EconPapers    
Economics at your fingertips  
 

Financialisation of the commodity markets. Conclusions from the VARX DCC GARCH

Karol Szafranek ()

No 8554, EcoMod2015 from EcoMod

Abstract: The global economy is highly dependent on the commodity prices, which are, by and large, the outcome of the market specific supply and demand fundamentals. As a result, driven by different determinants, financial assets and commodity prices should be negligibly correlated. However, systematically growing engagement of the noncommercial investors equipped with financial engineering innovations, generous inflow of capital resulting from the necessity for wider diversification of investment portfolios combined with the strengthening influence of purely financial and speculative motives have recently led to much stronger relation between the financial and commodity prices sparking a heated debate on the commodity markets financialisation. The main objective of this paper is to show that in the 2000’s commodity markets were under heavier influence of several macroeconomic, financial and speculative determinants, which resulted in the process of commodity markets financialisation. Secondary aim is to demonstrate the rising interdependence between the financial and commodity markets peaking during the outburst of the financial crisis as well as provide a proof of the changing strength of this connection in time and presumably the diminishing interplay between the markets after 2010. I use the daily time series starting from 1 I 2000 and ending at 31 XI 2013. Macroeconomic data with low frequency (monthly or weekly) are disaggregated and interpolated to daily data using the Denton-Cholette method. Dependent variables are the MSCI World Index (instrument for financial market) and five commodity indexes: S&P Goldman Sachs Energy Index, S&P Goldman Sachs Agriculture Index, S&P Goldman Sachs Industrial Metals Index, S&P Goldman Sachs Livestock Index and S&P Goldman Sachs Precious Metals Index (proxies for commodity sectors). Exogenous variables include: macroeconomic situation (composite leading indicators for the US, China and the world PMI), monetary policy (LIBOR USD and the yield of the US 10 year government bonds), foreign exchange market (carry trade index), investors sentiment, global risk aversion, idiosyncratic risk (measures as a sector implied volatility), global commodity demand (reflected in Baltic Dry Index), the demand from EMEs (year-on-year import from China), production costs, weather conditions and speculative behavior (aggregated volume and open interest in commodity sectors). With the use of a VAR model with imposed zero-restrictions on the matrix of exogenous variables I identify the factors influencing the price dynamics on the commodity markets. Then I use the DCC GARCH with asymmetry term and multivariate t distribution to show the changing market interdependence between the equities and commodities. The VARX lag is chosen by BIC minimization. Starting parameters for the DCC-GARCH are determined by estimated univariate GARCH models (optimal univariate GARCH models are chosen by minimizing BIC for GARCH(1,1) with ARIMA(1,0,1), a constant, no in-Mean effect witch changing error distribution and GARCH type). Two-step estimation is used: firstly fit a GARCH-Normal model to the univariate data and then proceed to estimate the second step based on the chosen multivariate distribution. I estimate 12 DCC(1,1) models allowing for the presence of the VARX component, the asymmetry of the DCC as well as different error distribution. I compute also 95% bootstrap confidence intervals for the CCC-GARCH in a sample of 10 000 simulations. All estimations are calculated in the periods of time: 2000-2004, 2005-201 and 2000-2013. For the check of parameter stability I use a recursive estimation in expanding window since the beginning of 2005. To evaluate the models I conduct also statistical tests. The empirical analysis supports the claim that since 2005 commodity markets have been under heavier influence of macroeconomic, financial and speculative determinants. The process, however, loses on strength since 2011. Results of the restricted VARX ADCC GARCH with multivariate t distribution of errors demonstrate, that the inclusion of the commodity markets growing sensitivity to macroeconomic conditions, financial markets turmoil and the impact of behavioral aspects alters the dynamic correlation between commodities and the financial markets markedly from 2005 to 2011 signaling the process of finacialisation. The bootstrap confidence intervals analysis shows that until 2005 commodity and financial markets showed no statistically significant correlations, whereas after 2005 the interdependence between them was much higher. The model is insensitive to changes in specifications (allowing for different error distribution as well as the inclusion of asymmetry does not have change the results considerably).

Keywords: Commodity sectors (energy; agriculture; industrial metal; livestock; precious metals) and global financial market.; Finance; Miscellaneous (search for similar items in EconPapers)
Date: 2015-07-01
References: View references in EconPapers View complete reference list from CitEc
Citations: Track citations by RSS feed

Downloads: (external link)
http://ecomod.net/system/files/Financialisation%20 ... modity%20markets.pdf

Related works:
Working Paper: Financialisation of the commodity markets. Conclusions from the VARX DCC GARCH (2015) Downloads
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ekd:008007:8554

Access Statistics for this paper

More papers in EcoMod2015 from EcoMod Contact information at EDIRC.
Bibliographic data for series maintained by Theresa Leary ().

 
Page updated 2019-07-08
Handle: RePEc:ekd:008007:8554