estimating Trade Flows: Trading Partners and Trading Volumes,” The Quarterly Journal of Economics, 05 2008, 123 (2), 441–487. Hoefele, Andreas, Tim Schmidt-Eisenlohr, and Zhihong Yu, “Payment Choice in International Trade: Theory and Evidence from Cross-country Firm Level Datae,” Canadian Journal of Economics, forthcoming. ICC, “Rethinking Trade Finance 2009: An ICC Global Survey,” ICC Banking Commission Market Intelligence Report, International Chamber of Commerce 2009. , “Global Risks - Trade Finance Report 2013,” Technical Report, International Chamber of Commerce April 2013. IFC, “Global Trade Finance Program Brochure,” Technical Report, International Finance Corporation January 2012. IMF, “Survey Among Banks Assessing Current Trade Finance Environment,” IMF-BAFT Trade Finance Survey, International Monetary Fund 2009. IMF-BAFT, “International Monetary Fund / BAFT - IFSA 6th Annual Trade Finance Survey,” Technical Report, International Monetary Fund 2011. 33
Klapper, Leora, Luc Laeven, and Raghuram Rajan, “Trade credit contracts,” Review of Financial Studies, 2012, 25 (3), 838–867. Love, Inessa, Lorenzo A Preve, and Virginia Sarria-Allende, “Trade credit and bank credit: Evidence from recent financial crises,” Journal of Financial Economics, 2007, 83 (2), 453–469. Malouche, Mariem, “Trade and trade finance developments in 14 developing countries post September 2008-A World Bank Survey,” Technical Report 5138 November 2009. Manova, Kalina, “Credit constraints, heterogeneous firms, and international trade,” The Review of Economic Studies, 2013, 80 (2), 711–744. Mayer, Thierry and Soledad Zignago, “Market Access in Global and Regional Trade,” Working Papers 2005-02, CEPII research center January 2005. McIntosh, Craig T. and Wolfram Schlenker, “Identifying Non-linearities In Fixed Effects Models,” Technical Report October 2006. Ng, Chee K, Janet Kiholm Smith, and Richard L Smith, “Evidence on the determinants of credit terms used in interfirm trade,” The Journal of Finance, 1999, 54 (3), 1109–1129. Niepmann, Friederike and Tim Schmidt-Eisenlohr, “International Trade, Risk, and the Role of Banks,” Staff Reports 633, Federal Reserve Bank of New York September 2013. and
, “No Guarantees, No Trade: How Banks Affect Export Patterns,” Staff Report
659, Federal Reserve Bank of New York 2013. Paravisini, Daniel, Veronica Rappoport, Philipp Schnabl, and Daniel Wolfenzon, “Dissecting the effect of credit supply on trade: Evidence from matched credit-export data,” The Review of Economic Studies, 2015, 82 (1), 333–359. Petersen, Mitchell A and Raghuram G Rajan, “Trade Credit: Theories and Evidence,” Review of Financial Studies, 1997, 10 (3), 661–91. Prete, Silvia Del and Stefano Federico, “Trade and finance: is there more than just trade finance? Evidence from matched bank-firm data,” 2012. mimeo. Schmidt-Eisenlohr, Tim, “Towards a theory of trade finance,” Journal of International Economics, 2013, 91 (1), 96 – 112.
34
Turkcan, Kemal and Veysel Avsar, “Investigating the Role of Contract Enforcement and Financial Costs on the Payment Choice: Industry-Level Evidence from Turkey,” 2014. mimeo, Akdeniz University and Antalya International University. Wilner, Benjamin S., “The Exploitation of Relationships in Financial Distress: The Case of Trade Credit,” Journal of Finance, February 2000, 55 (1), 153–178. Working Group on Trade, Debt and Finance, “Improving the availability of trade finance in developing countries: an assessment of remaining gaps,” Note by the Secretariat, World Trade Organization 2014.
35
FOR ONLINE PUBLICATION B
Tables Table B.1: Trade in services as control (1) log(# LCijt ) 0.539*** (0.0468) 0.408*** (0.139) 6.487* (3.899) -6.222** (2.678)
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it
(2) log(# DCijt ) 0.683*** (0.0496) 0.836*** (0.141) 22.14*** (5.575) -14.96*** (3.793)
rule of lawjt rule of law2 jt ln servicesijt Exporter × Year FE Importer × Year FE Observations R-squared
0.0955*** (0.0308) Yes No 16,221 0.726
0.142*** (0.0400) Yes No 13,654 0.689
(3) log(# LCijt ) 0.723*** (0.0467) 0.555*** (0.119)
(4) log(# DCijt ) 0.647*** (0.0576) 0.768*** (0.177)
15.87*** (4.922) -10.83*** (3.397) 0.116*** (0.0363) No Yes 16,207 0.735
20.17*** (5.809) -13.46*** (3.733) 0.124*** (0.0461) No Yes 13,665 0.673
Note: This table presents results when services trade is included as explanatory variable. log (# LC𝑖𝑗𝑡 ) and log (# DC𝑖𝑗𝑡 ) are the logs of the number of MT700 and MT400 messages sent from country 𝑗 to country 𝑖, respectively, in year 𝑡. Columns (1) and (2) include importing country variables and exporter-year fixed effects. Columns (3) and (4) include exporting country variables and importer-year fixed effects. All regressions control for the log of GDP per capita, the log of GDP per capita squared, LC requirements and the log of financial development in the destination and the source country, respectively, as well as bilateral controls. Standard errors are in parentheses. They are clustered by importing country in columns (1) and (2) and by exporting country in columns (3) to (4). *, ** and *** denote significance at the 10%, 5% and 1% level.
Table B.2: USA: related-party trade and estimation in ratios
log(exportsit ) log(distancei ) rule of lawit rule of law2 it Rel. intensityit Time FE Observations R-squared
(1) log(# LCit ) 0.954*** (0.0505) 0.536*** (0.172) 8.804*** (3.186) -7.642*** (2.515) -1.432*** (0.435) Yes 1,003 0.706
(2) log(# DCit ) 0.842*** (0.0673) 1.194*** (0.410) 13.99** (5.455) -6.857* (3.798) 0.105 (0.493) Yes 793 0.675
(3) LC count shareit 0.00569* (0.00309) 0.0155** (0.00748) 0.422** (0.204) -0.363** (0.182)
(4) DC count shareit -0.00104 (0.000801) 0.00448 (0.00279) 0.130** (0.0621) -0.0792** (0.0364)
(5) LC value shareit
(6) DC value shareit
0.0249 (0.0298) 0.783** (0.360) -0.694** (0.284)
0.000984 (0.00587) 0.0852 (0.0978) -0.0518 (0.0663)
Yes 790 0.191
Yes 591 0.117
Yes 180 0.267
Yes 134 0.093
Note: In this table, all regressions are based on data with the United States as the only exporter. Column titles show the dependent variable used in each column. log (# LC𝑖𝑡 ) and log (# DC𝑖𝑡 ) are the logs of the number of MT700 and MT400 messages sent from the importing country 𝑖 to the United States, respectively, in year 𝑡. LC count shareit and DC count shareit stand for the ratios of the number of MT700 and MT400 messages over the number of U.S. exports transactions, respectively. LC value shareit and DC value shareit are the ratios of the value of MT700 and MT400 messages over the value of exports, respectively. All columns control for the log of GDP per capita, the log of GDP per capita squared, LC requirements and the log of financial development in the destination country. Standard errors are clustered by importing country and are in parentheses. *, ** and *** denote significance at the 10%, 5% and 1% level.
1
FOR ONLINE PUBLICATION
Table B.3: Excluding India, China and South Korea (1) log(# LCijt ) 0.521*** (0.0273) 0.289*** (0.0954) 5.736*** (2.084) -5.474*** (1.528)
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it
(2) log(# DCijt ) 0.614*** (0.0376) 0.454*** (0.0991) 7.226** (3.198) -4.419* (2.295)
DC dummymijt DC dummymijt *rule of lawit log(GDP per capit ) 2
log(GDP per capit ) fin. developmentit LC requirementit
-1.095* (0.568) 0.0579 (0.0357) 0.539*** (0.134) 0.570*** (0.209)
-0.685 (0.619) 0.0444 (0.0390) 0.423*** (0.121) 0.279 (0.193)
(3) log(# messagesmijt ) 0.546*** (0.0238) 0.343*** (0.0761) 4.971** (1.917) -5.200*** (1.278) -2.582*** (0.352) 3.937*** (0.513) -1.038** (0.408) 0.0585** (0.0251) 0.491*** (0.106) 0.461*** (0.174)
rule of lawjt 2
rule of law jt
(4) log(# LCijt ) 0.662*** (0.0344) 0.359*** (0.0806)
(5) log(# LCijt ) 0.590*** (0.0364) 0.399*** (0.106)
0.483** (0.226)
8.502*** (2.383) -5.748*** (1.823)
8.121*** (2.733) -4.696** (2.023)
-1.712** (0.662) 0.0794** (0.0399) 0.446*** (0.143) 0.155 (0.186) 0.110 (0.159) -0.271* (0.146) 0.161 (0.128) 0.101 (0.0735) -0.0390* (0.0201) No Yes 44,392 0.584
-1.057 (0.671) 0.0437 (0.0412) 0.402*** (0.151) 0.0651 (0.203) 0.772*** (0.202) 0.0993 (0.165) 0.389*** (0.124) 0.299*** (0.0910) -0.0462* (0.0239) No Yes 31,689 0.577
DC dummymijt *rule of lawjt log(GDP per capjt ) 2
log(GDP per capjt ) fin. developmentjt LC requirementjt contiguityij common colonizerij common languageij
common legal originij time differenceij Exporter × Year FE Importer × Year FE Observations R-squared
0.169 (0.115) 0.112 (0.143) 0.448*** (0.113) 0.142** (0.0616) -0.0652*** (0.0191) Yes No 43,673 0.589
0.686*** (0.189) 0.189 (0.180) 0.660*** (0.114) 0.206** (0.0867) -0.0597** (0.0261) Yes No 31,516 0.584
0.382*** (0.105) 0.138 (0.118) 0.553*** (0.0862) 0.172*** (0.0548) -0.0610*** (0.0148) Yes No 75,189 0.565
(6) log(# messagesmijt ) 0.599*** (0.0320) 0.331*** (0.0812)
8.361*** (2.222) -5.209*** (1.676) -0.531 (0.328) -1.364** (0.595) 0.0606* (0.0362) 0.417*** (0.134) 0.118 (0.166) 0.357** (0.148) -0.144 (0.103) 0.269** (0.105) 0.178*** (0.0674) -0.0331* (0.0198) No Yes 76,081 0.509
Note: In this table, the baseline regressions shown in table 4 and table 5 are estimated, excluding India, China and South Korea. log (# LC𝑖𝑗𝑡 ) and log (# DC𝑖𝑗𝑡 ) are the logs of the number of MT700 and MT400 messages sent from country 𝑗 to country 𝑖, respectively, in year 𝑡. log (# SWIFT𝑚𝑖𝑗𝑡 ) is the log of the number of SWIFT messages of type 𝑚 ∈ {𝐿𝐶, 𝐷𝐶}. DC dummy𝑚𝑖𝑗𝑡 takes a value of 1 if the message refers to a DC and a value of 0 if the message is related to an LC. Columns (1) - (3) explore the role of importing country characteristics. Columns (4) - (6) study the role of exporting country characteristics. Standard errors are in parentheses. They are clustered by importing country in columns (1) - (3) and by exporting country in columns (4) - (6). *, ** and *** denote significance at the 10%, 5% and 1% level.
2
FOR ONLINE PUBLICATION
Table B.4: Baseline regression destination country rule of law, controlling for cost to import
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it
(1) log(# LCijt ) 0.567*** (0.0266) 0.315*** (0.0941) 4.952** (2.184) -5.513*** (1.563)
(2) log(# DCijt ) 0.612*** (0.0388) 0.438*** (0.103) 8.485** (3.317) -4.940** (2.482)
-0.000892*** (0.000215) 0.000000101*** (2.93e-08) Yes 37,921 0.633
0.0000533 (0.000292) -3.01e-08 (5.19e-08) Yes 27,134 0.593
DC dummymijt rule of lawit * DC dummy typemijt cost to importit cost to importit 2 Exporter × Year FE Observations R-squared
(3) log(# messagesmijt ) 0.572*** (0.0236) 0.362*** (0.0772) 5.007** (1.981) -5.463*** (1.359) -2.793*** (0.364) 4.080*** (0.540) -0.000529*** (0.000193) 5.02e-08* (2.79e-08) Yes 65,055 0.592
Note: This table presents a robustness check where cost to import and cost to import squared are added as additional controls. GDP per capita, GDP per capita squared, fin. development, LC requirement and bilateral controls are included in the regression but not reported. Standard errors are in parentheses and are clustered by importing country. *, ** and *** denote significance at the 10%, 5% and 1% level.
Table B.5: Baseline regression source country rule of law, controlling for cost to export
log(exportsijt ) log(distanceij ) rule of lawjt rule of law2 jt
(1) ln count 700 0.701*** (0.0302) 0.379*** (0.0755) 7.365*** (2.672) -5.619*** (2.134)
(2) ln count 400 0.621*** (0.0349) 0.435*** (0.101) 9.423*** (3.048) -6.079*** (2.266)
-0.00122*** (0.000321) 0.000000191** (7.78e-08) Yes 38,626 0.633
-0.00119*** (0.000330) 0.000000233*** (7.24e-08) Yes 27,390 0.595
DC dummymijt rule of lawjt * DC dummy typemijt cost to exportjt cost to exportjt 2 Importer × Year FE Observations R-squared
(3) log(# messagesmijt ) 0.632*** (0.0290) 0.364*** (0.0761) 8.259*** (2.519) -5.718*** (1.947) 0.282 (0.248) -0.463 (0.339) -0.00120*** (0.000285) 0.000000213*** (6.19e-08) Yes 66,016 0.545
Note: This table presents a robustness check where cost to export and cost to export squared are added as additional controls. GDP per capita, GDP per capita squared, fin. development, LC requirement and bilateral controls are included in the regression but not reported. Standard errors are in parentheses and are clustered by exporting country. *, ** and *** denote significance at the 10%, 5% and 1% level.
3
FOR ONLINE PUBLICATION
Table B.6: Baseline regression destination country rule of law, controlling for log(fin. development) squared
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it
(1) ln count 700 0.564*** (0.0268) 0.318*** (0.0948) 6.816*** (2.260) -6.192*** (1.678)
(2) ln count 400 0.605*** (0.0366) 0.418*** (0.0990) 9.154*** (3.236) -5.433** (2.353)
1.035* (0.587) -0.0566 (0.0778) Yes 49,404 0.626
-0.276 (0.577) 0.0863 (0.0779) Yes 35,437 0.600
DC dummymijt rule of lawit * DC dummy typemijt fin. developmentit fin. developmentit 2 Exporter × Year FE Observations R-squared
(3) log(# messagesmijt ) 0.567*** (0.0231) 0.348*** (0.0782) 6.209*** (1.940) -5.951*** (1.307) -2.776*** (0.361) 4.026*** (0.530) 0.606 (0.415) -0.0109 (0.0547) Yes 84,841 0.592
Note: This table presents a robustness check where log financial development squared is added as an additional control. GDP per capita, GDP per capita squared, LC requirement and bilateral controls are included in the regression but not reported. Standard errors are in parentheses and are clustered by importing country. *, ** and *** denote significance at the 10%, 5% and 1% level.
Table B.7: Baseline regression source country rule of law, controlling for log(fin. development) squared
log(exportsijt ) log(distanceij ) rule of lawjt rule of law2 jt
(1) ln count 700 0.711*** (0.0335) 0.386*** (0.0826) 9.374*** (2.703) -6.320*** (2.016)
(2) ln count 400 0.632*** (0.0358) 0.415*** (0.105) 9.227*** (2.907) -5.312** (2.106)
0.605 (0.625) -0.00280 (0.0827) Yes 50,181 0.624
0.367 (0.553) 0.0150 (0.0732) Yes 35,615 0.593
DC dummymijt rule of lawjt * DC dummy typemijt fin. developmentjt fin. developmentjt 2 Importer × Year FE Observations R-squared
(3) log(# messagesmijt ) 0.643*** (0.0313) 0.358*** (0.0813) 9.205*** (2.498) -5.732*** (1.834) 0.283 (0.240) -0.476 (0.328) 0.504 (0.512) 0.00274 (0.0685) Yes 85,796 0.539
Note: This table presents a robustness check where log financial development squared is added as an additional control. GDP per capita, GDP per capita squared, LC requirement and bilateral controls are included in the regression but not reported. Standard errors are in parentheses and are clustered by exporting country. *, ** and *** denote significance at the 10%, 5% and 1% level.
4
5
2003 9.028*** (2.156) -7.909*** (1.708) Yes 5,495 0.665
2004 8.119*** (2.327) -7.447*** (1.738) Yes 5,620 0.654
2005 6.288*** (2.402) -6.302*** (1.768) Yes 5,648 0.644
2006 8.547*** (2.364) -7.960*** (1.672) Yes 5,805 0.639
2007 7.412*** (2.453) -6.653*** (1.703) Yes 5,692 0.630
2008 7.515*** (2.170) -6.811*** (1.529) Yes 5,563 0.621
2009 8.486*** (2.307) -7.015*** (1.697) Yes 5,182 0.603
2010 8.145*** (2.436) -7.073*** (1.816) Yes 5,108 0.593
2011 6.790** (3.020) -6.494*** (2.391) Yes 2,669 0.501
2012 6.662** (3.011) -6.369*** (2.313) Yes 2,622 0.483
Note: This table presents results from regressions run on annual data. All baseline controls are included but not reported. Standard errors are in parentheses and are clustered by importing country. *, ** and *** denote significance at the 10%, 5% and 1% level.
Exporter × Year FE Observations R-squared
rule of law2 it
rule of lawit
Table B.8: Baseline LC regression for separate years, destination country rule of law
FOR ONLINE PUBLICATION
6
2003 8.436*** (3.033) -5.364** (2.186) Yes 3,992 0.632
2004 10.73*** (3.378) -6.989*** (2.333) Yes 4,047 0.625
2005 8.869** (3.403) -5.603** (2.453) Yes 4,054 0.622
2006 10.24*** (3.439) -5.795** (2.531) Yes 4,062 0.604
2007 8.314** (3.801) -4.639* (2.685) Yes 4,007 0.595
2008 7.357** (3.591) -3.587 (2.551) Yes 3,987 0.576
2009 9.218*** (3.469) -5.656** (2.615) Yes 3,646 0.574
2010 10.16*** (3.811) -6.391** (2.839) Yes 3,494 0.556
2011 9.374** (3.923) -5.674* (2.909) Yes 2,083 0.545
2012 9.205** (3.987) -5.920** (2.900) Yes 2,065 0.522
Note: This table presents results from regressions run on annual data. All baseline controls are included but not reported. Standard errors are in parentheses and are clustered by importing country. *, ** and *** denote significance at the 10%, 5% and 1% level.
Exporter × Year FE Observations R-squared
rule of law2 it
rule of lawit
Table B.9: Baseline DC regression for separate years, destination country rule of law
FOR ONLINE PUBLICATION
7
2003 11.30*** (2.295) -8.270*** (1.676) Yes 5,576 0.656
2004 11.62*** (2.478) -8.553*** (1.689) Yes 5,714 0.649
2005 10.00*** (2.682) -7.408*** (1.956) Yes 5,743 0.637
2006 10.16*** (2.511) -7.001*** (1.941) Yes 5,909 0.630
2007 10.31*** (2.645) -6.891*** (2.076) Yes 5,766 0.622
2008 8.773*** (2.809) -5.869*** (2.150) Yes 5,653 0.592
2009 10.39*** (2.814) -6.658*** (2.068) Yes 5,252 0.608
2010 9.213*** (2.839) -6.289*** (2.117) Yes 5,163 0.602
2011 9.832** (3.829) -6.559** (2.659) Yes 2,725 0.560
2012 6.218* (3.653) -4.377* (2.574) Yes 2,680 0.538
Note: This table presents results from regressions run on annual data. All baseline controls are included but not reported. Standard errors are in parentheses and are clustered by exporting country. *, ** and *** denote significance at the 10%, 5% and 1% level.
Importer × Year FE Observations R-squared
rule of law2 jt
rule of lawjt
Table B.10: Baseline LC regression for separate years, source country rule of law
FOR ONLINE PUBLICATION
8
2003 8.678*** (2.692) -5.390*** (2.049) Yes 4,012 0.631
2004 9.326*** (2.983) -5.721*** (2.190) Yes 4,058 0.625
2005 8.460*** (3.173) -5.205** (2.349) Yes 4,074 0.599
2006 8.426*** (3.099) -5.022** (2.293) Yes 4,074 0.593
2007 9.591*** (3.249) -5.253** (2.372) Yes 3,998 0.592
2008 9.353*** (3.317) -5.200** (2.347) Yes 3,987 0.570
2009 12.00*** (3.119) -7.220*** (2.211) Yes 3,676 0.578
2010 11.59*** (3.182) -6.800*** (2.369) Yes 3,531 0.562
2011 13.15*** (3.771) -9.022*** (2.473) Yes 2,104 0.553
2012 11.30*** (3.964) -7.084*** (2.681) Yes 2,101 0.536
Note: This table presents results from regressions run on annual data. All baseline controls are included but not reported. Standard errors are in parentheses and are clustered by exporting country. *, ** and *** denote significance at the 10%, 5% and 1% level.
Importer × Year FE Observations R-squared
rule of law2 jt
rule of lawjt
Table B.11: Baseline DC regression for separate years, source country rule of law
FOR ONLINE PUBLICATION
FOR ONLINE PUBLICATION
Table B.12: Baseline destination rule of law, including offshore centers
log(exportsijt ) log(distanceij ) rule of lawit rule of law2 it
(1) log(# LCijt ) 0.513*** (0.0248) 0.205** (0.0852) 6.460*** (2.087) -6.082*** (1.550)
(2) log(# DCijt ) 0.553*** (0.0345) 0.200** (0.0856) 7.771*** (2.846) -4.799** (1.989)
Yes 67,662 0.611
Yes 49,356 0.592
DC dummymijt rule of lawit * DC dummy typemijt Exporter × Year FE Observations R-squared
(3) log(# messagesmijt ) 0.519*** (0.0217) 0.202*** (0.0703) 5.679*** (1.860) -5.645*** (1.256) -2.594*** (0.350) 3.607*** (0.525) Yes 117,018 0.582
Table B.13: Baseline source rule of law, including offshore centers
log(exportsijt ) log(distanceij ) rule of lawjt rule of law2 jt
(1) log(# LCijt ) 0.632*** (0.0342) 0.291*** (0.0828) 9.484*** (2.591) -6.886*** (1.978)
(2) log(# DCijt ) 0.579*** (0.0334) 0.270*** (0.0979) 7.708*** (2.912) -4.808** (2.046)
Yes 68,574 0.592
Yes 49,400 0.571
rule of lawjt * DC dummy typemijt Importer × Year FE Observations R-squared
9
(3) log(# messagesmijt ) 0.580*** (0.0307) 0.258*** (0.0808) 8.688*** (2.427) -5.865*** (1.796) -0.409 (0.293) Yes 117,974 0.520
FOR ONLINE PUBLICATION
Table B.14: LC and DC intensities by importer and exporter, 2013 ISO 3 USA CHN DEU JPN FRA GBR NLD KOR CAN HKG IND ITA BEL MEX SGP ESP RUS BRA ARE AUS THA TUR CHE POL MYS IDN AUT SAU SWE CZE VNM ZAF HUN DNK NOR IRN SVK FIN UKR CHL PRT ROM ARG EGY IRL PHL GRC COL ISR DZA KAZ PAK IRQ VEN NGA PER MAR BLR NZL LTU OMN BGD
By importer LC share DC share 1.8% 4.9% 43.4% 0.7% 2.1% 1.0% 5.2% 1.2% 11.6% 1.5% 9.0% 2.0% 10.2% 1.3% 27.3% 1.7% 0.7% 0.6% 38.6% 3.1% 24.2% 1.8% 5.3% 1.7% 1.9% 1.0% 0.8% 0.1% 52.7% 4.0% 3.0% 1.6% 4.4% 0.0% 2.4% 0.3% 14.0% 1.3% 4.7% 1.2% 10.8% 0.8% 15.9% 4.0% 69.6% 2.6% 1.0% 0.4% 6.0% 1.1% 16.1% 0.4% 1.2% 0.6% 24.1% 2.3% 0.4% 0.4% 0.2% 0.1% 24.4% 2.5% 6.3% 1.5% 1.2% 0.0% 1.5% 1.1% 0.4% 0.4% 6.1% 0.0% 0.2% 1.6% 1.1% 0.6% 1.9% 0.0% 10.3% 0.4% 4.3% 1.3% 1.0% 0.1% 4.8% 0.4% 26.7% 7.0% 0.8% 0.9% 14.6% 1.0% 0.9% 2.2% 4.1% 0.0% 9.0% 3.9% 88.2% 1.3% 3.4% 0.0% 50.7% 0.4% 29.1% 0.0% 9.4% 0.0% 59.6% 1.0% 6.2% 1.2% 13.2% 3.4% 27.2% 0.0% 1.9% 0.9% 0.6% 0.0% 16.2% 1.5% >100% 0.1%
ISO 3 CHN USA DEU JPN NLD FRA KOR RUS ITA GBR CAN BEL MEX SAU CHE IND MYS ESP SGP AUS ARE BRA THA IDN POL AUT SWE TUR CZE NOR VNM IRL QAT ZAF DNK NGA HUN KWT VEN IRQ ARG HKG SVK CHL FIN PHL IRN KAZ UKR AGO ROM ISR PRT COL DZA OMN PER LBY NZL CRI EGY GRC
By exporter LC share DC share 7.7% 1.1% 7.6% 1.2% 6.5% 0.7% 14.1% 1.7% 8.9% 1.0% 10.7% 0.7% 14.4% 3.2% 14.5% 0.0% 6.7% 0.7% 24.3% 1.9% 4.3% 0.5% 5.2% 0.8% 0.7% 0.0% 3.1% 0.1% 37.7% 2.8% 23.4% 5.2% 9.5% 4.2% 4.3% 0.9% >100% 9.7% 25.9% 1.0% 21.0% 1.9% 5.7% 2.0% 12.0% 10.4% 9.9% 5.6% 0.5% 0.1% 6.4% 0.4% 3.5% 0.3% 11.9% 2.6% 0.7% 0.5% 5.1% 0.5% 9.1% 1.9% 4.6% 0.1% 11.9% 0.1% 8.6% 0.2% 2.8% 0.7% 12.8% 0.0% 0.2% 0.0% 6.0% 0.1% 0.3% 0.0% 24.8% 0.0% 1.4% 0.3% >100% 39.0% 0.1% 0.4% 13.9% 2.0% 5.6% 1.2% 2.1% 1.7% 0.0% 0.6% 22.7% 0.0% 2.1% 0.0% 0.5% 0.0% 0.9% 0.1% 4.4% 0.9% 4.9% 1.7% 4.1% 0.0% 0.6% 0.5% 21.1% 0.1% 7.2% 1.0% 31.9% 0.0% 16.6% 4.1% 0.1% 0.0% 18.4% 2.7% 3.8% 1.0%
Note: This table reports the ratio of the total value of LC (DC) messages sent through SWIFT over total imports and total exports by country, respectively.
10
FOR ONLINE PUBLICATION
Table B.15: LC and DC intensities by importer and exporter II, 2013 BGR SVN KWT ECU LUX QAT LBY TUN AGO JOR LBN HRV SRB MMR GHA CRI DOM SYR LVA EST LKA GTM KEN YEM URY LBR UZB TZA BHR CUB PRY ANT BHS AZE KGZ SLV CIV MOZ MAC TKM BOL SDN PAN HND ZMB BEN TTO KHM AFG BRN BIH ETH GEO COD LAO NPL CMR MNG PNG MKD CYP JAM
By importer 1.1% 0.2% 0.5% 0.2% 24.0% 0.5% 24.9% 0.0% 3.9% 0.4% 30.1% 1.0% 45.6% 24.4% 20.6% 5.3% 14.0% 0.6% 26.9% 2.8% 42.4% 2.3% 2.3% 0.1% 0.0% 0.0% 18.5% 0.0% 19.1% 4.3% 1.8% 0.1% 0.3% 0.0% 4.3% 1.3% 2.9% 0.1% 0.4% 0.1% 40.9% 8.6% 2.4% 0.0% 25.3% 0.6% 19.0% 0.4% 6.1% 0.6% 0.0% 0.0% 24.1% 0.0% 14.1% 0.0% 41.1% 4.0% 32.6% 0.3% 3.0% 0.3% 0.0% 0.0% 3.0% 0.0% 4.9% 0.0% 0.0% 0.0% 2.0% 0.0% 4.3% 1.4% 4.0% 0.3% 54.5% 3.7% 19.0% 0.0% 4.0% 0.8% 33.8% 0.2% 25.7% 1.8% 2.3% 0.0% 1.9% 0.0% 3.2% 0.2% 0.2% 0.0% 16.5% 0.0% 0.9% 0.0% 2.2% 0.3% 0.5% 0.0% 62.4% 0.1% 2.8% 0.0% 1.5% 0.0% 7.5% 0.0% 39.7% 0.0% 2.7% 0.8% 2.1% 0.1% 1.0% 0.8% 1.2% 0.0% 15.2% 2.2% 0.0% 0.0%
SVN AZE BGD PAK BGR ECU BLR LTU MAR TTO LUX TUN SYR EST GNQ CIV TKM LVA SRB HRV URY BRN BOL GTM MMR BHR COG YEM LKA KHM DOM GAB GHA HND JOR PRY ZMB PAN PNG KEN UZB COD MLT BWA CMR NIC SLV BIH CYP SDN ISL MOZ MKD LBN TZA MNG PRK LAO ANT MAC CUB MRT
By exporter 0.1% 36.2% 33.2% 29.9% 2.8% 4.3% 0.0% 1.2% 3.3% 0.8% 4.4% 14.9% 0.0% 1.6% 0.0% 0.0% 0.0% 5.8% 0.0% 0.6% 10.2% 0.1% 1.8% 0.4% 0.2% >100% 0.0% 0.1% 5.7% 1.6% 0.0% 0.0% 1.1% 0.2% 35.9% 0.6% 0.0% 31.8% 2.1% 6.2% 15.6% 0.0% 15.0% 0.0% 0.0% 0.0% 0.0% 0.7% 10.2% 0.7% 0.2% 0.0% 0.1% 77.1% 0.6% 0.1% 0.0% 0.3% 0.0% >100% 2.0% 1.4%
0.1% 0.0% 5.3% 7.2% 0.2% 0.8% 0.0% 0.0% 1.1% 0.0% 1.3% 21.4% 0.0% 0.1% 0.0% 1.6% 0.0% 0.3% 0.0% 0.3% 4.4% 0.0% 0.2% 0.2% 0.0% 5.3% 0.0% 0.0% 2.7% 0.8% 0.1% 0.2% 18.2% 0.6% 1.4% 0.9% 0.0% 0.9% 2.3% 2.1% 0.7% 0.1% 3.1% 0.0% 0.6% 0.3% 0.7% 0.0% 2.6% 2.0% 0.0% 0.1% 0.1% 9.5% 0.1% 0.0% 0.0% 0.0% 0.0% 5.3% 0.8% 0.6%
Note: This table reports the ratio of the total value of LC (DC) messages sent through SWIFT over total imports and total exports by country, respectively.
11
FOR ONLINE PUBLICATION
Table B.16: LC and DC intensities by importer and exporter III, 2013 SEN MLT NIC MDA PRK ZWE MUS ISL TJK HTI GAB ALB UGA COG ARM GIN DJI MRT BMU MDG GNQ NCL MLI FJI BFA TGO MNE BRB SOM MWI GUY NER LCA MDV TCD SLE RWA ABW SYC BLZ GMB FRO VUT
By importer 4.9% 35.6% 0.6% 0.0% 0.0% 0.6% 48.7% 0.7% 0.6% 1.4% 1.7% 0.9% 6.1% 0.1% 0.6% 11.8% 1.9% 20.0% 0.1% 11.5% 2.3% 3.4% 12.5% 0.8% 6.2% 8.8% 0.0% 0.0% 0.0% 11.6% 0.2% 4.3% 0.0% 7.5% 4.4% 1.0% 5.0% 0.0% 0.3% 0.0% 7.4% 0.0% 0.0%
2.7% 0.4% 0.0% 0.0% 0.0% 0.0% 9.7% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.6% 0.4% 0.4% 0.0% 1.0% 0.2% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 0.8% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
TCD BHS MUS ZWE GEO MDA SSD ALB NAM SEN ETH MDG UGA GIN SLE GUY TGO JAM ARM FRO SUR NCL SWZ KGZ MWI HTI TJK FJI GIB BEN SOM LBR NPL GRL BFA LSO ABW SLB AFG DJI SYC BRB BLZ
By exporter 0.0% 1.2% 67.0% 0.0% 0.4% 0.0% 0.0% 0.8% 0.0% 0.1% 52.3% 0.1% 3.7% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 3.9% 0.0% 0.0% 0.0% 0.0% 10.9% 0.0% 14.9% 0.0% 0.0% 0.0%
0.0% 2.3% 8.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.5% 1.8% 1.4% 3.9% 0.2% 0.1% 2.1% 0.0% 0.1% 2.5% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.4% 0.0% 0.1% 0.6% 4.5% 0.0% 0.0% 0.5% 0.0% 0.9% 0.0% 0.2% 0.0% 0.1% 1.7% 0.3% 1.3% 0.1%
Note: This table reports the ratio of the total value of LC (DC) messages sent through SWIFT over total imports and total exports by country, respectively.
12
FOR ONLINE PUBLICATION C
Figures Figure C.1: Average transaction size in the data
0
.2
Density .4
.6
.8
LCs DCs Other
8
10
12 ln USD
14
16
kernel = epanechnikov, bandwidth = 0.2747
Note: This figure depicts the kernel density estimates of the average transaction size by country in 2012. Densities are estimated for those 60 countries for which data is available on both the numbers and the total values of letters of credit, documentary collections and export transactions. Other is calculated as (total exports - LC exports - DC exports) / (#export transactions - #LCs - #DCs.) Tax havens and offshore centers are excluded.
13
FOR ONLINE PUBLICATION D
Details of the Payment Choice Model
In this appendix, we derive the full solution of the model discussed in the main text and provide formal propositions for the main results. We present a slightly more general version than in the main text, introducing interest rates 1 + 𝑟 and 1 + 𝑟* in the source and the destination country, respectively.
D.1
Setup and Solution
We begin by computing the expected profits of exporters for the pooling and the separating case for each of the four payment contracts OA, CIA, DCs and LCs and derive condition under which pooling occurs. We then focus on pooling equilibria and formally prove the relationship between optimal payment choice and destination country contract enforcement in Proposition 1. Proposition 2 shows how the optimal payment choice depends on the value of the sales contract. D.1.1
Cash-in-advance
Pooling case Under pooling, the exporter maximizes its expected profits, respecting the participation constraint of the importer. Expected profits between good and bad exporters differ because bad exporters only produce with probability 𝜆 < 1: ]︁ [︁ = 𝐶 𝐶𝐼𝐴,𝑝 − 𝐾, Good type: max E Π𝐶𝐼𝐴,𝑝 𝐸,𝑔 𝐶 ]︁ [︁ = 𝐶 𝐶𝐼𝐴,𝑝 − 𝜆𝐾, Bad type: max E Π𝐶𝐼𝐴,𝑝 𝐸,𝑏 𝐶
s.t. [︁ ]︁ 𝜂 + (1 − 𝜂)𝜆 𝐶𝐼𝐴,𝑝 E Π𝐼 = 𝑅 − 𝐶 𝐶𝐼𝐴,𝑝 ≥ 0 * 1+𝑟 (participation constraint importer), [︁ ]︁ 𝐶𝐼𝐴,𝑝 = 𝐶 𝐶𝐼𝐴,𝑝 − 𝐾 ≥ 0 E Π𝐸,𝑔
(10) (11)
(12)
(13)
(participation constraint good exporter). The optimal price 𝐶 𝐶𝐼𝐴,𝑝 and resulting expected profits of a good and bad exporter are: 𝜂 + (1 − 𝜂)𝜆 𝐶 𝐶𝐼𝐴,𝑝 = 𝑅, 1 + 𝑟* [︁ ]︁ 𝜂 + (1 − 𝜂)𝜆 Good type: E Π𝐶𝐼𝐴,𝑝 = 𝑅 − 𝐾, 𝐸,𝑔 1 + 𝑟* [︁ ]︁ 𝜂 + (1 − 𝜂)𝜆 𝐶𝐼𝐴,𝑝 Bad type: E Π𝐸,𝑏 = 𝑅 − 𝜆𝐾. 1 + 𝑟*
14
(14) (15) (16)
FOR ONLINE PUBLICATION Separating First, note that if a good exporter chooses CIA, a bad exporter always chooses it as well, imitating the good type. This is strictly preferable to revealing its type because the bad exporter can charge a higher price at no additional cost. If, however, good firms do not choose CIA, a bad firm might want to deviate and choose this contract. This case is considered in the following. Suppose a good exporter does not choose CIA. Given the ability to default on the contract, a bad firm might still consider it optimal to offer a CIA contract, even though this implies revelation of its type. In this case, the importer understands that it deals with a bad firm and adjusts its expected revenues downwards. The importer’s participation constraint becomes: [︁ ]︁ E Π𝐶𝐼𝐴,𝑠 = 𝐼
𝜆 𝑅 − 𝐶 𝐶𝐼𝐴,𝑠 ≥ 0. 1 + 𝑟*
(17)
The pre-payment that makes the participation constraint of the importer bind is: 𝐶 𝐶𝐼𝐴,𝑠 =
𝜆 𝑅. 1 + 𝑟*
(18)
The expected profits of a bad exporter in the separating case with CIA are thus: [︁ ]︁ E Π𝐶𝐼𝐴,𝑠 = 𝐸,𝑏
𝜆 𝑅 − 𝜆𝐾. 1 + 𝑟*
(19)
A sufficient condition for the bad exporter not to choose CIA is that its expected profits in the separating case are less than the expected profits of a good firm in the pooling case, that is, if: E
[︁
Π𝐶𝐼𝐴,𝑝 𝐸,𝑔
]︁
[︁ ]︁ 𝐶𝐼𝐴,𝑠 > E Π𝐸,𝑏 .
(20)
Substituting the expected profits into the inequality above and rearranging delivers: 𝑅 1 + 𝑟* > . 𝐾 𝜂
(21)
This condition ensures pooling. D.1.2
Open account
Now, the exporter can choose between a pooling and a separating strategy. Pooling refers to the case where good and bad importers accept the proposed contract. In the separating case, only bad importers agree to buy the goods at the offered price.
15
FOR ONLINE PUBLICATION Pooling case ]︀ [︀ 𝜂 * + (1 − 𝜂 * )𝜆* 𝑂𝐴 = 𝐶 − 𝐾, max E Π𝑂𝐴 𝐸 𝐶 1+𝑟 s.t. [︀ 𝑂𝐴 ]︀ E Π𝐼,𝑔 = 𝑅 − 𝐶 𝑂𝐴 ≥ 0
(22)
(23)
(participation constraint good importer). It is optimal for the exporter to choose 𝐶 𝑂𝐴 such that the participation constraint of the good importer binds. This implies: 𝐶 𝑂𝐴 = 𝑅, [︀ ]︀ 𝜂 * + (1 − 𝜂 * )𝜆* E Π𝑂𝐴 = 𝑅 − 𝐾. 𝐸 1+𝑟
(24) (25)
Separating The participation constraint of a bad importer is in the separating case: [︁ ]︁ 𝑅 − 𝜆* 𝐶 𝑂𝐴,𝑠 ≥ 0. E Π𝑂𝐴,𝑠 = 𝐼,𝑏 1 + 𝑟*
(26)
The participation constraint of a bad importer binds if: 𝐶 𝑂𝐴,𝑠 =
𝑅 . 𝜆*
(27)
The prepayment 𝐶 𝑂𝐴 exactly offsets the risk of non-payment by the importer. In expectation, the importer thus pays 𝑅 to the exporter. Expected profits of the exporter are, however, reduced as good importers reject the contract and only bad importers (share 1 − 𝜂 * ) accept it, implying: E
[︁
Π𝑂𝐴,𝑠 𝐸
]︁
*
= (1 − 𝜂 )
(︂
)︂ 1 𝑅−𝐾 . 1+𝑟
(28)
Comparing profits under pooling versus separating, an exporter strictly prefers a pooling contract if: 𝑅 𝜂 * (1 + 𝑟) > * . 𝐾 𝜂 − (1 − 𝜂 * )(1 − 𝜆* )
16
(29)
FOR ONLINE PUBLICATION D.1.3
Documentary collection
Pooling case The pooling case is very similar to that under OA. The only difference is a higher probability of payment and the additional fixed cost: 𝐶 𝐷𝐶 = 𝑅, [︀ ]︀ Λ(𝜑𝐷𝐶 ) E Π𝐷𝐶 = 𝑅 − 𝐹 𝐷𝐶 − 𝐾, 𝐸 1+𝑟
(30) (31)
with Λ(𝜑) = 𝜂 * (1 + 𝜑) + (1 − 𝜂 * (1 + 𝜑))𝜆* . Separating case The separating case differs from an OA only in that fewer firms are trying to get away without paying: [︁ ]︁ 𝑅 − 𝜆* 𝐶 𝐷𝐶,𝑠 𝐷𝐶,𝑠 E Π𝐼,𝑏 = ≥ 0. 1 + 𝑟*
(32)
The participation constraint of a bad importer binds if: 𝐶 𝐷𝐶,𝑠 =
𝑅 . 𝜆*
(33)
As before, the prepayment 𝐶 𝐷𝐶 exactly offsets the risk of non-payment by the importer. In expectation the importer thus pays 𝑅 to the exporter. Expected profits of the exporter are reduced because good importers and bad importers that do not cheat under a DC reject the contract, and only the remaining bad importers (share 1 − 𝜂 * (1 + 𝜑𝐷𝐶 )) accept it. Expected profits are: (︂ )︂ [︁ ]︁ 1 𝐷𝐶,𝑠 * 𝐷𝐶 𝐷𝐶 E Π𝐸 = (1 − 𝜂 (1 + 𝜑 )) 𝑅−𝐹 −𝐾 . 1+𝑟
(34)
Comparing profits under pooling versus separating, an exporter strictly prefers a pooling contract if: 𝑅 𝜂 * (1 + 𝜑𝐷𝐶 )(1 + 𝑟) > * . 𝐾 + 𝐹 𝐷𝐶 𝜂 (1 + 𝜑𝐷𝐶 ) − (1 − 𝜂 * (1 + 𝜑𝐷𝐶 ))(1 − 𝜆* ) D.1.4
(35)
Letter of credit
Pooling case The maximization problem is: [︀ ]︀ 𝐶 𝐿𝐶 max E Π𝐿𝐶 = − 𝐾, 𝐸 𝐶 1+𝑟 [︀ ]︀ 𝑅 − 𝐶 𝐿𝐶 s.t. E Π𝐿𝐶 − 𝐹 𝐿𝐶 ≥ 0 𝐼,𝑔 = 1 + 𝑟*
(36) (participation constraint importer).
17
(37)
FOR ONLINE PUBLICATION In the optimum, the participation constraint of the good importer binds. The price is optimally set to: 𝐶 𝐿𝐶 = 𝑅 − 𝐹 𝐿𝐶 (1 + 𝑟* ).
(38)
Plugging in the expression for 𝐹 𝐿𝐶 from equation (7) delivers: 𝐶 𝐿𝐶 = 𝑅 − 𝑚(1 + 𝑟* ) + (1 − Λ(𝜑𝐿𝐶 ))𝐶 𝐿𝐶 =
1 − (𝑚/𝑅)(1 + 𝑟* ) 𝑅. 2 − Λ(𝜑𝐿𝐶 ))
(39)
This implies the following exporter profits: ]︀ [︀ 1 − (𝑚/𝑅)(1 + 𝑟* ) = 𝑅 − 𝐾. E Π𝐿𝐶 𝐸 (1 + 𝑟)[2 − Λ(𝜑𝐿𝐶 ))]
(40)
Separating case The exporter chooses 𝐶 𝐿𝐶,𝑠 that makes the participation constraint of a bad importer bind: ]︁ [︁ 𝑅 − 𝜆* 𝐶 𝐿𝐶,𝑠 = E Π𝐿𝐶,𝑠 − 𝐹 𝐿𝐶 ≥ 0. 𝐼,𝑏 1 + 𝑟*
(41)
Plugging in 𝐹 𝐿𝐶 , we get: 𝐶 𝐿𝐶,𝑠 =
𝑅 − (1 + 𝑟* )𝑚 . 𝜆* + 1 − Λ(𝜑𝐿𝐶 )
(42)
The prepayment 𝐶 𝐿𝐶 depends on the probability that the importer does not pay and the LC fee 𝑚. The exporter’s expected profits are reduced as good importers and bad importers that do not cheat under an LC reject the contract and only the remaining bad importers (share 1 − 𝜂 * (1 + 𝜑𝐿𝐶 )) accept it. Expected profits are: (︂ [︁ ]︁ 𝐿𝐶,𝑠 * 𝐿𝐶 E Π𝐸 = (1 − 𝜂 (1 + 𝜑 ))
)︂ (1 − (𝑚/𝑅)(1 + 𝑟* )) 𝑅−𝐾 . (1 + 𝑟)[𝜆* + (1 − Λ(𝜑𝐿𝐶 ))]
(43)
Comparing profits under pooling versus separating, an exporter strictly prefers a pooling contract if: 𝑅 𝜂 * (1 + 𝜑𝐿𝐶 )(1 + 𝑟)(1 + 𝜅2 )(𝜆* + 𝜅2 )) > , 𝐾 (1 − 𝜅1 )[(𝜆* + 𝜅2 ) − (1 − 𝜂 * (1 + 𝜑𝐿𝐶 ))(1 + 𝜅2 )] with 𝜅1 =
𝑚 (1 𝑅
(44)
+ 𝑟* ) (fixed monitoring component of LC fee) and 𝜅2 = 1 − Λ(𝜑𝐿𝐶 ) (risk
component of LC fee). Discussion of pooling and separating cases We follow Schmidt-Eisenlohr (2013) and focus on the pooling cases for all contracts. This is reasonable for two reasons. First, the 18
FOR ONLINE PUBLICATION pooling case dominates for relatively weak conditions on the ratio 𝑅/𝐾. As long as revenues over production costs are sufficiently large, it is optimal to offer contracts also acceptable by good firms. Second and more importantly, the pooling case is intuitively the economically relevant one. In the pooling case, the agreed payment is lower or equal than the final sales price, that is, 𝐶 ≤ 𝑅. Also, the price is not so high as to generate adverse selection and leave only bad firms in the market. In the separating case, the exporter demands a payment that exceeds the final sales value of the goods 𝑅 because she knows with certainty that she is dealing with a bad counter-party. This generates adverse selection and only bad trading partners are left. Except for a few extreme countries in which contract enforcement is extremely weak and the fraction of good firms is very low, we should not expect to observe this pattern. Hence, we concentrate on the pooling cases and assume that the relevant conditions are satisfied. This is the case if there are sufficiently many good firms reflected in high 𝜂, 𝜂 * , and if the profit opportunities 𝑅/𝐾 in international trade are sufficiently large.
D.2
The role of destination country risk
Proposition 1 states how expected profits under each payment contract change with the probability 𝜆* that a contract is enforced in the destination country. Proposition 1 Expected profits from CIA, an OA, a DC and an LC change in destination country enforcement 𝜆* in the following way: [︀ ]︀ [︀ ]︀ [︀ ]︀ [︀ ]︀ 𝜕E Π𝑂𝐴 𝜕E Π𝐷𝐶 𝜕E Π𝐿𝐶 𝜕E Π𝐶𝐼𝐴 > > > = 0. 𝜕𝜆* 𝜕𝜆* 𝜕𝜆* 𝜕𝜆* Proof. First, note that only expected profit of good exporters are relevant for the choice of payment form. This is the case as we only consider pooling cases where bad exporters 𝜕E[Π𝐿𝐶 ] 𝜕E[Π𝐶𝐼𝐴 ] = 0 and that 𝜕𝜆* > 0, want to imitate good exporters. It is easy to see that 𝜕𝜆* 𝜕E[Π𝑂𝐴 ] 𝜕E[Π𝐷𝐶 ] 𝜕E[Π𝑂𝐴 ] 𝜕E[Π𝐷𝐶 ] > 0, and > 0. It therefore remains to be shown that > and * * * 𝜕𝜆 𝜕𝜆 𝜕𝜆 𝜕𝜆* 𝐿𝐶 𝐷𝐶 𝜕E[Π 𝜕E[Π ] ] that 𝜕𝜆* > 𝜕𝜆* . Begin by taking the difference between expected profits under OA and DC: [︀ ]︀ 𝜂 * + (1 − 𝜂 * )𝜆* Λ(𝜑𝐷𝐶 ) E Π𝑂𝐴 − Π𝐷𝐶 = 𝑅− 𝑅 − 𝐹 𝐷𝐶 . 1+𝑟 1+𝑟
(45)
Taking the derivative with respect to 𝜆* gives: [︀ ]︀ 𝜕E Π𝑂𝐴 − Π𝐷𝐶 𝜂 * 𝜑𝐷𝐶 = 𝑅 > 0. 𝜕𝜆* 1+𝑟 19
(46)
FOR ONLINE PUBLICATION Next, take the difference between the expected profits from DC and LC: ]︀ Λ(𝜑𝐷𝐶 ) [︀ 1 − (𝑚/𝑅)(1 + 𝑟* ) 𝑅 − 𝐹 𝐷𝐶 − 𝑅. E Π𝐷𝐶 − Π𝐿𝐶 = 1+𝑟 (1 + 𝑟)[1 + (1 − Λ(𝜑𝐿𝐶 ))]
(47)
Taking the derivative with respect to 𝜆* delivers: [︀ ]︀ 𝜕(E Π𝐷𝐶 − Π𝐿𝐶 ) 1 − 𝜂 * (1 + 𝜑𝐿𝐶 ) (1 − (𝑚/𝑅)(1 + 𝑟* )) 1 − 𝜂 * (1 + 𝜑𝐷𝐶 ) 𝑅 − 𝑅 = 𝜕𝜆* 1+𝑟 1+𝑟 [1 + (1 − Λ(𝜑𝐿𝐶 ))]2 It is easy to see that this term is positive since 𝜑𝐿𝐶 > 𝜑𝐷𝐶 and
(1−(𝑚/𝑅)(1+𝑟* )) [1+(1−Λ(𝜑𝐿𝐶 ))]2
(48)
< 1.
Corollary 1 follows directly from proposition 1:35 Corollary 1 Suppose that each contract type 𝐶 ∈ {𝐶𝐼𝐴, 𝑂𝐴, 𝐷𝐶, 𝐿𝐶} is used for some ¯* > 𝜆 ¯* > 𝜆 ¯ * , such that: 𝜆* ∈ [0, 1]. Then, there exist 𝜆 3
2
1
¯*. ∙ (i) CIA is used if 𝜆* ≤ 𝜆 1 ¯ * ). ¯*, 𝜆 ∙ (ii) An LC is used if 𝜆* ∈ (𝜆 2 1 ¯ * ). ¯*, 𝜆 ∙ (iii) A DC is used if 𝜆* ∈ [𝜆 3 2 ¯*. ∙ (iv) An OA is used if 𝜆* ≥ 𝜆 3 Proof. Follows immediately from proposition 1.
D.3
The role of transaction size
In the model, DCs and LCs imply fixed document handling, screening and monitoring costs. These give rise to increasing returns to scale. As a result, the model predicts that for small transaction sizes (low 𝑅), firms should rely on an OA or CIA terms to save on fixed costs. Transactions with intermediate values should be settled with DCs. When transactions are very large, LCs are most attractive; they imply the highest fixed costs but at the same time reduce payment risk the most. The trade-off between CIA and OA is independent of 𝑅. Let production cost 𝐾 be a constant fraction of revenues 𝑅, that is 𝐾 = 𝛾𝑅. Then, the following proposition can be derived: Proposition 2 Suppose that each contract type 𝐶 ∈ {𝑂𝐴, 𝐷𝐶, 𝐿𝐶} (or 𝐶 ∈ {𝐶𝐼𝐴, 𝐷𝐶, 𝐿𝐶}) ¯2 > 𝑅 ¯ 1 , such that: is used for some R. Then, there exist 𝑅 ¯1. ∙ (i) An OA (or CIA) is used if 𝑅 ≤ 𝑅 35
Assume, without loss of generality, that if indifferent between multiple payment contracts, an exporter chooses in the following order: OA, CIA, DC, LC.
20
FOR ONLINE PUBLICATION ¯1, 𝑅 ¯ 2 ]. ∙ (ii) A DC is used if 𝑅 ∈ [𝑅 ¯2. ∙ (iii) An LC is used if 𝑅 > 𝑅 Proof. Expected profits over revenues are: [︀ ]︀ 𝜂 + (1 − 𝜂)𝜆 E Π𝐶𝐼𝐴 /𝑅 = − 𝛾, 1 + 𝑟* [︀ ]︀ 𝜂 * + (1 − 𝜂 * )𝜆* − 𝛾, E Π𝑂𝐴 /𝑅 = 1+𝑟 [︀ 𝐷𝐶 ]︀ Λ(𝜑𝐷𝐶 ) E Π𝐸 /𝑅 = − 𝐹 𝐷𝐶 /𝑅 − 𝛾, 1+𝑟 [︀ 𝐿𝐶 ]︀ 1 − (𝑚/𝑅)(1 + 𝑟* ) − 𝛾. /𝑅 = E Π (1 + 𝑟)[1 + (1 − Λ(𝜑𝐿𝐶 ))]
(49) (50)
Now, take the limit of these expressions when 𝑅 → ∞ for DC and LC: [︀ ]︀ Λ(𝜑𝐷𝐶 ) lim E Π𝐷𝐶 /𝑅 = − 𝛾, 𝐸 𝑅→∞ 1+𝑟 [︀ ]︀ 1 − 𝛾. lim E Π𝐿𝐶 /𝑅 = 𝑅→∞ (1 + 𝑟)[1 + (1 − Λ(𝜑𝐿𝐶 ))]
(51) (52)
Expressions for CIA and OA do not change with 𝑅. Note that we can either be in case 1 (𝐶 ∈ {𝑂𝐴, 𝐷𝐶, 𝐿𝐶}) or case 2 (𝐶 ∈ {𝐶𝐼𝐴, 𝐷𝐶, 𝐿𝐶}). For given parameters 𝜂, 𝜂 * , 𝜆, 𝜆* , 𝑟, 𝑟* , either CIA dominates over OA or vice versa, independent of the transaction size 𝑅. Start with case 1, where, for some 𝑅, OA, DC or LC are used. Step 1: for 𝑅 sufficiently large, DC dominates OA. This directly follows from comparing equations (50) and (51). Step 2: for 𝑅 sufficiently large, LC dominates DC. For this, we need to compare equations (51) and (52). The following holds: (︀ [︀ ]︀ [︀ ]︀ )︀ lim E Π𝐿𝐶 /𝑅 − E Π𝐷𝐶 /𝑅 > 0 ⇔ Λ(𝜑𝐷𝐶 )(2 − Λ(𝜑𝐿𝐶 )) < 1. 𝐸
𝑅→∞
This can be rewritten as: (︀ [︀ ]︀ [︀ ]︀ )︀ lim E Π𝐿𝐶 /𝑅 − E Π𝐷𝐶 /𝑅 > 0 ⇔ Λ(𝜑𝐷𝐶 )(1 − Λ(𝜑𝐿𝐶 )) < 1 − Λ(𝜑𝐷𝐶 ). 𝐸
𝑅→∞
This always holds because 1 − Λ(𝜑𝐿𝐶 ) < 1 − Λ(𝜑𝐷𝐶 ) and Λ(𝜑𝐷𝐶 ) < 1. Step 3: for 𝑅 [︀ ]︀ sufficiently small, OA dominates DC and LC. This is easy to see: for 𝑅 → 0, E Π𝐷𝐶 /𝑅 𝐸 [︀ 𝐿𝐶 ]︀ and E Π𝐸 /𝑅 go to −∞. We have shown that for very small values of 𝑅, open account dominates the two alternatives and for very large values of 𝑅, LC is the best contract type. If DC is used for some 𝑅, this has to be the case for intermediate values of 𝑅. The proof of case 2 is analogous. DC and LC become relatively more profitable the higher 𝑅. Hence, for very low values of 𝑅, CIA is chosen. For very high values of 𝑅, LC dominates, and at 21
intermediate values, a DC is chosen.
E
Simulation Details
This appendix explains in detail how we calibrate the key parameters of the model. To simulate the model, we introduce a multiplicative error term to the profitability of each payment contract. For example, profits from cash-in-advance become: (︂
𝜂 + (1 − 𝜂)𝜆 𝑅−𝐾 1 + 𝑟*
)︂ * 𝜖,
(53)
where 𝜖 is drawn i.i.d. from a normal distribution with mean 1 and variance 0.02. For each contract we draw N = 2000 random draws from this distribution. For every level of 𝜆* ∈ (0, 1), we then calculate for each payment contract how often it dominates its alternatives and divide by N = 2000 to obtain its share in transactions. To compute the total share of letters of credit in overall transactions we then need to weigh the share of each contract at different levels of 𝜆* by the share of trade that goes to destinations with that level of rule of law. For this, we discretize the normalized rule of law measure in our data to obtain the distribution shown in the first row of table E.1. Given the number of degrees of freedom we set several parameters exogenously. These Table E.1: Share in total imports Rule of law decile
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
Share of imports
0
0.001
0.008
0.033
0.133
0.083
0.083
0.074
0.179
0.405
Share of imports (counterf.)
0
0.001
0.008
0.033
0.133
0.083
0.145
0.136
0.117
0.343
are: 𝜂 = 𝜂 * = 0.6, 𝜑𝐷𝐶 = 0.1, 𝜑𝐿𝐶 = 0.5, 𝑅 = 2 and 𝐾 = 1. We set 𝜆 equal to the average export-weighted rule of law of all source countries which is 0.734. We then simulate the model and select the fixed costs of using DC, 𝐹 𝐷𝐶 , and the letter of credit monitoring cost, 𝑚, to match the value shares of LCs and DCs in total transactions. As the DC share is an order of magnitude smaller than the LC share (1.8 percent versus 13 percent), we put more weight on matching the share of DCs. More precisely, we multiply the quadratic 𝐿𝐶 2 error of DCs by a factor of 4, that is we minimize: 𝐸𝑟𝑟𝑜𝑟 = (𝑠ℎ𝑎𝑟𝑒𝐿𝐶 𝑚𝑜𝑑𝑒𝑙 − 𝑠ℎ𝑎𝑟𝑒𝑑𝑎𝑡𝑎 ) + 4 * 𝐷𝐶 2 (𝑠ℎ𝑎𝑟𝑒𝐷𝐶 𝑚𝑜𝑑𝑒𝑙 − 𝑠ℎ𝑎𝑟𝑒𝑑𝑎𝑡𝑎 ) . Without this adjustment, the minimization procedure would pick
parameter values that generate zero use of DCs.
22
