Chapter 3 COMMODITY FUTURES AS AN INFLATION HEDGE. Commodity futures market performs two important functions. First, they establish the

Transcription

1 Chapter 3 COMMODITY FUTURES AS AN INFLATION HEDGE 3.1 Introduction Commodity futures market performs two important functions. First, they establish the relative prices of commodities used today to their expected value in future. Second, they help in risk management where the seller (owner of the commodity) takes a short position and buyer (user of the commodity) takes a long position to hedge their price risk. Speculators who takes both long and short positions based on their expectations about the futures prices to get a profit from the market movement are also an integral part of this process. Speculators serve the commodity market by providing liquidity and make the market less volatile - buy when the prices are low (short positions become larger) and sell when the prices are high (long positions become larger). Nevertheless, whenever the market deviates from normal and the price of commodities soar, speculators come under the lens of regulators and policy makers. According to them, speculators may distort the price of commodities - by excessive buying and selling - to fetch maximum profit from market movement, thus increasing the prices of commodities and fueling inflation. Increase in investment in commodity futures by institutional investors - the phenomenon called as financialization of commodity markets by academics - and sharp increase in the commodity prices in developing as well as the emerging economies of the world has raised concern about the role of speculators in commodity futures market. A number of investigations support the fact that such increase in investment led to excessive speculation in commodity futures, which causes unwarranted change in commodity 37

2 prices (Permanent Subcommittee on Investigation, 2006; 2007; 2009). Master (2008) produced a document to US senate /congress, which state that the long investment by the investment community (a special class of institutional investors called index speculators) in commodity futures is one of the major factors responsible for the then recent spurt in prices of commodities. Index speculator takes long position in commodity indices as part of their portfolio allocation decisions, which expand the market, and at the same time drives futures price higher. Higher futures prices attract more index speculators, and increase in allocation to commodities, which drive prices further. As commodity futures are the benchmark for the prices of physical commodities in future time period, increase in futures prices affects the spot prices. Spot price rises to a greater extent than expected and leads to inflation. Thus, institutional investors (index speculators) contribute to food and energy price inflation and to mitigate this problem government needs to take appropriate policy action (Master, 2008) and stringent regulations to make the trading of commodity futures more transparent. The policy measures that were suggested include lifting the position limit waivers for index traders and imposing a strict position limit per trader (Permanent Subcommittee on Investigation, 2009). Since the appearance of these initial reports there is a strong debate in the academic community, whether an increase in the trading of commodity futures and OTC derivatives is responsible for creating a speculative bubble in commodity prices that forced commodity prices higher than fundamental value. Increased speculation as a sole reason for the increase in prices of commodities was well supported by several wellknown international organizations and researchers (IATP 2008, 2009; Robles, Torero, and von Braun, 2009; Wahl, 2009; UNCTAD, 2009, 2011; UN Special Rapporteur 2010; 38

3 World Development Movement, 2010; Bafis and Haniotis, 2010; Herman, Kelly, and Nash, 2011). A number of academicians rejected the idea of increased speculation as a reason for high prices of commodities (Irwin, Sanders, and Merrin, 2009; Pirrong, 2010; Wright, 2011; Dwyer, Holloway and Wright, 2012; Knittel & Pindyck, 2013). On the other hand, they attributed the recent price rise in commodities to the fundamentals of supply and demand. Irwin et al. (2009) marked a number of flaws in Master s argument (referred to as the Masters Hypothesis (Irwin and Sanders, 2012)) and explained the dynamics of commodity futures markets where speculators could not be held responsible for the increase in price of commodities. The commodity futures market is a zero sum game market where for every long position there is a person/ investor who takes a short position. Therefore, a contract comes into existence when there are persons with opposite views about the futures prices. Theory posits that, there is no upper limit for the number of contracts that can be created at a given price level. Therefore, the money invested in the market to form new contracts does not represent demand/ supply and cannot affect futures prices. Any change in the futures price is a result of new information about the future supply and demand of the commodity that may be due to natural factors (such as flood, drought, El Nino, La Nina, Tsunami, Cyclones) or man-made artifacts (war, trade restrictions). Index investors/ speculators are not interested in physical delivery of the commodity. Therefore, they are unable to affect the long-term equilibrium prices that were discovered in the cash market. Thus, financialization of commodity market cannot directly account for the increase in spot prices of commodities (Baker, 2014). 39

4 Academic research tried to establish possible linkages between index positions and commodity prices. One of the evidence for the proposed financial speculation is the increase in correlation (increased co-movement) between commodity returns and financial returns in the past decade. Chong and Miffre (2010) and Büyükşahin, Haigh and Robe (2010) observed that the conditional correlation between commodity returns and stock returns decreased over time. Brunetti et al. (2011) found that speculators (including hedge fund and swap dealers) do not destabilize the financial market, but they reduce volatility and provide liquidity to the market. Similarly, Miffre and Brooks (2013) do not find evidence that speculators destabilize commodity prices by increasing volatility or increasing co-movements with traditional assets. Büyüksahin and Robe (2014) found that commodity-equity return co-movement are positively related to the amount of speculation in commodity market by speculators hedge funds who take long-short position in both commodity and equity market. However, they found no evidence of such positive relationship with Commodity Index Traders (CITs) and commodity swap dealers who take net long positions. Stoll and Whaley (2010) did not observe the impact of CITs on cross-commodity return correlation, while Tang and Xiong (2012) observed a positive relationship. Granger causality is employed (by researchers) to test the causality between index positions and commodity prices (Stephanie-Carolin Grosche, 2014). A large number of studies confirm the absence of any causal relation between index positions and returns (Stoll and Whaley, 2010; Brunetti et al., 2011; Sanders and Irwin, 2011a, 2011b; Capelle- Blancard and Coulibaly, 2011; Hamilton and Wu, 2014; Aulerich, Irwin, and Garcia, 2013) while few researchers observed a positive relationship (Gilbert, 2010; Gilbert and 40

5 Pfuderer, 2014). Lack of statistical power of Granger causality led researchers to find other avenues such as cross-sectional regression analysis, spreads tests (Irwin, 2013) and instrumental variables method (Gilbert and Pfuderer, 2014). Cross-sectional regression studies also provided evidence against Master s Hypothesis (Sanders and Irwin, 2010; Irwin and Sanders, 2012). Rolling of futures contract by index traders decreases the price of nearby contract and increases the price of the next contract. This leads to a new area of research to find the relationship between index positions and spread the difference between futures prices of different contractual maturities. Few studies reported that rolling of positions cause expansion of the spread (Mou, 2010; Aulerich, Irwin, and Garcia, 2013) while the majority of the research presented results that were not in favor of Master s Hypothesis (Irwin et al., 2011; Stoll and Whaley, 2010; Garcia, Irwin, and Smith, 2014; Brunetti and Reiffen, 2011; Hamilton and Wu, 2014). Looking at the mixed evidence it can be said that the increase in investment in commodity futures market cannot be the sole reason for the increase in prices of commodities. The Indian commodity futures market took a rebirth in November 2003 and soon after it has seen a rapid increase in the trading of commodity futures. The price of commodities, and food in general, has increased in the last few years and created a serious problem in the form of inflation. India has witnessed a sharp increase in commodity prices and inflation since The average inflation as measured by the rate of change in the Consumer Price Index (CPI) is above eight percent since 2008 and has touched double digits twice (2009 and 2010). High food grain production along with high agricultural commodity prices led to the belief that trading in commodity futures was responsible for the increase in spot prices and fueling inflation. Government of India formed an expert 41

6 committee to study the impact of futures trading on agricultural commodity prices. The committee did not find any clear evidence of either reduced or increased volatility of spot prices due to futures trading (Abhijit Sen Committee Report, 2008). Reserve Bank of India (RBI) conducted a similar study in 2010 and found similar results. Thus, both studies did not find any conclusive evidence that trading in agricultural commodity futures was responsible for the spot price rise or leads to food price inflation. Reserve Bank of India considers inflation as a major problem and takes adequate measures to reduce the prevailing high inflation. Inflation targeting as a measure to keep inflation within limits has been proposed (Urjit Patel Committee Report, 2014). Inflation reduces the purchasing power of common household and reduces the real rate of returns. Therefore, it affects consumers as well as investors. Inflation is an international phenomenon and nearly every country in the world has experienced it sometime in the past. In the developed financial world an obvious question that comes in mind of investors is that, whether financial market provides enough instruments to hedge inflation risk. If yes, which instrument can serve as an inflation hedge? A large amount of literature has looked at stocks, bonds, commodities, gold and real estate as an inflation hedge. Performance of stocks and bonds is inferior when compared to commodities during inflation. As the expected future price incorporates the expected inflation, any sudden rise in inflation observed is due to the unexpected component of inflation. It has been observed that stocks and bonds are negatively correlated whereas commodities are positively correlated to unexpected inflation (Gorton & Rouwenhorst, 2006). Thus, a 42

7 combination of commodity futures with stocks and bonds is expected to perform better during high inflation. At the same time, commodities have a direct link to inflation. The increase in spot price of commodities is one of the major sources of inflation. As commodity futures are the bet on future expected spot price, exposure to commodity futures can provide an inflation hedge as the increase in the spot price to some extent is captured as profit by the speculator who has long position in commodity futures. In such a case, a portfolio that incorporates commodity futures acquires the potential to avoid the ill effects of inflation on the portfolio. Assuming that a traditional investor holds only stocks and bonds, and commodity futures act as inflation hedge; a conventional portfolio with commodity futures must provide a better risk return tradeoff during high inflationary years. This chapter studies the inflation hedging potential of commodity futures by incorporating them in a conventional portfolio of stocks and bonds, and comparing the portfolio results during different inflationary regimes. The results show that commodity futures do possess necessary characteristics that they can be considered to be included in the conventional portfolio during high inflationary years. The chapter is structured as follows. The next section discusses the performance of stocks, bonds and commodity futures during inflation. Section 3.3 discusses the data and methodology used to select asset that possesses the best inflation hedging potential and development of inflation tracking portfolios. Section 3.4 discusses the results, section 3.5 concludes the findings, and section 3.6 lists the limitations. 43

8 3.2 Literature Review This section reviews the past studies that have studied the behavior of different assets during inflation Stocks The inflation episode of the 1970 s in the USA led to the search for assets that could serve as an inflation hedge. Common stocks represent claims on real assets whose real value assumed to be independent of the rate of inflation (Lintner, 1975). According to Fisherian assumption - real returns remain unaffected by current inflation or by the expectations of future inflation and nominal returns move in one to one relationship with expected inflation - stocks should provide a hedge against inflation. The results observed from the initial studies were against the conventional wisdom and show stocks to be a perverse inflation hedge (Johnson et al., 1971; Oudet, 1973). A number of researchers rejected Fisherian assumption and observe negative relationship of stock returns with expected inflation (Nelson, 1976; Jaffe & Mandelker, 1976; Bodie, 1976; Fama & Schwert, 1977; Gultekin, 1983) and unexpected inflation (Nelson, 1976; Jaffe & Mandelker, 1976; Bodie, 1976; Fama & Schwert, 1977). A majority of studies confirm a negative relation in the short run, but at the same time, a number of studies have shown that stocks can hedge inflation in the long term (Cagan, 1972; Jaffe & Mandelker, 1976; Boudoukh & Richardson, 1993; Ely & Robinson, 1997; Schotman & Schweitzer, 2000; Anari & Kolari, 2001; Hoevenaars et al., 2008; Alagidede & Panagiotidis, 2010; Bekaert & Wang, 2010). Barnes et al. (1999) observed a negative correlation between nominal 44

9 returns and inflation in low-to-moderate inflation economies and positive correlation in high inflation economies. Researchers put forward a number of theories to explain the negative relationship between stock returns and inflation. Modigliani and Cohn (1979) proposed inflation illusion hypothesis which pointed an error - investors incorporate expected inflation into their nominal discount rate to value equities - that investors commit while evaluating stocks. As a result, they were unable to price equities that could reflect their true economic value. Real after-tax hypothesis by Feldstein (1980) explains the relationship as a combined effect of tax laws, historical cost depreciation and taxation of nominal capital gains. He proposes that inflation increases the effective tax rate, reducing the net yield investors receive per unit of capital. Proxy hypothesis by Fama (1981) explains the negative relationship as spurious, which is induced due to the combination between real sector (explains the positive relationship between stock returns and real activity) and monetary sector (explains the negative relationship between inflation and real activity). Anticipating a reduction in real economic activity, a rational investor decreases demand for money that leads to excess money in the system causing inflation. Geske and Roll (1983) explained a similar phenomenon using money supply linkages. They propose that reduction in the stock price signals increase in inflation. This relationship reverses the causal relationship as explained by Fama (1981) and is thus called Reverse Causality Hypothesis. According to this hypothesis, anticipation of low economic activity affects stock returns, increases unemployment and decreases company earnings that lead to lower revenue collection by the government and increasing Treasury deficit. Treasury responds by increased borrowing from public. The central bank buys some of the 45

10 Treasury debt and increases the rate of base money in the system causing inflation. Kaul (1987, 1990) explains the stock return inflation relationship owing to equilibrium process and operating targets of the monetary sector. Money demand and counter-cyclical money supply leads to negative relationship, whereas money demand and pro-cyclical money supply leads to the insignificant or positive relationship between stock returns and inflation. Money demand and money supply factors systematically influence the relationship over time. The counter cyclical responses of central bank in presence of interest rate regimes make the relationship significantly stronger. Pindyck (1984) worked on the lines of Malkiel (1979) and attributed the relationship to the increased riskiness observed in net real returns on holding stocks, due to increase in the variance of inflation. According to Hess and Lee (1999) the stock inflation relationship is regime specific and depends on the relative importance of supply shock (real output) versus demand shock (monetary shock). These hypotheses were tested a number of times by researchers and they found mixed evidence. Based on the literature, the relationship between stock and inflation seems to be negative in the short run. However, some studies do posit a positive relationship in the long run Treasury bills and Bonds Nominal bonds are characterized by fixed future cash flows. In general, nominal bonds have negative correlation with inflation because high inflation is associated with higher bond yields and lower bond returns (Campbell et al., 2009). Higher bond yields in expectation of high inflation damages the current value of the nominal bond resulting in lower returns. During high inflationary environment short-term nominal bonds, notes and bills perform better than long-term bonds, as investors get the opportunity to reinvest 46

11 their proceeds in instruments that reflect high inflation expectations (Ruff and Childers, 2011). Treasury bills and short-term U.S. government bonds offer a complete hedge against expected inflation (Fama & Schwert, 1977). Treasury bills appear to be the best inflation hedge over time (Hoevenaars et al., 2008) as they represent the most dominant component of the inflation-tracking portfolio (Bekaert & Wang, 2010, Crawford et al., 2013). This results from the fact that Treasury bills are the best predictor of future inflation as they employ all available information about expected inflation (Fama, 1975) or due to the variance reduction property of treasury bills. Formation of inflation-tracking portfolio involves minimizing variance. As the majority of assets are more volatile than inflation, the positive weight of an asset in the inflation-tracking portfolio reflects its variance reducing property, rather than the ability to hedge inflation risk. Treasury bills being the least volatile among all assets got the highest weight in the inflation-tracking portfolio (Bekaert & Wang, 2010). Treasury bills are readily accessible, are a liquid investment vehicle, but their low yield limits their use in the investor s portfolio (Crawford et al., 2013). However, short-term investors should allocate a major component of their portfolio to Treasury bills (Brière & Signori, 2012). Variations in expected inflation and short-term real interest rates affect long-term nominal Treasury bonds and Treasury bills respectively (Campbell et. al, 2009). Therefore, these instruments are not safe for long-term investors. In this backdrop, inflation-linked bonds function as riskless investments that are suitable for long-term investors (Campbell & Viceira, 2001, Brennan & Xia, 2002, Campbell et al., 2009, Brière & Signori, 2012). The inflation-linked bond provides a long-run hedge against 47

12 inflation risk (Bodie, 1990) by providing sure returns that are well above inflation - by linking principal or coupon to realized inflation. U.K. government introduced these inflation-linked bonds in 1981 (Index-linked gilts). These bonds are quite popular among investors in U.S.A (Treasury Inflation-Protected Securities), Euro area, France, Australia, Canada and Japan (Bekaert & Wang, 2010). Reserve Bank of India introduced inflationlinked bonds linked to the Wholesale Price Index (WPI) and Consumer Price Index (CPI) during These bonds protect investors from the dangers of negative real returns that could prevail during high inflation. Rise in inflation may come along with the rise in real interest rate that could reduce the returns of inflation-linked bonds, but they still outperform the nominal bonds (Ruff and Childers, 2011). Past studies suggest that when inflation is expected to increase, indexed bonds (eg. TIPS) should be a dominant component of the portfolio and provide meaning diversification benefits (Roll, 2004; Kothari and Shanken, 2004, Lamm, 1998, Mamun and Visaltanachoti (2006a). However, the increase in volatility of TIPS and a high correlation with nominal bonds has decreased the diversification benefits (Briere and Signori, 2009) Commodity futures and inflation Literature pertaining to the relationship between commodity futures and inflation can be broadly divided into three categories. The first category deals with the negative correlation that is observed between commodities and stocks & bonds during inflationary periods. The second is related to the positive correlation between commodity futures return and unexpected component of inflation, while the third deals with the performance of portfolios incorporating commodity futures during high inflationary environment. 48

13 Commodity futures provide an inflation hedge. This has been documented in several studies (Greer, 1978; Bodie and Rosansky, 1980; Becker and Finnerty, 2000). During inflation, the performance of commodities is different when compared to stocks and bonds as inflation has a positive effect on commodity prices, but a negative effect on equity and bond markets (Anson, 2006). Annual returns of commodity futures are positively correlated with changes in inflation (Greer, 1978; Bodie and Rosansky (1980). Gorton and Rouwenhorst (2006) examined the correlation of stocks, bonds, and commodities to inflation at horizons ranging from one month to five years. They reported that commodities (stocks and bonds) were positively (negatively) correlated with inflation at all horizons. Kaplan and Lummer (1998) observed commodities as good diversifier and inflation hedge for long term investment, but find a positive correlation of commodities with traditional assets during short run. Kat and Oomen (2007) found that commodities performed well in the face of unexpected inflation, but that this varied significantly over individual commodities. However, Bjornson and Carter (1997) observed that expected commodity returns were inversely related to interest rates, economic growth, and inflation. Froot (1995) found that commodity futures act as an effective hedging tool against unexpected inflation. Research pertaining to the component of commodity futures return that provides a hedge against unexpected inflation is mixed. Ankrim and Hensel (1993) found that the spot return is the component of the commodity futures return that is most strongly correlated with unexpected inflation. Erb and Harvey (2006) advocated for the positive relationship between roll returns and unexpected inflation. However, Menzel and 49

14 Heidorn (2007) found that the positive relationship depends on roll return and collateral return and not on spot return component of commodity futures return. Becker and Finnerty (2000) examined the risk and return properties of equity/bond portfolios before and after the inclusion of a diversified portfolio of long commodity futures. Inclusion of commodities improved portfolio performance by enhancing the risk and return characteristics that are more substantial during the high inflation periods. They also reported that both equally weighted and production-weighted indexes of commodities were valuable portfolio components and proved to be more beneficial during high inflationary periods. Kaplan and Lummer (1997) observed that commodities performed better during inflationary periods while stocks and bonds performed poorly. Bodie (1983) found that a broad-based position in commodity futures supplements a conventional portfolio. Portfolios including commodity futures have a tendency to perform well in presence of unanticipated inflation. Menzel and Heidorn (2007) studied risk return relationship in equity and bond portfolios including commodities. Results showed that during high inflation and strong equity markets, inclusion of commodities is a sensible addition. Woodard (2008) formed optimal portfolios with and without commodities for low inflation and high inflation periods. Results revealed that the overall performance of the portfolio is improved by including commodities during high inflation period. 3.3 Data and methodology The data consists of the commodity futures traded on two of the largest commodity futures exchanges in India, namely Multi Commodity Exchange (MCX) and National 50

15 Commodity and Derivatives Exchange (NCDEX). Data was analysed on a yearly basis from 2005 to 2012 using monthly price series of the most liquid commodity futures. Trading volume data is used to sort the most liquid commodity futures contracts. The closing prices of these contracts are used to form monthly price series. Spot return is calculated as the percentage change in the spot price of the underlying commodity, i.e. the price used to mean the nearest (first nearby) future contract (the contract that is closest to maturity). Whereas, futures return is calculated as the percentage change in the futures price, i.e. the price used to mean the next nearest (second nearby) futures contract (Hafner and Heiden, 2008; Gorton et al., 2013). Price gap between different-maturity contracts (i.e. the price difference between the futures return and spot return) is called roll return or implied yield or futures basis. Consumer Price Index (CPI) is used to measure Inflation. Let denotes the CPI index value. The inflation rate is then defined as and unexpected inflation is equal to The difference between the inflation rate and unexpected inflation is considered as expected inflation. Expected inflation is already set up in the pricing of the securities. Therefore, what really matters is how the security responds to unexpected inflation (Kat & Oomen, 2006). 51

16 Risk to real returns stems from the uncertainty of the future price levels. Any risky asset tends to be an inflation hedge if it helps to reduce the risk of investors real return in the portfolio. Traditionally, investors deem stocks, bonds, commodities and real estate to possess inflation hedging property. The following section describes the methodology used by researchers to select assets that can provide an inflation hedge. The majority of research related to selecting assets that corresponds to inflation stems from Fisherian hypothesis which states that the real returns are independent of expected inflation and any change due to inflation is represented by nominal returns of the security. According to Fisher (1930), expected nominal interest rate can be expressed as a sum of expected real returns and an expected inflation rate. Fama and Schwert (1977) used this Fisherian assumption to measure the inflation hedging potential of an asset. Their equation involves regressing expected nominal returns on expected rate of inflation: Where represents the nominal return, a constant real return, a constant and an error term with mean zero and variance and unrelated to expected inflation. The inflation hedging potential of an asset is measured by. An asset with is a complete hedge, is a partial hedge and is a perverse hedge. 52

17 Fama and Schwert (1977) expand the above equation to include unexpected inflation Where represents unexpected inflation calculated as a difference between realized and expected inflation. Inflation hedging property of an asset is described by and. An asset with is a complete hedge against expected inflation and asset with is a complete hedge against unexpected inflation. Assets with and are partial hedges against expected inflation and unexpected inflation respectively. Assets with and are perverse hedges against expected inflation and unexpected inflation respectively. The above equations are used for short-term analysis up to one year. Long-term inflation hedging potential of assets was calculated using realized inflation Bodie (1976, 1982) and Schotman & Schweitzer (2000) used a different methodology to find the inflation hedging potential of an asset. They consider an investor who invests a fraction of his wealth (w) in risky asset and rest amount (1-w) in a riskless asset such as nominal risk-free discount bond. The variance of the risk-free discount bond is equal to the variance of inflation. If the inclusion of risky asset with risk-free asset decreases the variance and increases the returns of the portfolio to achieve global minimum variance portfolio, the risky asset can be considered as an inflation hedge. The weight of the risky asset in a mean-variance optimal portfolio consists of speculative demand demand arising from the real risk premium on the risky asset - and inflation hedging demand 53

18 demand arising from covariance between inflation and nominal return of risky asset. Inflation hedging demand is the amount of risky asset that needs to be added to the riskfree nominal bond to achieve global minimum variance portfolio. Where represents nominal returns on risky asset, represents risk-free return on nominal bond, represents variance of nominal returns on risky asset, represents covariance between nominal returns on risky asset and realized inflation and represents relative risk aversion parameter. Bodie (1976) defined inflation hedging potential of an asset as the ratio of the variance of the real returns on risky asset to the variance of unanticipated inflation. Lower variance ratio shows higher inflation hedging potential of an asset. Similarly, the ratio of the variance of the real rate of return on Minimum Variance Portfolio (MVP) to the variance of inflation is equal to (Bodie, 1982). where represents the ratio of variance of real returns on MVP to the variance of inflation (or variance of risk-free nominal bond) and is the correlation between inflation and nominal rates of returns on risky asset. The risky asset would be a perfect inflation hedge if (i.e. ) and the variance ratio is zero. An asset that is a complete hedge according to Fisher coefficient when added to the nominal bond might not reduce the real return variance of the portfolio due to the low 54

19 correlation between asset returns and inflation. In a similar way, an asset with a high correlation between asset returns and inflation might have a low inflation hedging demand. Such an asset, when added to nominal bonds can perform better and reduce the real return variance of the portfolio. Such instances show that Fisher coefficient and inflation hedging demand sometimes fail to account for the inflation hedging property of an asset. Spierdijk & Umar (2013) describe the relationship between Fisher coefficient and inflation hedging demand. Similarly, inflation hedging demand can be written as: The inflation hedging potential of an asset increases as Fisher coefficient and inflation hedging demand increased. Above equations show that inflation hedging potential is directly proportional to the correlation and the second part of both equations are reciprocal to each other. With a view that Fisher coefficient and inflation hedging demand can give contradictory results, use of correlation coefficient as a measure to calculate the inflation hedging potential of an asset has been proposed (Spierdijk & Umar, 2013). The inflation hedging potential of an asset increases when correlation increases from zero to one, and when it decreases from zero to negative one. Therefore, absolute value of correlation clearly represents the inflation hedging potential of an asset. 55

20 Correlation as a measure to explain the inflation hedging potential of an asset has been adopted by Hoevenaars et. al (2008). All the commodity futures were sorted based on the absolute correlation between realized inflation and futures return. Top six commodities with highest correlation coefficient were selected each year and used to form equiweighted and inflation targeted portfolios. These portfolios were later used to analyse whether commodity futures possess potential to act as inflation hedge or protect portfolios during high inflation. Inflation tracking portfolio is a type of economic tracking portfolio where the returns of the assets in the portfolio track inflation. Individuals who want to hedge inflation risk should take a position in the inflation tracking portfolio (Lamont, 2001). Weights of different assets in inflation tracking portfolio can be obtained by regressing realized inflation (dependent variable) with different assets (independent variable) in such a way that intercept is equal to zero and all the betas sum up to one. Betas (the slope coefficient) represent the weight of the asset in the portfolio. The regression equation used is as follows Where CPI t denotes realized inflation for month t, β represents slope coefficients and denotes the return for i th asset class for month t. The following regression is done by a simple optimization where the sum of squared error is minimized 56

21 Inflation tracking portfolios were formed each year using the top six commodities that were selected before based on highest absolute correlation. w represents weights of the commodities which sums up to one and no shorting constraint has been added. The inflation tracking portfolio thus formed shows the best composition of commodities that tracked the inflation from a historical perspective (Crawford et al., 2013). 3.4 Results Table 1 presents the correlation between stocks, bonds and commodity futures. The most liquid (based on volume) commodity futures were selected for yearly analysis. The time period for analysis has been divided into two parts low inflationary years ( ) and high inflationary years ( ). Results reported in Table 1a show that the average correlation between stocks and commodity futures reduces during high inflationary years. Similarly, the average correlation between bonds and commodity futures also reduces during high inflationary years. This decrease in correlation value had been observed for all the types of commodity futures returns futures, spot, and roll returns. Table 1a also reports the decrease in the average correlation between stocks and bonds during high inflationary years. This shows that the diversification benefits become better during high inflationary years. Table 1a: Correlation between stocks and bond Average Average S/B S- Stocks, B- Bonds 57

22 Table 1b: Correlation between stocks, bonds and commodity futures Average Average Commodity Futures Return Stocks Bonds Commodity Spot Return Stocks Bonds Commodity Roll Return Stocks Bonds Table 1b presents correlation between stocks, bonds and agricultural commodity futures. The results show that the average correlation between stocks and agricultural commodity futures (except stock and commodity spot return), as well as bonds and agricultural commodity futures decreases during high inflationary years. The above analysis shows that in general, during high inflationary years the correlation between stocks and commodity futures as well as bonds and commodity futures decreases. Thus, addition of commodity futures to the conventional portfolio of stocks and bonds during high inflationary environment could be beneficial. Table 1c: Correlation between stocks, bonds and agricultural commodities Average Average Commodity Futures Return Stocks Bonds Commodity Spot Return Stocks Bonds Commodity Roll Return Stocks Bonds

23 To further comprehend the notion as to whether incorporation of commodity futures during inflation may be useful, the correlation between stocks, bonds and commodity futures with realized and unexpected inflation was performed. Table 2 reports a positive correlation between futures and realized inflation, whereas stocks and bonds are negatively correlated with realized inflation. A sub-analysis of low and high inflationary years shows negative correlation between commodity futures and realized inflation during low inflationary years and a positive correlation during high inflationary years. Analysis of agricultural commodity futures with realized inflation yields similar results. Stocks and bonds have negative correlation with realized inflation during low as well as high inflationary years. Table 2 further reports negative correlation between bonds and unexpected inflation, whereas stocks and futures are positively correlated with unexpected inflation. A sub-analysis of low and high inflationary years shows negative correlation of stocks, bonds and commodity futures with unexpected inflation. However, all the assets have a positive correlation during high inflationary years that shows their inflation hedging potential. Analysis of agricultural commodity futures with unexpected inflation yields similar results. The agricultural commodity futures resulted in higher correlation values with realized and unexpected inflation when compared to all commodity futures. The decrease in the correlation between stocks & commodity futures, and bonds & commodity futures during high inflationary years, as well as positive correlations of commodity futures with expected and unexpected inflation makes commodity futures a perfect choice to be incorporated during high inflationary years in the conventional 59

24 portfolio of stocks and bonds. Table 3a reports the ex-ante performance of conventional portfolios with and without commodity futures. Table 2: Correlation between stocks, bonds, and commodity futures with inflation Realized Inflation Unexpected Inflation Stock Bonds Futures Futures (Agri) Stock Bonds Futures Futures (Agri) Average Low Inflation ( ) High Inflation ( ) One and the foremost question that arises while forming ex-ante portfolios to study the inflation hedging potential of commodity futures is on the choice of commodity futures to be made and determining their proportions. To answer this, first, the commodity futures (top six) that have highest correlation (positive or negative) with realized inflation were selected to form a portfolio (such portfolio is considered as a naïve portfolio). As the hedging potential of an asset increases with increase in correlation value (either positive or negative), such naïve portfolio possesses the best inflation hedging potential against inflation. Incorporation of equiweighted naïve portfolio (10%) with stocks (75%) and bonds (15%) results in higher Sharpe ratios during double-digit inflation years. The results also show that equiweighted naïve portfolios generally yield negative returns with 60

25 a very small standard deviation. Therefore, the risk-return tradeoff observed during high inflationary years would result from the risk reducing property of commodity futures. Subsequently, a similar analysis was performed using inflation tracking portfolios that consist of the same commodities that constitute naïve portfolio but differ in portfolio weights. The portfolio weights (slope values) were calculated by regressing inflation (dependent variable) with commodity futures (independent variable) in such a way that the intercept was zero and all the beta coefficients (slope) sum up to one. Such a portfolio best tracks the inflation. Incorporation of inflation tracking portfolio (10%) with stocks (75%) and bonds (15%) results in higher Sharpe ratios during high inflationary years (except 2011) which reveals the true hedging potential of the commodity futures. Table 3a: Ex-ante Returns and Sharpe Ratios of Naïve, Constrained and Inflation Tracking Portfolios Average Inflation 4.25% 6.18% 6.37% 8.35% 10.85% 12.01% 8.85% 9.34% Conventional portfolio Return 32.88% 28.32% 40.80% % 57.36% 22.20% % 14.76% Sharpe ratio Naïve Portfolio Portfolio Return -6.24% -2.52% % % 19.56% -0.48% -0.96% 2.52% Constrained Portfolio Portfolio Return 30.00% 26.16% 36.48% % 55.80% 20.64% % 13.80% Sharpe Ratio Inflation Tracking Portfolio Portfolio Return 31.20% 27.12% 38.52% % 55.32% 21.48% % 14.64% Sharpe Ratio Previous results (Table 2) show that agricultural commodity futures perform better when compared to all commodity futures. Table 3b reports the ex-ante performance of 61

26 conventional portfolios with and without agricultural commodity futures. Agricultural commodity futures (top six) that have highest correlation (positive or negative) with realized inflation were selected to form naïve portfolio. Incorporation of equiweighted naïve portfolio (10%) with stocks (75%) and bonds (15%) results in higher Sharpe ratios during double-digit inflation years. These results are similar to the results observed when all commodity futures were considered to form the portfolio. However, the results changed when further analysis was performed using inflation tracking portfolios that comprise of agricultural commodity futures. Incorporation of inflation tracking portfolio (10%) with stocks (75%) and bonds (15%) results in higher Sharpe ratios only two times during high inflationary years. Even though agricultural commodity futures seem to be highly linked to inflation, their performance as an inflation hedge is dubious. Such a performance could be on account of excessive volatility seen in the agricultural commodity futures market. Table 3b: Ex-ante Returns and Sharpe Ratios of Naïve, Constrained and Inflation Tracking Portfolios (agricultural commodities) Average Inflation 4.25% 6.18% 6.37% 8.35% 10.85% 12.01% 8.85% 9.34% Conventional portfolio Return 32.88% 28.32% 40.80% % 57.36% 22.20% % 14.76% Sharpe ratio Naïve Portfolio Portfolio Return 1.68% % -1.08% % 35.04% 15.24% % 4.68% Constrained Portfolio Portfolio Return 30.84% 24.48% 37.92% % 57.36% 22.20% % 14.04% Sharpe Ratio Inflation Tracking Portfolio Portfolio Return 31.20% 27.12% 38.52% % 55.08% 21.24% % 14.16% Sharpe Ratio

27 3.5 Conclusions Commodities have a direct link to inflation and their storage and trading come under scrutiny during high inflation. Generally, the speculators in the commodity futures market were held accountable for the increase in prices of commodities. Extensive research has been done to understand the increase in the price of commodities in recent years in developing as well as emerging economies. Few studies support Master s Hypothesis, but majority of them did not provide any convincing evidence that the participation of index traders or excessive trading of commodity futures can be the determinants of increase in commodity prices. The recent increase in commodity prices, can be attributed to the fundamentals of supply and demand. Inflation has been high in recent years and it has negative effects on both consumers and investors. The surge in inflation necessitates the search for assets with inflation hedging potential. During inflation commodities behave differently than stocks and bonds, and the presence of very less correlation between them makes commodities a perfect addition to the conventional portfolio of stocks and bonds. In this study, we attempt to make a case for incorporation of commodity futures in the conventional portfolio and assess whether its performance is better in a high inflationary environment. Based on the absolute correlation between commodity futures return and realized inflation, top six commodities were chosen yearly from the pool of liquid commodities that were traded on two nationalized Indian commodities futures exchanges viz. MCX and NCDEX (A short note on criteria adopted for forming inflation tracking portfolios is given in Appendix J). These six commodity futures were used to form an inflation tracking portfolio a portfolio that best track the inflation. Ex-ante analysis of 63

28 conventional portfolios with and without inflation tracking portfolios shows that commodity futures do possess the property to act as an inflation hedge. During high inflationary regime presence of the inflationary tracking portfolio by conventional portfolio resulted in higher Sharpe ratios. Similar analysis was also performed on agricultural commodity futures. The results show that the presence of an inflation tracking portfolio of agricultural commodity futures with conventional portfolio did not result in higher Sharpe ratios during high inflationary years. This could be due to the excessive price volatility observed in the agricultural commodities. These ex-ante results show that commodity futures seem to possess inflation hedging properties, but their incorporation in the portfolio as an inflation hedge should be done with due care. 3.6 Limitations The present research work shows that ex-ante inflation tracking portfolios of commodity futures when incorporated with conventional portfolio results in high Sharpe ratios during high inflationary years. The ex-ante results indicate the inflation hedging potential of commodity futures, but investors will be interested in knowing specific commodities to invest that can provide a hedge against inflation. For such analysis, tools are needed to forecast stocks, bonds and commodity futures along with inflation. If forecasting of inflation and other asset classes can be done accurately, it would be possible to form portfolios that can effectively hedge inflation. 64