Beyond Carry-Trading: New Insights into the Uses of Foreign-Denominated Bond Issuances by Latin American Firms
Rodrigo Pérez Artica
Joel Rabinovich
Nicolás Zeolla
After the 2008 Global Financial Crisis (GFC), corporate debt in emerging countries verified a substantial growth, becoming a key aspect of the global financial scene. Indeed, among developing and emerging market economies (EMEs), non-financial corporate (NFC) sector debt nearly doubled as a percentage of GDP between 2008 and the end of 2019, reaching 96% (Abraham et al., 2021). This trend was further amplified by the effects of the COVID-19 pandemic, which increased this ratio by an additional 11 percentage points by the end of 2020 (Barajas et al., 2021). Only during the post-pandemic, NFC debt nominally decreased, although marginally.
This growth in corporate debt modified the international financial integration of EMEs. Corporate issuances in foreign currency soared and became a major channel for the transmission of global liquidity, taking the place of international bank loans (Avdjiev et al., 2014; Caballero et al., 2016). Consequently, studies describe this trend as a defining characteristic of the global financial market since 2008, calling it the “second phase of global liquidity” (Aldasoro & Ehlers, 2018; Shin, 2014).
Latin America played a prominent role in this landscape. The Economic Commission for Latin America and the Caribbean (2019, p. 123) argues that Latin America is the region with the largest ratio of corporate bonds over total bonds in international markets. Although the continent is not the main issuer among EMEs, it was the most dynamic. The Bank of International Settlements (BIS) International Debt Statistics show that, while corporate debt outstanding denominated in foreign currency grew 190% in EMEs between 2007 and 2014, in Latin America the rise represented 276%. This is especially true for firms from Brazil, Chile, and Mexico, particularly due to the issuance of dollar-denominated securities (Abraham et al., 2021). The first two countries experienced the third and fourth largest increases in foreign currency debt between 2007 and 2014, a period that verified the largest growth in EME corporate debt, only surpassed by China and Türkiye (IMF, 2015).
When delving into the factors underlying the rise in NFC debt among EMEs, predominant research has underscored the prevalence of carry trade activities among EMEs firms. This entails EMEs firms borrowing cheaply from global financial markets, facilitated by quantitative easing policies (Shin, 2014), and subsequently directing these funds into domestic financial markets to exploit profitable investment opportunities. Such investments typically manifest as liquid financial assets (Bruno & Shin, 2017; Caballero et al., 2016) or as trade credit extended to local firms, effectively assuming the intermediary role (Hardy & Saffie, 2019; Shin & Zhao, 2013). However, a series of studies have cast doubt on this generalized speculative motive across EMEs countries emphasizing, on the contrary, precautionary motives (De Gregorio & Jara, 2024; Kaltenbrunner et al., 2024; Rabinovich & Pérez Artica, 2022).
Our article adds to this recent, more skeptical literature of carry-trading as the main motive for EMEs firms issuing bonds in foreign currency, focusing on the case of Latin American firms. We do so by introducing two innovations: one methodological in order to capture carry-trading, and other analytical regarding alternative uses for bond proceeds in foreign currency. In terms of the former, previous literature has measured the ex-ante incentives to engage in carry-trading using national and international-level data on interest rates differentials (Bruno & Shin, 2017; Calomiris et al., 2022; De Gregorio & Jara, 2024). In this paper, however, we use the ex-post gross interest income reported by firms. Carry-trades are ultimately an interest arbitrage and, if they exist, they should be reflected in the interest earned by the firm. In terms of the second contribution of the paper, we complement the analysis of asset side of the balance sheet (what has been typically studied in the literature) with the liability side. Hence, we analyze whether new debt issuances are associated with improvements in debt maturity, more specifically with the burden of short-term liabilities.
This paper confirms some previous findings, challenges others, and provides a complementary narrative for bond issuances by Latin American firms between 2000 and 2022. As it has been shown elsewhere (Bruno & Shin, 2017; Calomiris et al., 2022), we find that foreign currency bond proceeds are more strongly used to accumulate cash holdings when compared to local currency bond proceeds, particularly for firms with large or “index eligible” issuances, and during the immediate post-GFC period. However, we do not find consistent evidence of carry-trading in any of the periods we consider, neither for firms carrying-out large issuances. If any, firms earn positive interest income from domestic rather than foreign issuances. Finally, this article provides evidence for an alternative use of proceeds derived from issuance in foreign currency: liability management. In the second wave of global liquidity starting after the GFC firms have used part of the foreign proceeds to reduce short-term debt, therefore improving their financial position.
The paper is organized as follows. Section 1 offers an introduction to the research question. In Section 2 we provide an overview of the literature on the motives behind corporate bond issuance in the aftermath of the GFC. We organize the contributions according to whether they stress the demand, or instead, the supply side of corporate bonds markets. Then, we identify two gaps in the literature that inspire our main contributions. Section 3 discusses the regional and sectoral patterns of international debt and brings to light the relevance of Latin American firms in the global debt landscape. Section 4 presents our dataset for Latin American firms’ debt issuances and shows the main stylized facts regarding the nationality of issuers, currency of issuance, coupons, and maturities. In section 5 we introduce the econometric approach followed to gauge the uses of bond proceeds by issuing firms. Sections 6 and 7 show the main econometric results and a robustness analysis, respectively. Finaly, Section 8 concludes.
