A License to Issue (Anywhere)
Patterns and Drivers of Corporate Bonds in Latin America
Author: Svetlana Vtyurina, Adrian Robles1, and Mr. Bennett W Sutton
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Contributor Notes

Author’s E-Mail Address: svtyurina@imf.org; arobles@imf.org; bsutton@imf.org
Publication Date:
11 Jul 2017
eISBN:
9781484305911
Language:
English
Keywords:
WP; firm; market; Investment grade firm; firm level; Peruvian firm; sample firm; seasoned firm
Download PDF (2.8 MB)

This paper overviews patterns in bond issuance in local and external markets by firms in six large Latin American countries. Data suggest that despite rising issuance, local markets remain small and shallow in several countries. Nevertheless, since greater funding is available to many firms in both markets, we investigate the factors that may explain the firm’s choice on where to issue a bond. Using an unbalanced panel of firm and market-level indicators for years 1995-2015, we control for variables representing several theories of capital structure, and the results show that firm characteristics such as size and liquidity increase the likelihood of firms to issue externally. With respect to market characteristics, the market completeness hypothesis generates the most support, where market scale and depth are most important for the issuer’s choice of the market, suggesting that local markets will have to become deeper to draw more firms and investment.

Abstract

This paper overviews patterns in bond issuance in local and external markets by firms in six large Latin American countries. Data suggest that despite rising issuance, local markets remain small and shallow in several countries. Nevertheless, since greater funding is available to many firms in both markets, we investigate the factors that may explain the firm’s choice on where to issue a bond. Using an unbalanced panel of firm and market-level indicators for years 1995-2015, we control for variables representing several theories of capital structure, and the results show that firm characteristics such as size and liquidity increase the likelihood of firms to issue externally. With respect to market characteristics, the market completeness hypothesis generates the most support, where market scale and depth are most important for the issuer’s choice of the market, suggesting that local markets will have to become deeper to draw more firms and investment.

I. Motivation

Heeding lessons from crises in the 1990s, many emerging market governments have sought to create deeper and more liquid local bond markets to reduce the risk of the double mismatch of currencies and maturities, and to channel local savings into long-term domestic investment (Laeven, 2014, IMF (a)).

In Latin America, expanding the array of investment vehicles was seen necessary to expand the investor base domestically and abroad, to improve lending terms for corporate and sovereign borrowers, and to promote financial stability (Goldstein and Turner (2004), Borensztein et al (2008), Rodrigues Bastos et al (2015)). Policy makers have also eyed the accumulation of domestic savings to fund the region’s large infrastructure investments needed to raise potential growth (Cuevas et al (2015)). Long-term ambitions envisioned easier access to capital through the development of regional financial centers featuring best practices in financial infrastructure, and in regulatory and tax regimes. Increasing the absorptive capacity of local markets could also improve domestic monetary policy transmission.2

Efforts to attract investment, coupled with the Latin America’s rapid economic growth in the past decades, have brought a fresh wave of companies and investors into capital markets (Financial Times, 2014). Against this backdrop, this paper provides a granular look at the trends in corporate bond financing over the past two decades, especially after the Global Financial Crisis (GFC), in six of the most financially integrated economies in Latin America - Argentina, Brazil, Chile, Colombia, Mexico, and Peru (“LA-6” hereafter). In the context of the increased access to both local and external markets, and to contribute and expand on relevant research, this paper also examines the firm and market level factors influencing the choice of jurisdiction for bond placements. Guided by the outcomes, we offer some policy considerations on further development of local bond markets.

The rest of this paper is structured as follows. Section 2 overviews recent reforms, patterns in bond issuance and market structure in the LA-6. Section 3 presents a literature review, description of the selected empirical methodology, data and results. Section 4 concludes with some policy considerations.

II. Recent Reforms and the State of Latin American Bond Markets

Reform overview

Borensztein et al (2008) document that, in the early 1990s, Latin America had essentially no corporate bond markets (apart from Chile). The economic reforms of the 1990s, including privatizations and the introduction of private pension systems accelerated the demand for long-term debt instruments and deepening of the local markets (Jeanneau and Tovar (2006), de la Torre (2012), Tendulkar (2015)).3 Adoption of international best practices, like International Financial Reporting Standards (IFRS) and Basel bank supervisory regimes, signaled a strengthening of corporate governance and regulatory capacity, which, in turn, generated externalities such as more favorable credit risk assessments.

Governments also spurred the evolution of debt markets by easing restrictions on foreign investment, simplifying investment regulations, allowing pension funds to invest in a wider array of assets, and developing derivatives and repurchase markets. Concurrently, modern asset management strategies utilized by fund managers have increased demand for a more diverse universe of financial vehicles.

Governments also worked to make government debt instruments more attractive through greater financing of fiscal deficits on local markets, increased transparency with respect to the size, timing and participation in issuances, including by setting up the market makers groups, and the establishment of liquid local benchmarks.

Stylized facts

These efforts supported the growth and development of local bond markets, though the prominence of sovereign paper may have been an unexpected outcome. Government bonds constitute almost 60 percent of total stock, compared to 40 percent in Asia. Conversely, the role of corporate bonds is much smaller in Latin America. As a share of GDP, corporate bonds outstanding are about half the size of bonds in other emerging regions and advanced economies, and the flow of new issuances significantly lags other emerging regions (Figures 1 and 2).4 Among the LA-6, Brazilian firms have the most debt outstanding, with their liabilities accounting for nearly 60 percent of the regional corporate bond stock. Until 2016, quasi-sovereign firms (largely Brazilian and Mexican) represented about a third of corporate funds raised, with most of it occurring externally (Figures 2 and 3).5

Figure 1.
Figure 1.

Value of Bonds Outstanding in Local and External Markets, 2015

(In percent of GDP)

Citation: IMF Working Papers 2017, 157; 10.5089/9781484305911.001.A001

Sources: Bank for International Settlements, Securities Statistics; Dealogic; Ibero-American Federation of Exchanges; and IMF staff calculations.1 Includes AUS, AUT, BEL, CAN DEN, DEU, ESP, FIN, FRA, GBR, IRE, ITA, JPN, KOR, NLD, NOR, PRT, SWE, USA.2 Includes CHN, IDN, IND, MYS, PHL, THA.3 Includes CZE, CRO, HUN, LEB, RUS, PAK, POL, SAU, TUR, ZAF.
Figure 2.
Figure 2.

LA-6: Corporate Bond Issuance

(In percent of GDP)

Citation: IMF Working Papers 2017, 157; 10.5089/9781484305911.001.A001

Source: Dealogic.
Figure 3.
Figure 3.

LA-6: Corporate Issuance by Country1

(Millions of US dollars)

Citation: IMF Working Papers 2017, 157; 10.5089/9781484305911.001.A001

Sources: Dealogic; and IMF staff calculations.1 Currencies converted to US dollars at prevailing exchange rate on the day of issuance.

Efforts to develop local markets, coupled with macroeconomic stability, spurred domestic issuance (Table 1). However, more dramatic was the speed and degree to which corporate debt finance has moved offshore. In the early 2000s, close to 60 percent of corporate bonds were issued locally, but by 2013-15, the share had fallen to below 40 percent. Total issuance more than doubled in both value and number of issuances as external issuance exploded from U.S. 38 billion dollars to over U.S. 200 billion dollars.6 In additional to larger issuance amounts, Latin American firms were also attracted by longer maturities and lower interest rates in advanced economy markets where post-GFC quantitative easing programs exacerbated favorable financing terms. The trade-off has been a substantial increase in foreign currency liabilities, in contrast to the objective of reducing currency mismatches.7 Through much of the boom in foreign issuance (2009-13), the currency risks appeared to be contained by financial and natural hedges as well as by domestic currencies that began appreciating soon after the crisis ended. Just before the GFC, there was a spike in demand for local currency denominated debt issued abroad, however, the demand has since returned to pre-crisis levels (Figure 2).

Table 1.

LA-6: Summary of Corporate Bond Issuance

article image
Sources: Dealogic; and IMF staff calculations.

Average weighted by amount issued.

Within local markets, the major change has been the curtailed access for non-investment grade firms, while their external issuance doubled (Table 1). The result is, however, is highly influenced by Brazil, where a contraction took place in both local and external issuances for noninvestment grade firms as economic conditions deteriorated (Annex Table 1). Investment grade firms fared better despite the sovereign’s downgrade.8 However, in most countries, except Argentina, overall issuance declined after the 2013 Taper Tantrum episode, with non-financial companies suffering more than financials (Figure 3).

Relative liquidity conditions between local and external markets are also important indictors of market development. The level of market liquidity has many dimensions and cannot be captured by any single measure (IMF (2015)). Figures 4-6 provide some insights into general liquidity conditions in the LA-6 economies. Aside from in Chile (data for Mexico is not available), markets are characterized by low trading volumes. While data limitations hinder a more in-depth analysis of corporate versus sovereign trading conditions, the World Federation of Exchanges data on the value of bonds traded on exchanges point to stronger investor interest in sovereign paper than corporate, except in Brazil and Peru.9 Low trading volumes most likely encourage firms to cultivate demand from long-term institutional buyers and/or offer higher interest rates to compensate buyers for holding less liquid assets. These rigidities could push corporates to issue abroad where markets are more liquid.

Figure 4.
Figure 4.

Share of Trading of the 10 Largest Fixed Income Instruments, 2006-151

(Percent)

Citation: IMF Working Papers 2017, 157; 10.5089/9781484305911.001.A001

Source: Ibero-American Federation of Exchanges.1 Includes both sovereign and private instruments.
Figure 5.
Figure 5.

Corporate Bond Market Turnover, 2016

Citation: IMF Working Papers 2017, 157; 10.5089/9781484305911.001.A001

Sources: World Federation of Exchanges, Bolsa de Santiago; and IMF staff calculations.1 Ratio of the number of transactions to the number of bonds listed.
Figure 6.
Figure 6.

Corporate Share of Local Bond Trading, 2016

(percent of value traded)

Citation: IMF Working Papers 2017, 157; 10.5089/9781484305911.001.A001

Source: World Federation of Exchanges.

Characteristics of local bond markets

The largest markets are Chile, Brazil and Mexico. Chile has a well-developed local market that generally meets the needs of local firms as it provides size, tenor (average at 13 years), and the funding tailored to the local needs.10 The market’s buyside, however, is dominated by large pension funds, which only hold top-rated paper, subjecting the lower-rated firms to fund through banks. Brazil’s market is the largest (in nominal terms and by the number of issuances), and absorbs most local needs.11 However, it struggles to support long-term instruments as few tenors exceed 5 years, and, like Chile, its slate of corporate issuers is heavily dominated by investment-grade companies. Mexico boasts many issuers, but the buyside is concentrated. Pension funds and the insurance sector buy the longer dated corporate paper (7 to 10 years), and mutual funds, while mutual funds tend to buy the shorter floating rate notes with tenors of between 3 and 5 years.12 When buying local paper, pension funds are also limited to issuers rated AA- and above to guard against a forced sell-off if the debtor falls below investment grade.

Colombia’s local debt market is also dominated by high-grade issuers, which reflects conservative risk management among institutional investors that largely buy and hold. Tenors go up to 20 years, although the average is about 10 years. Peru’s local market is small, with most issuance dominated by financial institutions and a few large energy companies. As in other countries, the main players are very large pension funds, which are limited by prudential limits on lower rated firms. However, the largest obstacle preventing market growth is the limited number of corporate issuers and the small issuance amounts. Argentina’s market tends to feature shorter term issuances with tenors averaging 15-25 months given the country’s persistently high inflation, although volumes are large.12 Secondary trading is light, not least because investors tend to hold to maturity given short tenors.

In summary, patterns of issuance in local bond markets is not homogeneous in Latin America (Annex Table 1), but there are common features including the outsized role of pension funds and a strong preference for investment grade issuers.

III. Data, Methodology, and Results

Literature review

While the determinants of local market development are not explored in this paper, the topic underlies many of our priors and results. Studies by Burger and Warnock (2004), Eichengreen and Luengnaruemitchai (2004), Braun and Briones (2006), and Bae (2012), among others, examine the role of scale, institutional development, and macroeconomic policy in spurring growth of local bond markets across the globe. Chinn and Ito (2006) identify capital market openness, legal, institutional and accounting improvements when explaining the level of financial development.

We focus on the firm’s capital structure and motivation theories that could explain the firm’s decision regarding the jurisdiction of issuance. A comprehensive overview of those is found in Black and Munro (2010) and Mizen et al (2012). Studies by Rajan and Zingales (1995), Booth et al (2001), Gozzi et al (2012) find the size and strength of the firm’s balance sheet being the determining factors in financing choice decisions (local or external), across both developed and developing countries. Whereas, Myers and Majluf (1984) concluded that before issuing abroad the largest and strongest firms might first tap local savings akin to the pecking order theory.13 Lower-rated/smaller firms may attempt to go abroad where risk taking is more prevalent and the pool of investors is more diverse (Black and Munro, 2010).14

Some firms might also seek external markets for its completeness/depth, which relates to the capacity to absorb larger issuances and innovatively structured or tailored products available in a larger and more diversified investor pool. Firms may also try to maintain market presence to ensure market continued access ((Berger (2002), Faulkender (2005), Siegfried et al (2007)). Based on a sample of Asian economies, Mizen et al (2012) reaffirm that the depth of the markets, their size and liquidity, can affect corporate financing decisions, and highlight the importance of a large non-resident investment base and the exemption from withholding taxes.

Issuance decisions can also be driven by risk management considerations where firms look for natural hedges, in which the exposure to a foreign currency debt service risk is offset by foreign currency revenues. Issuers in less developed markets may tap external markets to realize lower costs and other considerations, such as lengthening the tenor or locking in a rate (timing the market for yield). In more sophisticated and active markets, price arbitrage/static trade off considerations may drive decisions where deviations in cost incentives are actively arbitraged through variations in interest rates in different currencies and proceeds are frequently swapped back into local currency (Black and Munro 2010).

The agency theory stipulates that while costs of disclosure and issuance fees rise when issuing aboard, this could be mitigated through collateral and the positive effects from greater transparency. Weak local indicators (namely, adverse macroeconomic conditions, inadequacy of local savings, tax regimes, underdeveloped local market infrastructure, information asymmetries, and barriers to non-resident investment) may also encourage firms in less developed local markets to issue externally (Burger and Warnock (2006), Chan et al (2011)).

Though these topics have been well covered in the literature, the strand of research looking together at the firm and market level factors influencing a choice of jurisdiction is not particularly large, with just a few studies examining the larger and more mature Asian markets. We contribute to the studies by examining these factors in the case of the LA-6 firms.

Data

We compile a dataset on issuances in local and global markets by LA-6 firms between 1995 and 2015, collecting both financial statements and issuance characteristics. The choice of variables for this analysis is guided by the findings in the previous literature but largely follows the approach of Mizen et al (2012), and adapted for data availability in our countries of interest. The data were sourced from Bloomberg for 2,163 companies and includes a total of 12,997 separate issuances (Annex Table 2).15 Compared to the analyses of Asian markets, which include advanced economies, the sample size is relatively small (Black and Mizen used between 35,000-45,000 observations). The pool of companies in the dataset is further restricted by the availability of financial statement information for each of the firm level explanatory variables for at least three years (Table 2). We then segment firms into financial versus non-financial and seasoned versus non-seasoned issuers to look for patterns in the structure and placement.16

Table 2.

Variables and Expected Outcomes

article image

Empirical methodology

We use a discrete choice (probit model), which is a natural empirical method to evaluate the probability of a firm to issue abroad, once a decision to seek financing has been made.

Pr(EXTBijt=1)=Φ(αi+Xijtβi+Zjtγ+∈ijt).

The variable EXTBijt takes the value 1 if the bond is issued externally, in year t, and zero if issued locally. We also include firm-specific regressors, Xijt for firm size, investment, years present on stock market, liquid assets, collateral, and total debt.17 These variables have been evaluated with and without a lag to check for robustness and to mitigate potential endogeneity concerns. Finally, the model includes dummies and a time trend to account for debt markets becoming increasingly international over time. In Table 2, we present the regressors and the expected signs of the estimated coefficients.

Data overview

In this section, we discuss summary statistics, including means and standard deviations for the individual factors in the choice of financing with respect to the jurisdiction. These are reported for all sample firms, then broken down into financial and non-financial, those that issue locally or externally, and for each country. Annex Table 3 shows that issuers in local markets are smaller and have lower capital expenditure needs, which suggests that their financing needs could be met in local markets. The results are similar to Mizen et al (2012) findings for the Asian economies.

At the same time, local firms are more leveraged (which maybe either a sign of vulnerability or a good rating since they could accumulate debt), less liquid (which could mean that they need funding) and possess more collateral (which could help improve borrowing terms). The fact that larger firms issue in external markets could be an indication of the lack of local market’s depth. Financial firms are smaller in size, and have much lower investment to total assets ratio, which is also in line with findings by Mizen et al (2012). Also, their assets are more liquid, which could be associated with the region’s substantial dependency on deposit funding or high levels of dollarization. Financial firms also maintain larger collateral. Seasoned firms are less leveraged than the non-seasoned ones, but non-seasoned entities are slightly larger in size.18 The rating dummy indicates that a large share of our estimation sample is composed of entities that have received a rating by at least one main rating agency.19

Annex Table 4 shows differences by country at the firm level. Issuance in local and external markets depict quite sizable variations by country. Unsurprisingly, Brazil has a big impact on the aggregate averages for most indicators. Brazil’s weight in the estimation sample increases after applying the selection criteria. Companies in Brazil, Colombia and Mexico are most indebted, but are also the most profitable and liquid. Aside from Argentine and Peruvian firms, the sizes of total assets are comparable, though investment ratios vary widely across countries. The ratio of rating is similar among countries apart from Chile, perhaps highlighting the depth of the local market.

Annex Table 5 shows the market level indicators. The averages display significant variation between each of the LA-6, with Brazil’s, Chile’s and Mexico’s markets having the biggest impact on regional averages. These have the deepest markets (TDSEC), while Peru’s market is small. Mexican and Brazilian firms dominate large issuances abroad (FCY) and in local markets (LCY) where the size of issuances are larger than their external placements, also indicated by size of the local market (ONSRT). Foreign Direct Investment (FDI) is similar across countries, with Chile having a higher level than average, indicating favorable domestic conditions for foreign investment.

Empirical results

As discussed, we aim to identify the factors behind the decision to issue in foreign jurisdictions. The decision is partly determined by firm characteristics (size, years on the market, etc.) and partly dependent on the level of development of the market (depth, liquidity, etc.). Our dependent variable is EXTB, which takes a value of 1 if the bond is issued externally.20

Firm level indicators

At the firm level, the results show that SIZE is statistically significant in all specifications (Table 3). This suggests that smaller firms are more inclined to issue in domestic markets where the investor base is more familiar with the issuer. Furthermore, international underwriters may be empowered to exercise a strong preference for large deals and thus de-prioritize issuance by smaller firms. Transaction costs of issuing externally could be higher and larger firms might have more capacity to absorb these costs. The statistical significance of size could be an indicator that the funding needs of larger firms exceed the available liquidity in local markets. While highly liquid (LIQ) firms may need less borrowing, they tend to issue externally (statistical significance in all specifications), perhaps owing to a need to maintain access to more liquid markets. At the same time, and contrary to expectations, firms with higher collateral (COL) seem to issue less offshore, as they most likely get better terms domestically as foreign lenders may be leery of the recovery values of assets residing abroad in countries with weak legal frameworks for creditor rights (statistical significance in all three specifications). Finally, the years of presence in the market (AGE) seem to have little impact, as the coefficient is small and not statistically significant in any of the specifications. More important seems to be the validation associated with having been rated by a major agency, as this variable (RATING) shows high statistical significance in all specifications. This suggest that even, “young” firms could enter the external markets if they are rated.

Table 3.

Choice of Market: Firm Level Data, with Time Trend

article image

Market level indicators

With respect to market characteristics, both, the relative size of a local market (ONSRT), and the overall size of the market (TDSEC) influence (negative coefficient) the jurisdiction choice, indicating support for the pecking order theory as firms will access a market if there is sufficient scale and depth. An incentive to issue abroad is also reflected in the interest rate differential (INTD), and consistent with expectations, it plays some role in the decision on issuing externally in both specifications (lower local rates reduce the probability of going abroad). Similarly, higher CEMBI spread makes external financing less attractive (although the effect is small), while a larger presence of sovereign external debt (EXGD) increases the probability of financing externally. Finally, higher net FDI inflows may be associated with supportive domestic conditions reducing the need of borrowing abroad. In sum, the results of the market indicators are consistent with the market depth theory.

Sectoral characteristics

When controlling for firm-specific characteristics, we find that factors affecting the choice of the jurisdiction vary depending on the firms’ business segment (financial or non-financial) and the presence on the market (seasoned or un-seasoned). In Annex Table 6, which shows the detailed breakdown for the choice of a market, non-financial entities display more significant variables primarily due to the small sample size of financials. The absolute size of the market (TDSEC) has a positive influence on the jurisdiction choice suggesting that the overall market depth is more important for all non-financial firms, seasoned and non-seasoned (most significance), while seasoned financial firms may be indifferent given several funding options at their disposal, including through deposits. The result for the relative size of the local market (ONSRT) for both groups of non-financial entities in the sample is similar to the model with a lag and interacted with firm level variables, while financial firms’ result may be once again explained by specificities of their funding structures. Interest rate differential (INTD) is most significant for non-financial seasoned firms. While CEMBI and external government debt have significant impacts, they either have very small coefficients or conflicting signs between specifications. Finally, non-seasoned, non-financial firms may be more likely to issue locally when there are strong FDI inflows.

In Annex Table 7, we provide full results, including both firm and market level indicators with seasoned dummy interactions, for all observations and non-financial firms. The financial corporations are not represented due to observation limitations in the sample. Most of the indicators behave as expected and consistent with previously reported model specifications, but we can clearly see the difference between seasoned and non-seasoned firms in both firm and market level variables.

We provide detailed results on goodness-of-fit tests, both intercept and full model, for all the specifications used in this exercise in Annex Table 8.

IV. Conclusions and Policy Considerations

The volume and the relative size of corporate bond issuances in both external and local markets increased significantly in the LA-6 over the past two decades. This was facilitated by greater macroeconomic stability and regulatory reforms. However, local markets remain relatively small compared to peers, not very liquid and dominated by government paper.

With the greater availability of funding in both foreign and domestic markets, we searched for evidence in support of several capital structure theories by examining the firm-level and market factors influencing the firm’s choice where to issue. Our results support the market completeness theory, where the choice of the jurisdiction depends on the markets’ scale and depth and their ability to accommodate the borrower’s needs. The size of the overall market was a significant factor in selecting the jurisdiction of issuance. At the firm level, size and liquidity were indicators of higher probability of external issuance, most likely driven by large financial and liquidity needs not being accommodated by the local market. This supports firm structure/scale and agency cost theories. The importance of interest rate differential supports the static trade off theory.

The analysis confirms that local bonds markets in several countries studied here will need to continue growing and developing to attract more issuers and provide a wider array of investment opportunities. However, this could be construed as a chicken and egg dilemma, as firms look for larger markets for funding, but markets will not become larger unless more firms enter. This is where the recommendations from other studies on the prerequisites for local market development become relevant.

Strong macroeconomic policies play an important role in spurring growth of local bond markets (Berger and Warnock, 2006). For example, in our country sample, recent macroeconomic imbalances resulting in high inflationary environments, like in Argentina, led to bond maturities of a very short nature, which are not attractive for long-term investors. Consistent with crowding out theory, a high level of government debt, as in Brazil, may have reduced the share of corporate bonds in the total stock.

Governments should continue to support local markets by establishing highly traded benchmark instruments against which private bond spreads can be valued. Domestic bond spreads provide traders and policy makers with market perceptions of credit risk, which can inform and improve the conduct of monetary policy. Also, the expansion of hedging instruments would help reduce currency risks and external funding dependence (Saxena and Villar, 2008). These are more available and diversified in the countries with larger capital markets (Mexico and Brazil) but are still scarce in countries like Peru. Ensuring continued participation of the country in emerging-market benchmarks and global portfolios is also an important factor for attracting global interest to the country.

Regulatory restrictions and reforms have also been found important in hindering or promoting local bond financing (Borensztein et al, 2008).21 For example, while Peru has achieved and maintained impressive macroeconomic stability, its local markets remain small, not least due to regulatory hurdles and institutional weaknesses. Overall, to foster greater issuer participation and investor confidence, it is necessary to further strengthen the corporate governance frameworks, streamline issuance processes and procedures, not least by reducing cumbersome registration requirements (IMF (2005)). Improving data collection and dissemination, and enhancing competitiveness of the market infrastructure (safer, more efficient payment and settlement systems) will also help achieve greater market efficiency and transparency (IOSCO, 2007).

Finally, as both firm and market size continue to be important obstacles to the development of local markets, consideration should be given to policies that widen the attractiveness of pooling vehicles that generate subsequent trades like mutual funds, money market accounts and index funds. (Borensztein et al, 2008). There is also room to consider greater cross-border integration to address the problem of small market size and liquidity, perhaps through the Integrated Latin America Market (MILA) initiative that aims to foster equity and bond market integration among Chile, Colombia, Mexico and Peru. Expanding pension and mutual funds not only creates demand for fixed income securities but also contributes to the increase in financial innovation, improved corporate governance, and enhances competition in the bond market (Silva, 2008).22

Building on the latter point, further research could consider the demand side factors, like the capacity of the domestic institutional investors to absorb the additional domestic bond issuance, although, as mentioned, the issue lies in part in the regulation and limits on investments guided by firms’ ratings, but also in the expansion employee participation in pension schemes. Another angle could be looking in more detail into the pecking order theory to gauge what types of firms first issue domestically or abroad (e.g. better rated firms, more liquid firms). Similarly, it could be explored how reliance/availability of bank financing factors into the decision on the firms’ financing choice.23

Annex Table 1.

Summary of LA-6 Corporate Bond Issuance by Country

article image
Sources: Dealogic; and IMF staff calculations.

Average weighted by amount issued.

Annex Table 2.

Corporate Bond Issuance in Estimation Sample

article image
Annex Table 3.

Summary Statistics for Firm-Specific Variables

article image
Notes: The table reports sample means with standard deviations in parenthesis. SIZEijt: Logarithm of total assets. INVijt: Investments over total assets. AGEijt: Numbers of years listed on the stock exchange. LEVERijt: Total debt to total assets. LIQijt: Current assets over total liabilities. COLijt: Tangible assets to total assets. RATINGijt: Rating provided by at least one rating agency. Equality of mean p-value is reported under Diff. with Ho: diff = o and Ha: diff !=0.
Annex Table 4.

Summary Statistics for Firm-Specific Variables, Country Level

article image
Notes: The table reports sample means with standard deviations in parenthesis. SIZEijt: Logarithm of total assets. INVijt: Investments over total assets. AGEijt: Numbers of years listed on the stock exchange. LEVERijt: Total debt to total assets. LIQijt: Current assets over total liabilities. COLijt: Tangible assets to total assets. RATINGijt: Rating provided by at least one rating agency. Equality of mean p-value is reported under Diff. with Ho: diff = o and Ha: diff != 0.
Annex Table 5.

Market Summary Statistics

article image

Local firm’s liabilities are calculated using only non-financial firm data for Colombia and calculated using only financial firm data for Argentina.

Notes: The table reports sample means with standard deviations in parenthesis. FCYjt: Firm’s external liabilities in millions of USD. LCYjt: Firm’s local liabilities in millions of USD. TDSECjt: Total bonds to GDP. ONSRTjt: Local bonds to total bonds. INTDjt: Short-interest differential between local and US nominal rates. CEMBIjt: Corporate EMBI sovereign spread. EXGDjt: External government debt. FDIjt: Foreign direct investment as percentage of GDP.
Annex Table 6.

Detailed Breakdown of Choice of Market, with Time Trend

article image
Annex Table 7.

Choice of Market: Full Results, with Seasoned Component

article image
Annex Table 8.

Post-Estimation Measures of Fit

article image

References

  • Bae, Kee-Hong, 2012, “Determinants of Local Currency Bonds and Foreign Holdings: Implications for Bond Market Development in the People’s Republic of China”, ADB Working Paper Series on Regional Economic Integration, No. 97.

    • Search Google Scholar
    • Export Citation

  • Berger, Allen, Marco A., Espinosa-Vega, W. Scott Frame, and Nathan H. Miller, 2005, “Debt Maturity, Risk, and Asymmetric Information”, Journal of Finance, Vol. 60, pp 2895–2923.

    • Search Google Scholar
    • Export Citation

  • Black, Susan, and Anella Munro, 2010, “Why Issue Bonds Offshore?”, BIS Working paper, No. 334 (Basel: Bank for International Settlements).

    • Search Google Scholar
    • Export Citation

  • Borensztein, Eduardo, Kevin Cowan, Barry Eichengreen, and Ugo Panizza, 2008, “Bond Markets in Latin America: On the Verge of a Big Bang?” Cambridge and London: MIT Press.

    • Search Google Scholar
    • Export Citation

  • Booth, Laurence, Varouj Aivazian, Asli Demirguc-Kunt, and Vojislav Maksimovic, 2001, “Capital Structures in Developing Countries”, The Journal of Finance, Vol. 56, 87–130.

    • Search Google Scholar
    • Export Citation

  • Braun, Matias, and Ignacio Briones, 2006, “The Development of Bond Markets around the World”, Mimeo, Anderson School, UCLA.

  • Burger, John, and Francis Warnock, 2004, “Foreign Participation in Local-Currency Bond markets”, International Finance Discussion Paper No. 794, Board of Governors of the Federal Reserve System.

    • Search Google Scholar
    • Export Citation

  • Burger, John, and Francis Warnock, 2006, “Local Currency Bond Markets”, National Bureau of Economic Research, Working Paper No. 12552.

    • Search Google Scholar
    • Export Citation

  • Chan Eric, Michael Chui, Frank Packer, and Eli Remolona, 2011, “Local Currency Bond Markets and the Asian Bond Fund II Initiative”, BIS Paper No. 63 (Basel: Bank for International Settlements).

    • Search Google Scholar
    • Export Citation

  • Chinn, Menzie, and Ito Hiro, (2006), “What Matters for Financial Development? Capital Controls, Institutions, and Interactions”, Journal of Development Economics, Vol. 81, pp 163–192.

    • Search Google Scholar
    • Export Citation

  • de la Torre, Augusto, Alain Ize, and Sergio L. Schumulker, 2012, “Financial Development in Latin America and the Caribbean”, The World Bank (Washington: The World Bank).

    • Search Google Scholar
    • Export Citation

  • Eichengreen, Barry, and Pipat Luengnaruemitchai, 2004, “Why Doesn’t Asia Have Bigger Bond Markets?”, National Bureau of Economic Research, Working Paper No. 10576.

    • Search Google Scholar
    • Export Citation

  • Faulkender, Michael, 2005, “Hedging or Market Timing? Selecting the Interest Rate Exposure of Corporate Debt”, The Journal of Finance, Vol. 60, No 2, pp 931–962.

    • Search Google Scholar
    • Export Citation

  • Fernández Roque B., Sergio Pernice, and Jorge M. Streb, 2007, “Determinants of the Development of Corporate Bond Markets in Argentina: One Size Does Not Fit”, Universidad Del Cema, Working paper No. 348.

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund: (a) Monetary Policy Implementation at Different Stages of Market Development, 2004;

    (b) Global Financial Stability Report (GFSR), 2015;

    (c) “Development of Corporate Bond Markets in Emerging Market Countries”, GFSR, 2005;

    (d) “Financial Integration in Latin America: A New Strategy for a New Normal”, edited by Charles Enoch, Wouter Bossu, Carlos Caceres and Diva Singh, 2017.

    • Search Google Scholar
    • Export Citation

  • International Organization of Securities Commission (IOSCO), 2007, “Factors Influencing Liquidity in Emerging Markets”, Report of the IOSCO Emerging Markets Committee.

    • Search Google Scholar
    • Export Citation

  • Jeanneau, Serge, and Camilo Tovar, 2006, “Domestic Bond Markets in Latin America: Achievements and Challenges”, BIS Quarterly Review (Basel: Bank for International Settlements).

    • Search Google Scholar
    • Export Citation

  • Goldstein, Morris and Philip Turner, “Controlling Currency Mismatches in Emerging Markets”, Peterson Institute for International Economics, 2004.

    • Search Google Scholar
    • Export Citation

  • Gozzi, Juan Carlos, Ross Levine, Maria Soledad Martinez Peria, and Sergio L. Schmukler, 2012, “How Firms Use Local and International Corporate Bond Markets”, Policy Research Working Paper, No. 6209 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

  • Hale, Galina and Mark Spiegel, 2012, “Currency Composition of International Bonds: The EMU Effect”, Journal of International Economics No. 88, 134–149.

    • Search Google Scholar
    • Export Citation

  • Laeven, Luc, 2014, “The Development of Local Capital Markets: Rationale and Challenges”, IMF Working paper 14/231.

  • Myers, Steward, and Nicholas S. Majluf, 1984, “Corporate Financing and Investment Decisions: When Firms Have Information That Investors Do Not Have”, Journal of Financial Economics, Vol. 13, pp 187–221.

    • Search Google Scholar
    • Export Citation

  • Mizen, Paul, Frank Packer, Eli Remolona and Serafeim Tsoukas, 2012, “Why do Firms Issue Abroad? Lessons from Onshore and Offshore Corporate Bond Finance in Asian Emerging Markets’, BIS Wor