Insights on...

commodity index design Commodit y Indexes: Design Matters

Institutional investors generally hold commodities either to diversify their equity and fixed-income holdings or to protect against inflation. One way to gain this exposure is through futures-based commodity indexes. Over the last few years, increasing amounts of assets have been benchmarked to commodity futures-based indexes1, whose performance is subject to variables such as changing spot prices and a futures-specific return component called “roll yield.” First-generation commodity indexes such as the DJ-UBS Commodity Index and S&P Goldman-Sachs Commodity Index (GSCI) were innovative when they were conceived because they provided investors with a broad, measurable and investable benchmark for a relatively Jordan Dekhayser, CFA Portfolio Manager, inaccessible asset class. They were developed using methodology similar to that of prevailing Global Index Management equity or fixed-income indexes: define the universe; choose a metric to weight each commodity [email protected] based on its relative importance in the marketplace; and create an index. Additionally, commodity indexes hold futures contracts that expire periodically, presenting another decision factor when creating an index. First-generation indexes generally did not explicitly incorporate other sources of return in commodity futures, such as momentum and alternative roll methodologies. Considering these features collectively, we arrive at a set of key variables to analyze a commodity index: commodity selection and weighting, dynamic versus static allocation and roll methodology. In recent years, a number of commodity indexes were created that seek to improve upon the design of first-generation indexes. One example is the ABCI, which is constructed to maintain exposure to the desirable characteristics of commodities and to provide a high correlation to firstgeneration commodity indexes. The latter feature may be important for investors using either the DJ-UBS or S&P GSCI as their stated commodities benchmark. Furthermore, the ABCI is designed to minimize exposure to undesirable characteristics of commodity indexes and ALTERNATIVE BENCHMARK COMMODITY INDEX to incorporate new and desirable features. Northern Trust uses the Alternative Benchmark Commodity Index (ABCI) to The actual, in-sample results – including offer clients exposure to commodities. This index, which tracks 19 different risk-adjusted returns, volatility and downenergy, metals and agricultural commodities, is well-aligned with our proprietary side protection – compare favorably with Intelligent Indexing® process that preserves wealth by managing risks, liquidity first-generation indexes. and transactions costs.

Pursuant to an exemption from the Commodity Futures Trading Commission in connection with accounts of qualified eligible persons, this brochure is not required to be, and has not been filed with the Commission. The Commodity Futures Trading Commission does not pass upon the merits of participating in a trading program or upon the adequacy or accuracy of commodity trading advisor disclosure. Consequently, the Commodity Futures Trading Commission has not reviewed or approved this trading program or this brochure.

northerntrust.com | Commodity Index Design | 1 of 12

Commodities: Key Terms for Investors ■■

Total return: The risk-free rate (typically the three-month U.S. Treasury bill rate) plus an excess return.

■■

Excess return: The spot return plus the roll yield.2

■■

Spot return: The return of the active futures contract. During a roll period, the active futures contract changes to the contract being rolled into.*

■■

Roll yield: The return generated from moving out of an expiring contract and into a new contract. This may be either positive or negative and is affected by borrowing costs, storage costs and convenience yield.

■■

Borrowing and storage costs: Charges related to holding the physical commodity. Also known as the cost-of-carry. All else being equal, futures prices will trade higher than spot prices due to these costs.

■■

Convenience yield: The benefit of having supplies available for commercial purposes. When supplies are tight, this benefit tends to be high. When supplies are ample, this benefit tends to be low. The convenience yield interacts with borrowing and storage costs to determine whether the futures curve is in contango or backwardation.

■■

Contango: A curve shape where the commodity futures price is higher than the spot price (Chart 1).

■■

Backwardation: A curve shape where the futures price is lower than its spot price (Chart 1). CHART 1: TERM STRUCTURE: CONTANGO VS. BACKwARDATION CONTANGO: FUTURES > SPOT

BACKWARDATION: FUTURES < SPOT

Futures Price

Futures Price

Current Spot Price

Current Spot Price Time to Expiration

Time to Expiration

Source: Northern Trust *In commodity indexes the spot return generally refers to a futures contract and not the commodity for immediate delivery. This is unlike indexes in other asset classes where the spot return typically refers to the physical securities.

northerntrust.com | Commodity Index Design | 2 of 12

Futures Curve Shapes

The futures curve’s shape is important to investors because it can create a hurdle that the spot return must overcome to create positive excess return. In some instances, the curve can be so steep that it is unlikely that the spot return will exceed this hurdle, making it questionable whether one wants to hold that specific commodity at full index weight. Furthermore, futures curves can change their shapes, moving from contango to backwardation (sidebar, p. 2) or vice versa, creating a key risk to certain strategies that seek to capitalize on the current shape of the curve. The spot market’s outlook also is critical. In fact, the correlations between the spot return and the roll yield tend to be negative, indicating that a commodity with a negative roll yield tends to have a positive spot return and vice versa (Table 1). Lastly, in addition to the curve’s slope, the degree of its concavity or convexity is important in determining where on the curve is the optimal point to hold futures positions in order to maximize roll yield. Merely holding longer-dated futures contracts does not necessarily increase investor roll returns. Regardless of whether a strategy is algorithmically positioned to maximize the roll yield or simply holds longer-dated futures contracts, the greatest liquidity and open interest generally concentrate in the front months, potentially reducing the size of roll-yield positioning opportunities. For investors, this ultimately means that the way a commodity or commodity index investment is made can significantly affect final investment results. For example, in 2012, the roll yield of natural gas was –42% in the DJ-UBS Index.3 Table 1: Correlation of Spot Return vs. Roll Yield (DJ-UBS Index) An index with a static allocation would be unable to incorporate this information into the Industrial Precious Agriculture Energy Metals Livestock Metals commodity’s weighting in the index. However, an index that is designed to incorporate this – 0.30 – 0.16 0.07 – 0.48 –0.13 information contemporaneously can help Sources: S&P Dow Jones, Bloomberg, Northern Trust. investors reduce the impact from a negative Correlations reflect 10-year monthly return data thru 12/31/2012. roll yield (high hurdle rate). Table 2: Commodity Index Correlations to Other Asset Classes 10-Year Correlation Matrix

ABCI

DJ-UBS

S&P GSCI

S&P 500

Barclays US Agg

ABCI

1.00

0.93

0.94

0.36

–0.03

DJ-UBS

0.93

1.00

0.89

0.49

0.06

S&P GSCI

0.94

0.89

1.00

0.42

–0.03

S&P 500

0.36

0.49

0.42

1.00

0.04

Barclays US Agg

– 0.03

0.06

– 0.03

0.04

1.00

Commodities – Diversification and Inflation Protection

The two commonly cited reasons for commodities investing are diversification and inflation protection. Diversification is provided through low correlations to other asset classes (equity and fixed income, Table 2). For example, the commodity indexes in Table 2 have a correlation between 0.36 and 0.49 to the Standard & Poor’s (S&P) 500. This diversification can help improve the efficient frontier of a multi-asset-class portfolio.

Sources: S&P Dow Jones, AIA, Bloomberg, Northern Trust. Correlations reflect 10-year monthly return data thru 12/31/2012.

northerntrust.com | Commodity Index Design | 3 of 12

Furthermore, individual commodity sectors have relatively low correlations to each other, underscoring the importance of maintaining broad-based commodity exposure in an index. As a frame of reference, the average pair-wise correlation between the five commodity sectors detailed in Table 3 is 0.25, as opposed to a value of 0.61 for S&P 500 sectors.4 Low sector correlations in commodities are due to the disparate risk exposures of each commodity sector, as well as the lack of a unifying source of return across all commodities such as exists for equities (market beta). For example, agricultural prices are driven by, among other factors, weather in their specific growing regions.5 The price of gold, however, relates more to global macroeconomic factors such as real interest rates and expected inflation rather than industrial demand and thus can help diversify against the risk of depreciating and potentially devalued currencies. Having a meaningful exposure to each commodity sector can thus add desirable diversification properties to an investor’s overall asset allocation and within the commodity allocation. Table 3: Commodity Sector Correlation Matrix (Spot Returns for Dow Jones-UBS Commodity Sector) Spot Return Correlations

Agriculture

Energy

Industrial Metals

Livestock

Precious Metals

Agriculture

1.00

0.28

0.41

–0.04

0.42

Energy

0.28

1.00

0.44

0.07

0.30

Industrial Metals

0.41

0.44

1.00

0.12

0.42

–0.04

0.07

0.12

1.00

0.11

0.42

0.30

0.42

0.11

1.00

Livestock Precious Metals

Sources: S&P Dow Jones, Bloomberg. Correlations reflect 10-year monthly return data 12/31/2002 – 12/31/2012.

Table 4: Index Correlations to CPI CPI

ABCI

0.65

DJ-UBS

0.61

S&P GSCI

0.68

S&P 500

0.27

Barclays US Agg

0.00

Sources: S&P Dow Jones, AIA, BLS, Bloomberg, Northern Trust. Correlations reflect 10-year monthly return data 12/31/2002 – 12/31/2012. CPI lagged 1 month.

Inflation protection is particularly relevant, given concerns that loose U.S. monetary policy and high commodities demand from developing countries could spur inflation. Commodity indexes have a high correlation to inflation, currently around 0.61 – 0.68 (Table 4), providing some level of protection against higher prices. The reason for this is rather straightforward. The U.S. Consumer Price Index (CPI) includes exposure to many of the raw commodities that comprise commodity benchmarks. For example, food and energy make up around 25% of CPI on a value-weighted basis but can account for more than half of the index’s volatility.6 Holding exposure to some of the inputs to CPI via a commodity index can help mitigate the deleterious impact of inflation on purchasing power, or more directly, on CPI-linked liabilities.

northerntrust.com | Commodity Index Design | 4 of 12

Investors can receive drastically different exposure to commodities, depending on the benchmark.

Commodity Selection and Weighting

Every benchmark methodology, regardless of asset class, must specify which positions to hold and in what weights. For equities and fixed income, there is a general consensus on construction of a beta index. The instruments included are representative of the asset class and are sufficiently liquid to maintain a substantial amount of assets that may be benchmarked to them. In this vein, commodity indexes look to include commodities that are economically significant and have enough open interest and trading volume to build investable benchmarks. However, commodity indexes differ dramatically from their equity and fixed-income brethren in that they can have substantially different weights for specific commodities and commodity sectors (Chart 2) while still claiming to represent the asset class. An individual stock’s weight in an equity index is typically driven solely by the stock’s free-float market capitalization as a percentage of the entire index’s free-float market cap (static weighting). However, to determine an individual commodity’s weight within a commodity index, most indexes use some combination of production and liquidity measures (static weighting). Furthermore, some commodity indexes impose artificial caps on the weight assigned to commodities and commodity sectors. These measures can result in drastically different commodity weights depending on how they are incorporated. For example, depending on which commodity index they choose, an investor would receive an energy allocation of either 69% (S&P GSCI) or 32% (DJ-UBS). Chart 2: Commodity Sector Weights 80% 69%

70% 60% 50%

42%

40% 30% 20%

32%

31% 23%

19%

16%

10%

19% 7%

10%

13% 4%

6%

5% 5%

0% Agriculture

Energy DJ-UBS

Industrial Metals S&P GSCI

Precious Metals

Livestock

ABCI (Max)

Sources: S&P Dow Jones, AIA, Bloomberg, Northern Trust. Data as of 12/31/2012.

northerntrust.com | Commodity Index Design | 5 of 12

Momentum in commodities is unlike that of other investable asset classes due to supply and demand dynamics.

The construction of the ABCI incorporates some of the uniqueness of commodities. Though the weight of a commodity in, for example, the S&P GSCI depends solely by its relative production, the weight of an individual commodity in the ABCI hinges not only on production and liquidity but also on recent momentum. Table 5 highlights key distinctions between index methodologies. Table 5: Index Methodology Highlights Commodity Weighting Method

Exposure Caps

Resulting Energy Weight as of 12/31/2012

DJ-UBS

Static. Liquidity (2/3 weight) and production (1/3 weight)7

Exposures capped at the commodity group, sector and single commodity level (33%, 25% and 15%, respectively

32%

S&P GSCI

Static. Production-weighted

None

69%

ABCI

Dynamic. Liquidity (2/3 weight) and production (1/3 weight) determine maximum weights

Sector minimum and maximum exist, but none currently affect weights

42%

Sources: S&P Dow Jones, AIA, Bloomberg, Northern Trust.

Dynamic vs. Static Allocation: The case for momentum in commodities

Momentum is important to all asset classes but is uniquely important to commodities because many commodities are consumable assets with both a limited shelf life and a limited supply. For example, adjusting wheat production to counter a supply shortage is impossible until the following year’s harvest. Thus, the price of wheat can be persistently strong in a market that is undersupplied in the short- to intermediate-term. Another example would be a market that is oversupplied, which could lead to a large amount of contango as supplies are placed in inventory. During 2012, this was the situation in the natural gas market, which had a highly negative roll yield. Depending on one’s outlook for the natural gas market, an investor could reasonably conclude that the expected return on a long futures position in natural gas may be negative. The momentum approach of the ABCI’s commodity-by-commodity dynamic allocation is a look-back strategy. To determine the sign of a momentum signal, the ABCI methodology looks to a futures index series incorporating the excess return (spot plus roll) of each commodity. The strategy will compare each index’s current value to three trailing periods and create an index position with a target weight equal to 40%, 60%, 80% or 100% of the maximum weights in Chart 2. The economic rationale discussed here for momentum in commodities is not relevant to, say, equity investors whose expected returns are generally positive. When commodity market dynamics create a hurdle rate that is either insurmountable or extremely high, less exposure to that particular commodity may make sense. The spot dynamics, however, must also be considered to determine whether to increase or decrease exposure to a commodity futures contract. By using the futures index series to determine the momentum signals, which ultimately determine the weight of individual commodities, the ABCI captures the interaction of spot return and roll yield. The index does not distinguish which source of return is dominant, but the result will be an index that, all else equal, will allocate away from commodities with negative roll yields (contango) and toward

northerntrust.com | Commodity Index Design | 6 of 12

commodities with positive spot returns. Relying on only spot or roll ignores the relationship of the two and can lead to a sub-optimal commodity allocation. The ABCI incorporates these momentum signals, acknowledging fundamental and technical reasons for the persistence of momentum in the short- to intermediate-term in spot and futures markets (Gunzberg and Kaplan [2007], Schneeweis et al. [2008]). From a technical standpoint, momentum is a popular strategy employed by active managers (Asness et al. [2012]). The ABCI seeks to capture the return associated with momentum in a rules-based approach, which may result in a more cost-effective solution for investors trying to capture this source of return. Incrementally, the dynamic asset allocation strategy that ABCI employs has added 2.5% to annual returns with a factor volatility of 3.6% (Sharpe Ratio of 0.75).8 Furthermore, the return attributed to dynamic weighting has been positive for 17 of the 19 commodities included in the index, suggesting these results are significant (Chart 3). Chart 3: Dynamic Asset Allocation – Investment Results 10 8.4

8 6.7 6 4.7 3.9

4

2.9

2.7

2.5

3.6 2.0 1.9

0

1.2

2.0 1.8 0.7

2.5 0.6

1.3

0.7

–0.1 Soybeans Sugar Wheat

Cotton Hogs

Coffee Corn

Agriculture Cattle

Silver

Gold

Nickel*

Copper

Metals

Aluminum*

Unleaded

WTI Crude

Natural Gas

Gasoil

Heating Oil

Energy

–0.9 Brent Crude

–2

4.7

4.4

3.7

2

Composite

The ABCI provides access to additional sources of return, such as momentum and relative roll, beyond traditional commodity indexes.

January 1991 – May 2012; *January 2002 – May 2012 Source: AIA.

Importantly, the dynamic asset allocation of the index is implemented concurrently with a cash position. At the aggregate level, the dynamic asset allocation and defensive cash position in the ABCI result in an index expected to have a long-run average of 70% exposure to commodities and 30% exposure to cash. Since 2005, ABCI has had an actual allocation to commodity futures and cash averaging 72% and 28%, respectively. Two-thirds of the time, the commodity exposure was between 58% and 85%. Rarely does the index go above 90% or below 50% (Chart 4, p. 8). Taken together, these features historically have resulted in lower volatility and improved riskadjusted returns (Table 6, p. 10). Investors who make asset allocation decisions on a risk-weighted basis may find benefits because they can increase risk-based allocations to other strategies.

northerntrust.com | Commodity Index Design | 7 of 12

Over long time periods, the rolling of futures can significantly affect the total return of a commodity portfolio. Therefore, understanding the roll methodology employed, as well as its effect on risk and return, is critical.

Chart 4: Percent of ABCI held in Commodities (versus cash)

100%

Two thirds of the time, the % allocation to commodities is between the green lines.

90% 80% 70% 60% 50% 40% 1/2/2006

1/2/2007

1/2/2008

1/2/2009

Futures Allocation

1/2/2010

1/2/2011

1/2/2012

12/31/12

Average Allocation

Sources: AIA, Northern Trust.

Roll Methodology

Roll yield is a very important concept for commodity index investors since the rolling of futures contracts directly affects their total return. Over short periods of time, the spot return tends to dominate due to its higher volatility. However, over long time periods, the rolling of futures can significantly affect the total return of a commodity portfolio. Therefore, understanding the roll methodology employed, as well as its effect on risk and return, is critical. DJ-UBS and S&P GSCI roll their contracts from the expiring contract to the next active contract during a concentrated five-day window near the beginning of the month. The index provider specifies these periods; therefore, with an accurate estimate of assets benchmarked to each index, one can predict the volumes that passive indexers must trade in order to track the benchmark. This feature can be a double-edged sword. On one hand, this volume advertising can allow for other participants to enter the market to offset the index supply and demand. On the other hand, non-index market participants may incorrectly forecast anticipated trade volumes and either exacerbate or even invert the expected index effect. Regardless, the fewer days in the roll period, the more impact a price event that unevenly affects contracts being rolled on one of those days will have on the index. The ABCI employs a daily roll methodology designed to completely allocate exposure away from the nearby futures contract before some of the common indexes start their rolls. Furthermore, by rolling a small amount of the position held each day, investors tracking the ABCI may have less market impact than with a concentrated roll cycle. The other result of the ABCI daily roll is that the index has a longer average maturity than an index that holds similar contracts. In other words, the ABCI relative to its peers becomes underweight the contract that is most active in other commodity benchmarks and overweight the next contract month. Historically, ABCI’s relative roll strategy has added 1.5%9 to its annual returns versus indexes with roll cycles like DJ-UBS or S&P GSCI. The volatility around this factor is 1.0%, leading to a Sharpe Ratio of 1.58 for the relative roll factor. The positive relative roll factor in 18 of the index’s 19 commodities indicates that this factor – as well as the momentum factor – is significant.

northerntrust.com | Commodity Index Design | 8 of 12

Chart 5: Relative Roll Factor – Investment Results 6

5.4

5 4.3 4 3 1.5

1.7 1.2

1 0.2

0

0.8

2.0

1.1

1.3 0.4

0.3

0.1

0.0

2.0

2.0 1.5

Wheat

Hogs Soybeans Sugar

Cotton

Cattle

Coffee Corn

Agriculture

Silver

Gold

Nickel*

Copper

Metals

Aluminum*

Unleaded

WTI Crude

Gasoil

Energy

Brent Crude

–0.6 Composite

-1

0.6

0.5

1.8

1.7

1.4

Natural Gas

2

Heating Oil

One important feature of all commodity benchmarks is that due to the expiration of their holdings, high turnover is common and expected. ABCI is designed with a unique feature, however, to keep turnover at a manageable level to control transaction costs.

January 1991 – May 2012; *January 2002 – May 2012 Source: AIA.

ABCI has delivered higher annual returns on both an absolute and risk-adjusted basis over the past five years, outperforming DJ-UBS and S&P GSCI by 6.4% and 9.3% per annum, respectively.

Dynamic Allocation and Roll Methodology – Managing Turnover

The interaction of dynamic allocation and roll methodology plays a unique role in the ABCI. Intuitively, one associates dynamic strategies as having higher turnover. One important feature of all commodity benchmarks is that due to the expiration of their holdings, high turnover is common and expected. ABCI is designed with a unique feature, however, to keep turnover at a manageable level to control transaction costs. To accomplish this, the index incorporates the changing asset allocations with the daily rolls of futures contracts. For example, on a given day, the index may be rolling five contracts of WTI crude oil futures out of the nearby contract and into the next active contract. On the same day, crude prices drop, requiring a smaller position in WTI crude oil. In this case, the index may roll out of the five nearby contracts but not buy five contracts in the next active month. In addition to this feature, not being a fully invested index also reduces the overall turnover of the ABCI, all else equal. A comparison of index turnover shows that ABCI has turnover in line with its peers and falls in between the turnover of DJ-UBS and S&P GSCI. ABCI historically has had about 10% more turnover than DJ-UBS and around 30% less turnover than the S&P GSCI. Commodity Indexes – Results

Over the last five years as of December 31, 2012, the ABCI outperformed DJ-UBS by 6.4% and S&P GSCI by 9.3% per annum while delivering lower risk over the same period (15.5% versus 21.9% for DJ-UBS and 27.4% for S&P GSCI). Taken together, this led to a Sharpe Ratio of 0.07 for ABCI versus –0.24 for DJ-UBS and –0.30 for S&P GSCI. Table 6 shows a snapshot of risk and return characteristics as of December 31, 2012, and Table 7 details historic returns beginning in 2002.10

northerntrust.com | Commodity Index Design | 9 of 12

Table 6: Commodity Index Characteristics as of 12/31/2012 Return 1-Yr

5-Yr

10-Yr

1-Yr

1.0%

1.2%

8.6%

–1.1%

–5.2%

0.1%

– 8.1%

ABCI DJ-UBS

Standard Deviation

S&P GSCI

Sharpe Ratio

5-Yr

10-Yr

9.2%

15.5%

4.1%

13.7%

2.7%

17.0%

Sortino Ratio

5-Yr

10-Yr

5-Yr

10-Yr

14.6%

0.07

0.48

0.11

0.93

21.9%

18.5%

–0.24

0.13

–0.30

0.31

27.4%

25.1%

–0.30

0.04

–0.37

0.15

Sources: S&P Dow Jones, Bloomberg.

Table 7: Annual Index Performance as of 12/31/2012 Year

Downside protection is a key feature of the ABCI. In 2008, the ABCI was only down 10.7% while DJ-UBS and S&P GSCI which were down 35.7% and 46.5%, respectively.

ABCI

DJ-UBS

2002

24.7%

25.9%

32.1%

2003

18.9%

23.9%

20.7%

2004

23.4%

9.2%

17.3%

2005

22.9%

21.4%

25.6%

2006

–2.5%

2.1%

–15.1%

2007

22.6%

16.2%

32.7%

2008

–10.7%

–35.7%

–46.5%

2009

11.6%

18.9%

13.5%

2010

10.3%

16.8%

9.0%

2011

–4.6%

–13.3%

-–1.2%

2012

1.0%

–1.1%

0.1%

10.0%

5.9%

5.1%

2002 – 2012

S&P GSCI

Sources: S&P Dow Jones, Bloomberg.

Evolving Commodity Index Design

Indexes for all asset classes have evolved dramatically over the past decade, which ultimately benefits investors by providing them with more choices to achieve specific investment goals. In the case of commodities and other non-capital asset classes, creating a benchmark can introduce complications and intricacies not evident in capital asset classes such as equities. The ABCI is an example of an index that preserves the desirable characteristics – diversification and inflation protection – for investing in commodities while significantly improving upon the design of firstgeneration commodity indexes. This enhancement is achieved by incorporating additional sources of return unavailable in statically weighted commodity indexes. The results – historically strong absolute and risk-adjusted returns, lower volatility and more effective downside protection relative to its peers – all are features that may improve investment outcomes.



northerntrust.com | Commodity Index Design | 10 of 12

END NOTES Asness, C. et al. (2012) “Value and Momentum Everywhere,” preprint. Gunzberg, J., and Kaplan, P., 2007, “The Long and Short of Commodity Futures Index Investing: The Morningstar Commodity Index Family,” in H. Till and J. Eagleeye (eds) “Intelligent Commodity Investing: New Strategies and Practical Insights for Informed Decision Making,” pp 241 – 274 (Risk Books). Greer, R. J., 2012, “Intelligent Commodity Indexing: A Practical Guide to Investing in Commodities,” McGraw-Hill. Schneeweis, T., Kazemi H., and Spurgin, R., 2008. “Momentum in Asset Returns: Are Commodity Returns a Special Case?” The Journal of Alternative Investments, 10 (4). Spurgin R., Schneeweis, T., Kazemi, H., and Martin, G., (Revised 2012), “The Alternative Benchmark Commodity Index: A Factor-based Approach to Commodity Investment,” AIA Research Report. Jordan Dekhayser, CFA, is a member of the Index Advisory Committee (IAC) of the ABCI. The methodology of, and intellectual property rights in, the Alternative Benchmark Commodity Index is owned by Alternative Investment Analytics LLC, and may be covered by one or more pending patent applications.

footntoes 1 As of 8/31/2012, there was $287 billion and $209 billion in gross and net notional index investment, respectively http://www.cftc.gov/ucm/groups/public/@marketreports/documents/file/indexinvestment0812.pdf. 2 For a detailed recap of the history of commodity research and how it relates to roll yield, see “Intelligent Commodity Investing: New Strategies and Practical Insights for Informed Decision Making”, Section 2: Research Perspective, H. Till and J. Eagleeye. 3 S&P Dow Jones, Bloomberg, Northern Trust. 4 Source: Northern Trust, S&P Dow Jones, Bloomberg 5 U.S. Department of Agriculture, National Agricultural Statistics Service 6 Greer, R. J., 2012, “Intelligent Commodity Indexing: A Practical Guide to Investing in Commodities,” McGraw-Hill. p. 41. 7 Production data is received from various sources depending on the individual commodity. For example, the source used by DJ-UBS for crude oil production is the Industrial Commodity Statistics Yearbook (ICSY) – see DJ-UBS Commodity Index Handbook. Liquidity is typically defined as open interest of futures contracts. 8 From January 1991 through May 2012 (the period for aluminum and nickel is January 2002 through May 2012). 9 From January 1991 through May 2012 (the period for aluminum and nickel is January 2002 through May 2012). See “The ABCI: A Factor-Based Approach to Commodity Investment,” Spurgin, Schneeweis, et al., (2012). 10 Alternative Benchmark Commodity Index returns prior to February 2007 are based on simulated or hypothetical performance that has certain inherent limitations. Unlike the results shown in an actual performance record, all index-related results presented do not represent actual trading performance. Also, because these trades have not been executed, these results may have under- or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity, trading expenses, management, administrative or other fees or costs associated with actual performance. No representation is being made that any account will or is likely to achieve profits or losses similar to the ones shown above as past performance is not indicative of future results.

northerntrust.com | Commodity Index Design | 11 of 12

Important Information There are risks involved in investing including possible loss of principal. There is no guarantee that the investment objectives of any fund or strategy will be met. Risk controls and models do not promise any level of performance or guarantee against loss of principal. This material is directed to eligible counterparties and professional clients only and should not be relied upon by retail investors. The information in this report has been obtained from sources believed to be reliable, but its accuracy and completeness are not guaranteed. Opinions expressed are current as of the date appearing in this material only and are subject to change without notice. This report is provided for informational purposes only and does not constitute investment advice or a recommendation of any security or product described herein. Indices and trademarks are the property of their respective owners. All rights reserved. Information intended for use with institutional investors only. Not to be distributed to or relied upon by retail investors. Past performance is not necessarily a guide to the future. Index performance returns do not reflect any management fees, transaction costs or expenses. One cannot invest directly in an index. Index performance is based upon information provided by the index providers. Indexes and trademarks are the property of their respective owners, all rights reserved. There are risks involved with investing, including possible loss of principal. For Asia Pacific markets, this material is directed to institutional investors, expert investors and professional investors only and should not be relied upon by retail investors. This information is provided for informational purposes only and does not constitute a recommendation for any investment strategy or product described herein. This information is not intended as investment advice and does not take into account an investor’s individual circumstances. Opinions expressed herein are subject to change at any time without notice. Information has been obtained from sources believed to be reliable, but its accuracy and interpretation are not guaranteed. Asset management at Northern Trust comprises Northern Trust Investments, Inc., Northern Trust Global Investments Ltd., Northern Trust Global Investments Japan, K.K., The Northern Trust Company of Connecticut and its subsidiaries, including NT Global Advisors, Inc., and investment personnel of The Northern Trust Company. Northern Trust Global Investments Japan, K.K. is regulated by the Japan Financial Services Agency. The Northern Trust Company has a branch in China mainly regulated by the China Banking Regulatory Commission, People’s Bank of China and State Administration of Foreign Exchange. The Northern Trust Company of Hong Kong Limited is regulated by the Hong Kong Securities and Futures Commission. In Singapore, Northern Trust Global Investments Limited (NTGIL), Northern Trust Investments, Inc. and The Northern Trust Company of Connecticut (NTCC) are exempt from the requirement to hold a Financial Adviser’s License under the Financial Advisers Act and a Capital Markets Services License under the Securities and Futures Act with respect to the provision of certain financial advisory services, and fund management activities. In Australia, NTGIL is exempt from the requirement to hold an Australian Financial Services License under the Corporations Act 2001 in respect to the provision of financial services. Issued by NTGIL. NTGIL is authorised and regulated by the Financial Services Authority under UK laws, which may differ from Australian laws.

© northern trust 2013 northerntrust.com | Commodity Index Design | 12 of 12 Q 53195 (3/13)