Commodities Total Return

Transcription

1 Commodities Total Return Commodities Total Return Ed Peters, Jesse Davis December Ed Peters, Partner, Co-Director of Global Macro Jesse Davis, CFA, Director, Research In this article we describe a new method for investing in commodities. This method combines a macro approach designed for generating long-term returns to hedge expected inflation with a fundamental view on how commodities should perform on a relative basis. The macro based allocation strategically weights commodities by their risk contribution and targets optimal diversification, and these weights vary depending upon the changing risk environment and the business cycle at the macro level. In the near-term, fundamental views based upon the long accepted Theory of Storage increase positions in commodities that are expected to outperform and reduce weight in those expected to lag. In addition, the very structure of the futures markets influences the way we need to trade. We make a compelling case for this Commodities Total Return (CTR) method as a far more investor appropriate construct than other commodity portfolios because it approaches commodity investing from all levels and from a market perspective. CTR seeks a total return that improves upon the long-term expected inflation hedging goal rather than accepting an arbitrarily developed index, and can add value from nearer-term inefficiencies in the commodity futures markets themselves. CTR is designed with the goal of offering these risk and return characteristics while still providing the same, if not better, expected inflation hedging and diversification with stocks and bonds than more conventional approaches to commodity management. 1FQ COMMODITIES 1 The Issues with Commodity Investing There is a strong consensus that commodities should be a part of any well diversified portfolio both to hedge bonds against the risk of expected inflation as well as participate in inflationary growth. While equities offer exposure to economic growth, inflation can substantially erode earnings. Commodities are able to provide an excellent counterpoint, as they hedge against the expected inflationary erosion of bond principal and equity earnings. Commodities can also significantly increase risk-adjusted returns when added to portfolios of stocks and bonds. However, traditional commodity investing can be problematic for many reasons. First, it is impractical for most investors to buy physical commodities, so the only alternatives are derivatives such as futures or swaps, or exchange-traded products such as ETFs and ETNs. Of these, futures offer the widest variety of possible investments, since swap and ETF exposure is limited to the most liquid commodities. ETFs and related instruments are cash intensive and are also a more expensive and less flexible way to gain exposure to commodities than either swaps or futures. The primary benefit they offer is that they are exchange-traded (like futures) instead of being traded over-the-counter (like swaps). Swaps, and those ETNs that are implemented via swaps, also have embedded credit exposure. In the end, futures offer the most flexibility and liquidity on the most continuous basis for the least cost, with no hidden credit exposure and low cash usage. Commodities trading involves substantial risk of loss. Past or simulated performance is no guarantee of future results. Potential for profit is accompanied by possibility of loss.

2 2FQ COMMODITIES 2 Second, commodities, unlike financial futures, are quite dissimilar from one another and developing a uniform approach to investing in commodities is difficult. For instance, some commodities, like agricultural commodities, are renewable while others, like gold, will always be in limited supply. Some are tied directly to the economy, such as energy and industrial metals, while others, such as livestock, have only a tenuous relationship to the broader economy. Some are difficult to store and are perishable, while others are inexpensive to store and last for decades, or in the case of gold, forever. So a common approach to investing in commodities based upon valuation, for instance, is a challenge. Third, due to these dissimilarities, there has not been consensus on how to create long-term strategic weights for a commodity investment. The most popular indices use production weighting which puts an undue weight on industrial commodities like energy rather than agriculture and precious metals. This makes the indices less likely to hedge against food inflation or economic uncertainty. Yet strategic weights are critical if inflation hedging and participating in economic growth are goals of the commodity investment. Unfortunately, active managers use these admittedly flawed commodity indices as the long-term strategic component of their returns. In addition, most commodity managers do not use a fundamental approach to commodities and instead depend upon momentum and various technical measures. While these approaches can work some of the time, they cannot deliver the confidence investors receive from an investment philosophy tied to fundamental aspects of the asset class. The Inefficiency of Conventional Commodity Indices It is common to look at the diversification of a portfolio or index through the number of assets and the capital weight of those assets. Two indices, the Goldman Sachs Commodity Index (GSCI) and the Dow Jones UBS Commodity Index (DJ UBSCI), are the most popular commodity indices used as benchmarks. The GSCI is frequently used because it has a liquid cash settled futures contract, so it offers an easy way to passively invest in commodities. The GSCI weights commodities according to their worldwide production which gives it a 71% capital weight to energy. This raises questions about the diversification of the index, so the DJ UBSCI modifies the weighting scheme by placing maximums on sectors and individual commodity as well as taking into consideration the liquidity of commodities markets. As a result, the DJ UBSCI has a 33% weight in energy and so looks on the surface to be more diversified. Both indices have a large number of commodities in them. The GSCI has 24, while the DJ UBSCI has 19. However, if we look at the risk budgets of each index, neither looks diversified at all. There are five primary commodity sectors: agriculture, energy, industrial metals, precious metals, and livestock. The capital weights for both GSCI and DJ UBSCI are determined once a year. Exhibit 1 shows the capital weights of the GSCI and DJ UBSCI on the left while the right shows the risk budget weights. EXHIBIT 1: ALLOCATING TO COMMODITIES: Capital versus Risk Capital Allocation Sources: First Quadrant, LP, Global Financial Data (GFD) Risk Allocation 2% 1% 17% 4% 2% 17% % 71% 15% 4% 7% 3% 2 11% 33% 29% % 61% S&P GSCI DJ UBSCI S&P GSCI DJ UBSCI Agriculture Energy Precious Metals Industrial Metals Livestock DJ UBSCI: Dow Jones UBS Commodity Index S&P GSCI: S&P Goldman Sachs Commodity Index The GSCI is almost entirely an energy index. Not only is 71% of its capital invested in energy, over 9 of its risk budget is in energy. The other sectors make a small contribution to the index. Precious metals, for instance, are virtually non-existent, so those who consider gold an important investment receive virtually no exposure through the GSCI. The DJ UBSCI is better, but still 61% of its risk budget is given to energy with the balance primarily in agriculture and industrial metals. Precious metals and livestock again receive virtually no weight. Exhibit 2 shows the return contribution of the different sectors from 1/88 12/11, and we can see that the return contribution of each sector is very similar to their risk weights. This is no coincidence. Research has shown that there is a strong relationship between risk and return contribution in a portfolio. From this analysis we can see that the most common commodity indices do not really give true diversification, but are energy heavy both in their returns and risk contribution. As such, they are inadequate as the long-term strategic weights for a commodity strategy. Commodities Total Return December 2011 Page 2 of 10

3 3FQ COMMODITIES 3 EXHIBIT 2: ATTRIBUTION Risk Allocations 2% 17% 95% S&P GSCI 61% 2% 17% DJ UBSCI 1% 4% Sources: First Quadrant, LP, Global Financial Data (GFD) Excess Return Contribution S&P GSCI DJ UBSCI Agriculture Energy Precious Metals Industrial Metals Livestock Commodities Total Return 5% 81% 9% 2% 3% 14% 1 53% 2 Most commodity managers use one of the above two indices as a benchmark for investment. That is, they believe that their job is to add value to a passive investment in the index. We take a different view. Instead of trying to enhance something that is fundamentally flawed, we created a strategy that is driven in all respects by the inner workings of the commodity markets. A common approach to investing outside of commodities is to first study all aspects of the relevant assets and then use this knowledge to combine the assets into a portfolio. But this is, in fact, a relatively new idea in commodities at an institutional scale for the following reasons. First, commodities as an institutional investment are a relatively new phenomenon. Twenty years ago, most institutional funds were not invested in commodity markets, whereas today it is commonplace. Flows into the asset class have been extraordinary. In the past decade alone the cumulative investment has increased by almost 10 times. Due to this explosive growth, the race among investment managers has been one of building assets and capacity rather than creating strategies to best capture the unique aspects of the asset class. Second, the existing standard indices described above originally came at the market from the perspective of an aggregate descriptive index and not from the perspective of an optimal investment. Although indices often become investment benchmarks, they do not necessarily attempt to combine assets in a way that might be considered optimal for an investor. For example, the common equity index 3% methodology of market capitalization weighting or issue weighting in bond indices is identical to the production weighting tenet used by most commodity indices. In all cases, assets receive a weight proportional to the weight of the total available amount of that asset. Other critical investment features, such as risk, correlation, predictability, or expectation of return, are not considered. Indices offer a shortcut for describing an asset class as a whole, while offering limited guidance as to what an investor should be doing to capture the returns inside that asset class. Third, commodity markets have been evolving rapidly in the past decade. Until recently, there was not enough liquidity to make a market driven approach feasible. But the recent growth of the asset class itself has increased the number of potential strategies that can be applied to the asset class. The almost 10 fold growth of flows turns into an almost 10 fold growth of strategy capacity. For managers that run a more blunt, index style strategy, this is good news, but it is even better news for investors. It means investors now have access to a new set of diversified and uncorrelated strategies which could not previously exist in any meaningful way. Investors can now choose strategies that better suit their investment needs and more optimally achieve their goals. Due to the complacency which has led to the dominance of traditional commodity indexing, commodity markets have not been studied as an investment as thoroughly as other more traditional asset classes. When we take the time to understand this asset class, we find that there are many unique features of commodities that don t exist in other asset classes. We gain different insights, for example, by focusing on the characteristics of commodity markets over different investment horizons. Once these features are recognized and understood, we have the tools necessary to create a strategy for investment. The Supply/Demand Horizon: Fundamental Effects First, we examine a near-term horizon on the order of six to eighteen months. While commodities are diverse and uncorrelated as a group, they respond broadly to macroeconomic risks (a topic we will discuss below), but at a nearer-term horizon, commodities respond to their own bottom up fundamentals of supply and demand. Within this time-scale, we find that the features most relevant to investors are ones that drive commodities relative to each other. In other words, from a bottom up stand point, it is most effective to analyze how particular commodities respond relative to other commodities, rather than how they respond in an absolute context. It is at this investment horizon that the major published academic theories of commodities apply. Page 3 of Past 10 performance Commodities Total is no Return guarantee December of future 2011 results. Potential for profit is accompanied by possibility of loss.

4 4FQ COMMODITIES 4 The primary centerpiece of academic commodity theory is The Theory of Storage. This almost century old theory describes several fundamental features of commodity markets. Aspects of the theory have been developed and tested by a veritable army of researchers over the decades up to the present from Keynes (1930), through Kaldor (1939), Working (1948 and 1950), Brennan (1958), and Fama and French (1988). It has become an enduring and consistent underpinning of our understanding of the inner-workings of commodity markets. The three main conclusions of the theory relevant to this discussion are: 1) Commodity prices rise as supply falls relative to demand, 2) The volatility of commodity prices rises as supply falls relative to demand, and 3) The basis of commodity prices rises as supply falls relative to demand. (The basis of a commodity is the difference between its current and future values.) This theory gives us tools to make statements about expected commodity prices as they relate to the current or future supply/demand balance. Putting this together in a relative context, suppose we had information about the supply/ demand balance for wheat, and we knew that as a whole it was in under-supply relative to corn. We can now make a statement about the relative behavior of wheat and corn. In this case, we would expect wheat to out-perform corn. We would not necessarily know the absolute direction of wheat and corn prices (that would depend on a host of other macroeconomic factors as well), but we could say with some degree of certainty that wheat should relatively out-perform corn until the fundamentals changed. The Strategic Horizon: Macro Effects Commodities also respond to macro-economic effects at longer horizons, varying broadly with the major portions of market cycles and also display top-down structure across market cycles. At first thought, we might not expect to find useful information over a full market cycle other than a general trend, but in fact, there are two important features of commodities that jump out. At this longer-term horizon, first, commodities have a weak asset class identity. Commodities are an exceptionally diverse group of assets that have a lot less self-similarity than other assets such as equities. For example, from October December 2011, the average pair-wise correlation between equity sectors is near 0.5, but the average pair-wise correlation between commodity sectors is near 0.1. Looking at this over time in rolling threeyear windows, the minimum pair-wise correlation between equity sectors is about 0.3 whereas the maximum pair-wise correlation between commodity sectors is about 0.4. So while equity sectors are decent proxies for each other, commodity sectors are not. This means an investor must take more care when determining appropriate weightings for a commodity portfolio. A second important observation is that while commodity returns are difficult to forecast over the business cycle, commodity risks are surprisingly stable from period to period. For example, if we rank the commodity sectors each year by their returns and separately by their risks, we find that from , the correlation of the year-to-year ranked returns is 1%, whereas the correlation of the ranked risks is 67%. In other words, although we cannot count on the stability of returns, we can count on the stability of risks. We also find that commodities respond to the same global macroeconomic risks as equities. Commodities are often classified as risk assets (as are equities and certain currencies), which means that they suffer during times of heightened market risk. Risk here can be defined in a variety of ways, but our research has concentrated on four primary measures: 1) The VIX index, 2) Corporate credit spreads, 3) Global PMI (Purchasing Managers Index), and 4) Global monetary policy. Roughly, if the VIX or credit spreads are high or global PMI is low and global monetary policy is tight, we define that as a high macroeconomic risk environment, and in those high risk environments, we find that risk assets, including commodities, do poorly. Conversely, if the VIX or credit spreads are low, or global PMI is high and global monetary policy is loose, we find that risk assets, including commodities, do well. The effects are more subtle inside the commodity asset class as well. Sectors like the energies and industrial metals respond more to the business cycle than sectors like agriculture and livestock. In a hierarchy of needs, a person must eat before they must buy a new car or build a new house. We have previously documented this effect in Using Volatility Regimes and we combine these measures into an indicator we call the FQ Market Risk Index (MRI). It has five levels of risk. In another paper, Peters (2011), we show that from December 1934 to February 2010 the DJ AIG Spot Index (a commodity index) volatility was influenced by equity volatility regimes ranging from an average volatility of 9.39% in low volatility regimes to an average volatility of 15.84% in high volatility regimes with an average volatility of 13.01% over the entire Commodities Total Return December 2011 Page 4 of 10

5 5FQ COMMODITIES 5 EXHIBIT 3: SHARPE RATIO FOR INDICES S&P GSCI DJ UBSCI Overall High Volatility Regime Low Volatility Regime Sources: First Quadrant, LP, StyleAdvisor, Bloomberg LP 76 year period. In Exhibit 3, we show performance for a more recent period for the GSCI and DJ UBSCI. Annualized Risk (%) Annualized Risk (%) Here we define the regimes as periods when the three month moving average of the VIX is above or below its long-term median of 19. As we can see from the chart on the right, the risk of the commodity indices is affected by the VIX regimes which really measure levels of economic uncertainty. On the left, we also see that the levels of excess return per unit of risk (the Sharpe ratios ) are also influenced by these volatility regimes. In high volatility regimes, the Sharpe ratios are negative, while in low volatility regimes, they are positive. So commodities, like equities, give the most return per unit of risk when investors are passively holding the least amount of risk if they hold their capital allocation static. This implies that a dynamic allocation based upon the risk regime will not only control volatility more effectively, but enhance return as well. The Trading Horizon: Futures Market Micro-structure Finally, at a trading horizon of zero to six months, commodity markets are a bit more complex than other asset classes due to the structure of the commodity futures markets themselves. Whereas futures markets for assets such as equities, bonds, or currencies have liquidity concentrated in the nearest contract to maturity, commodities have liquidity spread out over many maturities. This makes sense from a fundamental stand point since a producer of finished goods may have enough of a raw input to last through a time period like the next six months, but may also want to lock in the price of that raw input after that. Another producer may carry only one month of inventory of raw input, and so may want to lock in the price just one month out. This spreading out of liquidity allows some interesting features to develop in the markets involving the term structures of price, liquidity, and risk S&P GSCI DJ UBSCI First, prices of nearer-term futures may be higher or lower than prices of further out futures. These effects are described as a commodity being in backwardation (near > far) or contango (near < far). If one expects the shape of this price term structure to hold, and say is taking a long position, it is advantageous to have a backwardated market since we could expect the futures price to naturally roll up the curve as a future gets nearer to expiration. This is called a positive roll yield. Contrarily, contangoed markets offer a negative roll yield, so if we would like to take a long position, we can expect to lose the value of the roll yield. That would be the end of the story if backwardation and contango were linear, but they are generally non-linear with price curves generally getting steeper as we approach expiration. This means that if we took a long position in a backwardated market, we would prefer being closer to maturity where we benefit the most from positive roll yield. In a contangoed market, we would rather be farther from maturity where we would suffer the least from negative roll yield. Second, liquidity of commodity futures exists out many months and even years in certain cases, but generally decays as we get further from maturity, and we must also take this into consideration when deciding which contract to trade. For example, if we would like to hold a long position in a contangoed market, we would prefer a future that is farther from maturity. However, we have to take into account the lower liquidity of the more distant contracts which may generate excessive transaction costs or concentration in a market and maturity. There are nuances in this liquidity term structure as well. For example, in the northern hemisphere mid-summer, a future on wheat or corn may have more liquidity than a late spring maturity due to common crop cycle effects. Additionally, an end of the year future in crude may have more liquidity than a mid-fall future due to behavioral effects of market participants. A deep understanding of this dimension is clearly important. Finally, the risk of different commodity contracts also has a term structure. This risk term structure is famously known as the Samuelson effect after Paul Samuelson (1965), the MIT and Nobel prize winning economist who first described it. Samuelson showed that for purely stochastic reasons, the volatility of a contract farther from maturity is lower than the volatility of a contract nearer to maturity. This effect is straightforward to measure. Across a wide set of commodities, the average volatility of a future 12 months from maturity is about 2 less than a future one month to maturity. In other words, if the annualized volatility of a one month to maturity future is 15%, the annualized volatility of a 12 month to maturity future would be roughly 12% on average. Understanding this effect is critical to properly Page 5 of 10 Commodities Total Return December 2011

6 6FQ COMMODITIES 6 construct a commodity portfolio in a risk managed context. Following on our previous example, a long position farther out into the future may require a larger notional position to achieve the same risk. Tying the price, liquidity, and risk term structures together is a complex balancing act, and as such is a critical element to creating a Commodity Total Return portfolio. Constructing the CTR The Commodity Total Return (CTR) strategy takes the above three primary dimensions of commodity markets into account to create a market driven approach to commodity investing: 1) The Strategic Horizon: Allocations dynamically adapt to changes in the risk environment and macro effects over the business cycle, 2) The Supply/Demand Horizon: Near-term views based on the laws of supply and demand as incorporated in The Theory of Storage further adjust positioning, and 3) The Trading Horizon: Implementation management that considers the liquidity, risk, and price term structures when purchasing or selling individual commodity futures is applied. We believe each dimension improves the performance of the portfolio. The methodology is straightforward. First, an overall risk target is set for the portfolio. One of the issues with a commodity portfolio is that the low correlation among the commodities limits the amount of risk that a portfolio can achieve if we limit the weights to a total of 10. A true risk balanced commodity portfolio without leverage would produce a risk of 8% or less. That could, in fact, be a reason that other less diversified weighting schemes are commonly used. However, Commodity Total Return is a cash and futures portfolio, so we are not limited to a total weight of 10. For instance, a risk balanced portfolio with a 12% risk target requires a leveraged weight of about 111%. However, this 12% portfolio can be more diversified and achieve lower total risk than the GSCI or DJ UBSCI as we will see below. For the purposes of this analysis we will use a 12% risk target. Once our risk target is chosen, we risk balance within the sectors and across the sectors with target weights depending upon the current level of the FQ Market Risk Index. This is where the leverage comes in. Precious metals and livestock are two of the more diversifying commodity sectors because they have low and often negative correlations with the other three. Yet, they are also the lowest risk sectors, so the only way to risk balance the portfolio is to leverage the precious metal and livestock sectors. Total leverage is determined by the current MRI level with the most leverage during the Very Low Risk regime of 148% and a de-levered portfolio in the Very High Risk regime of 89%. It is important to remember that the portfolios in both regimes are designed to deliver 12% annualized volatility in their respective regimes. Next, we modify the strategic positions by taking into account the relative attractiveness of the various commodities using fundamental supply/demand models. There are many perspectives on a market s fundamentals, but we group them into three major themes: 1) Market implied fundamentals, 2) Projected fundamentals, and 3) Current fundamentals. The market implied fundamentals estimate the fundamentals currently being priced into the market. We filter out the noise caused by speculators in order to observe the fundamentals implied by the commercials, the companies that are actively a part of the production and consumption supply chain of a commodity. Projected fundamentals are provided by both governmental and industry group sources for numerous markets. These projections are readily available because they are updated and published on a regular basis. Finally, we examine the current fundamentals of a market commonly reported by industry groups or aggregated by data providers. Taking these different perspectives into account adds diversity to our market assessment and ultimately leads to more accurate market measurements. Finally, when trading we look at the attractiveness across the various maturities to determine the best contracts to buy and sell to take advantage of the various term structure opportunities. Simulation¹ The Commodities Total Return was simulated using data from January 1988 to December 2011 targeting a total risk level of 12%. What follows is a discussion based on the observations made from the Commodities Total Return simulation (henceforth CTR ). Exhibit 4 updates Exhibit 1 to include the return attribution of the CTR s Macro Effects along with the GSCI and DJ UBSCI and shows the benefits of diversification that comes with long-term risk balanced allocation as opposed to the production weight focus of the two indices. Commodities Total Return December 2011 Page 6 of 10

7 EXHIBIT 4: ATTRIBUTION Agriculture Energy Precious Metals Industrial Metals Livestock Sources: First Quadrant, LP, Bloomberg LP EXHIBIT 5: COMMODITIES TOTAL RETURN: Impacts on Risk Annualized Risk (%) Annualized Risk (%) 2% S&P GSCI 5 0 Risk Allocations 2% 17% 4% 95% 61% 17% DJ UBSCI 1% 21% 15% 19% 22% 23% FQ CTR Macro Effects Simulation 1 Excess Return Contribution 10 2% 3% 5% 9 14% 17% 3% % 6 25% 5 81% 4 53% 3 27% % 9% S&P DJ FQ CTR GSCI UBSCI Macro Effects Simulation 1 S&P GSCI DJ UBSCI FQ CTR Simulation¹ Overall High Volatility Regime Low Volatility Regime EXHIBIT 6: COMMODITIES TOTAL RETURN: Impacts on Risk-Adjusted Return, Overall High Volatility Regime Low Volatility Regime S&P GSCI DJ UBSCI FQ CTR Simulation¹ Sources: First Quadrant, LP, StyleAdvisor, Bloomberg LP EXHIBIT 7: COMMODITIES TOTAL RETURN: Adding Value Step-by-Step Simulation: S&P GSCI Step 2: Macro Effects Step 3: Fundamental Effects 1.0 FQ Commodities Total Return DJ UBSCI Progression presents the additive 7FQ COMMODITIES 7 Sources: First Quadrant, LP, StyleAdvisor, Bloomberg LP The CTR has a much more even return attribution by sector reflecting its more diversified risk composition. Exhibit 5 shows the stabilizing effect of using volatility regimes to manage risk. The blue bars show the average annualized risk over the period, the yellow bars show the average annualized risk in the high volatility regime, while the green bars show the average annualized risk in the low volatility regime. In the case of the GSCI and DJ UBSCI, we can see that realized risk follows the regimes, while the CTR shows more stable risk across the two regimes. Exhibit 6 shows the Sharpe ratios across regimes and illustrates that whatever excess return the GSCI and DJ UBSCI earn, it comes during the low volatility regime but tends to give it up in the high volatility regime. CTR has positive Sharpe ratios in both regimes, though it is still higher in the low volatility regime. Sources: First Quadrant, LP, StyleAdvisor, Bloomberg LP Finally, Exhibit 7 shows how the fundamental and macro portions of CTR each improves the Sharpe ratio of the portfolio. We can see that the CTR has more favorable risk/return potential than the GSCI and the DJ UBSCI, but does it still have the diversifying effects that investors expect of commodities? Exhibit 8 shows the correlations with stocks and bonds for all 3 strategies and we can see that the CTR has a similar small positive correlation with stocks and negative correlation with bonds that the other two indices have. So from a correlation standpoint, the CTR can offer similar diversifying characteristics. Aside from diversification to stocks and bonds, does CTR preserve other benefits of an allocation to commodities, namely inflation hedging? We can measure inflation in multiple ways, but the Consumer Price Index (CPI) is a Page 7 of 10 Commodities Total Return December 2011

8 8FQ COMMODITIES 8 EXHIBIT 8: DIVERSIFICATION BENEFIT Correlation Sources: First Quadrant, LP, Bloomberg LP EXHIBIT 9²: INFLATION HEDGING US Inflation S&P GSCI DJ UBSCI FQ CTR Simulation¹ Correlation % 28% World Stocks 13% Downside Correlation Sources: First Quadrant, LP, Bloomberg LP -13% -16% -14% World Bonds S&P GSCI DJ UBSCI FQ CTR Simulation¹ Upside Correlation common indicator. Exhibit 9² shows upside and downside correlations of commodity strategies to CPI. We see that CTR in fact enhances the correlation to positive moves in CPI and drastically reduces the correlation to negative moves in CPI; i.e. CTR gives stronger inflation hedging when it s needed on the upside, but doesn t suffer nearly as much as other commodity strategies on the downside. Finally, Exhibit 10² shows data for the last primary feature of a commodity allocation that we would like to preserve. Commodities are expected to provide direct participation in global growth, and so one expects that commodities will be positively correlated to global growth. What we find, however, is that common commodity strategies are much more strongly correlated to growth when it is declining, than when it is rising. Although correlation to growth is pointed to as a beneficial feature of commodities, if the correlation is coming dominantly from the downside, it is not actually all that beneficial. CTR enhances the way in which a commodity allocation can be more beneficially correlated to growth. As we see from Exhibit 10, CTR drastically decreases the EXHIBIT 10²: ECONOMIC GROWTH US GDP 0.70 Correlation correlation to global growth on the downside at the expense of giving up only a small amount of correlation to growth on the upside. The amount of downside protection received is approximately 6 times the amount of upside benefit given up, and so the tradeoff is clearly beneficial. When would the Commodities Total Return underperform the GSCI and the DJ UBSCI? Whenever energy far outstrips the other sectors, then portfolios more concentrated in energy, like the GSCI and the DJ UBSCI, will likely outperform depending upon the performance of Commodities Total Return s near-term opportunity signals. At other times, we can expect that diversification and an approach that considers the market s unique characteristics will win out. Summary S&P GSCI DJ UBSCI FQ CTR Simulation¹ Downside Correlation Upside Correlation Commodities are accepted as important components of a diversified asset allocation portfolio. Most commodity managers offer an index beta coupled with an alpha based upon over and underweighting individual commodities relative to that index, and this is often driven by technical signals like momentum. The index still generally accounts for a substantial amount of the total return. A more rational approach is to construct a better long-run portfolio, take into account changes in risk over the business cycle, and then adjust these weights based upon nearer-term Sources: First Quadrant, LP, Bloomberg LP Commodities Total Return December 2011 Page 8 of 10

9 opportunities driven by fundamentals. Commodities Total Return does just that. A strategic portfolio is constructed that is designed to be well diversified across all sectors of the commodity markets. These weights take into account changes in macro market conditions relating to the business cycle and volatility. Positions are modified for near-term opportunities that come from fundamental supply/demand considerations. Finally, a multi-factor assessment of the term structures of commodity futures markets is undertaken in order to implement these strategies in an efficient manner. Combined together, as demonstrated earlier in the paper, the Commodities Total Return is designed to achieve strong risk-adjusted returns combined with significant diversification of traditional stock and bond holdings and superior hedging characteristics. Endnotes 1 Simulation is supplemental Information. Please see Commodities Total Return Simulated Performance (Gross of Fees) and Commodities Total Return Strategy Composite Information and Commodities Total Return Strategy disclosures found at the end of this document for information concerning the simulation, the live composite, and the effect of fees on the performance. ² US GDP was used due to the historical data availability at a quarterly granularity. US inflation data was used for consistency with the GDP data. References Brennan, M. The Supply of Storage, American Economic Review (1958), 47(1), pp Darnell M., Peters E., Ye J. Rethinking Beta, FQ Perspectives (December 2008) Erb C., Harvey C., The Strategic and Tactical Value of Commodity Futures, Financial Analysts Journal (March/April 2006) Fama, E F. and French, K R. (1988). Business Cycles and the Behavior of Metals Prices Journal of Finance (1988), 43(5), pp Goldwhite, P. Diversification and Risk Management: What Volatility Tells Us, FQ Perspectives (October 2008) Keynes, J M. A Treatise on Money, Volume II: The Applied Theory of Money, London: Macmillan, 1930, pp Kaldor, N. Speculation and Economic Stability, Review of Economic Studies (1939), 7(1), pp Peters, E. Balancing Betas: Essential Risk Diversification, FQ Perspectives (February 2009) Peters, E. Using Volatility Regimes: the FQ Market Risk Index, FQ Perspectives (September 2009) Peters, E. The Balanced Risk Commodity Strategy and the Market Risk Index, FQ White Paper, 2011 Samuelson, P. Proof that Properly Anticipated Prices Fluctuate Randomly, Industrial Management Review, 6 (1965), pp Working, H. Theory of the Inverse Carrying Charge in Futures Markets, Journal of Farm Economics (1948), 30(1), pp Working, H. The Theory of Price of Storage, American Economic Review, 39(6), (December 1949), pp FQ COMMODITIES 9 This material is for your private information. The views expressed are the views of First Quadrant, LP only through the period ended December 2011 and are subject to change based on market and other conditions. All material has been obtained from sources believed to be reliable, but its accuracy is not guaranteed. Updated results on analyses available upon request. First Quadrant, LP 800 E. Colorado Boulevard, Suite 900, Pasadena, California Marketing Services: info@firstquadrant.com Copyright by First Quadrant, LP, 2011, all rights reserved. Past performance 800 is no E. Colorado guarantee Boulevard, of future Suite results. 900, Pasadena, Potential California for profit is accompanied telephone by possibility of loss. Page 9 of 10 Commodities Total Return December 2011

10 Commodities Total Return Simulated Performance Unless otherwise noted, performance figures do not reflect the deduction of investment advisory fees. These fees are described below. The returns shown will be reduced by the advisory fees and any other expenses the advisor may incur in the management of an investment advisory account. Simulated performance is no guarantee of the future results in a live portfolio using the strategy. Potential for profit is accompanied by possibility of loss. General Disclosures: Hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not actually been executed, the results may under or over compensate for the impact, if any, of certain market factors, such as lack of liquidity or security positions that need to be rounded based upon contract size when live futures trades are executed. Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. Further, backtesting allows the security selection methodology to be adjusted until past returns are maximized. No representation is being made that any account will or is likely to achieve profits or losses similar to those shown. Unless otherwise noted, performance returns for one year or longer are annualized. Performance returns for periods of less than one year are for the period reported. The simulated performance used in this presentation will differ from live performance experienced using the strategy because the simulated performance was derived from the backtesting or the retroactive application of First Quadrant s current proprietary model and the simulation assumes that the strategy guidelines are constant through the life of the portfolio, whereas, the guidelines for live portfolios may have changed over the life of each portfolio. Further, with respect to the Balanced Risk Commodities sub-strategy (herein referred to as Macro effects) simulation : The simulation assumes that we adjust the risk and capital allocated to each sector or sub-sector on a monthly basis after the close on the last day of each month, whereas the live product may not adjust the allocations exactly at that time due to intra-month market movement and risk regime shifts. The simulation assumes fixed transaction costs whereas live portfolio transaction costs will be variable The simulation assumes all trading takes place once a month (on the last day of the month) whereas live portfolios may trade often during the month. With respect to the Commodities Long/ Short sub-strategy (herein referred to Fundamental effects) simulation: The simulation assumes all trading takes place once a day, executed at the previous day s closing price minus the assumed transaction costs, whereas live trading would occur throughout the day The simulation assumes simplistic transaction costs, which is a function of trade volume, calibrated to recent market data. The transaction costs ignore market impact, whereas in live trading, liquidity will affect transaction costs. The simulated guideline risk target does not take into account risk-capital changes and extreme market conditions whereas when the strategy is deployed in a live portfolio the guideline risk target may change as a client s risk-capital is scaled into the strategy and the risk target will be reduced in certain extreme market events. The simulation assumes implementation of the strategy via commodity futures, whereas a live portfolio may use other instruments (i.e. options, options on futures) with a different return or cost. Disclosures Specific to Simulation: This simulation was created in June 2011 and updated every month end or quarter end. It is comprised of 65% Macro effects sub-strategy and 35% Fundamental effect sub-strategy. Allocations to each sleeve are rebalanced monthly basis. The Macro effects sub-strategy simulation is constructed with the goal to diversify risk in a portfolio by strategically allocating risk to several commodities. Allocations are made to commodities within the following sectors: 1) Energy: WTI Oil, Brent Oil, Natural Gas, RBOB Gasoline, Heating Oil, 2) Agriculture: Wheat, Corn, Soybeans, Coffee, Sugar, Cocoa, 3) Industrial Metals: Copper, Aluminum, Lead, Zinc, Nickel, 4) Precious Metals: Gold, Silver and 5) Livestock: Live Cattle, Lean Hogs. The simulation balances risk across these so that each commodity has an equal risk footing in the portfolio. The simulation also attempts to balance risk relative to commodity weightings. The simulation targets overall portfolio risk allocations based on pre-determined indicators of market risk which may change over time. The simulation reflects an approximate 12% risk level. All income is reinvested monthly, no external cash flows are assumed. The Fundamental effects sub-strategy simulation is a combined forecast mix of multi-factor models, all of which would be used to manage a live portfolio. The simulation uses current model s historical signals on active commodity futures (corn, soybeans, wheat, cocoa, coffee, crude oil (enters simulation 8/19/85), copper, sugar, and cotton). The overall risk management system threshold includes: a +/- 125% limit on each grain instrument (wheat, corn, soy), a +/- 66.7% limit on each individual non-grain instrument, a +/- 83.3% aggregate limit on long or short positions, and a +/- 12 on the aggregate level of grains. No cash returns or cash flows are assumed in the simulation. Capital gains are assumed to be fully reinvested. Investment Management Fees: Net performance returns presented assumes the following asset-weighted fee schedule to each sub-strategy: The Macro effects sub-strategy fee schedule is as follows $1 $100, 0.5; $100-$350, 0.3; and more than $350, 0.15%. The Fundamental effects sub-strategy fee schedule is 2% with a 2 incentive fee. Market Impact On Returns: Performance in 2010 was unusually strong as market movements in Commodities were significant, particularly with increases in Cotton, Corn, Copper, Soy, and Sugar during the Fall and Coffee and Wheat during the Summer. The portfolio was well positioned to take advantage of these market movements. FQ COMMODITIES Commodities Total Return Strategy Composite Total Return Gross Total Return Net Composite Dispersion (%) Total Composite Assets 1 (Millions USD) Total Firm Assets 1 (Millions USD) *Actively Managed AUM 1,2 (Millions USD) *Total Firm Assets (Including Notional Values) 1,3 (Millions USD) Number of % of Firm Portfolios 1 Assets (Jun-Dec)** -5.4% -5.9% < ,941 16,725 16,725 See additional disclosures for important information concerning this composite and the effect of fees. *Supplemental Information. **All performance and AUM data is preliminary. 1 At end of period reported. 2 Includes market values for fully funded portfolios and the notional values for margin funded portfolios, all actively managed by First Quadrant and non-discretionary portfolios managed by joint venture partners using First Quadrant, LP investment signals. First Quadrant is defined in this context as the combination of all discretionary portfolios of First Quadrant, LP and its joint venture partners, but only wherein FQ has full investment discretion over the portfolios. 3 Includes market values for fully funded portfolios and the notional values for margin funded portfolios managed by First Quadrant and non-discretionary portfolios managed by joint venture partners using First Quadrant, LP investment signals. First Quadrant is defined in this context as the combination of all discretionary portfolios of First Quadrant, LP and its joint venture partners, but only wherein FQ has full investment discretion over the portfolios. Commodities Total Return Strategy GENERAL DISCLOSURES First Quadrant, L.P. claims compliance with the Global Investment Performance Standards (GIPS ) and has prepared and presented this report in compliance with the GIPS standards. First Quadrant, L.P. has been independently verified for the periods and The verification reports are available upon request. Verification assesses whether (1) the firm has complied with all the composite construction requirements of the GIPS standards on a firm-wide basis and (2) the firm s policies and procedures are designed to calculate and present performance in compliance with the GIPS standards. Verification does not ensure the accuracy of any specific composite presentation. First Quadrant ( FQ or the Firm ) is defined as the combination of all discretionary portfolios of First Quadrant, L.P. and its joint venture partners but only wherein FQ has full investment discretion over the portfolios. First Quadrant L.P. is a registered investment adviser and is an affiliate of Affiliated Managers Group, Inc. A complete list and description of the Firm s composites is available upon request. COMPOSITE DETAILS Composite Description: (Creation Date: August 2011) The portfolios in this composite invest in the Commodities Total Return Strategy, which is composed of two sub-strategies which when combined target a volatility level between and including 1 to 15%: the Commodities Long/Short (herein referred to as Fundamental effects) strategy and the Balanced Risk Commodities (herein referred to as Macro effects) strategy. The Fundamental effects sub-strategy seeks returns independent of general market direction and employs a quantitative allocation process to allocate assets among various underlying long/short commodity positions. FQ has implemented each position using futures contracts. The Macro effects sub-strategy seeks to generate total returns in excess of inflation over a market cycle. The strategy intends to invest in a broad range of commodities that balance risk along three dimensions: across sectors, within sectors and across time. This is a total return strategy which is not managed against any benchmark or universe. Presenting the composite returns with no benchmark demonstrates clearer accountability by removing the distortions caused by blending strategy specific total and benchmark returns. Additional strategies may be developed and used by FQ in managing the strategy. Portfolio Criteria: There is no minimum balance requirement for a portfolio to be included in a composite. The strategy utilizes leverage at FQ s discretion. The returns presented reflect this leverage. Calculation Methodology: Valuations and returns are computed and stated in U.S. dollars. Individual contributions and withdrawals are only permitted to occur on a set schedule four times a month. The investment results use a dollar-weighted rate of return formula. Monthly strategy returns are calculated by linking the four intra month period returns using the linked internal rate of return formula. Annual strategy returns are calculated by linking the monthly returns using the linked internal rate of return formula. The dispersion of a composite is calculated using the asset-weighted standard deviation formula. Only portfolios managed for the full calendar year are included in the dispersion calculation. As this composite contains five or fewer portfolios, a measure of dispersion is not statistically representative and is therefore not shown. Additional information regarding the Firm s policies and procedures for calculating and reporting performance returns is available upon request. Derivatives: The strategy utilizes derivative instruments in both long and short positions to achieve desired returns. A derivative is defined as a financial asset or liability whose value depends on (or is derived from) other assets, liabilities or indices. Futures contracts are derivatives that specify a purchase or sale of an asset at a specified price on a specified date in the future. Forward contracts allow the purchase or sale of currency in the future at a currently agreed upon rate of exchange. Put and call options contracts are derivatives, which permit the owner, depending on the type of option held, to purchase or sell an asset at a fixed price on or until a specified date depending on the contract. An option to purchase an asset is a call, and an option to sell an asset is a put. There is a risk that a derivative may not perform as expected, thereby causing a loss or amplifying a gain or loss for the portfolio. With some derivatives, there is also the risk that the counterparty may fail to honor its contract terms, causing a loss for a portfolio. Investment Management Fees: The gross performance returns presented represent the strategy s rate of return using set period valuations and include the deduction of operating expenses but do not include the deduction of management fees and incentive fees. Net performance returns presented represent the strategy s rate of return using set period valuations and include the deduction of operating expenses, and assumed the following asset-weighted fee schedule to each sub-strategy: The Macro effects sub-strategy fee schedule is as follows $1 $100, 0.5; $100-$350, 0.3; and more than $350, 0.15%. The Fundamental effects sub-strategy fee schedule is 2% with a 2 incentive fee. All performance results presented include trading commissions Dow Jones Indexes is a licensed trademark of CME Indexes. Dow Jones, Dow Jones Indexes, DJ, UBS, Dow Jones-UBS Commodity Indexes, DJ-UBSCI and all other index names listed above are service marks of Dow Jones Trademark Holdings, LLC ( Dow Jones ) and UBS AG ( UBS AG ), as the case may be, and have been licensed for use by CME Indexes. CME is a trademark of Chicago Mercantile Exchange Inc. The Dow Jones-UBS Commodity Indexes SM are jointly promoted and marketed pursuant to an agreement between CME Indexes and UBS Securities. All other names are the trademarks or service marks of their respective owners. The S&P GSCI provides investors with a reliable and publicly available benchmark for investment performance in the commodity markets. The S&P GSCI is widely recognized as the leading measure of general commodity price movements and inflation in the world economy. The S&P GSCI is proprietary data of Standard & Poor s, a division of The McGraw-Hill Companies, Inc. All rights reserved. Citigroup World Government Bond Index (WGBI) is a Standard and Poor s /Citigroup Indices are proprietary data of Standard & Poor s, a division of The McGraw-Hill Companies, Inc. All rights reserved. This material is for your private information. The views expressed are the views of First Quadrant, LP only through the period ended December 2011 and are subject to change based on market and other conditions. All material has been obtained from sources believed to be reliable, but its accuracy is not guaranteed. Updated results on analyses available upon request. 10 Past performance is no guarantee of future results. Potential for profit is accompanied by possibility For Advertising of loss. Use Commodities Total Return December 2011 Page 10 of 10