WHITE PAPER GLOBAL LONG-TERM UNCONSTRAINED FEBRUARY 217 FOR PROFESSIONAL CLIENTS ONLY Martin Currie s Asia Long-Term Unconstrained (ALTU) strategy has, since inception in 28, been successful in delivering long-term returns to clients, with lower volatility than the broader equity market. This has been achieved by making a long-term capital commitment to companies with high returns that are sustainable into the future. This paper looks at the potential efficacy of implementing such a philosophy on a truly global basis, with the creation of a Global Long-Term Unconstrained (GLTU) strategy. Executive summary State Street Global Exchange was commissioned to undertake an independent, quantitative analysis of the premise behind the Global Long-Term Unconstrained strategy (GLTU). That is, committing capital to high quality businesses that can consistently create value, delivers superior returns with less volatility over the long term. Stocks were screened for a return on invested capital (ROIC) that exceed their cost of equity (COE) over each of the last 1 years and a minimum market cap of US$5 billion. The analysis indicates: The strategy universe outperforms the MSCI AC World over the backtest period on a risk-adjusted basis The stocks which pass the factor screen are relatively balanced across sector and geographic regions In general, any statistically significant exposures to value, size and momentum factors appear to be sporadic The range of potential returns within the strategy universe could be enhanced by stockpicking martincurrie.com
THE PHILOSOPHICAL PREMISE Stocks with a long track record of high economic returns converted to strong cash flow generation, supported by sustainable business models, deliver the most consistent and attractive returns to shareholders. The market is, by its nature, short term and speculative, consequently it fails to adequately capture the true worth of such stocks as they compound their value quarter after quarter, year after year. The GLTU strategy aims to capture this value. TO SUPPORT THIS RESEARCH INTO A GLTU STRATEGY, STATE STREET WAS COMMISSIONED TO UNDERTAKE AN INDEPENDENT, QUANTITATIVE ANALYSIS OF THE PREMISE. THE ACADEMIC PROOF To support this research into a GLTU strategy, State Street Global Exchange was commissioned to undertake an independent, quantitative analysis of the premise. More specifically, they were asked to screen all stocks that belong to the MSCI AC World index as a proxy for global listed equities including emerging markets. The analysis period stretched over the 2 years from September 1995 to December 214, with fundamental data extending back a further 1 years. The screen was for stocks that meet the following criteria: Value creation: Companies where the last reported return on invested capital (ROIC) exceeded cost of equity (COE) as at 3 June each year during the most recent 1 years. Minimum size: Companies with a market capitalisation of US$5 billion or more as at 31 December 214. As at 3 June each year State Street dynamically scaled this minimum market capitalisation threshold to earlier backtest periods in proportion to the growth of total MSCI AC World index market capitalisation. Due to the global nature of the MSCI AC World index and the lag at which accounting data for stocks becomes public information for the corresponding time period, the analysis accounted for relevant lags in a uniform manner across all stocks. Specifically, it assumed that the latest ROIC data available corresponding to the end of the previous financial year was available as at 3 June each year. A lag of three months for accounting data for stocks in countries where the financial year ends at the end of March was incorporated, as was a lag of six months for accounting data for stocks in countries where the financial year finishes at the end of December. The screen was applied at 3 June each year. At this point stocks that passed the factor screen were rebalanced back to 1/n equal weights. Further specific detail on the formulae used, data sources and any cleaning or adjustments to that data is included as Appendix 1. 2
EXAMINING THE SCREENED UNIVERSE The number of stocks passing the screen increased over the backtest period. The increase in the number of stocks in the MSCI AC World index through the period contributed to this. Perhaps, more meaningfully, the reduction in the COE over time as interest rates have fallen positively impacted the number of stocks meeting the criteria (chart 1). Chart 1: Number of stocks that pass the factor screens 35 3 25 Number of stocks 2 15 5 1994 1995 1996 1997 1998 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. The results showed a consistent breadth of exposure in stocks passing the screen, by sector (chart 2) and region (chart 3 overleaf) over the previous 2 years. There is no evidence that any backtest results are dominated by any single country or sector. Chart 2: Breakdown of stocks that pass the factor screen by MSCI GICS sector 8 Sector (%) 6 4 2 1994 1995 1996 1997 1998 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Utilities Telecommunications Materials Information technology Industrials Healthcare Financials Energy Consumer staples Consumer discretionary Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. The resources (materials and energy) and financials sectors offer up relatively fewer opportunities, than the consumer and healthcare sectors. However, all major sectors have some representation throughout the last 2 years. GLOBAL LONG-TERM UNCONSTRAINED 3
Chart 3: Breakdown of stocks that pass the factor screen by geographical region 8 Region (%) 6 4 2 1994 1995 1996 1997 1998 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Africa Australasia South America Asia Europe North America Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. North America is the largest region in the MSCI AC World universe and, perhaps unsurprisingly, dominates. Similar analysis at the country level shows a wide spread of opportunity. HISTORIC PERFORMANCE AND BEHAVIOUR Having identified the companies that meet the criteria each year over the last 2 years the next step was to analyse the returns that would have been achieved by investing in this universe. Both return and risk data were calculated on an equal weighted basis as described above. The results for the equally weighted universe are shown below (chart 4). Chart 4: GLTU strategy universe cumulative performance 6 5 MSCI AC World Equally weighted universe Cumulative return index 4 3 2 1994 1995 1996 1997 1998 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. The market in the chart above is represented by the MSCI AC World index and the returns for both the market and the screened universe exclude dividends. Summary results over the 2-year period are given in the table overleaf (table 1). All returns are in US dollar. 4
Table 1: GLTU strategy universe annualised performance Strategy universe equally weighted MSCI AC World index Annualised return (%) 8.7 5. Annualised risk (%) 14.4 17.5 Return/risk 8 Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. The initial results show that a focus on businesses that have been able to sustain high value creation would have delivered significant excess returns over the sample period. In addition, the universe of stocks used to generate these returns over the last 2 years maintained a broad representation of countries and sectors. However, it was important to test if these returns were simply a reflection of some other risk factor inherent in the screened universe. State Street looked for statistically significant exposures which could indicate that the strategy s excess performance could be attributable to factors other than the intended stock selection, based on the proprietary measure for sustained value generation. They were specifically asked to choose their preferred method for doing this. They opted to test for beta, value, size and momentum factors using their own methodology and then to repeat the tests using the market, value, size and momentum factors from the Carhart four-factor model. The definitions of the factors they used are given in the table below (table 2). Table 2: Factor descriptions Factor Description State Street equity style flow indication factors Beta Value Market cap Momentum High market beta stocks minus low market beta stocks High book-to-market stocks minus low book-to-market stocks Small market capitalisation minus large market capitalisation stocks Winners minus losers Carhart four-factor model Market excess return Value Market cap Momentum Excess return High book-to-market stocks minus low book-to-market stocks Small market capitalisation minus large market capitalisation stocks Winners minus losers Source: State Street Global Exchange, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html. GLOBAL LONG-TERM UNCONSTRAINED 5
Tests of significance were carried out using an ordinary least squares univariate regression over rolling 24 months for the longest available common sample period from the data. The results of these rolling regressions using the State Street factors are shown in the charts below (charts 5 8). The factor coefficient is displayed if the factor is statistically significant at the 1% level; otherwise no coefficient value is plotted. The R-squared value for each rolling univariate regression is also plotted. The results shown below are for an equal weighted universe. Chart 5: Value Coefficient R 2 1.5.5 (.5) () 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange: as at 31 December 214. Coefficents displayed are significant at the 1% level. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Any exposures to the value factor appear to be sporadic, and in most cases, insignificant. Chart 6: Market capitalisation Coefficient (.5) () (1.5) (2.) R 2 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange: as at 31 December 214. Coefficents displayed are significant at the 1% level. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Any exposures to the market capitalisation factor appear to be sporadic, and in most cases, insignificant. 6
Chart 7: Momentum Coefficient.5 (.5) () R 2 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange: as at 31 December 214. Coefficents displayed are significant at the 1% level. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Any exposures to the momentum factor appear to be sporadic, and in most cases, insignificant. Chart 8: Beta Coefficient R 2 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange: as at 31 December 214. Coefficents displayed are significant at the 1% level. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. The strategy universe does have persistent exposure to the market beta factor. This is not surprising as the universe reflects a long-only portfolio of stocks. In terms of direction, the expectation was that these higher-quality businesses would in aggregate deliver excess returns with less volatility than the market. This is a feature of the ALTU strategy. The positive impact of this lower volatility is clearly greater in periods of market stress (e.g. the bear markets of 2 23, 28 and 211) as can be seen from the chart above (chart 8). However, the strategy universe has also delivered absolute returns in excess of the market in periods when the market has been rising. Results for the Carhart model factors yield similar results and are included as Appendix 3. GLOBAL LONG-TERM UNCONSTRAINED 7
RANKING AND GREATER RETURN OPPORTUNITY FROM WITHIN THE QUALITY UNIVERSE The next question to be answered was whether or not additional value can be found in emphasising the highest-quality businesses from within those that make the cut. Specifically, did stocks which rank the highest within the universe that pass the screen outperform those which rank lowest? At 3 June each year for the last 2 years, stocks that passed the screen by their factor score were ranked and equal size groupings were formed based on these rankings. The performance of stocks which belonged to each quintile and tercile throughout the backtest period were measured. The results provide supporting evidence that during the backtest period, the stocks with higher ranks outperform those with lower ranks, in general. The results hold for both equally weighted portfolios sorted by quintile and those sorted by tercile. Charts 9 and 1 and table 3 below show the results by quintile over the backtest period. Quintile 5 reflects the highestscoring stocks within the screened universe and quintile 1 reflects the lowest-scoring stocks. Chart 9: Absolute return by quintile 9 8 Q1 Q2 Q3 Q4 Q5 Cumulative return index 7 6 5 4 3 2 1994 1995 1996 1997 1998 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. Chart 1: Relative return by quintile 35 3 Q1 Q2 Q3 Q4 Q5 Cumulative return index 25 2 15 5 1994 1995 1996 1997 1998 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. 8
Table 3: GLTU equally weighted universe and absolute performance by quality (quintiles) (%) Strategy Index Q1 Q2 Q3 Q4 Q5 Annualised return 8.7 5. 6.9 6.6 8.4 1.5 1.3 Annualised risk 14.4 17.5 17.6 14.5 15.2 14.4 16. Risk/return 8.39 5.55.73 4 Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. The ability to improve the return outcome by subselecting stocks from within the universe is significant. The final step was to further explore the range of opportunity from stockpicking, by considering the behaviour of concentrated global stock portfolios drawn from this universe. PORTFOLIO CONSTRUCTION AND THE RISK/RETURN OPPORTUNITY For this final stage the range of risk and return outcomes that could be expected from a distribution of equal weighted 35-stock portfolios drawn from the universe of high-quality companies was examined. To do this 1, randomly simulated, 35-stock portfolios were created from the universe. To make this test realistic, turnover was constrained in the portfolios. This was done by randomly replacing 2% of the stocks in each simulated portfolio each year, rather than randomly replacing all of the stocks in the portfolio each year. So for each simulation, 35 stocks were randomly selected from the universe of stocks that pass the factor screen as of 3 June 1994. As at the end of June each year thereafter, a randomly selected 2% of the stocks within the portfolio were replaced by new stocks, which in turn were randomly selected from the remaining universe of stocks that passed the factor screen for the respective periods. The idea was to simulate a concentrated portfolio with limited annualised turnover but with stocks chosen at random. The results are shown below. Chart 11: Annualised return distribution 12 Mean: 8.7% Standard deviation:.7% Market 5.% 8 Frequency 6 4 2 5 6 7 8 9 Annualised return (%) 1 11 Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. Annualised turnover calculations incorporate the turnover from the annual portfolio stock selection in addition to the quarterly rebalancing trades back to target weights. GLOBAL LONG-TERM UNCONSTRAINED 9
Chart 12: Annualised risk distribution 8 Mean: 14.8% Standard deviation:.9% Market: 17.5% Frequency 6 4 2 11 12 13 14 15 16 17 18 19 Annualised risk (%) Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. Annualised turnover calculations incorporate the turnover from the annual portfolio stock selection in addition to the quarterly rebalancing trades back to target weights. Chart 13: Market beta distribution 8 Mean:.77% Standard deviation: 5% Market: Frequency 6 4 2.55 5.7.75 5.9.95 Beta Source: State Street Global Exchange, Datastream, Worldscope, Thomson Reuters: as at 31 December 214. These numbers do not represent the performance of an actual or model portfolio and should not be considered as an indication of future returns. Theoretical results without transaction costs and with estimated trade dates. Annualised turnover calculations incorporate the turnover from the annual portfolio stock selection in addition to the quarterly rebalancing trades back to target weights. 1
In general, the distribution of outcomes from the randomly simulated 35-stock portfolios indicate that a portfolio of 35 stocks selected at random from the universe of stocks that pass the factor screen should: Generate positive excess annualised returns versus the MSCI AC World index Generate lower annualised risk versus the MSCI AC World index Exhibit a market beta which is less than 1 It is important to note that these distributions display the range of outcomes that are likely to occur for a randomly selected portfolio containing 35 stocks from the universe of stocks that pass the factor screen, derived from 1, simulations. This does not reflect the range of potential outcomes where the manager tasked with stockpicking possesses skill. Therefore, the annualised return attainable for a strategically selected 35-stock portfolio from the universe of stocks which pass the factor screen, by a manager with skill, could reside outside of this distribution. Similarly, the results from the quintile ranking of stocks earlier show that in general, stocks with higher factor scores outperform stocks with lower factor scores in the corresponding period. It is noteworthy that the absolute annualised return distribution for randomly selected portfolios would likely shift to the right (i.e. higher returns) if State Street had replaced 2% of the portfolio s randomly selected stocks with stocks which had relatively high factor ranks, based on these findings. CONCLUSION State Street was asked to carry out this research to test the strategy s investment belief. That is, committing capital to high quality businesses that can consistently create value will deliver superior returns with less volatility over the long term, and this works for global as well as Asian stocks. The findings from this analysis indicate: The strategy universe outperforms the market over the backtest period on a risk-adjusted basis. The stocks which pass the factor screen are relatively balanced across MSCI GICS level 3 industry classifications and distributed across geographic regions similarly to the MSCI AC World index (the market). In general, any exposures to value, size and momentum factors appear to be sporadic. In most cases, overall, they were insignificant, which provides supportive evidence that the strategy is generating a unique source of returns to the factors modelled. The long-only nature of the strategy can explain the persistent market beta factor exposure observed throughout the backtest. The results from the stock grouping backtests provide supporting evidence that in general, during the backtest period, the stocks with higher ranks outperform those stocks with lower ranks. Again, in general, the distribution of outcomes from the randomly simulated 35-stock portfolios indicate that, on average, a portfolio of 35 stocks selected from the universe of stocks that pass the screen should: Generate positive excess annualised returns versus the market Generate lower annualised risk versus the market Exhibit a market beta which is less than 1 These distributions display the range of outcomes that are likely to occur for randomly selected portfolios. While this can provide the range of potential returns realised depending on random stock selection, this does not reflect the range of potential outcomes where the manager tasked with stockpicking possesses skill. Therefore, the annualised return attainable for a strategically selected 35-stock portfolio from the universe of stocks which pass the factor screens by a manager with skill could be superior. GLOBAL LONG-TERM UNCONSTRAINED 11
APPENDIX 1: DATA DESCRIPTION Data type Source Description Return on invested capital (ROIC) Worldscope, Datastream ROIC is derived from the following equation: NI BL + ((INT ICAP) * (1 T)) (TC t + STD t ) + (TC t-1 + STD t-1 ) 2 * Cost of equity (COE) Market beta (BETA) State Street Global Exchange State Street Global Exchange where: NI = Net income, BL = Bottom line, INT = Interest expense on debt, ICAP = Interest capitalised, T = Tax rate, TC = Total capital, STD = Short-term debt and current portion of long-term debt. ROIC calculation uses restated data for last year s values, where available. Cost of equity is estimated from the following equation: COE = r ƒ + ß.RPM where: COE = cost of equity, r ƒ = risk-free rate, ß = market beta, RPM = risk premium. Risk premium is assumed to equal 4.3% for all stocks across all time periods, as specified by Martin Currie Investment Management. Regression coefficient in the regression: r = a + ß.r m where: r = stock price return, ß = market beta, r m = return of the MSCI AC World index. Market beta is estimated using daily data over a 2-year rolling window. For time periods where we are unable to compute the market beta for a given stock (due to data availability reasons), we assume its market beta is equal to the market beta of the stock s designated MSCI GICS industry for the corresponding time period. As agreed with MCIM, we make the simplifying assumption that each stock s market beta prior to 1995 is equal to its market beta observed on 3 June 1995 (or its industry market beta on 3 June 1995, if not available). Risk-free rate (RF) State Street Global Exchange Thomson Reuters, Datastream 1-year benchmark government bond redemption yield for the geographical classification of the stock, where available. For countries with insufficient history, the yield is estimated from the equation: r ƒt = r ƒus + CRED where: r ƒt is the risk-free rate for the country with missing data, r ƒus is the 1-year U.S. government benchmark bid yield and CRED is the average yield spread between the country with missing data and the U.S over the common sample available over the full backtest period. For countries with no data available, riskfree rates for proxy countries are used. Please see tables 2 and 3 for further information on assumed risk-free rates. Market capitalisation (MV) Datastream Market value is the share price multiplied by the number of ordinary shares in issue. The amount in issue is updated whenever new tranches of stock are issued or after a capital change. For companies with more than one class of equity capital, the market value is expressed according to the individual issue. Price index (PI) Datastream The price index expresses the price of a stock as a percentage of its value on the base date, adjusted for capital changes. Source: State Street Global Exchange, Datastream, Worldscope. 12
APPENDIX 2: RATIONALE FOR PROXY RISK-FREE RATE ASSUMPTIONS Geographic region Proxy Rationale Abu Dhabi United States Currency peg to US dollar Channel Islands United Kingdom Currency Dubai United States Currency peg to US dollar Jordan United States Currency peg to US dollar Luxembourg Europe Currency Peru United States Geographic proximity and exchange rate stability Qatar United States Currency peg to US dollar Source: State Street Global Exchange. APPENDIX 3: GLTU STRATEGY FACTOR EXPOSURES (CARHART FACTORS) Chart A: Market excess return R 2 Coefficient 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Chart B: Value Coefficient R 2 2. () (2.) 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange: as at 31 December 214. Notes: Coefficients displayed are significant at the 1% level. GLOBAL LONG-TERM UNCONSTRAINED 13
APPENDIX 3 (CONTINUED): GLTU STRATEGY FACTOR EXPOSURES (CARHART FACTORS) Chart C: Market cap Coefficient R 2 2. () (2.) 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Chart D: Momentum Coefficient R 2 2. () (2.) 1999 2 21 22 23 24 25 26 27 28 29 21 211 212 213 214 Source: State Street Global Exchange: as at 31 December 214. Notes: Coefficients displayed are significant at the 1% level. 14
IMPORTANT INFORMATION The back test results presented in this document are based on simulated performance results. Please be aware these have certain inherent limitations. Backtested performance returns do not represent the impact of trading. The trades in the backtesting have not been executed and may not fully reflect the impact of market factors such as liquidity. Martin Currie makes no representation that any account will or is likely to achieve returns similar to those illustrated as a result of the backtesting presented in this document. This information is issued and approved by Martin Currie Investment Management Limited ( MCIM ). It does not constitute investment advice or represent an inducement to invest. Market and currency movements may cause the capital value of shares, and the income from them, to fall as well as rise and you may get back less than you invested. The information contained in this presentation has been compiled with considerable care to ensure its accuracy. But no representation or warranty, express or implied, is made to its accuracy or completeness. Martin Currie has procured any research or analysis contained in this presentation for its own use. It is provided to you only incidentally, and any opinions expressed are subject to change without notice. The document may not be distributed to third parties and is intended only for the recipient. The document does not form the basis of, nor should it be relied upon in connection with, any subsequent contract or agreement. It does not constitute, and may not be used for the purpose of, an offer or invitation to subscribe for or otherwise acquire shares in any of the products mentioned. Investors should also be aware of the following risk factors which may be applicable to the strategies. Investing in foreign markets introduces a risk where adverse movements in currency exchange rates could result in a decrease in the value of your investment. Emerging markets or less developed countries may face more political, economic or structural challenges than developed countries. Accordingly, investment in emerging markets is generally characterised by higher levels of risk than investment in fully developed markets. The strategies holds a limited number of investments. If one of these investments falls in value this can have a greater impact on the portfolio s value than if it held a larger number of investments. Smaller companies may be riskier and their shares may be less liquid than larger companies, meaning that their share price may be more volatile. For Investors in the USA, the information contained within this document is for Institutional Investors only who meet the definition of Accredited Investor as defined in Rule 51 of the United States Securities Act of 1933, as amended ( The 1933 Act ) and the definition of Qualified Purchasers as defined in section 2 (a) (51) (A) of the United States Investment Company Act of 194, as amended ( the 194 Act ). It is not for intended for use by members of the general public. Any distribution of this material in Australia is by Martin Currie Australia Limited ( MCA ). Martin Currie Australia is a division of Legg Mason Asset Management Australia Limited (ABN 76 4 835 849). Legg Mason Asset Management Australia Limited holds an Australian Financial Services Licence (AFSL No. AFSL 24827) issued pursuant to the Corporations Act 21. GLOBAL LONG-TERM UNCONSTRAINED 15
Martin Currie Investment Management Limited Registered in Scotland (no SC6617). Registered office: Saltire Court, 2 Castle Terrace, Edinburgh EH1 2ES Tel: 44 () 131 229 5252 Fax: 44 () 131 228 5959 www.martincurrie.com Authorised and regulated by the Financial Conduct Authority. Please note that calls to the above number may be recorded.