Please enable JavaScript.

Coggle requires JavaScript to display documents.

Spillover & Connectedness: Commodity Prices (CP) - Coggle Diagram

- - - - Chen & Rogoff 2003: Pointed out that changes in WCPs are exogenous to small economies
      - Kilian 2009: Price of some commodities like crude oil are even exogenous to the US economy in short-term
      - Basher et al. 2012, 2016: Price of some commodities have important influences on EXR
      - Dangl & Halling 2012, Rapach et al. 2010: Forecast combination approaches have been widely employed in asset price forecasting studies (FAVG)
      - Verdelhan 2010: Proposes a habit model which advocates that domestic investor is more risk-averse than foreign counterparts when the economy is a recession so the author expects a positive currency excess return during bad times at home, creating the predictability of excess return
      - Lustig et al 2014: Found countercyclical behavior of expected return on a short position in the dollar and long position in a basket of foreign currencies
    - - This portfolio is fully exposed to EXR risk
      - Dynamic portfolio formed by commodity forecasts of currency return has a higher annualized return 1.6% - 3.1% than the portfolio formed by benchmark of historical average forecasts
      - World CP fluctuations potentially explain a major component of their terms-of-trade fluctuations in these countries
      - World commodity price is exogenous to EXR to a large extent
      - Commodity factor addresses the model uncertainty which causes worse forecasting performance of economic models
    - - None
    - - In-sample evidence shows significant Granger causality
      - Out-sample indicate the univariate predictive regression with commodity factors produces significantly more accurate exchange rate forecasts than benchmark of random walk model for all 4 currencies
      - Detected predictive content of CPs for currency excess return and significant predictability in-sample and out-sample
      - Currency predictability can be explained that CP provide useful info regrading future business changes
      - Kilian 2009, Wen et al. 2012, Zhang 2017, Salisu et al. 2018, Wang et al 2013, 2019: Oil price shocks have essential impacts on the variables reflecting economic activities
    - - Dependent variable: EXR
      - Currencies: AUD, CAD, NZD, ZAR
      - Reason: 4 countries primmary commodities account for significant fraction of their exports
    - - Investigates the predictive content of CPs for exchange rates (EXR)
      - 17 Popular commodities including crude oil
      - Average commodity returns can successfully predict the level and excess returns to EXR of currencies in Australia, Canada, New Zealand and South Africa for in-sample and out-sample perspectives
      - Excess currency returns is economically significant
      - Mean-variance allocating wealth between domestic and foreign bonds can improve the portfolio performance by using commodity forecasts of currency return instead of benchmark of historical average forecasts.
  - - - Cody and Mills 1991: A close relationship between CP and CPI
      - Clarida et al. 1991: CP strongly inform movements in inflation, interest rates and output (GDP)
      - Bhar and Hamori 2008: CP can be informative for formulating monetary policy, due to the role played as input factors to industrial production
      - Stock and Watson 2003: Argued that CP are useful as predictors for inflation and output growth
      - Holtemoller and Mallick 2016: Employed both an estimated VARX and small open economy New Keynesian model to demonstrate that global food price shocks constitute an important contributor to cost-push inflation in India
      - Mallick and Sousa 2013: Estimated a battery modern dynamic macroeconometric models in order to robustly identify the importance of commodity price shocks, which lead to rise in inflation and demand a more aggressive behavior from central banks towards inflation stabilization
      - Bernanke 2008: Calls for the development of a better understanding of the factors that underpin commodity prices
    - - Evidence demonstrating how daily measures of liquidity observed across different commodity markets exhibit a tendency to co-move in accordance with the existence of a common liquidity factor
      - Shown with such a common liquidity bellwether may be underpinning much of observed co-movement of commodity prices, by presenting evidence from a series of co-integrating regressions and Granger causality test
      - Statistically establish the existence of significant daily price co-movement among different commodity classes
      - Liquidity can drive aggregate price shock, which amplify inflation risk
      - Additional tests further suggest that the impact of liquidity on CP co-movement is stronger in a post-2008 financial crisis sub-sample of our complete data set
    - - Dynamic linear regression carried out on lagged liquidity variables & commodity market CRB index to control for changes in overall macroeconomic conditions
      - Co-integrating regression analysis similar from Leybourne et al. 1994
      - Establishing the existence of a co-integrating & long-run equilibrium relationship between CP co-movements and common liquidity factors
      - Liquidity effect on prices is not only found to spill over across different commodity sector boundaries, but also to propagate across different maturities of commodity futures securities
    - - Establish existence and usefulness of common commodity liquidity bellwether as an important market-derived
      - Real-time predictor of cross-sectional CP co-movements
      - Presence of a series of long-run co-integrating relationships between common liquidity bellwether and array of CP while establishing predictive utility
  - - - GARCH: univariate volatility processes
      - 1.a. Oil prices and 4 CP
      - Period: 1980
      - VAR: multivariate for each country
      - Monthly data
      - Monthly log changes in output, interest rates, CPs, EXR & price of each country's important commodity
      - ARMA method
    - - Each country behaves idiosyncratically. EXR & inflation volatility are linked to CP fluactuations in more cases rather than output and interest rates
      - Examined individual results being Chile the 'commodity currency'
      - Chilean economy influences copper prices indicating traders recognize the country's strong ties to metal
    - - Van der Ploeg and Poelhekke 2009: Both oil and non-oil price volatility reduce economic growth and contributed to 'resource curse'
    - - CP volatility spillovers to country's EXR or output but also policymakers can react and impacting interest rates and inflation
      - Make policymakers in making difficult decisions
      - Not all central banks must be as sensitive to CP swings as others
      - Defining countries by categories: Commodity economies, Diversified exporters, insulated oil exporters (less affected by oil price shocks)
      - In the long run, diversification into manufacturing may help protect countries from wild swings in CPs
      - In short run, countries must manage risk either through capital controls or inflation targeting monetary policy to avoid transmission to EXR and prices
    - - Price decrease cause capital outflows
      - Deteriorations in balance of payments
      - Reduction in competitiveness
      - This paper studies the volatility processes of 6 major CPs
      - Price increases lead to currency appreciation
      - Before applying Multivariate GARCH to examine spillovers among the variables
      - Global CPs has large increase during Global Financial Crisis & steep decline after the crash
      - Chile is most closely tied to copper price, Indonesia to oil and tin
      - Brazil & Philippines are less affected
      - Russia is highly insulated from fluctuations in world oil prices
    - - Fluctuations in CPs can have strong linkages with entire macroeconomy, currency are most closely connected
      - Commodity exports bring in foreign exchange, any volatility can ahve direct impact on balance of payments
      - Drop in CP can cause a drop in GDP
  - - - Random effects logit model approach by Mundlak 1978 & Chamberlain 1982
      - 60 LICs
      - Period; 1963 - 2015
    - - Found that volatility of CPs is a significant predictor of banking crises in a sample of 60 low-income countries
      - Shown that major channel through which CP movements can affect the real economy is through their effect on banks' balance sheets and financial stability
      - In contrast to findings on advanced, emerging economies, credit booms and capital inflows do not play a significant role in predicting banking crises
      - Historical data 40 peripheral economies between 1848 and 1938
      - Effect of CPs volatility on banking crises is concentrated in LICs with fixed EXR regime and high share of primary goods in production
      - CPs volatility is likely to trigger financial instability through a reduction in government revenues and shortening of sovereign debt maturity, which weaken the banks' balance sheets
    - - Hardy & Pazarbasioglu 1999: Banking system distress are differ across economies with different structural characteristics
      - Fernandez et al 2017: Since 1960s global shocks to CPs account for about one quarter of output fluctuations in LICs, they are comparable with richer countries and has significantly increased over past 15 years with financialization of commodity markets
      - Reinhart & Rogoff 2013: Graduation from banking crises has so far proven illusive
      - Koren & Tenreyro 2007, Fernandez-Villaverde at al 2011: Shown the importance of volaitlity shocks for economic growth and development
      - Bloom 2009: Found uncertainty shocks including CPs have real effects on firms' hiring and investment decisions
      - Bleaney and Greenaway 2001, Blattman et al 2007, Williamson 2012, Cavalcanti et al 2015: Shown the volatility of CPs more than growth are matters for output fluctuations
      - CPs fluctuations through Financial Channels: Due to effect on balance sheets and financial stability
        
        Cespedes and Velasco 2012, Agarwal et al 2020: Shown a fall in CPs reduces bank lending for commodity exporter of developing countries
        
        Kinda et al 2018: Found that negative CP shocks are associated with higher non-performing loans and lower bank profitability
        
        Caprio and Klingebiel 1996a, 1996b: Documented a number of banking crises in 1980s and early 1990s and show that volatile terms of trade as associated with systemic crises, specifically on countries with a concentrated export base
      - Terms of trade fluctuation leading to financial instability and crises
        
        Sharp drop in prices translate into revenues for exporting firms which is more difficult to service their debt obligations, with potential negative effects on banks asset quality
        
        Caprio & Klingebiel 1996b: Large variations in prices increase asymmetric information and make it more difficult to select good from bad borrowers
      - a. Emphasized key role that CPs play in triggering financial sector stress. So, this paper help whether fluctuations in CPs can help predict banking crises
      - a.i. Agarwal et al 2020: Shown declining CPs are associated with worsening bank health and lead to a contraction of bank lending in LICs
      - b. Most literature looking at advanced and emerging markets so focused in this paper towards LICs.
    - - None
    - - 1980s & 1990s: Period of financial instability, most concentrated among commodity exporters. Handful on developing countries have been hit by systemic banking crises over past 2 decades.
      - Factors contributed, an extended period of sustained economic growth, financial deepening and favorable external conditions, period of stable and high CPs
      - Since 2014, increasing number of LICs have been experiencing financial distress, declining bank profitability and deterioration in bank asset quality
      - Mendoza 1997, Deaton 1999, Kose 2002, Raddatz 2007, Cespedes & Velasco 2012: CPs is motivated by the observation that they are one of the most important factors driving economic aggregates in developing countries
  - - - Focusing on investigating whether there exist economic fundamentals linked to EXR fluctuations
      - Focusing on short-horizon predictive ability
    - - Little systemic relation between CP changes and EXR changes at monthly and quarterly frequencies
      - CP related to daily nominal EXR of commodity currencies
      - Relationship is statistically and economically significant
      - Predictive ability of lagged realized CP changes is more ephemeral (temporary)
      - Allowing time variation in relative performance is crucial to show that lagged CPs can be statistically significant predictors of EXR out-of-sample
    - - Buyuksahin & Robe 2014: Studied commodity future markets & financialization. Their evidence condirms role of speculators in driving cross-market correlations between equity returns and commodity returns
      - Chen et al 2010: Found EXR predict CPs at quarterly frequency and CPs do not predict EXR. Focusing on CP indices, which average across several commodities, not just individual commodities
      - Faust et al 2003: Predictive ability easier to find in a real-time data
      - Amano & Van Norden 1998a,b, Issa et al 2008, Cayen at al 2010: In-sample explanatory variables fro real EXR. Considering out-sample predictive ability for nominal exchange rates
    - - Effects of oil price changes on EXR changes are short-lived & frequency of data is crucial to capture
      - Conditional on knowing the future value of CPs and forecsting the EXR well
      - Good model to forecast oil prices, could exploit to forecast future EXR
      - Stronger at daily frequency
      - Found predictive ability using daily data which disappears at longer horizon
    - - Existence of short term relationship at daily frequency between changes in price of a country's major commodity export and changes in nominal EXR
      - Relationship appears to be robust & to hold when using contemporaneous CP changes in regression
    - - Limitation: The existence of out-of-sample correlation is not informative regarding economic causality in the data
      - Monetary fundamentals do not help forecast EXR out-of-sample not even in terms of out-of-sample fit, at the monthly or quarterly frequencies
      - Past values of CPs were not good forecasts of future values of CPs cause ended up rejecting the predictive ability CPs
    - - Daily frequency
      - Short run co-movement
      - Canadian-US Dollar EXR
      - Oil prices
      - Uses real-time data but uses economic fundamentals that is very different from traditional ones
      - Country; Canada due to crude oil export high, sufficient long history of market-based floating EXR, small open economy and small size in world oil market (price taker)
      - Canadian - US interest rate
  - - - Shown that oil market exhibit higher coherence with copper, natural rubber, fuel oil futures
      - Shown low coherence with corn, soybeans, soybean meal, white sugar futures on long term scale
      - Investigations towards oil and Chinas commodity futures markets
    - - Total risk connectedness at extreme lower quantile level 0.01 is higher than conditional mean and median level
      - The main spillovers net-transmitters: WTI oil, Brent oil, soybean oil and copper futures
      - DECO model: point to time-varying with low average equi-correlations between oil and commodity futures
      - Risk spillover net-recipients; White sugar, soybean, soybean meal, cotton, corn, aluminium, natural rubber and fuel oil futures
      - Dynamic extreme negative risk spillovers are highly volatile & vulnerable to major international events (GFC, oil price plunge and Covid 19 pandemic)
    - - Investigate time-frequency dependence
      - Extreme spillovers
      - Dynamic linkages
    - - Brent oil offer better portfolio diversification benefits than WTI oil
      - Optimal-weighted portfolio illustrates the highest risk and downside risk reduction effectiveness
      - Extreme spillovers between oil and China's commodity futures markets should be monitored constantly since affecting the hedging and portfolio diversification
      - Investors in China's commodity futures should be alert to the extreme risk spillovers transmitted from oil markets and hold oil assets in their portfolios to diversify portfolio risk
      - Chinese govt should establish a forward-looking extreme spillover warning mechanism to avert systemic financial risks
      - Investors and portfolio managers should develop dynamic and alternative portfolios and regularly recalibrate their portfolio strategies according to specific market conditions
      - China should strengthen its crude oil reserves and seek more cleaner and sustainable energy resources to reduce high dependence on crude oil imports and minimize the negative impacts caused by the significant oil price fluctuations
      - Emergency response mechanism for major crisis events should be further enhanced
    - - Included Global Financial Crisis 2008-2009, European Sovereign Debt Crisis 2010-2012, Oil prices collapse 2014-2016, China's stock market crisis 2015 and ongoing Covid-19 pandemic
    - - Quantile connectedness approach (Risk net-pairwise connectedness network)
      - DECO-FIAPARCH model
      - Wavelet coherence
      - Period; Jan 2006 - Dec 2020
      - Correlation
  - - - crude and corn futures prices
      - To obtain empirical volatility response functions & time-varying correlation coefficient
      - Multivariate GARCH model
      - Short term and long term futures exhibit shift in mean and volatility
      - Trigonometric function
      - Volatility do not manifest in the same manner for different maturities
      - Indicating that term structure of futures volatility changes over time
    - - BRIC countries: Brazil, Russia, India, China
      - low-frequency trigonometric functions into conditional variance equations in order to capture the growth of BRIC
      - Multivariate GARCH framework
      - Variance impulse response function suggested by Hafner & Herwartz 2006
      - Daily settlement prices of corn and crude oil futures
    - - Hertel & Beckman 2011, Tyner 2010, Muhammad & Kebede 2009: Argued the increased reliance on biofuels is a key factor contributing to the increased linkages between the grain and petroleum markets.
      - Enders & Holt 2012: Suggested rapid income growth in emerging economies (BRIC) countries is one of the primary drivers of the price boom, citing increased demand for both agricultural and energy products.
      - Summer 2009, Wright 2011: From 2006 through mid 2008 grains experienced one of the largest percentage price increases in history and that volatility increases sustained
      - Enders & Holt
        
        2012: Examine the behavior of real petroleum and agricultural prices over 50 year period and identify structural changes in each by estimating shifting-mean autoregression
        
        2014: Generalized the procedure and estimate the prices as a mean-shifting VAR
        
        Both of these papers treat the conditional variance of each market as a constant, so they ignore the possibility of volatility shifts spillovers
      - In contrast
        
        Wetzstein and Wetzstein 2011: Strong connection between oil and agricultural prices is a myth because of creation of biofuel capacity entails adjustment costs non-reversibilities and uncertainties
        
        Tyner 2010, Hertel & Beckaman 2011, Saghaian 2010, Zhang et al 2010: Found that changes in government policy and non-petroleum input price changes often govern large movements in grain prices
        
        Myers et al 2014: Use common trend-cycle decompositions and find that co-movements between energy and agricultural feedstock prices tend to dissipate in the long run.
      - Trujillo-Barrera et al 2012: Argued that underinvestment in agriculture, low inventory levels, supply shocks in key producing regions, fiscal expansion and lax monetary policy in many countries, and depreciation of the US dollar, have also contributed to increased commodity price volatility
      - Zhang et al 2009, Trujillo-Barrera et al 2012: Model the agricultural and petroleum prices in a multivariate GARCH setting and explore volatility spillovers, but do not allow for structural breaks in mean or variance
      - Perron 1989; It is important to allow mean & volatility breaks as demonstrated that neglected structural change in mean and variance equations typically result in suprious existence
      - Gallant 1981: Gallant's Flexible Fourier Form to allow for breaks in VAR with GARCH effects and show means and variance-covariance matrix of corn and crude oil futures exhibit structural change
      - Baillie & Myers 1991, Brunetti & Gilbert 2000, Jin & Frechette 2004: Found that CP exhibit long memory in that conditional volatilities appear to be fractionally integrated
    - - Price discovery is important feature of exchange-traded futures. Informed traders will prefer futures over spot transactions due to low margin requirements and lack of physical products.
      - Brooks 2012: crude oil & corn futures are characterized more as unarbitraged rather than fully arbitraged markets so futures prices may reflect future expected spot prices
      - Future markets permit the analysis of maturity time dimension and calendar time dimension so volatility shocks & spillovers can be studied from both short term and long term perspectives
    - - Biofuel production is not likely to be highly responsive to short-run oil price changes
      - Corn and oil future prices often move together and that both volatilities are far larger in the latter third of the sample than during the 1990s.
      - Structural change in volatility presents in both short-term & long term futures contracts
      - Variance impulse response function to assess whether controlling for volatility breaks leads to reduction in volatility persistence