Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Financial Fertility Among Hungary And Other Visegrad Countries: Co-Movement And Causality In Stock And Bond Market Return And Volatility

Summary

Bond markets linkage can have an impact on the yield curve, exchange rate therefore the co-movement can affect the monetary policy, the transmission mechanisms and so the financial stability.

The stock market returns have strong interdependence (positive dependence)on short term at 1 day to 5 days frequency just like stock market volatility and it is constant after the crisis.

In bond markets, we detected price return interdependency before the first quarter of 2017, which is likely to be caused by the self-financing program (by the NBH) on long term.

As a result of the program the Hungarianmonetary policy became less transparent, while the domestic banking systemcaptured the state, however in the meantime certainly became more independenton a price level.

In spite of the independent price return on bondmarkets, we can still detect interdependent volatility within the market whichmeans volatility on international markets continues to affect the Hungarianmarket. Turbulency in foreign bond market causes turbulency in the Hungarianmarket as well so volatility can be identified as one of the certainfertilization channels.

Many economists concerned about the causal influence of one country performance on other countries. In this paper, I focus on the Visegrad countries especially on Hungary and examine whether the Czech, Polish, Slovakian financial markets affect the Hungarian economy. Contrary to the majority of literature, both stock and bond markets are examined from various aspects i.e. price and volatility.

The topic of price co-movement among the international stock markets has been analysed recently and became a widely discussed topic in international finance. The knowledge of financial markets linkage between countries’ financial markets is very important for the investors’ hedge strategy hence they would be able to build optimal portfolios which can fully reflect their risk appetite.

Bond market interdependency is less popular research topic in the financial literature however the global financial crisis pointed out the importance of global financial networks. To be aware of the linkage between bond markets i.e. sovereign government bond markets is very important for the policy makers including fiscal and monetary policy decision makers. Bond markets linkage can have an impact on the yield curve, exchange rate therefore the co-movement can affect the monetary policy, the transmission mechanisms and so the financial stability. The bond market interdependency is also very important for fiscal policy as it can affect the cost of financing and the risk of refunding (liquidity risk).

The analysis is mainly based on wavelet coherence method and copula methods. The co-movement of the analysed variables is done by wavelet method by transforming time series into frequencies. Chui (1992), and Strang and Nguyen (1996) are good introductions to wavelets. Gencay et al. (2002) discusses and illustrates how wavelets can be applied in economics and finance. Copulas have recently become a sophisticated modelling asset in many fields where multivariate dependence is of interest. In finance, these models are primarily used for asset pricing, credit scoring, risk modelling, and risk management (Bouye et al. 2000; Embrechts et al. 2003; etc).

Two kind of dataset were used for this paper. The dataset consists of daily stock and bond market data. Polish, Slovakian, Czech and Hungarian daily returns and volatility were consisted for both stock and bond market. Daily return means the logarithmic difference of the indices or price while the volatility is the difference of daily maximum and daily minimum price or index value from Reuters database.

Stock and long-term government bond market are different in structure as they are order and dealer driven market respectively they are different length time series. For bond markets, only the common and continuous periods were used for analysis.

The analysis was coded in R software package (R Core Team, 2016) which is a leading, open source software facility for data manipulation, calculation and graphical display. Table 1 shows the correlation matrix including the Pearson correlation coefficients related to stock market data in which the highlighted cells are the ones with significant values lower than 10%. As we can see, Hungary, Czech and Poland data show significant linear correlation while Slovakia stock indice seems to be lineary independent from the others. Returns and volatility have positive correlation with other countries’ respective data while returns have negative correlation with own volatility.

1. Table: Correlation matrix for stock market data

Hungary return

Czech return

Poland return

Slovakia return

Hungary volatility

Czech volatility

Poland volatility

Slovakia volatility

Hungary return

1

,501

,522

-,007

-,148

-,146

-,155

,014

Czech return

1

,553

,016

-,129

-,210

-,163

-,026

Poland return

1

-,018

-,170

-,142

-,250

-,024

Slovakia return

1

-,006

-,012

-,031

,045

Hungary volatility

1

,471

,465

,031

Czech volatility

1

,443

,061

Poland volatility

1

,014

Slovakia volatility

1

Source: author’s calculation

Bond market linear dependency was also checked with correlation for both price changes and daily volatility shown in table 2. We found out that Hungarian bond price movements is significantly correlated with the bond price movements in Poland and Slovakia. Hungarian volatility is also found to be significantly correlated with the Hungarian price movements and also with the other countries bond market volatilities.

2. Table: Correlation matrix for bond market data

Hungary return

Czech return

Poland return

Slovakia return

Hungary volatility

Czech volatility

Poland volatility

Slovakia volatility

Hungary return

1

,012

,232

,063

-,064

-,013

-,038

-,027

Czech return

1

,027

,039

,030

-,015

-,017

,034

Poland return

1

,103

-,052

-,015

-,101

-,067

Slovakia return

1

-,008

,005

-,008

-,013

Hungary volatility

1

,124

,404

,219

Czech volatility

1

,026

,076

Poland volatility

1

,290

Slovakia volatility

1

Source: author’s calculation

These findings confirm our previous assumption regarding the relevance of using an asymmetric model for obtaining marginal distribution.

In order to evaluate the lag-lead relationship and co-movement level between each stock market index on a longer horizon, the paper will utilize a wavelet analysis. This wavelet coherence approach is applied when we want to capture the interdependence through time and frequencies. The frequencies in this instance stand for the duration in days within which a movement in one variable affects the other variable through a specific time period. The dataset and wavelet coherence used in this paper are carried out solely in pairs with Hungary.

The horizontal axis stands for the time period, while the vertical axis for the frequency. The bar on the right side represents the strength of the dependence between two variables. Red shows that the dependency is strong, while deep blue signifies low dependency between the variable pairs. Correspondingly, the black thick line scattered around in the red area represents strong coherency at the 5% significance level with respect to that frequency and time period.

The stock market returns have strong interdependence on short term at 1 day to 5 days frequency just like stock market volatility. The interdependence between the data pairs seem to be constant over the examined period. Copula methods confirmed positive tail interdependency which means that massive changes in return and volatility can be reflected in other country’s markets while the low values have no fertilization effect. According to the lagged models, we cannot identify the dominant country within the pairs.

The bond market seems to be complicated than the stock market. Czech bond data seems to be non-continuous and not as liquid like the others. As it is illustrated on figure 1, Hungary seems to be co-moving with Poland and Slovakia in terms of bond prices until the beginning of 2017. After the first quarter of 2017, the Hungarian bond price movements have been getting independent from Poland and Slovakia. The price changes impact can be measured for up to 64 for days as it is reflected on the upper maps of following graph. The wavelet analysis of Hungary – Poland co-movement is illustrated on the left while Hungary – Slovakia is illustrated on the right graphs. The volatility is visualised on the bottom graphs. As we can see in spite of the independent price movement after 2017, significant volatility co-movement was detected in whole period.

Copula methods confirmed positive tail interdependency which means that massive changes in return and volatility can be reflected in other country’s markets while the low values have no fertilization effect. According to the lagged copula models, we can identify Slovakia’s bond market is dominating the Hungarian bond market as changes on the Slovakian bond market can have protracted impact on the Hungarian market.

Conclusion and final thoughts

As the result of our wavelet and copula approach, we can conclude the stock markets have been moving together with positive tail dependency but no dominant market was detected within the region under study. We can infer that stock markets are integrated to the global markets, therefore all of the countries within the region are following the same pattern and they react to global events instead regional ones. In case of bond markets, we detected price return interdependency after the first quarter of 2017, which is likely to be caused by the self-financing program (by the National Bank of Hungary) on long term bonds. The central bank’s program aimed to make the Hungarian bond market more or less independent from international financial turmoil. As a result of the program the Hungarian monetary policy became less transparent, while the domestic banking system captured the state, however in the meantime certainly became more independent on a price level. In spite of the independent price return on bond markets, we can still detect interdependent volatility within the market which means volatility on international markets continues to affect the Hungarian market. Turbulency in foreign bond market causes turbulency in the Hungarian market as well so volatility can be identified as one of the certain fertilization channels.

As a result of the symmetric copula tests we can conclude Slovakian market dominates the Hungarian market. It can be caused on the one hand, by the shared set of investors the countries of the CEE region, on the other by the euro denominated debts in Slovakia which allows it to enjoy the advantage of broader acceptance of its currency without FX risk to the investors. Hungary seems to be the bottleneck within the region, as Hungary is the only country offered ‘junk’ or not recommended sovereign debt by credit agencies. We can infer that the investors, who invest into Hungary also invest into the region, while the CEE investors might not invest into Hungary due to their risk appetite or investment policy. As a result of this phenomenon if something happens with the CEE investors it certainly affects the Hungarian bond market, while the Hungarian bond market investors are just a smaller portion for the CEE investors.

Bibliogrpahy

Chui, C. K. (1992). An Introduction to Wavelets. San Diago: Academic Press

Bouye, E., Durrleman, V., Bikeghbali, A., Riboulet, G., Rconcalli, T., (2000): Copulas for finance – A reading guide and some applications. Working paper, Goupe de Recherche Op´erationnelle, Credit Lyonnais.

Gencay, R., F. Selcuk and B. Whitcher (2002): An Introduction to Wavelets and Other Filtering Methods in Finance and Economics. London: Academic Press.

R Development Core Team (2016) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna

Strang, G. and T. Nguyen (1996): Wavelets and Filter Banks. Wellesley, MA: Wellesley–Cambridge Press.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.