Global Tactical Asset Allocation October 2003 Giorgio De

Global Tactical Asset Allocation October 2003 Giorgio De Santis Managing Director Quantitative Strategies Group

Discussion outline A B C Overview of GTAA i) Definition and objectives ii) A brief history of (G)TAA Motivating GTAA i) Theoretical justifications ii) Some empirical evidence Implementing a GTAA program i) Portfolio construction ii) An example: Risk budget and optimal portfolio iii) Benefits of using futures and forwards D Concluding remarks Global Tactical Asset Allocation 1

Overview of GTAA

GTAA is a global strategy designed to capture relative value opportunities across countries, currencies and asset classes · GTAA is like security selection, where the securities are country stock markets, bond markets and currencies. · In contrast to stock and bond selection strategies, which focus on individual stocks and bonds, GTAA focuses on countries and broad asset classes. 1 2 3 Traditional TAA 4 Global Tactical Asset Allocation 3

A timeline of asset allocation The evolution of modern portfolio theory 1990: Black’s “Universal Hedging” defines global market equilibrium 1964: Sharpe’s “Capital Asset Pricing Model” 1986: Brinson, Hood and Beebower · In equilibrium, optimal portfolios are combinations of T-bills and the market capitalization portfolio · Strategic asset allocation explains more than 90% of institutional portfolio risk 1952 1960 Harry Markowitz publishes “Portfolio Selection” · Given a set of expected returns and risk estimates, portfolios with maximum return per unit of risk can be constructed 1970 1980 Mid 1970 s: · Tactical Asset Allocation (TAA) first used by US. pensions to capitalize on changing opportunities in a single country’s stock, bond and cash markets 1990 Late 1980 s: · Interest in TAA increases significantly following 1987 market crash · GTAA first used by US pensions to capitalize on changing opportunities at the country and global asset class level 1990 s: · Value-oriented TAA managers experience difficult environment · Top-tier GTAA managers continue to deliver promising results 2000: Renewed interest in GTAA strategies due to: · Increasing liquidity in global derivative markets · Increasing evidence of return predictability across and within asset classes · Recognition that quantitative models can manage risk effectively 1991: Black-Litterman propose practical model to overcome shortcomings of traditional mean-variance analysis Global Tactical Asset Allocation · Increased familiarity of institutional investors with GTAA 4

Improving the overall risk/return profile A GTAA process seeks to: · Reduce unintended asset allocation risk (Step 1): Completion · Generate alpha (Step 2): Overlay Result: Improved information ratio Global Tactical Asset Allocation 5

Managing portfolio risk Rebalancing frequency and methods of implementation The frequency of portfolio rebalancing and the method of portfolio rebalancing will affect the portfolio in terms of unintended asset allocation risk and transactions costs. Below, we simulated annual, quarterly and monthly rebalancing for a $2 billion portfolio from Jan-85 to Sep-02 and provide annualised tracking error figures as well as annualised transaction cost estimates. It assumes that the strategic benchmark resets monthly. Annual Rebalancing Quarterly Rebalancing Estimated annualised tracking error to the benchmark: TE = 0. 68% TE = 0. 28% Annualised rebalancing costs: Monthly Rebalancing Estimated annualised tracking error to the benchmark: TE = 0. 02% Annualised rebalancing costs: Physicals Futures 0. 03% 0. 06% 0. 07% 0. 02% Source: Goldman Sachs Asset Management Note: Simulated performance results do not reflect actual trading and have certain inherent limitations. Please see appendix for further disclosures. Global Tactical Asset Allocation 6

Theory and evidence in support of a GTAA program

Theoretical and intuitive motivation for GTAA In theory, asset class and country returns should be predictable · Valuations can drift away from fair · Investors may be slow to incorporate new information · Risk premia change over time · Structural barriers exist (market segmentation) · Market participants may not be motivated by profits (central banks) Global Tactical Asset Allocation 8

Some evidence that country returns are predictable Over time, inexpensive countries and high-momentum countries have provided significantly higher returns. Annualized gross returns on simulated three-way sorts of country and currency returns (1980 – 2001) Valuation theme Momentum theme Global Stock Markets: Global Bond Markets 1: Currencies: Global bond market tritile sorts cover period: January 1985 – December 2001 Note: Simulated performance results do not reflect actual trading and have certain inherent limitations. Please see appendix for further disclosures. 1 Global Tactical Asset Allocation 9

Using value and momentum in long/short portfolios GTAA Long/Short Portfolio Summary Statistics Equity Country Selection Long/Short Portfolios Bond Country Selection Long/Short Portfolios Currency Selection Long/Short Portfolios Book/Price 1 -Year Momentum Yield Curve Slope 1 -Year Momentum 5 -Year Reversal 1 -Year Momentum Mean Annual Return 4. 9% 13. 2% 1. 0% 0. 4% 3. 9% 3. 5% Annualized Volatility 11. 9% 13. 1% 4. 8% 3. 9% 7. 9% 9. 1% Information Ratio 0. 41 1. 01 0. 21 0. 12 0. 50 0. 38 T-statistic 1. 94 4. 74 0. 92 0. 51 2. 33 1. 78 2. 68% 0. 00% 17. 87% 30. 36% 1. 02% 3. 78% Statistic Probability Value for Mean > 0 The performance results stated above are backtested based on a methodology that is derived from an analysis of past market data with the benefit of hindsight. These results do not reflect the performance of a GSAM managed account or composite and are being shown for informational purposes only. If GSAM had managed your account during the period shown above it is highly improbable that your account would have been managed in a similar fashion due to differences in economic and market conditions. The performance results disclosed herein do not represent the results of actual trading using client assets. The backtested performance results depicted above do not reflect the deduction of advisory fees, brokerage or other commissions or exchange fees or any other expenses a client would have to pay. Global Tactical Asset Allocation 10

The benefits of combining factors Backtested cumulative excess return on Global Equity Country Long/Short portfolios 1980 – 2001 · Due to the low correlation between value and momentum, a strategy that combines the two signals outperforms each of the two strategies based on a single signal Source: Goldman Sachs Asset Management Note: The performance results stated above are backtested based on a methodology that is derived from an analysis of past market data with the benefit of hindsight. These results do not reflect the performance of a GSAM managed account or composite and are being shown for informational purposes only. If GSAM had managed your account during the period shown above it is highly improbable that your account would have been managed in a similar fashion due to differences in economic and market conditions. The performance results disclosed herein do not represent the results of actual trading using client assets. The backtested performance results depicted above do not reflect the deduction of advisory fees, brokerage or other commissions or exchange fees or any other expenses a client would have to pay. Global Tactical Asset Allocation 11

A simple stock/bond valuation timing model Backtested cumulative excess return on a U. S. Stock/Bond Timing strategy 1926 – 2001 · Overweight in U. S. stocks versus bonds is proportional to the spread between trailing S&P 500 earnings yield and intermediate-term bond yield. Source: Goldman Sachs Asset Management, Ibbotson Associates Note: The performance results stated above are backtested based on a methodology that is derived from an analysis of past market data with the benefit of hindsight. These results do not reflect the performance of a GSAM managed account or composite and are being shown for informational purposes only. If GSAM had managed your account during the period shown above it is highly improbable that your account would have been managed in a similar fashion due to differences in economic and market conditions. The performance results disclosed herein do not represent the results of actual trading using client assets. The backtested performance results depicted above do not reflect the deduction of advisory fees, brokerage or other commissions or exchange fees or any other expenses a client would have to pay. Global Tactical Asset Allocation 12

Implementation of a GTAA program

How should a Global Tactical Asset Allocation (GTAA) program be implemented? · A Global Tactical Asset Allocation (GTAA) program can generally be thought of as four separate strategies trading more than 40 assets: Strategy Assets 1. Asset class timing (3 assets) Stocks, bonds, cash 2. Equity country selection (23 assets) DAX, FTSE, S&P, Nikkei, . . 3. Fixed income country selection (9 assets) Eurobund, Gilt, US Treas, JGB, . . 4. Currency selection (11 assets) Euro, pound, yen, dollar. . · GTAA is usually coupled with a rebalancing strategy to seek to reduce unintended asset class risk in the underlying portfolio · GTAA is best implemented using liquid derivative instruments like futures and forward contracts · A GTAA overlay strategy can be implemented with minimal capital outlay and limited disruption to existing managers Global Tactical Asset Allocation 14

From Views to an Optimal Portfolio: The Black-Litterman Model The ‘output’ of Black-Litterman is a vector of mixed expected returns. Historical Data Equilibrium E(R)s (we use CAPM) Covariance Matrix Additional Views (e. g. anomalies) Black-Litterman Model Black-Litterman ‘mixed’ E(R)s Benchmark Risk Aversion Transaction Costs Mean-Variance Optimization Optimal Portfolio Weights 15

Example: A US Healthcare Company overall risk budget Targets Strategic Weights Capital Allocation ($mm) Excess Return (bps) Information Ratio Risk Allocation 0 0. 00 0% 200 500 0. 40 39% 39 300 750 0. 40 5% 2. 8% 19 500 1, 000 0. 50 2% Russell 2000 Value 2. 8% 19 500 1, 000 0. 50 2% International EQ MSCI ACWI ex US 11. 1% 77 360 600 0. 60 14% US FI Lehman Gov / Credit 44. 4% 278 80 100 0. 80 6% GTAA * N/A 30 80 100 0. 80 31% Total ** 100% 693 244 179 1. 37 100% Product Strategic BM US Large Cap Index Equity S&P 500 11. 1% 77 0 US Large Cap Active Equity Russell 1000 22. 2% 154 US LC Growth Russell 1000 Growth 5. 6% US Small / Mid Cap Growth Equity Russell 2500 Growth US Small Cap Value Equity Tracking Error (bps) · GTAA uses 4% of plan assets carved out of US fixed income. · In this case, the alpha is ported onto the overall portfolio (thus, an “overlay”) ¾ Completion portfolio equitizes/bondizes cash from overlay as well as other unintentional asset allocation risks in total portfolio ¾ GTAA overlay generates market-neutral alpha on the total plan * Risk-adjusted and assumes zero cross correlation. ** Assume 100% notional exposure. Note: For illustrative purposes only. There can be no assurance that the targets stated above can be achieved. Please be advised that the targets shown above are subject to change at any time and are current as of the date of this presentation only. Targets are objectives and should not be construed as providing any assurance or guarantee as to the actual returns or tracking error that will be realized in the future from investments in any asset or asset class described herein. Please see appendix for further information. Global Tactical Asset Allocation 16

US Healthcare Company GTAA portfolio snapshot As of July 31, 2002 [A] Our completion trades seek to minimize the expected tracking error to the benchmark portfolio. [C] = [A] + [B] [D] [E] = [C] + [D] Benchmark Weights Underlying Weights Completion Trades Completion Portfolio Total Weights 1. 1% 0. 7% 0. 5% 4. 2% 2. 4% 3. 6% 40. 9% 1. 2% 0. 7% 0. 5% 4. 2% 2. 5% 5. 5% 32. 9% 0. 0% -0. 3% 0. 1% 0. 2% 0. 0% -0. 3% 0. 4% 5. 8% 1. 2% 0. 4% 0. 6% 4. 4% 2. 2% 2. 1% 5. 8% 38. 7% 0. 0% 1. 0% 0. 1% -1. 6% 1. 3% -0. 3% 0. 0% 2. 4% 1. 2% 1. 4% 0. 7% 2. 8% 3. 6% 1. 8% 5. 8% 41. 0% 55. 56% 49. 67% 5. 88% 55. 55% 2. 85% 58. 40% 44. 4% 0. 0% 42. 8% 0. 0% 1. 7% 0. 0% 44. 4% 0. 0% 18. 2% 14. 5% -13. 0% -12. 2% 62. 6% 14. 5% -13. 0% -12. 2% 44. 44% 42. 75% 1. 69% 44. 44% 7. 47% 51. 92% Currency selection Dollar Block 2 Europe Japan 1. 2% 6. 6% 2. 2% 1. 2% 6. 7% 2. 2% 0. 0% -0. 1% 0. 0% 1. 2% 6. 6% 2. 2% 1. 6% -0. 5% 3. 4% 2. 7% 6. 1% 5. 6% Total Currency 10. 00% 10. 12% -0. 12% 10. 00% 4. 46% 100. 00% 92. 43% 7. 57% 100. 00% 10. 32% 110. 32% Regional Summary Equity country selection Emerging Markets Asia ex Japan Canada Europe Japan United Kingdom US Small Cap US Large Cap 1 Our overlay trades seek to maximize expected return subject to risk and other constraints. [B] Total Equity Fixed income country selection Dollar Block 2 Europe Japan United Kingdom Total Fixed Income Total Exposure Estimated Annualized Expected Return Estimated Annual TE Expected Information Ratio For illustrative purposes only. 1 2 -0. 35% 1. 20% -0. 29 Overlay Deviations Total Portfolio Weights 0. 00% 0. 29% 0. 00 Completion reduces unintended asset allocation risk 2. 58% 1. 22% 2. 12 Overlay maximizes expected return while taking a measured amount of tactical risk North America equity weight includes a 1. 8% equity overweight from our asset class timing strategy. Dollar Block fixed income weight includes a 13. 7% fixed income overweight from our asset class timing strategy. Note: Portfolio holdings are subject to change without prior notice. The above portfolio holdings may vary for each client in the strategy based on market conditions, client guidelines and diversity of portfolio holdings. Expected returns are statistical estimates of hypothetical average returns of economic asset classes, derived from statistical models. Actual returns are likely to vary from expected returns. Expected return models apply statistical methods and a series of fixed assumptions to derive estimates of hypothetical average asset class performance. Reasonable people may disagree about the appropriate statistical model and fixed assumptions. These models have limitations, as the assumptions may not be consensus views, or the model may not be updated to reflect current economic or market conditions. Accordingly, these models should not be relied upon to make predictions of actual future account performance. Goldman Sachs has no obligation to provide recipients hereof with updates or changes to such data. Please see appendix for further information. 17

Why we use futures Average cost of trading cash vs. futures in 15 equity markets (US$25 mn Trade) 80 Price Impact Spread Commissions & Fees 70 Cost (bps) 60 50 40 30 20 10 0 Stocks · Liquid futures are available in most major markets and asset classes. Futures · Futures are cheaper to trade than cash (baskets of stocks). · Commissions and spreads are approximately 90% lower than physicals. · Price impact is approximately 75% lower. Source: Goldman Sachs Asset Management Global Tactical Asset Allocation 18

Summary

Comparative advantages of a GTAA strategy FEATURE BENEFIT The modern GTAA program is a well diversified strategy unlike traditional TAA Potential for more consistently positive excess return The strategy is implemented using low cost derivative instruments Transaction costs are approximately 80% lower than for traditional strategies GTAA can reduce unintended asset allocation risk and add potential alpha Can significantly improve a fund’s overall information ratio Low correlation with other active risks Average correlation between active manager returns and GTAA manager returns is approximately 0. 01 Source: Goldman Sachs Asset Management Global Tactical Asset Allocation 20

A Quantitative Approach to Portfolio Management Global Tactical Asset Allocation 21

General notes This material is provided for educational purposes only and should not be construed as investment advice or an offer to sell or the solicitation of offers to buy any security. Opinions expressed herein are current opinions as of the date appearing in this material only. This presentation does not constitute an offer or solicitation to any person in any jurisdiction in which such offer or solicitation is not authorized or to any person to whom it would be unlawful to make such offer or solicitation. Prospective investors should inform themselves and take appropriate advice as to any applicable legal requirements and any applicable taxation and exchange control regulations in the countries of their citizenship, residence or domicile which might be relevant to the subscription, purchase, holding, exchange, redemption or disposal of any investments. Past performance is not a guide to future performance and the value of investments and the income derived from those investments can go down as well as up. Future returns are not guaranteed and a loss of principal may occur. Opinions expressed are current opinions as of the date appearing in this material only. No part of this material may be i) copied, photocopied or duplicated in any form, by any means, or ii) redistributed without Goldman Sachs Asset Management's prior written consent. Simulated performance results have certain inherent limitations. Such results are hypothetical and do not represent actual trading, and thus may not reflect material economic and market factors, such as liquidity constraints, that may have had an impact on the Adviser’s actual decision-making. Simulated results are also achieved through the retroactive application of a model designed with the benefit of hindsight. The results shown reflect the reinvestment of dividends and other earnings, but do not reflect advisory fees, transaction costs and other expenses a client would have paid, which would reduce return. No representation is made that a client will achieve results similar to those shown. Effect of fees on performance: The following table provides a simplified example of the effect of management fees on portfolio returns. For example, assume a portfolio has a steady investment return, gross of fees, of 0. 5% per month and total management fees of 0. 05% per month of the market value of the portfolio on the last day of the month. Management fees are deducted from the market value of the portfolio on that day. There are no cash flows during the period. The table shows that, assuming that other factors such as investment return and fees remain constant, the difference increases due to the compounding effect over time. Of course, the magnitude of the difference between gross-of-fee and net-of-fee returns will depend on a variety of factors, and the example is purposely simplified. Period Gross Return (%) Net Return (%) Differential (%) 1 year 6. 17 5. 54 0. 63 2 years 12. 72 11. 38 1. 34 10 years 81. 94 71. 39 10. 55 In the event any of the assumptions used in this presentation do not prove to be true, results are likely to vary substantially from the examples shown herein. These examples are for illustrative purposes only and do not purport to show actual results. The strategy discussed herein may include the use of derivatives. Derivatives often involve a high degree of financial risk in that a relatively small movement in the price of the underlying security or benchmark may result in a disproportionately large movement, unfavorable as well as favorable, in the price of the derivative instrument. References to market or composite indices, benchmarks or other measures of relative market performance over a specified period of time (each, an “index”) are provided for your information only. Reference to the indices shown herein does not imply that the portfolio will achieve returns, volatility or other results similar to such indices. The composition of the index may not reflect the manner in which a portfolio is constructed in relation to expected or achieved returns, portfolio guidelines, restrictions, sectors, correlations, concentrations, volatility or tracking error targets, all of which are subject to change over time. Tracking Error is one possible measurement of the dispersion of a portfolio’s returns from its stated benchmark. More specifically, it is the standard deviation of such excess returns. Tracking error figures are representations of statistical expectations falling within “normal” distributions of return patterns. Normal statistical distributions of returns suggests that approximately two thirds of the time the annual gross returns of the accounts will lie in a range equal to the benchmark return plus or minus the tracking error if the market behaves in a manner suggested by historical returns. Targeted tracking error therefore applies statistical probabilities (and the language of uncertainty) and so cannot be predictive of actual results. The tracking error that will actually be achieved may inherently lie outside of the range suggested by a "normal " statistical distribution of returns. The actual tracking error is the result of many factors (including but not limited to market volatility, company specific anomalies, instability of correlation between benchmark holdings, timing differences between the calculation of the portfolio value and the valuation of the benchmark by the index provider. In addition, past tracking error is not indicative of future tracking error and there can be no assurance that the tracking error actually reflected in your accounts will be at levels either specified in the investment objectives or suggested by our forecasts. © Copyright 2003 Goldman, Sachs & Co. All rights reserved. Member SIPC/NASD. Global Tactical Asset Allocation 22