Our baseline specification used a very simple definition of crisis periods. We assumed that all countries in the sample faced turbulence between 2009-2013, and all other periods were tranquil. As a robustness check, we now use the crisis timing in Laeven
2The difference in signs comes from the fact that indebtedness means a negative NFA and a positive gross debt position.
Table 7: State dependence
(11) (12) (13) (14)
Lagged NFA/GDP -0.0123*** -0.0107*** -0.00837*** -0.0387***
[0.00310] [0.00313] [0.00319] [0.00692]
Infl. diff. -0.274*** -0.224*** -0.208** -0.183**
[0.0819] [0.0834] [0.0837] [0.0823]
Relative GDP -6.479*** -5.588*** -5.916*** -5.603***
[1.576] [1.599] [1.591] [1.573]
NEER volatility 11.48** 11.49** 9.912** 12.06**
[5.035] [5.007] [4.987] [4.924]
Current acc. -0.0100 -0.00469 -0.00150 0.000499
[0.0177] [0.0177] [0.0178] [0.0178]
Reserves -0.00891 -0.00800 -0.00892 -0.00177
[0.00908] [0.00903] [0.00898] [0.00896]
Budget bal. -0.0718** -0.0596** -0.0569** -0.0605**
[0.0290] [0.0292] [0.0290] [0.0290]
NFA x Crisis -0.00644*** -0.00869*** -0.0116**
[0.00230] [0.00244] [0.00541]
NFA squared 1.11e-05
[1.47e-05]
NFA squared x Crisis 6.35e-05***
[1.90e-05]
NFA x Rel. GDP 0.0272***
[0.00574]
NFA x Crisis x Rel. GDP 0.00608
[0.00464]
Constant 5.265*** 4.733*** 4.989*** 4.232***
[1.001] [1.014] [1.008] [1.008]
Country FE yes yes yes yes
Year FE yes yes yes yes
Observations 883 684 684 684
R squared 0.036 0.140 0.272 0.281
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01
Country Years Country Years
Angola 2015 Luxembourg 2008-2012
Austria 2008-2012 Malaysia 1997-1999
Belgium 2008-2012 Moldova 2014-2017
Brazil 2015 Myanmar 2012
Cyprus 2011-2015 Nepal 1984, 1988. 199
Czech Republic 2000 Netherlands 2008-2009
Denmark 2008-2009 New Zealand 1984
Ethiopia 1993 Norway 1991-1993
Finland 1991-1995 Philippines 1997-2001
France 2008-2009 Portugal 1983, 2008-2012
Germany 2008-2009 Russia 2000, 2008-2009, 2014
Ghana 2009, 2014 Slovakia 2000-2002, 2008-2012
Greece 2008-2012 South Africa 1984-1985, 1993, 2015
Honduras 1990, 1992 Spain 1980-1981, 1983,
2008-2012
Hungary 2008-2012 Sweden 1991-1995, 2008-2009
Iceland 2008-2012 Switzerland 2008-2009
Ireland 2008-2012 Thailand 1999-2000
Italy 1981, 2008-2009 Trinidad and Tobago 1986, 1989 Jamaica 1983, 1990-1991,
Uganda 1980-1981,
1996-1998 1988, 1993
Japan 1997-2001 United Kingdom 2007-2011
Korea 1997-1998 United States 1988, 2007-2011
Latvia 2008-2012 Venezuela 2002, 2010, 2017
Table 8: Crisis events from Laeven and Valencia (2018)
and Valencia (2018), and code a country-year cell a crisis event according to their classification. A crisis event for a country occurs if there was a banking, currency, or sovereign debt crisis as in Laeven and Valencia (2018). Alternative crisis definitions are available in Eichengreen and Gupta (2018) or Cavallo, Powell, Pedemonte and Tavella (2015). We work with the classification of Laeven and Valencia because of its comprehensiveness.
Table 8 lists the various crisis events from Leaven and Valencia (2018) that overlap with our baseline sample. The majority of the country-years observations overlap with the global financial crisis, although the timing is somewhat different from our previous assumption. There are additional crisis events such as the East Asian crisis starting in 1997, and other episodes that affected individual countries.
Table 9 presents the results with this alternative definition of crisis events. We omit the basic regression without state-dependence or nonlinearity, since it is inde-pendent of how we define crises. The results are again very similar to the baseline.
The NFA coefficient is much larger during turbulent years, now defined at the coun-try level. There is evidence for significant nonlinearity, but only in crisis years. State dependence is detected conditional on the level of development, and it becomes even stronger during turbulent times. Overall, we conclude that while the precise coeffi-cients depend on the crisis definition, there is very strong evidence for the kinds of state dependence we were looking for.
5 Conclusion
The paper studied the relationship between measures of indebtedness and the inter-est premium on government bonds. In particular, the main quinter-estion was whether such a relationship is dependent on time, the state of economy, and the types of coun-tries studied. The answer is yes to all three questions. Whether we look at tranquil of turbulent periods, and the relative development of the countries, all influence the magnitude and significance of the debt-premium relationship.
The estimated elasticity is in line with both previous empirical work and es-timates from DSGE models. Linear models, however, have to be calibrated such that they take into account the type of the country (emerging or advanced) they model. When the time period under study includes the global financial crisis (or other important global events), regime switching models might need to be used.
There are empirical problems that arise mostly from the fact that data is patchy.
Ideally, one would like to use debt instruments denominated in the same currency.
Unfortunately widespread interest rate data is not available for such instruments.
Moreover, selection of both entry to international financial markets and the matu-rity and denomination of debt may not be random. Nevertheless, we think that
Table 9: State dependence
(15) (16) (17)
Lagged NFA/GDP -0.00554** -0.00522** -0.0197***
[0.00240] [0.00242] [0.00553]
Infl. diff. 0.346*** 0.369*** 0.384***
[0.0610] [0.0606] [0.0596]
Relative GDP -5.395*** -5.173*** -4.862***
[1.188] [1.176] [1.155]
NEER volatility 13.78*** 14.74*** 13.69***
[4.061] [4.025] [3.940]
Current acc. 0.0553*** 0.0588*** 0.0549***
[0.0142] [0.0142] [0.0138]
Reserves -0.00168 0.00184 0.00457
[0.00794] [0.00789] [0.00787]
Budget bal. -0.0557*** -0.0547*** -0.0837***
[0.0207] [0.0204] [0.0204]
NFA x Crisis -0.0207*** -0.00896** -0.0687***
[0.00336] [0.00419] [0.00776]
NFA squared 9.02e-06
[1.33e-05]
NFA squared x Crisis 0.000210***
[4.63e-05]
NFA x Rel. GDP 0.0130***
[0.00475]
NFA x Crisis x Rel. GDP 0.0699***
[0.00989]
Constant 4.332*** 4.043*** 3.400***
[1.069] [1.059] [1.052]
Country FE yes yes yes
Year FE yes yes yes
Observations 1,013 1,013 1,013
R squared 0.359 0.374 0.398
Standard errors in parentheses
* p<0.10, ** p<0.05, *** p<0.01
our study provides useful findings to understand the complex interactions between indebtedness and the risk appetite of international financial markets.
References
[1] Agca, S. and O. Celasun (2012). “Sovereign debt and corporate borrowing costs in emerging markets.”Journal of International Economics, 88: 198-208.
[2] Brzoza-Brzezina, M. and J. Kotlowski (2016). “The nonlinear nature of country risk and its implications for DSGE models.” NBP Working Papers 250, National Bank of Poland.
[3] Blagov, B. (2018). “Financial crises and time-varying risk premia in a small open economy: a Markov-switching DSGE model for Estonia.”Empirical Economics, 54: 1017–1060.
[4] Bencz´ur, P., and I. K´onya (2016). “Interest Premium, Sudden Stop, and Adjust-ment in a Small Open Economy.”Eastern European Economics, 54: 271-295.
[5] Cavallo, E., A. Powell, M. Pedemonte and P. Tavella (2015). “A new taxonomy of Sudden Stops: Which Sudden Stops should countries be most concerned about?” Journal of International Money and Finance, 51: 47-70.
[6] Chinn, M. and S. Quayyum (2012). “Long Horizon Uncovered Interest Parity Re-Assessed.” NBER Working Papers 18482.
[7] Comelli, F. (2012). “Emerging Market Sovereign Bond Spreads; Estimation and Back-testing.” IMF Working Papers 12/212, International Monetary Fund.
[8] Csont´o, B. (2014). “Emerging market sovereign bond spreads and shifts in global market sentiment.”Emerging Markets Review, 20: 58-74.
[9] Dell’Erba, S., R. Hausmann and U. Panizza (2013). “Debt levels, debt com-position, and sovereign spreads in emerging and advanced economies.”Oxford Review of Economic Policy, 29: 518-547.
[10] Eichengreen, B. and P. Gupta (2018). “Managing Sudden Stops.” In: E. Men-doza, E. Past´en and D. Saravia (ed.), Monetary Policy and Global Spillovers:
Mechanisms, Effects and Policy Measures, ed. 1, vol. 25, ch. 2, p. 009-047, Central Bank of Chile.
[11] Garcia-Cicco, J., R. Pancrazi, and M. Uribe (2010). “Real Business Cycles in Emerging Countries?”American Economic Review 100: 2510–31.
[12] Jakab, Z. and I. K´onya (2016). “An Open Economy DSGE Model with Search-and-Matching Frictions: The Case of Hungary.”Emerging Markets Finance and Trade, 52: 1606-1626.
[13] Lane, P. and G. Milesi-Ferretti (2018). “The External Wealth of Nations Re-visited: International Financial Integration in the Aftermath of the Global Financial Crisis.”IMF Economic Review, 66: 189-222.
[14] Laeven, L. and F. Valencia (2018). “Systemic Banking Crises Revisited.” IMF Working Papers 18/206, International Monetary Fund.
[15] Passari, E. and H. Rey (2015). “Financial Flows and the International Monetary System.”The Economic Journal, 125: 675-698.
[16] Rey, H. (2013). “Dilemma not Trilemma: The Global Financial Cycle and Mon-etary Policy Independence.” Proceedings - Economic Policy Symposium - Jack-son Hole, Federal Reserve of Kansas City Economic Symposium, p.285-333.
[17] Schmitt-Grohe, S. and M. Uribe (2003). “Closing small open economy models.”
Journal of International Economics,61: 163–185.