RESEARCH GROUP ADDRESSING INVESTMENT GOALS USING ASSET ALLOCATION

M A Y 2 0 0 3 STRATEGIC INVESTMENT RESEARCH GROUP ADDRESSING INVESTMENT GOALS USING ASSET ALLOCATION

T ABLE OF CONTENTS ADDRESSING INVESTMENT GOALS USING ASSET ALLOCATION 1 RISK LIES AT THE HEART OF ASSET ALLOCATION 1 ASSET CLASSES SHARE KEY CHARACTERISTICS 5 HOW WE DEVELOP ASSET ALLOCATIONS 6 THE DIFFERENCE BETWEEN STRATEGIC AND TACTICAL ASSET ALLOCATION 8 BENEFITS OF ASSET ALLOCATION AS A STRATEGY 8

A DDRESSING I NVESTMENT G OALS U SING A SSET A LLOCATION For most people, investing is a means to an end a way to achieve their financial goals whether they be retirement or college funding, purchase of real estate, or accumulating an estate. An asset allocation policy identifies a mix of investments that is best suited to meet investor goals, given their risk tolerance and time horizon. This paper explains the principles behind asset allocation, including the role risk plays in developing an investment strategy, the importance of diversification, how asset allocations are created, and why, for most people, following a strategic asset allocation is an effective approach to meeting long-term financial goals. R ISK LIES AT THE HEART OF ASSET ALLOCATION 1 Much of the modern theoretical framework for constructing portfolios was developed by Harry M. Markowitz in the early 1950s. In fact he won a Nobel Prize in economics in 1990 for proposing a way to measure portfolio risk and introducing new concepts of diversification and efficient portfolios. His major contribution was developing a portfolio selection process that addresses the needs of all investors regardless of their appetite for risk. RISK AVERSION Since investment outcomes are generally not known in advance, all investing involves risk. Most investors are averse to risk, meaning they prefer prospects with less risk to those with more risk. A riskaverse individual faced with a sure prospect and a risky alternative (a gamble) with the same expected payoff should always choose the sure prospect. A risk-averse investor won t accept a risky prospect unless he or she feels sufficiently rewarded for bearing the risk. For example, a risk-averse individual will prefer receiving $100 with certainty to a gamble that gives a 50% probability of winning $150 and a 50% probability of winning $50. The same individual, however, may be inclined to choose a risky alternative if the terms were altered so that a reward is given for bearing more risk. That person might prefer a gamble paying $80 or $150 with equal probability as an alternative to receiving $100 with certainty.

ASSET ALLOCATION BEGINS WITH AN HONEST ASSESSMENT OF RISK 2 It s important that investors discover their preferred trade-off between risk and return, which is usually done by filling out an investment questionnaire with their financial professional. This is a key step that often determines the success or failure of a chosen asset allocation strategy. An incorrectly specified risk tolerance frequently leads to changes in portfolio allocations at the worst possible time, as illustrated in the following example. Imagine an investor named Jim who was given a comprehensive investment risk questionnaire in 1995 that indicated a moderate risk appetite. Based on that profile, the recommended portfolio may, for simplicity s sake, have included 60% equity and 40% fixed income assets. After three years, Jim became increasingly dissatisfied with the performance of his portfolio, not because it was doing poorly it had increased almost 85% during those three years but because the S&P 500 Index had increased by 125% over that period. Everyone else seemed to be getting rich and he was getting left behind. Jim called his financial professional, who reminded him that one cannot get equity market returns without assuming equity market risk. But he was determined to participate in the new economy and demanded to be 100% invested in equities. Two years later the equity-only portfolio had earned another 50% and Jim congratulated himself on his investment acumen. What did his financial professional know? Asset allocation was obsolete! Then came 2000 and his portfolio lost 9% of its value. Still he expected things to get better; it was the new economy after all a time of perpetual growth. He waited another year and lost another 12%. Still Jim believed that the recovery was at hand. He knew there hadn t been three consecutive years of declining equity markets since 1941. Then, by the end of 2002, he lost an additional 22%. When Jim called his financial professional, he was informed that if he had kept his original portfolio, he would have been ahead of his current portfolio by 35 percentage points. And it would not have been subjected to such wide volatility swings. What did Jim learn? He had either become more risk tolerant or he had not described his original risk appetite truthfully. If he had become more risk tolerant, then he was unlucky because his appetite for risk increased as the market peaked. In effect Jim experienced only a small benefit of the bull market while suffering all the consequences of the subsequent bear market. The other possibility was that his original risk appetite was correctly determined, but the initial gains blinded him to the increased risk inherent in selling his bond holdings and buying additional equity.

SETTING REALISTIC EXPECTATIONS FOR PORTFOLIO RETURNS When portfolios are developed based on an investor s risk aversion and investing preferences, returns should be evaluated in that context. If, for example, a risk profile matches an investor with a 50%/50% (stock/bond) portfolio, the returns probably won t beat an equity market index in a bull market since it doesn t contain 100% equity market risk. Lower return is the price the investor pays for lower risk. Conversely, when equity markets decline, the portfolio won t decline as much since its bond component will help offset some of the loss. Similarly, if an investor wants only domestic securities in the portfolio, it wouldn t make sense to compare the performance of the equity fund investments to the best performing global equity fund. It s the wrong comparison because a conscious decision was made to exclude foreign securities. STANDARD DEVIATION A WAY OF MEASURING RISK Since investors face choices involving risk, it s important to find a way to measure risk. There are many ways to quantify investment risk, but the one used most often defines risk in terms of the volatility of investment returns. The way we measure this volatility is with a statistic known as standard deviation. Standard deviation measures the dispersion of a series of returns around the average (or mean) return. A high standard deviation indicates that few of the returns are near the average, while a low standard deviation implies that the returns are concentrated near the average. Standard deviation is itself a type of average: it is the average amount by which the individual periodic returns deviate from the mean return, hence the name standard deviation. Standard deviation is a particularly useful concept when associated with the normal (bell-shaped) distribution. For example, when equity returns have an annual standard deviation of 18% and a mean return of 10%, we re able to estimate the likely ranges of possible annual returns. In this example, about 68% of all observed annual returns would fall between 8% and +28%, which is within one standard deviation from the mean. Similarly, two standard deviations around the mean return capture approximately 95% of the returns; in this example, it is a range between 26% and +46%. 3 ONE STANDARD DEVIATION TWO STANDARD DEVIATIONS Mean Return Mean Return 68% of Observations 95% of Observations 1 Standard Deviation +1 Standard Deviation 2 Standard Deviation +2 Standard Deviation

PORTFOLIO DIVERSIFICATION AND CORRELATION Standard deviation can be applied to one asset or to a whole portfolio of assets. At the portfolio level, it is possible to reduce overall risk through diversification, but only when asset returns do not move up and down together in perfect unison. Thus, the benefits of diversification depend less on the number of assets included in the portfolio than on the way individual assets perform relative to one another. A statistic known as the correlation coefficient measures the tendency for asset returns to move together relative to their average values. Positive correlation indicates that when the return for one asset is high (or low), the return for the other asset will also tend to be high (or low). Negative correlation, on the other hand, occurs when one asset return tends to be high when the other is low. PERFECT POSITIVE CORRELATION (+1) PERFECT NEGATIVE CORRELATION ( 1) 4 Asset A Return Asset B Return Asset B Asset A Time Period Time Period Correlation is measured on a scale from 1 to +1. Three special cases are of particular interest. Correlation equal to +1 indicates perfect positive correlation (or simply perfect correlation ) and implies that the asset returns move in lockstep with each other. Perfect negative correlation, a value of 1, implies that the assets always move in opposite directions. Lastly, two assets that have a correlation of 0 are uncorrelated their movements are not related in a predictable fashion. The lower the correlation between assets, the more risk can be reduced through diversification. Perfectly correlated assets provide no diversification since they always move together. Combining uncorrelated assets results in substantial risk reduction because, on average, those that perform poorly will be offset by others that perform well. Negatively correlated assets provide even greater risk reduction because they almost always offset one another. In the special case of perfect negative correlation, it is possible to eliminate all risk. In practice, however, most asset returns exhibit positive correlation.

EFFICIENT PORTFOLIOS Markowitz used the principles of correlation and standard deviation in devising a formal methodology called mean-variance analysis to construct efficient portfolios. These are portfolios of assets that result in the highest possible expected returns at a given level of risk. Targeted Return 20 18 16 14 12 10 8 6 4 2 0 Efficient Frontier Portfolio 2 Portfolio 1 More return, same risk Current Portfolio Same return, less risk 100% Bonds 100% Stocks 0 2 4 6 8 10 12 14 16 18 20 Standard Deviation (Risk) 5 The set of efficient portfolios is often represented in a graph as the efficient frontier, such as the one above. The efficient frontier shows the risk/return trade-offs facing an investor considering various portfolio strategies. Along the efficient frontier, the investor cannot increase expected portfolio return without simultaneously assuming a higher level of risk. Any portfolio that plots below the efficient frontier is considered inefficient and can be improved upon, as shown on the graph. By definition, no portfolio can plot above the efficient frontier. Inefficient portfolios are not effectively meeting their investment potential and should be adjusted. A SSET CLASSES SHARE KEY CHARACTERISTICS Asset classes divide the universe of securities into sets of mutually exclusive groups. Securities within the same asset class share similar economic characteristics. Careful definition of asset classes is important since ambiguity may lead to double counting and misallocation of resources. Therefore there are some general rules that should be satisfied when distinguishing asset classes. First, an asset class must represent a large enough portion of the universe of securities to be able to accommodate a meaningful portion of the portfolio. Second, each asset class should have distinct long-term expected returns and standard deviations. Finally, an asset class should not have perfect pair-wise correlation with another asset class or combination of asset classes; otherwise, it provides no diversification benefits to the portfolio. Asset classes are generally split into two major categories: equities and fixed income. Equities are further classified by capitalization level (large cap, midcap, and small cap); investment style (value and growth); or market (domestic, developed, and emerging). Fixed income securities are divided by issuer (Treasuries, municipals, and corporates); maturities (long-term, intermediate-term, and short-term); and credit quality of the issuer. The following table illustrates a typical breakdown of asset classes.

The choice of asset classes is often driven by some practical considerations, such as the size of the portfolio, trading costs, or other investment management factors. For an individual investor, the choice of asset classes may also be contingent on the available fulfillment options, that is, the number and type of available funds may determine the asset classes used in an asset allocation analysis. COMMONLY USED ASSET CLASSES EQUITIES FIXED INCOME OTHER 6 Large-Cap Value Growth Mid-Cap Value Growth Small-Cap Value Growth Foreign Equities Developed Markets Emerging Markets All Maturities Government Corporate Mortgages Municipal High Yield Foreign Limited Maturities Cash Equivalents Convertibles Real Estate Private Equity Hard Assets H OW WE DEVELOP ASSET ALLOCATIONS The process of combining different asset classes into portfolios and identifying the best alternative for an investor constitutes the core of asset allocation. The diagram on the next page outlines the process. There are three key inputs: (1) the investor s risk preferences, (2) the menu of asset classes, and (3) our Capital Market Assumptions, i.e., estimates of expected returns, standard deviations, and correlations for the selected asset classes. Given these inputs, how do we choose the best portfolio? Even for a very short list of potential asset classes, there are many different ways to combine them into portfolios. As discussed above, however, only those on the efficient frontier are worth considering. Recall that these are the portfolios having the highest expected return for a given level of risk. They can be identified using a mathematical technique called mean variance optimization. Armed with a measure of the investor s willingness to accept additional risk in exchange for additional return, the optimizer can also identify the best portfolio for that investor. Note that the efficient frontier becomes flatter as the risk level increases, i.e., as we move from left to right on the graph. This means that each increment to expected return costs more and more in terms of additional risk. As we move up the curve, we will eventually reach a point where this cost just matches the investor s willingness to accept more risk in exchange for a higher expected return. The portfolio at that point on the frontier is the best asset allocation for that investor. As time goes by, investors need to evaluate their financial circumstances and reassess the risk they are willing to bear. A significant change in either would imply a need to update their asset allocation policy. Investors also need to monitor changes in their actual portfolio allocations brought about by the relative market performance of various asset classes. Over time, the portfolio will tend to become more concentrated in asset classes that have recently done well. Typically, this implies an increase in risk and necessitates rebalancing the portfolio to re-establish the desired risk/return characteristics.

T HE A SSET A LLOCATION P ROCESS When building asset allocations, we begin by grouping securities that share certain characteristics into asset classes, such as large-cap stocks, long-term bonds, etc. For each asset class we make Capital Market Assumptions, which include estimates of expected returns, risk, and correlation among the various asset classes. This data is then processed with information gathered in the client profile (using our Mean Variance Optimizer program) to yield the investor s optimal asset allocation. The portfolio resulting from this process contains the mix of investments with the best chance of yielding the client s desired return at the least amount of risk. Asset Classes Capital Market Assumptions Client Profile Objectives Investment preferences Investment amount Risk tolerance Time horizon 7 Mean Variance Optimizer LOWER TO HIGHER REWARD Investor s Optimal Allocation LOWER TO HIGHER RISK

T HE DIFFERENCE BETWEEN STRATEGIC AND T ACTICAL ASSET ALLOCATION 8 Asset allocation approaches can be characterized as either strategic or tactical. Strategic asset allocation (SAA) is a long-term investment discipline designed to achieve stated goals. It explicitly takes into account an investor s risk tolerance and investing preferences in an attempt to identify an optimal mix of investments. Tactical asset allocation (TAA), on the other hand, covers shorter time horizons, usually a month or a quarter, with the goal of exploiting changes in near-term market conditions using timely and accurate forecasts. In short, it is a market timing strategy. TAA is frequently used as a strategy of over- or underweighting allocations relative to strategic asset allocation policy. These temporary departures from a long-term allocation policy are often determined using sophisticated and proprietary quantitative techniques. A TAA recommendation would normally be implemented in conjunction with a particular strategic asset allocation portfolio. If, for example, SAA establishes an optimal portfolio of 50% stocks, 40% bonds, and 10% Treasury bills, a TAA recommendation for the next month or quarter might be to modify this portfolio to 55%/40%/5%. Some investors implement TAA as if it were just another asset class or an investment management style. In these instances, the TAA strategy simply receives some portion of the total portfolio allocation. B ENEFITS OF ASSET ALLOCATION AS A STRATEGY In the uncertain world of investing, asset allocation remains the best approach to helping investors achieve their long-term financial goals since it offers the following benefits: A disciplined way to invest. Investors are not subject to the whims of the market. Instead they follow their asset allocation policy recommendations until a new policy is warranted, whether due to a change in risk appetite or because of the dynamic nature of the asset allocation strategy itself. Risk reduction through diversification. Asset allocation exploits the interrelationships among different asset classes to provide portfolios that offer the minimum level of risk for the selected level of expected return. Portfolio risk matched with investor risk appetite. Using the efficient frontier concept, asset allocations can be constructed for specific risk levels while accounting for investment goals and investing preferences. From a practical standpoint, asset allocation offers unique solutions for unique needs rather than promoting a one size fits all solution.

Securities products and services are distributed by Prudential Investment Management Services LLC (PIMS), Three Gateway Center, 14th Floor, Newark, NJ 07102-4077. PIMS is a member SIPC and Prudential Financial company. Prudential Financial is a service mark of The Prudential Insurance Company of America, Newark, NJ, and its affiliates. Investments are not insured by the FDIC or any federal government agency, are not a deposit of or guaranteed by any bank or any bank affiliate, and may lose value. IFS-A080127 SIRG109 Ed. 05/01/2003