ALTERNATIVE RISK PREMIA INFORMATION FOR INVESTMENT PROFESSIONALS ONLY A GUIDE TO ALTERNATIVE RISK PREMIA. (alt)

Transcription

1 ALTERNATIVE RISK PREMIA INFORMATION FOR INVESTMENT PROFESSIONALS ONLY A GUIDE TO ALTERNATIVE RISK PREMIA (alt) β

2 A GUIDE TO ALTERNATIVE RISK PREMIA

3 CONTENTS I. HISTORICAL BACKGROUND 1 II. UNIVERSE DIVERSITY 3 III. CONSTRUCTING AN ALTERNATIVE RISK PREMIA STRUCTURE 5 IV. ACCESS TO ALTERNATIVE RISK PREMIA 6 V. LOW CORRELATIONS 7 VI. APPLICATIONS OF AN ALTERNATIVE RISK PREMIA APPROACH 9 VII. DISTINCTION FROM OTHER FACTOR-BASED APPROACHES 11 VIII. INVESTOR QUESTIONS AND UPTAKE 12 THE COLUMBIA THREADNEEDLE APPROACH 13 PORTFOLIO PARAMETERS AND TARGETS 15 COLUMBIA THREADNEEDLE ALTERNATIVE RISK PREMIA TEAM 16 LIQUID RISK PREMIA DEFINITIONS 17-22

4 ALTERNATIVE RISK PREMIA AN INTRODUCTION For years, academics and investors have been researching sources of return to understand what drives outperformance. Increasingly, this research has been focused on alternative risk premia strategies, a form of factor-based investing which is now entering the mainstream. Alternative risk premia seek to systematically capture return streams embedded in financial markets, each with strong academic underpinnings or resulting from persistent and well-documented investor behaviour. While alternative risk premia may be able to be accessed through liquid and transparent low-fee vehicles, what appeals most to institutional investors in a world of highly correlated markets, is their minimal market directionality and low correlations with traditional asset classes. This not only makes alternative risk premia genuine portfolio diversification tools but may also help to limit drawdowns when used as part of a broader portfolio.

5 I. HISTORICAL BACKGROUND With the increasing separation by investors of alpha (the return from manager skill), from market beta (the return derived from a passive exposure to the broader market), there has been a fundamental reappraisal of what constitutes alpha. In particular, returns that were previously thought to be alpha can, in many cases, be decomposed at low cost, into investible, liquid risk premia, or alternative risk premia. While alternative risk premia have always been embedded in financial markets, their identification began with Rolf Banz and his research into the small cap effect in Then in 1993, the so-called Fama/French three-factor model showed that market beta alone was not enough to explain varying, or abnormal, stock returns. In addition to the small cap effect, Eugene Fama and Kenneth French also discovered that value stocks, those with a low book-value-to-price ratio, tended to generate excess returns. As a result, Fama and French determined that a stock s return depended on three factors market beta, size and value. 2 In 1997, this was augmented by Mark Carhart s four-factor model, which added momentum to the other three factors explaining a stock s return. This followed on from prior research that had shown stocks with strong shortterm price momentum tended to carry on outperforming over the next 12 months. 3 The evolution of understanding portfolio return is depicted in Exhibit 1. Sources: 1 Rolf W. Banz. The Relationship Between Market Value and Return of Common Stocks. Journal of Financial Economics. November Eugene F. Fama and Kenneth R. French. University of Chicago. Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics 33 (1993) Narasimhan Jegadeesh and Sheridan Titman. Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. The Journal of Finance. Vol 48, No 1 (March 1993) pp

6 Exhibit 1: The framework of understanding portfolio return has evolved Manager skill ALPHA Higher cost and elusive Style Momentum Value Carry Curve Volatility Equity Credit Commodities Rates Currency ALTERNATIVE RISK PREMIA TRADITIONAL BETA Lower cost and harder to replicate Lower cost and prevalent Source: Columbia Threadneedle Investments 2

7 II. UNIVERSE DIVERSITY Alternative risk premia are systematically captured across multiple asset classes including equities, fixed income, credit, currencies and commodities, through non-traditional investment strategies such as rulesbased long/short, relative value and arbitrage. This is shown in Table 1. A portfolio management team can take varying approaches to capturing these diverse return streams: focusing on a single asset class, multiple asset classes or solely on capturing specific risk premia. This decision may consider the factor exposures and gaps in an investor s existing portfolio and the investor s overall diversification goals. As noted earlier and demonstrated by the breadth of asset classes and styles covered in Table 1, alternative risk premia have the potential to provide meaningful diversification for institutional investors. 3

8 Table 1: Alternative Risk Premia are available across multiple asset classes and styles Equity Fixed Income Credit Currency Commodity Momentum Momentum Momentum Momentum Momentum Implied v. Realised Volatility Implied v. Realised Volatility Implied v. Realised Volatility Implied v. Realised Volatility Carry Carry Carry Carry Curve Curve Curve Value Value Value Beta Size Quality Liquidity Liquidity Liquidity Alpha Strategies Source: Columbia Threadneedle Investments 4

9 III. CONSTRUCTING AN ALTERNATIVE RISK PREMIA STRUCTURE Constructing a liquid alternative risk premia structure with limited market directionality is, in principle, relatively straightforward. For instance, an investor seeking to systematically capture the value risk premium in equity markets ie, the tendency of cheaper stocks to outperform expensive stocks over time must first establish both the measure of value to be employed and the universe of stocks, or equity benchmark, from which to systematically capture this pricing anomaly. Exhibit 2 describes the methodology for creating such a structure using price-to-book as the measure of value and the MSCI ACWI as the benchmark. By ranking the MSCI ACWI top-to-bottom based on price-to-book, from the cheapest to the most expensive stocks, a long position is taken in the top quintile of cheapest stocks and a short position in the bottom quintile of most expensive stocks. No investment is made in the middle 60% of the universe. The result is a structure that captures the value risk premium. Crucially, because long and short exposures are typically balanced, and the structure does not invest in the middle 60% of the universe, there is limited exposure to directional market moves, or market beta. In contrast, the returns from a long-only smart beta approach to capturing risk premia are dominated by market beta. Exhibit 2: Creating an Alternative Risk Premia Index using Price-to-Book Lowest p/b Quintile 1 Buy Long Quintile 2-4 Price-to-Book Highest p/b Quintile 5 Sell Short Source: Columbia Threadneedle Investments 5

10 IV. ACCESS TO ALTERNATIVE RISK PREMIA Alternative risk premia can be accessed either through internally or externally manufactured structures. The former requires a strong internal quantitative research team, while the latter swap-based approach is predicated on strong counterparty relationships and deep quantitative analysis. In our view, there is no benefit to maintaining a dogmatic adherence to one approach; we believe investors should be agnostic as to how liquid risk premia are sourced. The key is to access the most cost-effective and capital-efficient array of risk premia structures available with complete transparency around the composition of the structure and portfolio positions. Indeed, we believe the emphasis should be on assembling the most advantageous diversified portfolio of risk premia possible; what we term the principle of best ingredients. Of course, active management is integral to the process of selecting and implementing alternative risk premia as is recognising the limits to the scalability of particular strategies applied to certain asset classes. Therefore, manager skill and breadth of expertise is of paramount importance when selecting an alternative risk premia manager. 6

11 V. LOW CORRELATIONS A key attraction of alternative risk premia is their low correlation to general equity and fixed income market movements, and to each other. This makes them especially appealing to investors seeking to diversify portfolio returns against the backdrop of highly correlated asset classes and markets. Indeed, our research found that the correlation of returns from a diversified basket of alternative risk premia to the market overall was barely positive at an average of 0.02 over a near 13-year period which captured the global financial crisis. (Exhibit 3 for the period 31 March 2005 to 31 March Integral to this result is the multiple negatively correlated pair-wise relationships that exist amongst many alternative risk premia, each of which can be explained by an economic and/or behavioural rationale. Equally compelling is the ability of a diversified alternative risk premia allocation to reduce portfolio drawdowns. Again, this is the result of the multiple negative correlations that exist between many alternative risk premia and the low correlation of alternative risk premia to general equity and fixed income market movements. However, acknowledging that individual risk premia can experience large drawdowns and underperformance for considerable periods of time is critical to the success of an alternative risk premia strategy, as is diversifying across liquid risk premia and actively managing positions. 7

12 Exhibit 3: Historical average pair-wise correlations have demonstrated stability through time /1/2005 8/1/2005 Global Financial Crisis 2/1/2006 8/1/2006 2/1/2007 8/1/2007 2/1/2008 8/1/2008 2/1/2009 Rolling 24-Month Average Correlation 8/1/2009 2/1/2010 8/1/2010 2/1/2011 8/1/2011 2/1/2012 8/1/2012 2/1/2013 8/1/2013 2/1/2014 8/1/2014 Average 2/1/2015 8/1/2015 2/1/2016 8/1/2016 2/1/2017 Source: Columbia Management Investment Advisers. Percentages shown indicate rolling 24 month correlations among risk premia over 31 March March Correlation ranges from +1 to -1. Positive correlation indicates returns moving in the same direction, negative correlation indicates returns moving in opposite directions, and a correlation of 0 would indicate no relationship between the movement of the two returns. Please see appendix for alternative risk premia correlation sources. 8

13 VI. APPLICATIONS OF AN ALTERNATIVE RISK PREMIA APPROACH With its strong and long-held academic underpinnings, alternative risk premia is an increasingly applied investment approach. Indeed, we have seen several broadly defined objectives emerge among investors who have demonstrated interest in alternative risk premia. These include adding alternative risk premia to an existing alternatives allocation; replacing those alternative investments that are delivering beta and pricing it as if it were alpha; and employing alternative risk premia in an effort to smooth out the volatility of an existing portfolio. Table 2 describes the incremental or complementary manner in which investors may employ alternative risk premia. 9

14 Table 2: Alternative Risk Premia can be used for different client solutions Application Multi-Strategy/Multi-Asset Manager Replacement Alpha Beta separation Completion portfolio to an existing portfolio structure Complementary portfolio to a traditional multi asset portfolio Potential Benefit nexposure to risk premia that drive much of traditional multi-strategy performance nlower cost, greater liquidity and transparency than many multi-strategy investment structures nefficient and liquid exposure to alternative market betas at low cost nallows asset owners investment manager selection process to be focused on true alpha/skilled managers naddresses gaps and concentrations in existing factor exposures nprovides a more balanced/diversified overall portfolio that can adapt as market regimes change nintroduces alternative risk premia to a portfolio of traditional betas nmay provide greater diversification and reduce market drawdown during periods of high asset market correlations Source: Columbia Threadneedle Investments 10

15 VII. DISTINCTION FROM OTHER FACTOR-BASED APPROACHES As factor-based offerings have multiplied in the marketplace, many approaches have been established. Notable among these are a longonly approach dubbed smart beta. Smart beta equity strategies provide exposure to the return streams associated with traditional, long-only equity investing, but use non-market-cap-weighted approaches to gain that exposure. While traditional indices are constructed using a marketcap weighted approach, smart beta indices may be constructed by equalweighting the stocks in an index, or weighting the stocks based on factors such as value or momentum. However, because smart beta employs longonly techniques, it typically carries more market directionality in its return profile. By contrast, alternative risk premia strategies employ structures that comprise both long and short positions, creating return streams with limited exposure to market movements. 11

16 VIII. INVESTOR QUESTIONS AND UPTAKE While the academic evidence behind alternative risk premia has a long history, many strategies in the marketplace have track records with less than five years of history. This is principally a result of the time and technology that was needed to construct and execute on the algorithms employed to access alternative risk premia return streams. Commercial interest in the approach is also a somewhat recent phenomenon. As investors examine fees, restrictions on liquidity, and the limited transparency associated with many other alternative investment strategies, we expect interest in alternative risk premia approaches to continue to grow. Some have questioned whether increasing interest and investment in alternative risk premia will result in risk premia being arbitraged away, reducing the efficacy of the approach. Two structural arguments suggest not. Firstly, we believe that much of the investment to date in alternative risk premia has been a repositioning of capital, not new capital coming into the space. Money allocated to liquid alternative risk premia is frequently that re-allocated away from the illiquid alternatives space. Secondly, the long-held and academically-verifiable economic and behavioural rationale behind these risk premia would suggest these long-standing and well-established risk factors should continue to exist. However, it is important to be cognisant of these pressures and to actively manage the exposures within a portfolio of risk premia. The interest in alternative risk premia varies globally from region to region. Despite being an increasingly applied investment approach, alternative risk premia implementation must still prove its ability to generate consistent returns with limited risk. Therefore, investors are continuing to study the concept in a desire to understand more about the structure, and how liquid risk premia might fit within a multi-asset portfolio. As investor understanding of the approach evolves, we believe that alternative risk premia will likely become mainstream allocations either in standalone form, as part of a diversified multi-strategy portfolio, or as a completion portfolio. 12

17 THE COLUMBIA THREADNEEDLE APPROACH Our Diversified Alternative Risk Premia strategy is a liquid, diversified, transparent, and low-cost approach which invests across all five major asset classes (equity, fixed income, credit, currency, commodities) and employs all major long/short and arbitrage techniques (style, curve, carry, short volatility and liquidity). We construct a portfolio of approximately 30 to 50 alternative risk premia using a combination of risk parity, risk targeting and tactical asset management. Starting with a risk parity position across all the risk premia, and drawing on the macro insights of a 20-person asset allocation team, we take active, macro views about the appropriate tilts to the various risk premia in which we invest, reflecting real world dynamics. In addition, our active approach to portfolio construction allows us to implement hedges where appropriate if we see over-concentrations developing. Our portfolio construction process is illustrated in Exhibit 4. All risk premia employed in the portfolio must be well established in the academic literature or be the result of persistent and well documented investor behaviour. When investing in externally constructed alternative risk premia structures, we require access to the counterparty s algorithm construction methodology and rule book for creating the structure. Ultimately, we seek to construct a portfolio of liquid, low cost, and transparent alternative risk premia that are uncorrelated to the broader market, and to each other. 13

18 Exhibit 4: Stages of portfolio construction an active process Liquid risk premia from recommended universe (30-50) Risk parity portfolio Risk targeting n Strategic portfolio constructed using equal risk weights n Asset class weights adjusted to account for idiosyncratic risks Capital Market Risk Mitigation n Discretionary macro viewpoints developed and risk mitigation measures effected to avoid major market dislocations Unlevered portfolio levered to target volatility FINAL PORTFOLIO There is no guarantee that return and volatility expectations will be met. 14

19 PORTFOLIO PARAMETERS AND TARGETS Performance Target Volatility Leverage Hedging Individual Positions Our Diversified Alternative Risk Premia strategy aims to generate a return of 7-10% gross over a typical market cycle (three years) given a 7.5% volatility target. The targeted volatility of the strategy is 7.5%, but can vary for separately managed accounts based upon individual investor preference. For a 7.5% volatility target portfolio, leverage would range from 3x to 6x at the portfolio level with higher leverage applied to individual low volatility risk premia in order to size their risk contribution. The portfolio will not be hedged at the aggregate level. However, the team may employ currency hedges based upon the Multi-Asset team s review of the prospects of a country s currency. There are no stated limits on individual position sizes. However, our risk parity approach seeks to balance all of the individual risk/positions/ exposures, and the team monitors position, size and risk exposures on a continual basis. There is no guarantee that the investment objective will be achieved or that return expectations will be met. Portfolio parameters are internal guidelines used by the investment team and are subject to change without notice. Formal investment parameters are set forth in the offering documents or investment management agreement. 15

20 COLUMBIA THREADNEEDLE ALTERNATIVE RISK PREMIA TEAM A team of seven investment professionals drawn from portfolio management and the firm s independent risk management team support our Alternative Risk Premia strategy. Senior portfolio management members of the team include: William Landes, Ph.D. Deputy Head of Global Investment Solutions and Alternative Investments William Landes joined the firm in 2014 and is head of alternatives for Columbia Threadneedle Investments. Previously, Dr. Landes has been in the investment community since 1985, and received a B.S. in economics from the University of Findlay and a Ph.D. in finance from University of Cincinnati. Jeffrey Knight, CFA Head of Global Investment Solutions, Co-Head of Global Asset Allocation Jeffrey Knight joined the firm in 2013 and was named head of global investment solutions and co-head of global asset allocation in Mr. Knight has been a member of the investment community since 1987, and received a B.A. from Colgate University and an MBA from the Amos Tuck School of Business at Dartmouth College. Mr. Knight holds the Chartered Financial Analyst designation. Joshua Kutin, CFA Senior Portfolio Manager, Global Investment Solutions Joshua Kutin joined the firm in 2015 and is a senior portfolio manager for the Global Investment Solutions Group at Columbia Threadneedle Investments. He has been a member of the investment community since 1998, and received a B.S. in economics and a B.S. in mathematics with computer science from MIT, as well as a masters in finance from Princeton University. He holds the Chartered Financial Analyst designation. Marc Khalamayzer, CFA Analyst, Global Investment Solutions Marc Khalamayzer joined the firm in 2014 and is an analyst for the Global Investment Solutions Team at Columbia Threadneedle Investments. He has been a member of the investment community since 2006, and received an M.S. in finance and a B.S. in economicsfinance from Bentley University. He holds the Chartered Financial Analyst designation. 16

21 LIQUID RISK PREMIA DEFINITIONS Liquid Risk Premia: Investors have historically generated returns by taking on market directional risk and capturing the associated payoff (premia) with directional risks (betas). Alternative Betas (also known as Liquid Risk Premia) are also designed to generate returns by capturing the payoffs associated with market risks, but these premia are generated independently of market direction. Liquid Risk Premia (LRP) are typically market neutral in their return profile. They generate returns by isolating the payoffs driven by investor behavioural tendencies and preferences. This creates a return opportunity that has little to no correlation with traditional markets and cannot be explained by traditional market behaviour or direction. The returns are a function of the investment styles (ie, value, quality, momentum etc.) or market dis-equilibriums (ie, carry, short volatility etc.). Examples of LRP are listed below. These strategies are available as indices and accessible via swaps or rules-based trades. EQUITY Equity Momentum: Equity Value: Equity Momentum captures trends in stock prices. Once a momentum algorithm is defined, the investable universe is ranked from highest to lowest momentum stocks. A long position is established in the top 20% of the universe positions and a short position in the bottom 20% of the same universe. This basket of long and short positions is then beta, country and sector adjusted to isolate the momentum premia and associated payoff. The strategy tends to perform best in markets with well-defined trends. Equity Value captures the tendency of cheap stocks to outperform expensive stocks. Once a value algorithm is defined, the universe is ranked from cheapest to most expensive stocks. A long position is established in the top 20% of the value ranked equity universe and a short position in the bottom 20% of the value ranked universe. The basket is then beta, country and sector adjusted to isolate the value premia and its associated payoff. The strategy tends to do best when investors are more prepared to take on risk. 17

22 Equity Quality: Equity Low Volatility: Equity Multi- Factor: Equity Volatility Carry: Equity Quality captures the tendency of high quality companies to outperform low quality companies. Once a quality algorithm is defined, the universe is ranked from highest to lowest quality. A long position is established in the top 20% of the quality ranked equity universe and a short position in the bottom 20% of the quality ranked universe. This basket of long and short positions is then beta, country and sector adjusted to isolate the Quality premia and associated payoff. The strategy tends to do best in risk off markets, when investors shy away from risk. Equity Low Volatility captures the tendency of low volatility stocks to outperform high volatility stocks over time. Once a volatility capture algorithm is defined, the universe is ranked from the lowest to highest volatility stocks. A long position is established in the top 20% of the volatility-ranked equity universe and a short position in the bottom 20% of the volatility-ranked universe. This basket of long and short positions is then beta, country and sector adjusted to isolate the low volatility premia and its associated payoff. The strategy tends to react best to volatility risk off and stable to declining interest rate environments. This approach aims to maximise exposure to the five equity factors (momentum, value, quality, low volatility and size). The strategy uses a generalised risk parity approach which aims to allocate equal risk to each equity factor. The process takes the correlations between factors into account, and emphasises factors that are inversely correlated, eg, value and momentum. Country and regional industry exposure are adjusted. The strategy takes advantage of the tendency of investors to overpay to hedge their equity portfolio against capital losses (overpay for insurance). The strategy calculates the spread between implied volatility and realised volatility across a wide range of equity option markets. It then goes short implied volatility and hedges that exposure by owning realised volatility. The strategy is typically constructed using short-term option strategies (delta hedged strangles being the best example). The strategy reacts best to stable or rallying equity markets and suffers when equity volatility spikes. 18

23 FX FX Value: FX Carry: FX Momentum: FX Volatility Carry: FX Value captures the tendency of cheap currencies to outperform expensive currencies. Designed to capture the return premia associated with being long currencies ranked highest by purchasing power parity (PPP) or real effective exchange rates (REEF) and short currencies ranked lowest by the same measure. Can be long by any number of currencies and short by any number of currencies on a given day. Can be G10 only or a mixture of G10 and EM currencies. The strategy reacts most acutely to relative currency volatility and is subject to idiosyncratic currency market surprise. FX Carry captures the tendency of higher-yielding currencies to provide higher returns than lower-yielding currencies. Takes advantage of the implied carry rate (spread between FX Forward vs FX Spot) for any number of currencies (G10 & EM). Strategy ranks currencies based on the spread between FX Forward and FX Spot. The strategy goes long single currencies with the highest carry and short single currencies with the lowest carry. The strategy reacts best in risk-on markets. FX Momentum takes advantage of trends in currency prices. The strategy evaluates the recent performance (trend) of a number of currencies against the USD based on a momentum algorithm. The strategy then takes either a long or short position on each of the currencies based on its momentum ranking. The strategy tends to perform best in markets with well-defined trends. The strategy takes advantage of the tendency of investors to overpay for currency hedging (ie, insurance). The strategy calculates the spread between implied volatility and realised volatility across a wide range of currency option markets. It then goes short implied volatility and hedges that exposure by owing realised volatility. The strategy is typically constructed using short-term option strategies (delta hedged strangles being the best example). The strategy reacts best to stable or rallying currency markets and suffers when currency volatility spikes. 19

24 FIXED INCOME Rates Momentum: Interest Rate Curve: Yield Curve Reshaping: Rates Volatility Carry: The strategy takes advantage of the trends in bond prices. The strategy evaluates the moving average of the daily returns of specific bond futures and ranks them according to the sign of their momentum. The strategy then takes either a long or short position on the bond futures depending on whether the sign of the momentum factor is positive or negative. Rates momentum reacts to trends in interest-rate markets and suffers the most at volatile inflection points. The strategy seeks to capture the difference between returns at the short end of global yield curves and the returns at the long end of global yield curves. The strategy recognises that risk adjusted excess returns at the shortend of global yield curves tend to be higher than risk adjusted returns at the longend of global yield curves. Via the futures markets the strategy takes long positions at the short and intermediate ends of global yield curves while taking short positions at the long end of global yield curves. The exposure to all futures markets is adjusted to be duration-neutral. The strategy reacts to changes in short rates as well as volatility in long-term interest rates. The strategy seeks to take advantage of the reshaping of global yield curves, either flattening or steepening. The strategy aims to extract excess returns from a steepening/ flattening trend in the USD Swap curve. The strategy may be implemented through rolling investment in duration-neutral 2-year 10-year curve positions, with the decision to switch from default steepener to flattener position determined by a dynamic signal constructed from evolution of 3-month overnight index swap rates. When the U.S. Federal Reserve is expected to raise rates, the strategy is to adopt a flattener; otherwise the default steepen is present. Takes advantage of the tendency of investors to overpay for fixed income hedging (ie, insurance). The strategy calculates the spread between implied volatility and realised volatility across a wide range of fixed income option markets and then shorts implied volatility and hedges that exposure by owning realised volatility. The strategy is typically constructed using short-term option strategies (delta hedged strangles being the best example). The strategy reacts best to stable or rallying fixed income markets and suffers when fixed income volatility spikes. 20

25 CREDIT High Yield Carry: COMMODITY Commodity Carry: Commodity Curve: The strategy seeks to capture the risk premium differential between high-yield debt and investment-grade debt. The strategy captures the premia by going long high-yield bonds and/or associated derivatives and short investment-grade bonds and derivatives in both US and non-us markets. Long and short exposures are ratio-adjusted to account for relative volatility. The strategy is highly impacted by fluctuation in credit spreads and performs best when spreads are tightening. The strategy is a sector neutral/market neutral investment strategy that seeks to isolate and capture carry returns in the commodity markets. The strategy goes long instruments that are most backwardated (spot price trading above futures price) and short instruments that are least backwardated within correlated groups of commodities (ie, energy, metals, agricultural, etc.). Unlike less constrained commodity indices which can take directional bets on sectors, energy, etc., this sector-constrained index should be far less susceptible to exogenous commodity market issues. This strategy tends to be uncorrelated with almost all traditional directional markets. Designed to exploit tendency for deferred futures contracts to outperform nearer dated contracts due to producers hedging further out than consumers and passive investors investing near the front of the curve. The strategy goes long selected points in the curve of each commodity, equally weighted amongst commodities, and short the nearest contract of each commodity. The short positions are betaadjusted. The curve strategy includes multiple exposures to each commodity in the Curve Basket. The Curve strategy simultaneously takes long and short exposures in various commodities. The aggregated long exposure of the various commodities is equal to 100%. The aggregated short exposure is scaled such that its previous 3-month realised volatility is equal to the previous 3-month volatility of the long position. This strategy tends to be uncorrelated with almost all traditional directional markets. 21

26 Contrarian: Congestion: Backwardation L/S: The strategy seeks to exploit the behavioural tendencies of commodity market speculators. The strategy uses publicly available market positioning data with the aim of systematically identifying a diversified set of under-valued and over-valued commodities. It identifies commodities with significant net-long or net-short positioning from speculators relative to past positions. Index construction allocates positions that are opposite in direction to speculator positioning. This strategy tends to be uncorrelated with almost all traditional directional markets. The strategy attempts to capitalise on a congested roll period by taking long exposures on a version of the commodity benchmark index that rolls during a different time window, and taking short exposure on the benchmark index. For each commodity, the expected congestion/liquidity alpha returns may vary across the futures curve. By seeking to take exposure on the optimal (most crowded) part of the commodity futures curve, the strategy aims to maximise the expected absolute return from establishing the long/short exposure. The strategy is rebalanced monthly and the selected commodities are equally weighted. This strategy tends to be uncorrelated with almost all traditional directional markets. The strategy seeks to maximise commodity alpha by each month taking long positions in the six most backwardated (or least contango) commodities and short positions in the six most contango (or least backwardated) commodities. Backwardation is measured using the 12-month slope in the commodity term structure, calculated as the slope from the first contract in the futures curve and the contract one year ahead. This strategy tends to be uncorrelated with almost all traditional directional markets. Alternative Strategies Investment Risk. An investment in alternative investment strategies (Alternative Strategies) involves risks, which may be significant. Alternative Strategies may include strategies, instruments or other assets, such as derivatives, that seek investment returns uncorrelated with the broad equity and fixed income/debt markets, as well as those providing exposure to other markets (such as commodity markets), including but not limited to absolute (positive) return strategies. Alternative Strategies may fail to achieve their desired performance, market or other exposure, or their returns (or lack thereof) may be more correlated with the broad equity and/or fixed income/debt markets than was anticipated, and the investment may lose money. Some Alternative Strategies may be considered speculative. Diversification does not assure a profit or protect against loss. Derivatives Risk. Derivatives could result in losses if the underlying asset or reference does not perform as anticipated. The value of derivatives may move in unexpected ways and may result in unlimited losses for the portfolio. Derivatives may be leveraged (creating leverage risk) and are subject to counterparty risk, hedging risk, pricing risk and liquidity risk. Non-investment-grade (high-yield) securities present greater price volatility and more risk to principal and income than higher rated securities. Short positions (where the underlying asset is not owned) can create unlimited risk. Foreign investments are subject to political, economic, market, and social risks, as well as to currency instabilities. 22

27 To find out more visit columbiathreadneedle.com Important Information: For Professional and/or Qualified Investors only (not to be used with or passed on to retail clients). Past performance is not a guide to future performance. The value of investments and any income is not guaranteed and can go down as well as up and may be affected by exchange rate fluctuations. This means that an investor may not get back the amount invested. This material is for information only and does not constitute an offer or solicitation of an order to buy or sell any securities or other financial instruments, or to provide investment advice or services. The research and analysis included in this document has been produced by Columbia Threadneedle Investments for its own investment management activities, may have been acted upon prior to publication and is made available here incidentally. Any opinions expressed are made as at the date of publication but are subject to change without notice and should not be seen as investment advice. Information obtained from external sources is believed to be reliable but its accuracy or completeness cannot be guaranteed. Issued by Threadneedle Asset Management Limited. Registered in England and Wales, Registered No , Cannon Place, 78 Cannon Street London EC4N 6AG, United Kingdom. Authorised and regulated in the UK by the Financial Conduct Authority. Columbia Threadneedle Investments is the global brand name of the Columbia and Threadneedle group of companies. columbiathreadneedle.com J26930