Value Partners Center for Investing Newsletter - The Potential of Hong Kong’s Commodity Derivatives Market

Abstract

 香港科大商學院 惠理集團 HKUST BUSINESS SCHOOL 香港科大商學院 Value Partners Center for Investing The Potential of Hong Kong’s Commodity Derivatives Market February 18, 2014 Sponsored by: 2 The Potential of Hong Kong’s Commodity Derivatives Market Executive summary In light of HKEx Group’s (HKEx) acquisition of the London Metals Exchange (LME), this newsletter analyzes the mismatch between the volume of physical commodities and the volume of commodities futures traded in Asia. We compare the commodity derivatives markets of the US and China and estimate the potential of a China-centric global commodity exchange. Even though commodity trading is big business in Asia, the majority of commodity derivatives trading is still conducted in trading centers in the US and Europe. Given the importance of physical commodity trade in Asia and especially China which is the largest consumer of commodities, there is huge potential for an Asian commodity derivatives bourse. We estimate the potential futures trading volumes on a China-centric commodity futures exchange by applying the ratio of the futures market size to the physical trade size in the US to the size of China’s current physical commodity trade. We look at current market dynamics and find that the greatest opportunities for developing benchmark futures contracts for an internationally accessible China-centric exchange exist in cotton, wheat, soybean, sugar, corn, copper, aluminum and crude oil. We also believe there is potential to create smaller yet successful regional futures markets in rice and palm oil. LME provides a well-established infrastructure and platforms for HKEx to leverage the existing metals business and to rapidly establish a world-class multi-commodity futures trading exchange in the region. HKEx should also (examine/try to emulate?) the micro-structure of successful energy and agriculture derivative markets run by CME Group and IntercontinentalExchange Group, Inc. Issues such as physical delivery, fungibility of contracts, clearing and margining, relative value trading choices and market data are important considerations for attracting and retaining trading volumes. We also suggest which futures contracts might attract the most liquidity in the region and how Hong Kong can position itself to secure a share of the global commodity derivatives trading volume. Background of commodity derivatives markets around the world Historically the major commodity derivative exchanges are located in London, New York and Chicago. Nowadays, Asia (China in particular) dominates the physical trade in many commodity markets. In 2012, for instance, the Asia-Pacific region accounted for 40% of global energy consumption1, Asia/Oceania accounted for approximately 33%, 41% and 45% of the global consumption of corn, soybean oil and wheat respectively2 and Asia accounted for approximately 66% of the global consumption of both copper and aluminum3. 1 BP Statistical Review of World Energy 2013 2 USDA statistics 3 World Bureau of Metal Statistics 4 Singapore has an active OTC commodity derivatives market, but as it is a private market, its volume is difficult to quantify accurately. Asian commodity consumers and producers generally rely upon the benchmark contracts in London, New York and Chicago for hedging purposes due to the lack of liquid alternative markets in Asia4. This gives rise to both basis risk and liquidity risk during Asian trading hours The commodity exchanges in Asia are regional in nature and do not come close to those in the West in terms of the depth of liquidity or product range. HKEx’s acquisition of LME brings together the world’s leading metals exchange and one of the largest operators of 3 exchanges and clearing houses in Asia, enabling Hong Kong to meet the need for a global commodity derivatives exchange located in Asia to close the gap between physical and derivative markets. The volume of commodity derivatives in Asia is practically insignificant compared to that (traded?) in the US, where nine of the top twenty futures contracts by trading volume across all asset classes in 20125 were commodity contracts6. We can see that among the most liquid commodity futures in 2012 traded on the Dalian, Shanghai, Zhengzhou, Tokyo and Sydney exchanges—where direct comparisons with US commodity futures can be made—only soybean meal, copper and sugar had trading volumes that were higher than those in the US7. However this is an absolute comparison and does not take into account the relative size of domestic physical commodity supply/demand. 5 CRB Yearbook 2013 6 Crude oil, natural gas, corn, soybeans, gold, gasoline, heating oil, soybean oil and wheat 7 See Appendix 1 8 http://us.spindices.com/index-family/commodities/sp-gsci Data and benchmark futures selection We retrieve data from Bloomberg, the CRB Commodity Yearbooks 2012 and 2013, and US government statistical sites. We then compare the Chinese physical commodities market with the US one based upon the relative size of the US market in order to project potential futures trading volumes on an internationally accessible China-centric commodities exchange. We suggest Asian commodity benchmark futures contracts that HKEx can focus on initially to enable rapid growth of their commodity derivatives trading platform. We regard the constituents of the S&P GSCI8, recognized as the leading commodity futures market index, as global benchmark futures. This index is world production weighted and is the most widely tracked commodity index by investors, as it employs a liquidity filter that only allows inclusion of the most actively traded commodities contracts. We do not include gold and silver, due to their quasi capital asset status. We also exclude the three livestock commodities due to their localized nature, i.e. they are not typically seaborne-traded commodities. Finally, we ignore the Europe-based Brent Crude Oil and Gasoil contracts, as WTI Crude Oil and Heating Oil are better indicators for the US market. We have also excluded cocoa and coffee as we do not have access to reliable production and consumption data. Commodities future trading volume vs physical demands in the US The US monthly energy production and consumption data is sourced from the EIA and is compared with total monthly futures volumes. The US agricultural annual production and consumption data is obtained from the USDA and is compared with total futures volumes for the same period. For metals, we take world monthly production and consumption data from the World Bureau of Metal Statistics and compare that with both the total monthly three-month forward contract and the total futures volumes for each metal. World data is considered because LME has a global warehouse system to facilitate delivery. This implies that no one country is a good representation with which to compare physical production and consumption data. Although COMEX copper is not primarily part of the S&P GSCI, we compare its futures trading volume with US consumption and production data. Figures 1 through 6 show the ratio of total futures trading volume to production/consumption volume for different commodity groups for the period from Jan 1995 to Feb 2013. Figure 1 shows the ratios for energy contracts. For crude oil, domestic supply has been much lower than domestic demand, hence the ratio of crude supply to futures trading has been consistently higher. It is worth noting that until around 2005, the futures volumes (excluding crude oil supply) have been 5 to 10 times higher 4 that systematic investing in commodity futures allows one to diversify a portfolio effectively. Commodity futures’ high correlation with inflation was another reason for their popularity among institutional investors. Hence the increase in the ratio was primarily driven by the increase in futures volumes, which was in turn due to the growth of hedge funds and Commodity Trading Advisors who implemented discretionary or systematic futures trading. With advances in trading technologies, algorithmic trading gained increasing volumes at the expense of floor traders, as demonstrated in the 2009 documentary FLOORED. 5 Figure 2: Metals futures & forwards trading volume vs. global physical supply and demand Figure 3 shows a very pronounced upward trend in the ratio for COMEX copper, even though this metal is not included in the S&P GSCI. The ratio of COMEX copper to domestic supply/demand, even though the futures contracts are included in various other commodity indexes, has been driven by both an increase in trading volumes and a reduction in physical supply and demand, the latter due to the US economic downturn beginning around the time of the subprime crisis. US production was relatively stable at ~100,000 MT per month between 2003 and 2008, decreasing steadily to around 83,500 MT per month in 2012. Meanwhile monthly contract volumes increased from about 260,000 contracts per month in 2003 to about 1.3 million contracts per month in 2012. Figure 3: COMEX copper futures trading volume vs. physical supply and demand in the US 6 As can be seen in Figure 4, the ratios of US agricultural commodities futures volumes to physical trade have shown upward trends. Production ratios (except for sugar) are slightly lower than consumption ratios due to America’s status as a large agricultural commodities exporter. Corn ratios have been relatively low because consumption has almost kept pace with production due to domestic use and the need for ethanol in auto fuel. Sugar has been the most actively traded, with the futures to physical ratio averaging 115 over the last 10 years. It is interesting to note that, with the exception of sugar, the ratios all went up in the 2006/2007 marketing year. Indeed taking the aggregate futures volumes for the five agricultural commodities we see the largest year-on-year percentage change is between 2006/2007 and 2005/2006 at +38%. This again supports the idea that institutional money (became more prevalent in/poured into?) the futures market after the concept of using commodity indexing as a portfolio diversifier gained widespread acceptance. In aggregate, we see futures volume growth every year since 2001/2002, with the exception of 2008/2009 where volumes dropped 13%, most likely due to investor deleveraging and withdrawal as a response to the global financial crisis. Across our five contracts, we also saw a 1% reduction in volumes year on year in the 2011/2012 marketing year, as increasing regulation started to impact the derivatives markets in the US. A growing ratio of derivatives trading volumes to physical trading volumes indicates higher levels of investor and speculator trading turnover, assuming producer and consumer hedgers are not becoming more active. Figure 4: Agricultural futures trading volume in comparison with physical supply and demand in U.S. Volatile relationships in China In the Chinese agricultural and metal futures markets, there has been no clear trend in the ratio of futures volumes to underlying production and consumption. This may be because onshore Chinese markets are mostly detached from international markets, implying there has been little growth in institutional or hedge fund trading participation. Intriguingly, the ratios of futures volumes to underlying supply/demand for sugar, cotton, copper and zinc are all very high. Figure 5 shows that the volume of sugar futures has been on average 227 times larger than the physical supply/demand volume in the last six years. It’s also 53 times larger for cotton, 43 times larger for soybeans, 7 times larger for wheat, and 5 times larger for corn on average. Since China is often a large agricultural product importer, the ratios are much higher against domestic production in sugar, cotton and soybean. For wheat and corn, China has had a more balanced domestic supply/demand situation, so the futures volume ratios are at similar levels for production and consumption. 7 Figure 5: Agricultural futures trading volume in comparison with physical supply and demand in China In Figure 6, we can see that the Shanghai Futures Exchange volumes (metals) have been very volatile while domestic consumption and production have been growing steadily. Copper has traditionally been the most popularly traded contract and has always dominated aluminum in terms of futures volumes and also as a ratio against the underlying physical trade. Since its introduction in 2007, the zinc contract has been averaging 92 times consumption and 97 times production. In July 2010, zinc futures volume exploded and reached nearly 45 million contracts pushing the ratios to production and consumption to 550 and 487. Since the beginning of 2012, zinc futures volumes have averaged just over 3 million contracts per month. Figure 6: Metals futures trading volume in comparison with physical supply and demand in China 8 Implication to HKEx’s commodity futures market Given the high trading volume compared to physical demand10 and the fact that China is the biggest consumer of commodities in the world, HKEx can leverage the logistical infrastructure and onshore trading liquidity supplied through the three onshore commodity exchanges in Dalian, Shanghai and Zhengzhou. In order to generate liquidity, HKEx can introduce contracts based on commodities that can be globally arbitraged and that already have benchmark contracts in current liquidity centers. This means that physical traders could lock in prices as they move shipments from one pricing center to another, and that speculators can trade relative value trades. 10 We use the terms “demand” and “consumption”, as well as the terms “supply” and “production”, interchangeably, even though some people define demand as consumption plus exports and supply as production plus imports. In agriculture we suggest that HKEx focus on cotton, wheat, soybeans, sugar, and corn. The US already exports large quantities of grains to China, and as seen above has liquid futures in these markets. Additionally, because onshore markets exist for each of these commodities, a Hong Kong-based China contract would open up arbitrage trading to all international investors. Currently, only big trading houses trade the US–onshore China arbitrage because most international investors don’t have the onshore licenses required to do so. Additionally, for a wheat contract there are additional arbitrage opportunities with Europe and Australia, where wheat futures markets also exist, which helps to supply liquidity. For sugar there are liquid futures markets in London, and to an extent, also in Brazil. In metals, HKEx should focus on increasing liquidity in the Asian trading hours for copper and aluminum. Again, the London-Shanghai markets arbitrage is currently only available to a select few traders who have onshore presence in China. If LME created a lookalike contract for Shanghai copper and Shanghai aluminum (and potentially zinc and lead), more traders would be attracted to the London-Shanghai arbitrage due to more efficient margin requirements. In energy, a contract in crude oil clearly would make sense. Until the prices of Chinese refined products become fully deregulated, product futures such as gasoline and diesel would not work. Although fuel oil has a futures contract in Shanghai, it is a niche market with the global fuel oil derivatives market based on OTC swaps, and so it may not be able to attract sufficient liquidity. According to media reports, the Shanghai Futures Exchange is set to launch a crude oil futures contract in the near future. Until recently, natural gas was a regional product and transportation was by pipeline. However, with an increase in global supplies of natural gas due to liquefaction technology, international LNG markets are slowly growing allowing natural gas to be seaborne traded. A natural gas arbitrage trade with US Henry Hub futures contract could emerge in the future, but the prices in the Asian natural gas markets are still predominately set by contract pricing and not spot pricing. However, US Henry Hub natural gas futures and UK National Balancing Point natural gas futures have relatively low correlation (unlike global crude oil, metals and agricultural benchmark prices, which have higher correlations due to the physical systems in place for global arbitrage) meaning the economics of global arbitrage in natural gas are still not prevalent. Although Asia is a huge producer and consumer of coal, listed futures markets for this commodity have not taken off anywhere in the world despite the efforts of various exchanges. The derivatives market in coal is based on cash-settled OTC swaps, with little speculator activity due to its low liquidity. There is no reason why HKEx 9 cannot be the first exchange to launch a successful coal futures contract. Within energy, a crude oil contract should take priority depending on how the LNG markets develop. The same goes for iron ore, as currently there is only a lightly traded futures market in Singapore, a growing onshore futures market in Dalian, and a relatively illiquid but growing OTC swaps market, despite the large size of the underlying market. Lookalike contracts Lookalike contracts have been relatively successful in the crude oil markets, with NYMEX launching a lookalike Brent contract and ICE introducing a lookalike WTI crude oil contract <when?>. The reason for this lies in the popularity of the Brent-WTI arbitrage spread trade. Originally, trading the spread required trading on two different exchanges, implying that, despite the high correlation, two sets of margins were required for mark to market daily settlements. Normally, when a trader trades a highly correlated offsetting contract, he will be given a certain amount of relief on margin against the offsetting trade, which makes trading more capital efficient. However, this only happens when the contracts are on the same exchange, and highly correlated, as is the case with time spreads and product spreads. To win local traders from other derivatives market’s trading, the WTI-Brent spread, NYMEX and ICE introduced lookalike contracts which have exactly the same contract specifications as the original contract, but trade on their own exchanges. This means relative value traders can obtain the benefits of margin offset and trade more efficiently. HKEx could introduce this feature once it becomes clear its clients were spread trading their contracts against rival exchanges’ contracts. Niche markets of Asia: Rice and Palm Oil Neither rice nor palm oil has made it into the S&P GSCI, but both China and Asia have huge physical markets in these two commodities, as they are regional staples. In the 18th century rice futures in Osaka were said to be the first futures contracts traded and two rice futures contracts still trade on the Osaka Dojima Commodity Exchange now albeit with low liquidity. Illiquid rice contracts are also listed on the Agricultural Futures Exchange of Thailand with market participants bemoaning the political risk in domestic Thai rice prices, which the government manipulates to keep farmers happy. Indeed, rice trades on the Zhengzhou Commodity Exchange also suffer from relatively low liquidity and are off-limits to most international investors. A Rough Rice contract trades on the Chicago Board of Trade, which could be a basis for hedging, although this contract is not as liquid as the more established benchmark agricultural contracts. Therefore, there is a desperate need for a liquid and credible Asian rice contract for producer and consumer hedging. Palm oil does not trade on any of the big commodities exchanges in the West. However, there is a successful crude palm oil contract denominated in Malaysian Ringgit on the Bursa Malaysia. There is also a palm oil contract on the Dalian Commodity Exchange. Given that Indonesia and Malaysia are the world’s two major producers and India and China are the world’s two largest importers, a USD or CNH contract out of Hong Kong would have wide appeal across Asia and would be able to compete with the Malaysia contract. Currency domination of contracts Contracts could be denominated in USD, even though it would restrict some (but certainly not all) onshore liquidity, in order to capture market share. A CNH market may be more attractive to onshore traders, but CNH liquidity is still relatively low and this would discourage offshore traders, who are accustomed to trading mostly in USD. The solution may be to have each contract listed in both USD and CNH. Although this would divide liquidity across two contracts, it would also add liquidity because arbitrageurs (with FX expertise) would trade 10 both contracts simply to keep them in line in theory, although bid-offer spreads might make this difficult in practice. If the contracts were fungible, overall liquidity would be expected to increase. Expected size of China-focused contracts Assuming an highest-trading volume for each commodity, no growth in China’s own consumption and production from 2012 levels, and contract sizes equivalent to those on the US exchanges or LME, we could potentially see annual trading volumes as follows: 1. Crude Oil: 93m contracts vs. benchmark 141m 2. Cotton: 36m contracts vs. benchmark 5m 3. Wheat: 91m contracts vs. benchmark 30m 4. Soybeans: 43m contracts vs. benchmark 52m 5. Sugar: 37m contracts vs. benchmark 25m 6. Corn: 50m contracts vs. benchmark 78m 7. Copper (based on COMEX/US numbers): 88m contracts vs. benchmark 16m 8. Copper (based on LME/world numbers): 18m contracts vs. benchmark 49m 9. Aluminum (based on LME/world numbers): 42 contracts vs. benchmark 95m

The ingredients to a successful commodity futures exchange Commodity derivatives markets are successful if natural sellers, natural buyers and speculators all participate in them. Contracts that are successful tend to stay successful (for the most part), and contracts that do not take off tend to die sooner or later. It is important that HKEx launches the suggested contracts because these commodities are established on other major exchanges and so can be arbitraged against their natural spread “partners” on the other exchanges. To reduce directional risk and/or to take advantage of a particular informational or logistic edge, traders often like to trade spreads (long one market vs. short a related market). It could be intra-market spreads (aka time spreads) such as winter vs. summer, or inter-market spreads such as the “crack spreads” crude vs. gasoline and crude vs. heating oil, or gasoline vs. heating oil, which may reflect a view on refining dynamics. Other popular inter-market spreads include the “crush spread” which refers to soybean vs. a combination of soymeal and soy oil. Other popular trades include inter-commodity trades such as soybeans vs. grains, because although soybeans and wheat/corn cannot be used directly interchangeably, they have common supply and demand characteristics due to their crop cycle, protein content, use in foods and animal feed, etc. Initially, HKEx should focus on the most established delivery centers for physical delivery; probably replicating the current onshore markets because physical flows are established so there is already hedging demand. With time, additional markets can be introduced with different grades or geographic delivery points, in order to introduce inter-China spread plays. These would have to replicate common trade flows, where the hedging of point A vs. point B would be useful for traders looking to lock in prices and the spread before transporting the commodities from A to B. In the US, the fact that the ICE corn contract failed to attract liquidity away from the incumbent CBOT contract shows the importance of a critical mass of liquidity and the difficulty of attracting liquidity away for essentially the same contract to a new exchange. This is where a dual currency contract may help because if traders can switch contracts easily from RMB to USD, they can then trade a spread against international exchanges for international arbitrage. 11 With a 24-hour market, spread trades vs. US or European contracts can be executed more easily, but liquidity may be spread out throughout the day, and not concentrated in more liquid windows of time. The western markets have various catalysts to spur trading activity, for example USDA crop reports for agriculture, or the DOE numbers for energy, in addition to the general raft of economic data. For China, if credible market data can be introduced on a regular basis, by either the government or the private sector, it could result in a window of regular liquidity directly after the relevant statistics are released. HKEx can commission a data vendor to calculate and supply regular credible inventory or supply and demand numbers for whichever commodities they introduce in order to stimulate this liquidity boosting feature of the market. Conclusion HKEx’s purchase of the London Metals Exchange (LME) provides puts it in the perfect position to satisfy the huge demand for commodity futures trading in Hong Kong and Asia. This demand is largely driven by the fact that China has become the biggest consumer of commodities in the world. HKEx can leverage LME’s well-established infrastructure and platform to establish a world-class commodity futures trading exchange in the region in no time. 12 Value Partners Center for Investing The Value Partners Center for Investing of the Hong Kong University of Science and Technology Business School is an academic and intellectual center supporting research and training on investing with an emphasis on China and Hong Kong financial markets. It aims to promote Hong Kong's role as the regional asset management center. The center is sponsored by Value Partners Group Limited. Value Partners Group Value Partners is one of Asia’s largest asset management firms. Since its establishment in 1993, Value Partners has been a dedicated value investor with a focus on the Greater China region. The Group manages absolute return long-biased funds, long-short hedge funds, exchange-traded funds, quantitative funds, and private equity funds for institutional and individual clients in Asia Pacific, Europe and the United States. Note to readers ICE introduced lookalike contracts for crude oil (WTI) and natural gas, although we have not accounted for their volumes. We assumed they are predominately traded as spreads against ICE Brent crude oil and ICE UK natural gas, and thus do not represent consumer/producer physical hedging. We have not accounted for “mini” contracts which target retail traders. We have also not included the new ICE corn contract, which is a lookalike of the CME corn contract. For RBOB Gasoline, we used NYMEX Unleaded Gasoline contract prior to RBOB’s introduction, and the sum of the two futures’ volumes in 2006, which was the year when <what was handed over to what?>. For wheat, we combined Kansas and Chicago volumes. Appendix 1—2012 volumes, Asia vs. US To demonstrate the size of liquidity in Asia, we look at the liquid contracts on Asian exchanges and compare them with the global benchmark. We have made the adjustments for contract size and measurement unit for comparability. Appendix 2—Commodity Sectors Commodities broadly fall into two groups: hard and soft. Hard commodities consist of energy and metals. Soft commodities, which are usually weather dependent, can be broken down into agriculture and livestock. Author Contact Information Dr. Samuel Xin Liang Associate Director Email: samliang@ust.hk Mr. Thomas Streater Research Assistant Email: vpts@ust.hk Value Partners Center for Investing Tel: +852 2358 8204 Fax: +852 2358 1749 Supporting Researchers: Ms. Cheuk, Man Yin Ms. Fang, Lei