Search

ortfolio Management activities are the basic tools of allocating resources in the financial markets of the free market economies. Institutional investors in the developed countries perform this (financial) resource allocation. But in countries that do not have established institutional investors, individuals perform this resource allocation, usually in an ad hoc way. Turkey is one of these countries. In addition, Turkey has huge deficits in its social security system, which makes things worse. These factors require a more scientific and disciplined approach to portfolio management so as to enable more efficient resource allocation. Recent government plans for establishing pension funds in Turkey is an attempt to increase the efficiency in resource allocation and this proposal increased the importance of portfolio management. This study will focus on the most important aspect of portfolio management process: asset allocation. We aim to give theoretical information on dynamic asset allocation strategies, their place in portfolio management process and apply some trading rules based on these strategies using Turkish data. 1. PORTFOLIO MANAGEMENT PROCESS AND ASSET ALLOCATION 1.1. Risks In Financial Assets Risk, in the financial sense, can be defined as the uncertainty in getting the expected return in the financial assets. A more universal and analytical definition is as follows: risk is the standard deviation of returns of the financial assets. Financial risk can be explained in two broad categories. First category is the systematic risk that arises out of external factors rather than firm-specific ones. Some examples of these common factors areinflation, unemployment, political shocks etc. Systematic risk further can be categorised into market risk and interest rate risk. Unsystematic risk is the risk that depends on firm-specific factors such as management quality, consumer preferences and so on. Business risk, accounting risk, liquidity risk are sub-categories of unsystematic risk. 1.2. Passive Management - Dynamic Management A distinction between passive management and dynamic management is often made within the investment industry. Passive management requires for holding securities for relatively long periods.1 Implicit in this strategy is the assumption that the capital markets are efficient. Their decisions are in line with the acceptance of consensus market estimates of risk and return. Passive portfolio managers do not try to beat the market. They hold, so-called, market portfolio in the form of index funds or a portfolio, which is tailored according to the needs of the individual investors, and hold this portfolio for long periods. Dynamic (active) managers, on the other hand, start with the assumption that there are mispriced securities and/or groups of securities from time to time. They use these deviations from the equilibrium prices as trading opportunities and try to make a profit out of it. Above discussion requires us to give a more detailed explanation of market efficiency because philosophy underlying portfolio manager's strategy is largely affected with the efficiency assumption. Fama2 describes an efficient market as below: In an efficient capital market, the market actors digest information into the security prices in a fast and efficient way. So, security prices reflect all available information at all times. 1 Sharpe, W., Gordon J.Alexander ve David LFowler (1993); Investments, First Canadian Edition; Prentice Hall Canada Inc.; s.704 > 2 Fama, Eugene F. (1972); `Components of Investment Performance`; The Journal of Finance; 27, 3, pp.551-567Fama explains market efficiency in three categories. Following table gives a short summary of these categories. Most parts of the capital market in the US is weak-form efficient.3 We obtained a similar result for Istanbul Stock Exchange (ISE) by testing if there is any relationship between the current and past index values. This is achieved by constructing a model like this: ri,t=a + Pri,t-i+ei,t In the above expression r shows the return for period t, s shows the error term and a and P are the regression coefficients. We performed this regression for the period between December 1998 - January 2000 using monthly returns and one-month lag. Resulting regression equation is given below: rIt = 0.069 + 0.096rIt_! 1.3. Dynamic Portfolio Management Process Portfolio management process for a securities portfolio can be divided into three phases. In the first phase, portfolio manager must determine investment objectives and policy. Investment objectives should be stated in terms of both risk and return. Therefore, factors like risk tolerance of the investor along with sector constraints and transaction costs are important. This step concludes with the determination of potential financial asset categories for consideration. Second step is the portfolio construction and optimization. This phase involves determinnig the asset classes to invest and proportions of each class in the portfolio. Expected returns, volatilities and; correlations 3 Sharpe, W., Gordon J.Alexander ve David J.Fowler (1993); Investments, First Canadian Edition; Prentice Hall Canada Inc.; s.72of the asset classes are analysed and optimum portfolio composition is derived with risk tolerance, expected return constraints of the investor. Optimization is performed using a method like Markowitz mean-variance optimization, dynamic programming or more simple methods. This will depend on sophistication of the portfolio manager as well as time and budget limitations. Most well-known of these optimization methods is the Markowitz mean-variance optimization which tries to find the optimum investment level in the portfolios. Markowitz states that a portfolio is mean-variance efficient if it has the highest return for a given level of risk or minimum risk for a given level of return.4 Markowitz optimization tries to find the dominant portfolio in terms of minimum risk - highest return criteria. The third, and last, step of the process is performance evaluation. The factors considered during this stage include the level of return obtained for the assumed risk, return comparisons with benchmark portfolios. Portfolio composition may be rebalanced according to the findings of the performance evaluation stage. Performance evaluation also plays a crucial role in determination of the compensation packages of the portfolio managers. Sharpe Ratio, Treynor Ratio and Jensen Alpha are the basic methods for portfolio performance evaluation. We compare values of these measures for the investment portfolio with those of benchmark portfolios. These performance measures are based on Capital Asset Pricing Model and they use the same assumptions as CAPM. Validity of these assumptions are largely questioned by several authors5. Therefore, use of basic performance measures requires careful treatment and attention to the validity of assumptions for the market in question. Measurement of market timing ability is another aspect of performance evaluation. Market timing is a conscious startegy used by portfolio managers which changes the portfolio composition in accordance with the return expectations of portfolio managers. These expectations, in turn, are determined by market 4 Markowitz, Harry M. (1959); Portfolio Selection`.; Yale University Press; New Haven, Connecticut 5 Admati, A.R. ve S.A.Ross (1985); `Measuring Investment Performance in a Rational Expectations Equilibrium Model`; Journal of Business; 58:1: s.1-26conditions. Portfolio managers forecast the expected returns for the related asset classes and adjust the portfolio composition with giving more weights to assets which are expected to outperform the other asset classes. Market timing ability can be measured using techniques like quadratic regression model and Henriksson-Merton Regression Model. Studies of market timing generally suggests that portfolio managers have negative6 or no timing skill at all.7 Regressions used for detecting market timing ability are also prone to statistical errors. We correct these errors using Bayesian adjustment, accounting for heterocedasticity and autocorrelations. In addition, expected betas may be calculated instead of historical betas to prevent chance events distorting the relationship between portfolio and benchmark index. 1.4. Asset Allocation and Its Place in Portfolio Management Process Asset allocation is the distribution of available funds among `main` asset classes.8 Asset classes include broad categories such as stocks, bonds and cash. Volatility of returns in the investment portfolios largely arises out of asset allocation activities. Asset allocation decision might be a strategic one which has a long investment horizon as well as a tactical one which tries to take advantage of mispricing in the financial markets and represents deviations from the long-term strategy. `Dynamic` Asset Allocation aims for changing portfolio composition when a change occurs in macroeconomic factors, expected returns and correlations of financial asset classes and risk tolerance of the investor. Therefore, techniques such as buy-and-hold and indexing do not belong to dynamic asset allocation strategy classes. 6 Henriksson, R.D. ve R.C.Merton (1981); `On Market Timing and Investment Performance. Il.Statistical Procedures for Evaluating Forecasting Skills`; Journal of Business; 59:2; s.217-235. ` 7 Treynor, J.L. ve K.K. Mazuy (1966); `Can Mutual Funds Outguess the Market?`; Harvard Business Reviews AA,A ; s. 131-136 V; 8 Sharpe, William F. (Winter: 1992); `Asset Allocation : Management Style and Performance Measurement`; Journal Of Portfolio Management; s. 7-19Asset allocation decision is the most important decision for a portfolio manager. There are studies showing that 80% of the volatility of returns in the investment portfolios occur as a result of asset allocation decisions.9 Financial markets include different asset classes and pricing of these financial assets are made through consensus judgements of market participants. Asset allocation decision is made using risk tolerance, investment horizon and return expectations of the investor and it is a systematic attempt to expose the return expectations of market participants. `Dynamic` Asset Allocation strategy may be;. strategic, which is based on long-term equilibrium relationships,. insured, which is based on risk tolerance of the investor,. or tactical, which attempts to make use of deviations from long term equilibrium relationships. 2. STRATEGIC ASSET ALLOCATION Strategic asset allocation aims to get the optimum return using long term asset composition decisions in a portfolio.10 Strategic asset allocation studies simulate long term expected returns and constructs the portfolio by taking risk tolerance of the investors into consideration. Long term capital market conditions, expected returns, risks and correlations are held constant during the analysis period. Strategic asset allocation strategies have long term investment horizons. Models change only when there is a lasting change in the assets and liabilities of the investors or in capital market conditions. This long term orientation makes it necessary to study the long term historical expected returns in the financial markets. We calculated the long term historical returns and volatilities for stocks, bonds and USD/TL exchange rate between the period January 1996 - September 1999 using MKB-100 index and State Domestic Debt Index returns as proxies for stocks and bonds. Results show that stocks have the highest return and the highest 9 Hammer, David A. (1991); Dynamic Asset Allocation; First edition; John Wileyand Sons Inc.; New York; s.62 10 Sharpe, William F. (September-Ocober 1987); `Integrated Asset Allocation`; Financial Analysts Journal; pp.25-32volatility. One interesting finding is that volatility of stock returns is higher than expected return. This is opposite of what we observe in the developed financial markets. Strategic Asset Allocation assumes that financial markets are efficient. There is no attempt to `time` the market to exploit market anomalies. Objective function and risk tolerance of the investor is determined rationally for the long run and model is rebalanced only occasionally. Another advantage of Strategic Asset Allocation is its ease of explanation to the top management. Complex market timing strategies usually deter the performance of the investment portfolios11. That's why Strategic Asset Allocation has the potential to add value in the long run. 3. TACTICAL ASSET ALLOCATION Tactical Asset Allocation (TAA) includes those trading techniques, which aim to exploit market inefficiencies and mispricing in the financial markets.12 Investment horizon of TAA strategies is much shorter than Strategic Asset Allocation (SAA) strategies. Despite their widespread use, TAA does not have a common definition. In general, every strategy that changes the portfolio composition when there is a change in the expected return of individual asset classes can be categorised as tactical. Forming the return expectations is the critical part of TAA and this activity is performed in a disciplined manner. TAA strategies can also be viewed as techniques to adjust the risk and return profiles of long term SAA schemes. Portfolio managers try to forecast deviations from the long run expected returns and they exploit these deviations by buying cheap and selling expensive. TAA assumes that expected returns of the asset classes may be observed from financial markets directly. These returns represent the consensus view of the market participants on relative attraction of the 11 Arnott, Robert D.ve Frank J.Fabozzi (1992); Active Asset Allocation; Irwin Professional Publishing; Chicago; Illinois; s.51 12 Sharpe, William F. (September-Ocober 1987); `Integrated Asset Allocation`; Financial Analysts Journal; pp. 25-32asset classes. Expected returns of the asset classes are also viewed as the long term equilibrium returns. If future returns deviate from this long term equilibrium returns, financial markets bring them back to their equilibrium level.13 TAA strategies use the equilibrium returns observed in the markets as inputs in their analysis. By using some objective model, TAA attempts to determine future expected returns. If model returns are different than equilibrium returns then TAA model exploits that deviation. This usually requires doing the opposite of what the market is doing, i.e. buying when the market is going down and selling when the market is going up. In this sense, TAA strategies are contrarian in nature. One may question the use of TAA rules in order to increase portfolio returns without increasing the portfolio variance on grounds of market efficiency. The answer to this question is provided by the utility theory. When markets are going up portfolio assets increase in value. But this is achieved at the cost of decreasing returns during the following periods. Users of SAA do not respond to the market ups and downs because their risk tolerance stays the same. On the other hand, TAA investors see a market rally as a selling opportunity because they expect to buy the assets back when the returns decrease in the following periods. Clearly, this strategy calls for a high risk tolerance. TAA may improve the portfolio returns in the long run. But increasing returns is not the same thing as increasing the `utility`. Because of its contrarian nature, application of a tactical strategy requires very strong nerves. Psychologically, most investors would want to decrease their risks when the market goes down and tactical strategies advise exact opposite of this will. Optimum TAA decision can be made in a multi-period or one-period setting. Realistic analysis of optimum portfolio choices can only be made in a multi period setting because expected returns, risks and variances change with time. But multi-period analysis is too costly in terms of time and money. It uses complex analytical models and very high computing power. In today's global world, there are so many financial assets in many countries. Simulating all returns in different countries for several periods into future 13 Arnott, Robert D. ve Roy D.Henriksson (May/June 1989); `A Disciplined Approach To Global Asset Allocation`;quickly becomes a nightmare. Therefore TAA managers prefer using one-period (myopic) analysis in determining optimum portfolio composition. Actually multi-period and myopic analyses give the same result if one assumes that investors are not involved with intertemporal hedging. But avoidance of intertemporal hedging is possible if following conditions are met14. Non-stochastic Investment Opportunity Set : If parameters of the expected return distributions of financial assets in the investment opportunity set stay constant over the periods then there is no need for intertemporal hedging.. Logarithmic Utility Function : If the optimum portfolio is constructed by using logarithmic utility function then investment opportunity set for this portfolio is constant over the periods. Most important feature of the TAA process is the forecasting of asset class returns in the investment horizon. TAA managers invest in those classes, which are expected to outperform equilibrium returns. In a way, TAA strategies try to `time` the market. Investor's relative risk tolerance is constant in the investment horizon and is not affected with the changes in the value of assets and liabilities of investor. Portfolio composition is rebalanced as the expected returns of the asset classes change. Some sophisticated TAA programs also consider the risks and correlations of asset classes in portfolio construction and rebalancing. TAA (Market timing) strategies have their own risks. First of all, tactical strategies may not work after some time. It is easily conceivable that one market timing strategy will be obsolete after other investors in the market are aware of that particular strategy because contrarian strategies cannot be employed if the majority of the market attempts to do the opposite of what the market is doing. Second risk is the psychological pressure to give up when the portfolio assets decrease in the implementation of the strategy. Financial Analysts Journal 14 Merton, Robert C. (1969); `Lifetime Portfolio Selection Under Uncertainty:The Continuous-Time Case`; Review of Economics and Statistics; s.247-25710 The critical step in tactical management is the construction of the objective model, which is used for forecasting expected returns. If portfolio manager makes a mistake in this phase TAA program will give false signals. Another danger is the fact that model will give logical results based on objective inputs but markets will behave illogically and mostly based on psychological factors. Judging the direction of behaviour of the market participants may be more important than return forecasts. Return forecast models usually use historical data to form expectations. Tactical managers should also be aware of the fact that these historical relationships may not be valid in the future. In this case, model results will not give correct results. On the other hand, if the models are too responsive to the market conditions then TAA performance will be spoiled by trend reversals. Especially, markets with sideway pattern and high volatility will cause large losses. Quality of model inputs and validity of assumptions are vital considerations in tactical models. 4. PORTFOLIO INSURANCE Portfolio insurance is an investment management strategy which is inspired by the option pricing theory. Portfolio insurance procedures aim to improve the fit between long term performance of the portfolio and investor's objectives.15 They try to achieve this without timing the market. Early forms of portfolio insurance involved dynamic changes in asset allocation to change the composition of the portfolio in order to replicate certain option positions. For example, one can replicate the outcomes obtained by holding a portfolio plus a one-year put option written on that portfolio. An approach of this type is called option-based portfolio insurance. Option-based portfolio insurance strategies usually allocate assets in two major classes (such as stocks and t-bills.) They provide a trading rule that relates the appropriate asset mix to the excess of the investor's net worth over a desired `floor`. The difference between asset value and floor is named as Sharpe, William F. (September-Ocober 1987); `Integrated Asset Allocation`; Financial Analysts Journal; pp. 25-3211 `cushion`. In the option-based portfiolio insurance procedurethe floor at any given time is the present value of the desired floor at the horizon date. Option-based portfolio insurance strategies are time-variant. They require the amount invested in the risky asset to be a function of both the current value of the asset cushion and the time remaining until the horizon. In this context, horizon is the option's expiry date. As the horizon approaches, extreme changes in the asset mix are necessary. Just before the horizon, assets are invested entirely in one class or the other. This allocation will depend on whether assets are equal to the floor or above it. Many portfolio managers objected to this time-variant nature of option-based portfolio insurance.16 This feature of option-based insurance was inappropriate for institutional investors with very long investment horizons such as pension funds. A more recent approach to portfolio insurance is time-invariant. The curve relating the amount invested in the risky asset to the size of the cushion remains constant from period to period. This approach is called Constant Proportion Portfolio Insurance (CPPI).17 In CPPI, amount invested in the risky asset is a constant times the size of the cushion as shown below: Amount invested in the risky asset : m (Assets - Floor) In the above formula m is a multiplier which is greater than one. Portfolio insurance strategies assume that expected returns, risks and correlations among the asset classes remains unchanged during the investment period. Use of portfolio insurance strategies largely depends on the risk tolerance of the investor. Leland describes the users of portfolio insurance strategies as follows18:. Portfolio insurance can be used by those users with average return expectations and whose risk tolerance increases faster than average investor as the portfolio assets increase.Classical example of this type of 16 Ibid. 17 Black F. and R. Jones (Fall 1987); `Simplifying Portfolio Insurance`; The Journal of Portfolio Management; pp.48-5112 investor is pension funds. Pension funds usually aim to set a floor level that is equal to the present value of their future liabilities. Remaining funds over the floor is invested in risky assets.. Investors with average risk tolerance and more than optimistic expectations about asset returns will want to use portfolio insurance. Institutional investors with well-diversified portfolios and who want to improve returns with superior security selection are good candidates for this use. Portfolio insurance promises to replicate option positions by dynamic trading. Most practitioners have been puzzled by the complexity of the process.19 Underlying option-pricing theory involves using sophisticated models. Besides this feature, the strategy also demands buying more stock when the market is going up and selling stock when the market is going down. This nature of the strategy is contrary to the intiution of many portfolio managers. 5. APPLICATIONS OF ASSET ALLOCATION STRATEGIES 5.1. Strategic Asset Allocation : Constant Mix Strategy In this section we perform an ex-post simulation to show how a strategic asset allocation strategy like Constant Mix works in Turkish financial market setting. We selected IMKB stock index as the risky asset and ISE G-Bond Performance Index as the riskless asset. Using data from December 1995 to November 1999 we run the simulation by holding a constant stock/bond proportion of 60/40. We rebalanced the portfolio every month to restore the initial 60/40 proportion. Choice of rebalancing period, initial mix and asset classes are arbitrary. Optimization of these parameters are possible but outside the scope of this study. Constant mix strategy resulted in buying the asset when its price falls and selling the asset when its price rises. The strategy is contrarian in nature and payoff curve (value of risky assets in the portfolio charted against value of the risky asset) is concave. During the simulation period we observed 3 bull and 2 bear 18 Leland, Hayne E. (1980) ; `Who Should Buy Portfolio Insurance?`, The Journal of Finance`; 32, 2, pp.581-59613 market periods. Rapid falls in the stock market value is not passed directly to the simulation portfolio because riskless asset provided some protection. Volatility of the markets is the most important factor to watch in this strategy. Because of its nature, Constant Mix is a bet on market reversals. So, it performs best when there is a sideways pattern with frequent reversals in trend. Constant Mix is not likely to work well in long bull or bear markets. This result may be generalised for all contrarian strategies. 5.2. Tactical Asset Allocation : Moving Average Rule In this part, we defined a variation of the famous moving average rule and simulated portfolio results for the period 31.01.1991 to 31.01.2000. We selected IMKB-100 Stock Index (risky asset) and US Dollar (again risky, used as cash instrument) as two general asset classes to include in the portfolio. Initial mix of the portfolio is again 60/40 for stock market and US dollar respectively. We defined the moving average rule as below : 1 (Buy) (0.05*PV), if (r, - MA(12)) > %10 0 (Wait), if % 10 > (r, - MA(12)) > -% 10 -1 (Sell) (0.05*PV), if (rt - MA(12)) < - %10 In the above expressions; PV = Portfolio value MA(12) = 12-ömonth moving average of IMKB-100 stock index rt = IMKB-100 index return in month t. So, we use 10% difference between stock indek return and 12-month MA as a buy or sell signal. We do nothing for the values in between. We buy or sell risky asset amounting to 5% of the total portfolip value 19 O'Brien, Thomas (Spring 1988); `The Mechanics of Portfolio Insurance`;7%e Journal of Portfolio Management, pp.40-4714 when a buy or sell signal is produced. In short a 10% signaling percentage and a 5% trading percentage is used. Choice of these numbers will depend on transaction costs and risk tolerance of the investor and should be optimised. We chose arbitrary numbers because we want to show the workings of the rule in Turkish markets in order to get a general idea of the more general tactical asset allocation strategies. Our simulation resulted in a final portfolio value of 29,806 TL for 100 TL initial portfolio value. This is a better result when compared to a buy-and-hold strategy with the same initial mix. But this particular kind of MA strategy performed poorly compared to a 60/40 constant mix strategy. This means, MA rule had a negative timing performance. We calculate the loss from timing activities and found 2,342 TL which is 23 times the initial portfolio value. Tactical strategies try to `time` the market to enhance portfolio return without increasing portfolio volatility. Tactical asset allocation rules are countless and very difficult to describe in terms of payoff profiles. But we usually expect to see a contrarian strategy with concave payoff profile. Our simulation resulted in a similar payoff profile. Consequently, they buy assets as their price go up and sell assets as their price go down. They perform well in markets with frequent reversals and with trend. 5.3. Portfolio Insurance : Constant Proportion Portfolio Insurance (CPPI) strategy In this part we used CPPI rule to get an idea as to the workings of portfolio insurance strategies in general. We used two asset classes : IMKB-100 Stock Index (risky asset) and ISE G-Bond Performance Index (Riskless Asset). Using data for the period December 1995- November 1999 we simulated the rule according to the following definition : Proportion invested in stocks = m (Total assets - Floor) Or; Proportion invested in stocks = m (Cushion) In the above expression `m` is a multiplier which is greater than one.15 If the multiplier is high enough it is possible for the rule to call for an investment in stocks which is higher than the total portfolio value. In this case one can assume that the extra amount will be borrowed at the risk-free rate or this leverage is not allowed. In our first simulation we did not allow leverage. Simulation results showed that low floor levels and high multipliers result in higher final portfolio values. In the second simulation leverage was allowed and the best results are obtained using smaller multipliers. This shows that increasing riskiness of the portfolio through leverage is not a good option for volatile stock markets like Istanbul Stock Exchange. Therefore, if leverage is possible then portfolio manager should choose a smaller multiplier. We also examined direct and indirect costs of the CPPI strategy. In order to calculate direct costs we assumed a 0,5 % transaction fee for stock trades. In this case, final portfolio values go down further. As the multiplier increases and floor decreases stock trades tend to increase and transaction costs rise accordingly. We found that transaction costs can be as high as 5,54 % of the initial portfolio value for the 2,5 year period between January 1996 to July 1998. If the multiplier-floor combination goes to extreme values strategy becomes a static buy stock-and-hold strategy and direct transaction costs decrease. In order to get an idea about the indirect costs of portfolio insurance we calculated `upside participation ratio`. This is the ratio of the portfolio value to the benchmark portfolio value, which is the buy- abd-hold portfolio of the highest yielding asset class included in the portfolio. As expected, indirect costs are higher for high floor-low multiplier combinations. This is similar to buying insurance and foregone upside potential is the price paid for the insurance. One last observation is that indirect costs are more sensitive to floor value rather than multiplier. 6. CONCLUSION We believe Portfolio Management activities will gain increasing importance in Turkey in the near future. Asset allocation activity is the most important part of the portfolio management process and we tried16 to emphasize this importance in this study in a Turkish context. In a developing financial market markets are more volatile and the need for sound portfolio management activity is necessary. Asset allocation strategies mentioned in the study should also be a useful guide to insurance companies and pension funds which are likely to emerge or be popular in Turkey. Legislation is planned for setting up private retirement funds (pension funds) in our country and management of the pension portfolio will require sound asset allocation decisions and portfolio management practices rather than random and emotional investment decisions. In managing the investment portfolio, prime activity to concentrate should be asset allocation. We covered active (dynamic) asset allocation strategies in our study and disrgarded passive techniques like buy-and-hold and indexing. It emerged that each strategy perform well in different market settings. We summarize these results below:. Constant Mix Strategy (which is an example of Strategic Asset Allocation) performs well in volatile markets with sideways pattern and frequent marekt reversals.. Moving Average strategy which is used as a tactical strategy performs best in markets with market reversals and with trend.. Constant Proportion Portfolio Insurance strategy peforms best in rising markets. In the above expressions `market` usually refers to a risky asset class rather than riskless asset. They show that describing the right market conditions and forecasting trend and volatility of the asset classes are activities of prime importance. Turkey is a country with volatile financial markets. This imposes higher risk but also provides higher profit opportunities. If the portfolio manager constructs the portfolio in the above manner and gets the right composition, he is likely to get a competitive advantage over his rivals. Another point we would like to make is the fact that financial markets are not so complete and efficient in Turkey compared to developed countries. This fact points to opportunities in following tactical17 strategies to enhance portfolio returns. Tactical strategies may be difficult to employ in UK or US but not so difficult in a country like Turkey which does not have sophisticated financial markets. We believe portfolio management shoul get increasing attention from the general public as well as from investment professionals in Turkey. This requires that studies in this field must increase to construct and develop the necessary knowledge base. This study is a small attempt in this purpose. 128