Investigating the determinants of domestic bonds: the role of socio-economic and institutional factors

ABSTRACT

This article investigates the determinants of the size of domestic bond market using economic, social and institutional factors. We expand the body of existing literature by suggesting that economic and social environment as well as institutional settings vary between developed and emerging economies. The article uses recent data from a wide range of countries, incorporates a variety of macroeconomic variables, social indicators and institutional factors to reassess the determinant of domestic bond markets. Robustness of the empirical analysis is established through both two-stage least squares and generalized method of movements techniques. The results of this article show that the size of the economy, breadth and depth of the banking system, the monetary policy stance, the degree of openness, the level of corruption, the degree of civil liberty and status of market access to investors, all play a crucial role in the determination of the size of the domestic bond market. We also find differences across developed and emerging market samples. The results are robust to different specifications and the corresponding estimation techniques.

KEYWORDS:

• Emerging economies
• domestic bond market
• financial development
• social and institutional indicators
• panel estimation

JEL CLASSIFICATION:

• C33
• E44
• F21
• F33
• F34
• G15

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ See Obstfeld (2009) for a detailed survey on the benefits of reforms and globalization.

² See Shin (2000), Kim (2001), Shah and Thomas (2002), Park and Park (2003), Jiang and McCauley (2004) and Herring and Nathporn (2006), Khalid (2006) and Borensztein et al. (2008) for more details.

³ For more details, see Harwood (2000) and World Bank (2001).

⁴ Further readings on this issue include De la Torre, Gozzi and Schmukler (2008), Batten and Kim (2000), Cheung and Cham (2002), Eichengreen and Luengnaruemitchai (2004), Herring and Chatusripik (2000), Mihaljek, Scatigna and Agustin (2002), Mohanty (2002), Turner (2002) and Plummer and Click (2005).

⁵ This follows the empirical evidence provided in Burger and Warnock (2006b), Geanne and Guscina (2006) and Abbas and Christensen (2007). However, Mihaljek, Scatigna and Agustin (2002) found the relationship to be weak during the period 1995–2000. For instance, they found Korea and Malaysia to have large domestic debt markets along with low inflation (Yartey, 2008; Reddy, 2002).

⁶ See Berth et al. (2004) for more details.

⁷ Hsiao (1986) shows that the generalized least squares estimator for the random effects model, under an assumption of independence between the fixed effects and the explanatory variables, is biased.

⁸ Arellano and Bond (1991) and Judson and Owen (1999) provide more detail about one-step and two-step GMM estimators. Bond (2002) highlights that many applied work using these GMM estimators has focused on results for the one-step estimators rather than the two-step estimators. The simulation study by Arellano and Bond (1991) shows that a very modest gain in efficiency is observed using the two-step procedure, even in the presence of heteroscedasticity.

⁹ Also see Rose and Spiegel (2015).

¹⁰ The results are not the same when GMM is used.

¹¹ This is especially the case for Asian countries. See Mihaljek, Scatigna and Agustin (2002) for details. It may be worth mentioning that the number of countries switching to inflation-targeting as a monetary policy tools increased significantly since this publication. The above article notes that inflation-indexed bonds are prominent in Chile and Israel.

¹² Negative first-order serial correlation is expected in the first-differenced residuals if ${\epsilon }_{it}$ is serially uncorrelated.