Abstract
In this report, two different articles associated with the topic of commodity financialization were chosen for exploration. The advantages and disadvantages of each will be pointed out. The first article to be analyzed is Silvennoinen and Thorp’s (2013) “Financialization, crisis and commodity correlation dynamics,” and the second one is “Financialization and de-financialization of commodity futures: A quantile regression approach” by Bianchi, Hua Fan and Todorova (2020). Both articles are peer-reviewed and available through the Elsevier database. Because the Saudi Arabia market can explode into turbulence, the studies analyzed in the paper point to the need to apply simple approaches to portfolio diversification among commodities.
Introduction
The topic of the current research is the Risk Management and Portfolio Optimization for Building Portfolios with Commodity Futures in Saudi Arabia. It focuses on the construction of optimal portfolios by enabling efficient diversification by using commodities. The problem that prompted the research pertains to the appointment of various assets for gaining maximum returns and minimum risks at an appropriate time. This has generated the issue of multi-objective portfolio optimization, which should be solved in the nearest future. One of the terms used for explaining the performance of commodities is their financialization. The financialization view is that the increased trading within commodity futures markets is associated with the boost in the growth rate and volatility of commodity spot prices (Chari and Christiano, 2017). This view gained recognition because, in the 2000s, the volumes of trading increased significantly, leading with many commodity prices to rise and become more volatile (Watugala, 2015). The issue was selected for analysis to advance the exploration of how portfolio management can accommodate the paradox linked to balancing the need for establishing control and stability to establish favorable conditions within portfolio risk management. Previously, several research studies have aimed at addressing the problem.
Critique
In their article, Silvennoinen and Thorp (2013) searched for evidence of closer integration between a conventional asset and commodity feature returns to test the hypothesis that such a connection is not influenced by financialization. The authors suggested that a stronger interest of investors in commodities would create a closer integration with regular asset markets. They estimated that both gradual and sudden shifts would occur in the correlation between stocks, bonds, and commodity futures returns that have been initiated by changes in financial variables that could be observed over time. The researchers focused on the market connection between commodity and equity and the identification of indicators of financial market conditions that drive the dynamics of such a correlation.
The data included in the analyses comprised of Wednesday-Wednesday log returns to future contracts on twenty-four commodities between May 1990 and July 2009 (Silvennoinen and Thorp, 2013). The commodities are related to such categories as agriculture, metals, energy, and financials. In the agriculture category, commodities included “corn, soybeans, soybean oil, wheat, live cattle, pork bellies, coffee, cotton, orange juice, and sugar” (Silvennoinen and Thorp, 2013, p. 3). In the metals category, commodities included gold, platinum, aluminum, silver, copper, nickel, lead, tin, and zinc. In the energy category, commodities included Brent oil, crude oil, heating oil, and natural gas. Finally, the financial commodities included short rate, yield spread, USA stocks, German stocks, France stocks, USA bonds, volatility, and USA exchange rate index.
According to the findings of the article, most of commodity features returns series point to the lower return/risk ratios than stocks, and in some cases, mean returns are negative. Academic studies usually view ‘non-commercial’ traders as financial investors because the category includes mainly money managers, hedge funds, or speculators. The non-commercial sub-category contains primarily hedge funds, which are represented by managers pulling funds from less significant investors and taking either short or long positions in the markets of futures. The prevalence of long open interests aligns in growth with speculators that accept risks from commodity suppliers as well as institutional investors that take long positions in commodity futures directly as a means of alternative asset class or through commodity indexes (Silvennoinen and Thorp, 2013).
The authors suggested that most conditional relationships between the returns of commodity futures the indexes of US stock returns increased. Financial shocks to the market showed to possess a dramatic influence on predicting the dynamics of the correlation. For half of the equities analyzed, the expected high levels of the volatility of the stock market increased correlations with equities (Silvennoinen and Thorp, 2013). It was surprising to reveal that the pattern of correlations change for oil futures and stocks took place during the global financial crisis (GFC) that took place in 2008 (Silvennoinen and Thorp, 2013). The concentration of such an effect later in the data sample points to the rising price of the commodity and the gradual integration of the stock market. Besides, correlation structure breaks occur for the majority of metals, some foods, and some grains around the beginning of 2010s, during which both financial investor interests and fundamentals were increasing their power. It seemed that the evidence of both early and gradual increases within conditional correlations was the most apparent for base meals and pre-dated the most impactful financialization period of 2003 and 2004 (Silvennoinen and Thorp, 2013).
Compared to other examinations of returns of commodity futures available before 2013, the results of Silvennoinen and Thorp (2013) do not show a decrease in power relationship among bond returns and conventional stock as well as commodities. The research is valuable for presenting quantitative data that favored closer commodity and financial market integration. The authors of the article put forth several innovations associated with the modeling of correlation dynamics that can somewhat help explain the differences in the results of their study and previously published research. For example, they extended the sample to include data from the second part of 2008 and the first part of 2009, which allowed to include a significant amount of new data variations (Silvennoinen and Thorp, 2013). Besides, the researchers included detailed modeling of both common and idiosyncratic factors in the variances and means to include relevant predictions regarding currency fluctuations as well as seasonal changes in means.
The introduction of the double smooth transitional correlation (DSTCC-GARCH) models allowed the researchers to identify possible changes in the regime of correlation occurring over time, influenced by observable transition variables. This enabled to illustrate the existence of structural breaks within the process of correlation that are not included in the DSTCC-GARCH models. In addition, it was possible to offer evidence regarding the relevance of the variable of financial transition that had not been included in the correlation studies available up to that point.
The study by Bianchi et al. (2020) employed the approach of quantile regression to study the financialization of commodity futures. The significant level of growth within the sphere of commodity training has encouraged a new strand of research aimed to understand the magnitude and the nature of the financialization of the commodity market. The researchers set the goal of understanding the dynamics of the market by exploring both the degree and structure of the dependence between global stocks and commodity futures. Through using a quantile regression approach, the scholars aimed to measure the impact of financialization as well as possible de-financialization of commodity futures (Bianchi et al., 2020). Quantile regression is a category of regression analysis within which the method of least squares helps estimate the conditional mean of the response variable within the values of the predictor variable (Le Cook and Manning, 2013). As an extension of linear regression, the quantile regression approach is introduced in data analysis when the conditions of linear regression are not being met. Compared to Silvennoinen and Thorp’s (2013) research that included 24 commodities, Bianchi et al. (2020, p. 3) included half of the list. Namely, the twelve commodity futures markets included “Western Texas Intermediate crude oil, heating oil, natural gas, gold, silver, aluminum, copper, soybeans, wheat, corn, coffee, and cattle.” Even though the number of the commodities included in the analysis is lower, it is enough for offering a wide-range coverage of commodity markets by investigating commodities in the energy, precious metals, base metals, and agricultural categories. Importantly, Bianchi et al. (2020) did not include the financial type of market commodities in their analysis.
For the pre-financialization period marked between 1991 and 2003, it was found that the estimated coefficients were insignificant for all specifications for energy-related commodities (Bianchi et al., 2020). Gold showed both significant and negative sensitivities related to equity market developments, originating from intermediate quantiles, as found in the analysis. On the other hand, however, such categories as industrial metals, commodities related to agriculture, cattle, and coffee all showed a positive association with the stock market. Copper was identified as an exception to the rule as it demonstrated to have an overall even relationship. In addition, the base-case coefficients related to copper are the most significant across all commodities under the consideration of researchers, followed by aluminum. These findings are overall consistent with the notion that the market of industrial metals is generally associated with the state of the economy of a country. In general, higher coefficients were observed at higher quantiles; therefore, commodity futures show to be more significantly correlated to equity markets at time points that are characterized by high returns compared to other days. For the financialization period marked 2004-2013 was considered next, with the researchers finding potential spreading of trends from stock markets to commodity markets during the period.
Since both Silvennoinen and Thorp (2013) and Bianchi et al. (2020) covered the GFC of 2008, it is essential to mention the findings of the latter as acquired through quantile regression. For example, energy-associated commodities were found to exhibit an increase in the average coefficient of the interaction term across quantiles around 0.2 (Bianchi et al., 2020). Commodities in the agricultural category, such as livestock, also exhibited positive average effects even though they were lower in their magnitude. The highest observable results were found for wheat, which showed an average increase of 0.306 across quantiles (Bianchi et al., 2020). Such figures are evidence of the contagion effect, which suggests that the turbulent effects in the stock market are being transferred to the commodity futures market. The contagious quality implies adverse effects, which means that the transmission of turbulent impacts is more likely to increase in lower quantiles. The identification of contagion is important to the study because it is usually related to falling prices of stocks and negative returns, so it is more effectively characterized by an increase in the lower quantile dependence than its upper quantile dependence. Importantly, both precious and base metals paint a different picture within the GFC. The coefficient average for all markets in the category is low, which is indicative of the metals decoupling from the stock market.
Finally, in the de-financialization period of 2014 to 2017, energy commodities exhibit consistency in terms of their positive marginal sensitivities, with the strongest indicator reaching 1.796 for crude oil, thus making it possible to track a change in the dependence degree compared to the previous period (Bianchi et al., 2020). The marginal sensitivities of crude oil in the last years of the sample exceed the indicators during the financialization period of 2004 to 2017 by 0.6 and 1.33 for the 5% and 95% quantiles, respectively, and by 0.201 at its median (Bianchi et al., 2020). Again, a different picture emerges for precious and industrial metals, for which the post-2013 marginal effects were generally lower compared to the decade prior. Gold reports negative coefficients across all quantiles, while other metals were positive. In the de-financialization period, the remaining commodities show a sustained positive relationship to the equity market, although the picture is more mixed compared to the 2004-2013 period, with the average coefficient across quantile serving as an indicator of a potential dependence change, documenting a strengthening for such commodities as cattle and a weakening in the co-movement of wheat, corn, and coffee within the equity market.
Discussion
The study by Bianchi et al. (2020) examines the effects of commodity market financialization since the beginning of the twenty-first century. Previous literature, including the article by Silvennoinen and Thorp (2013), focuses on volatility and correlation for identifying and measuring the level of dependence between stocks and commodities. Therefore, the study by Silvennoinen and Thorp (2013) was limited because the mentioned conventional and time-varying measures were restrictive and were only useful for measuring the form of volatility and correlation. Compared to previous research, the article by Bianchi et al. (2020) used a quantile regression approach for examining the initial points of both futures of commodity and international stocks. Importantly, the approach taken by the scholars in the more recent study provided a broad informational context regarding the twelve commodities and the dependence between them and stocks along with the entire distribution of returns. This allowed measuring both the degree of dependence and the structure of dependence over the past two and a half decades. The scholars hypothesized that the possible decoupling of markets of the commodity from equity markets after 2013 led to the lower dependence degree to the 2004-2013 period and a change in the structure of such dependence.
The findings of Bianchi et al. (2020) are more relevant to the topic of the proposal regarding risk management insights and portfolio optimization for building portfolios with commodity futures in Saudi Arabia because the scholars found a distinct effect of financialization. Such a degree of dependence showed to be more robust in terms of commodities in the energy category, somewhat less significant for metals, and significant but with the lowest impact for the remaining number of commodities. The problem identified in the proposal is associated with the issue of multi-objective portfolio development, and there is the need for gaining control and stability for enabling favorable conditions in portfolio risk management.
Bianchi et al. (2020) suggested that the findings were consistent with the notion that energy markets play the most significant part in the inputs of production in the global economy, then followed by metals and agricultural markets, which is important in the context of Saudi Arabia, which is a market relying heavily on the energy sector (Shabaneh and Schenckery, 2020). Besides, co-movements within equity markets in the 2014-2017 period continued to increase for energy-related commodities and decreased significantly for metals, which is essential to consider with the development of a multi-objective portfolio and considering the risk management process.
Conclusion
Therefore, it is essential to allow for not only the degree of dependence changing but also the shifts in the structure of such a dependence. The empirical findings of both articles analyzed in the paper show that aggregate dependence between commodity markets was not the same across a range of quantiles. The effect of financialization appeared to be more dominant and impactful than imagined initially, with the reduced ability of some commodities to diversify the portfolio of equities. Because the Saudi Arabia market can explode into turbulence, the studies analyzed in the paper point to the need to apply simple approaches to portfolio diversification among commodities.
Reference List
Bianchi, R., Hua Fan, J. and Todorova, N. (2020) ‘Financialization and de-financialization of commodity futures: a quantile regression approach’, International Review of Financial Analysis, 68, pp. 1-15.
Chari, V. and Christiano, L. (2017) Financialization in commodity markets. Web.
Le Cook, B. and Manning, W. (2013) ‘Thinking beyond the mean: a practical guide for using quantile regression methods for health services research’, Shanghai Archives of Psychiatry, 25(1), pp. 55-59.
Shabaneh, R. and Schenckery, M. (2020) ‘Assessing energy policy instruments: LNG imports into Saudi Arabia’, Energy Policy, 137, pp. 1-12.
Silvennoinen, A. and Thorp, S. (2013) ‘Financialization, crisis and commodity correlation dynamics’, Journal of International Financial Markets, Institutions & Money, 23, pp. 42-65.
Watugala, S. (2015) Economic uncertainty and commodity futures volatility.