Abstract
This paper examines the time varying nature of European government bond market integration by employing multivariate GARCH models. We state that unlike other bond markets, in euro markets the default(credit) risk factor and other macroeconomic and fiscal indicators are not able to explain the sovereign bond yields after the beginning of monetary union. This fact might be counted as a signal for perfect financial integration. However, we also find that the global shocks affect Germany and the rest of euro bond markets in various levels, creating particular discrepancies in asset prices even we take into account the market specific factors. Different level responses of each euro market to the global shocks reveal that euro bond markets are not fully integrated with each other unlike the recent literature claimed. Besides, we explore that the global factors are effective for the volatility of yield differentials among euro government bonds.
Similar content being viewed by others
Notes
“Higher risk” members are euro members that are in relatively vulnerable fiscal positions and have higher current account deficits. When we check Fig. 1 it is clearly observed the diversity on the current account positions of EMU members. Besides, the “high risk” definition refers to either higher default risk of the euro members or poor credit rated members.
It is observed from historical data that Portugal, Spain, Italy, and Greece government bonds have higher default risk premia to attract cross border portfolio holders.
However, an important argument used by the Stability and Growth Pact opponents is that governments were not sufficiently forced to dispose of extreme government service debt and deficits. Since the beginning of monetary union, government bond yields have not effectively shown the various degrees of default (credit) risk associated with the sovereign debt issued by euro-zone central governments. It is observed that since the beginning of the common currency, there is – surprisingly – only a modest difference in the risk premia for euro-denominated central government bond yields, despite the fact that the total debt of each country differs enormously, ranging from around 30% for Ireland to over 100% for Italy, Belgium and Greece. Remaining euro countries are gathered around the Maastricht “ceiling,” especially after 1999. Figure 3 illustrates the relationship between the average bond yield differentials for each market benchmark bond relative to the German benchmark government bond. Although we can see clear relationship between the ratings and yields, there are significant exceptions.
Similar results could be extracted from the credit ratings announced by the Standard and Poor’s and MTS groups in the last five years. Fiscally vulnerable countries are expected to bear lower credit ratings, and they would have higher yield outlays, but the results are not as expected.
Not only do the government debt obligations have credible differences, but also the current account risks are implicitly “bailed out” in the markets, though the head of central bank argued that it is not convincing to think European Central Bank (ECB) will guarantee each government’s debt service obligations.
Even though the dataset starts from 1999, for some of the markets, MTS does not contain bid and asking price data before 2001, the rest of the missing data is obtained from the Bloomberg Cooperation.
For more info about the dataset check http://www.mtsgroup.org/newcontent/data/.
Corporate bond spreads are calculated by subtracting the corporate bond index from the benchmark government bond yield.
The economic indicators used in this paper are listed in Appendix.
When the announcements are released in U.S., the Euro markets are generally closed. Therefore, we utilize the effect of announcements in euro markets one business day after they were released.
In the model, we employed some number of lag variables of the announcements to find the best model; however, by using the general to specific model—known as GS model—we ended up that the lag variables of the announcements are strongly insignificant, thus, we decided not to use the lag variables of the announcements. To neglect the asymmetric shock effect and make it simpler, we got the absolute values of the macroeconomic announcements.
Table 9 provides the stationary test results of the dependent variables. ADF test results documented that not for all yield differential variables are stationary.
Similar stationary test has been performed for the control variables as well. We performed that the control variables are stationary.
Since the government bond yield is the risk free return rate and the AAA corporate bond yield index contains the market risk excluding the specific factors. The difference surrogates for the U.S. market risk factor.
Codogno et al. (2003) have employed a similar dataset and found that for almost all euro countries, U.S. market risk has significant and positive coefficient.
We used the latter variable, the spread between U.S. swap rates and U.S. 10 year government bond yields, and found similar results.
The motivation of this paper is to extract the effect of global shocks on euro markets, and modeling the yield differentials between U.S. and Germany, and not being affected by any other member country are really strong assumptions. Therefore in Eqs. 5.a and 5.c the spillover coefficients are not zero, but expectedly these coefficients are statistically insignificant.
The coefficients μ iu and μ iu are not presented in the tables. The coefficients are infinitesimal and statistically insignificant. These results indicate that the yield differentials between Germany and US benchmark government bonds are not affected by the pairwise yield spreads within any euro market.
The swap interest rates are deducted from the bond yields for Germany and non-euro Bond market yields in order to eliminate the exchange rate fluctuations.
Euro bond bias refers to the euro share of international bond portfolios in the total volume of the international portfolio.
It is assumed that the effect of the announcements on the market will be simultaneously. The announcements are generally made before 11.00 am. The announcements realized in US will be effective on the following business day in euro markets.
References
Aggarwal R, Lucey B, Muckley C (2004) Dynamics of equity market integration in Europe. Evidence of changes over time and with events. Discussion Paper Series:019, The Institute for International Integration Studies
Barr DG, Priestley R (2004) Expected returns, risk and the integration of international bond markets. J Int Money Financ 23:71–97
Beckaert G, Harvey CR, Ng A (2005) Market integration and contagion. J Bus 78:39–69
Berndt ER, Hall BH, Hall RE, Hausman JA (1974) Estimation and inference in nonlinear structural models. Ann Econ Soc Meas 3:653–666
Blanco R (2001) The Euro-area government securities markets: recent developments and implications for market functioning. Working paper 0120, Banco de España, Servicio de Estudios
Bollerslev T (1986) Generalized autoregressive conditional heteroskedasticity. J Econ 31(5):307–327
Bomfim AN (2003) Pre-announcement effects, news effects, and volatility: monetary policy and the stock market. J Bank Finance 27:133–151
Christiansen, C (2007) Volatility-spillover effects in European bond markets. Eur Financ Manag 13:923–948
Codogno L, Favero C, Missale A (2003) EMU and government bond spreads. Econ Policy 18:503–532
Driessen J, Melenberg B, Nijman T (2003) Common factors in international bond returns. Discussion Paper, Tilburg University, Center for Economic Research
Dungey M, Martin VL, Pagan AP (2000) A multivariate latent factor decomposition of international bond yield spreads. J Appl Econ 15:697–715
Engle RF (1982) Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Econometrica 50:987–1006
Engle RF, Li L (1998) Macroeconomic announcements and volatility of treasury futures. Working paper, University of California San Diego
Erb CB, Harvey CR, Viscanta TE (2004) Understanding emerging market bonds. Emerg Mark Q 4:7–23
Favero CA, Pagano M, Von Thadden E-L (2005) Valuation, liquidity and risk in government bond markets. Working paper no:281, IGIER
Flannery MJ, Protopapadakis AA (2002) Macroeconomic factors do influence aggregate stock returns. Rev Financ Stud 15:751–782
Folkerts-Landau D, Mathieson DJ, Schinasi G (1997) International capital markets developments, prospects, and key policy issues. Working paper, IMF Staff, November 1997
Fratzscher M (2002) Financial market integration in Europe: on the effects of EMU on stock markets. Int J Financ Econ 7(3):165–193
Geyer A, Kossmeier S, Pinchler S (2004) Measuring systematic risk in EMU government yield spreads. Rev Financ 8(2):171–197
Kim SJ, Moshirian F, Wu E (2006) Evolution of international stock and bond market integration: influence of the European monetary union. J Bank Financ 30(5):1507–1534
Pagano M, Von-Thadden EL (2004) The European bond markets under EMU. Oxf Rev Econ Policy 20:531–554
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Macroeconomic announcementsFootnote 22 used in paper, are listed as,
United States: From the large range of U.S. economic announcements, we employ only seven, namely changes in non-farm payrolls (USNFP), NAPM (USNAPM), CPI (USCPI), PPI (USPPI), unemployment (USUNEMP), hourly earnings (USHRLYE), industrial production (USINDP), trade in goods and services (USTRDGS), final gross domestic product (USGDPF), housing starts (USHSES), and U.S. retail sales (USRSL). We have followed Fleming and Remolona (1999) and selected these indicators based on their paper. However, only four of them are statistically significant and illustrated in the tables (Tables 9, 10, and 11).
Germany: Unemployment Rate Producer Price Index, Consumption Price Index
Euro Area: Trade Balance, Money Supply (M3), Consumer Price index (Harmonized), Producer Price Index, Unemployment Rate
Rights and permissions
About this article
Cite this article
Balli, F. Spillover effects on government bond yields in euro zone. Does full financial integration exist in European government bond markets?. J Econ Finance 33, 331–363 (2009). https://doi.org/10.1007/s12197-008-9029-3
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12197-008-9029-3