Where Smart Money Goes.®

 Asset Allocation and Portfolio Performance

Studies have been performed to identify which factors contribute the most to a portfolio's overall performance.  The studies analyzed the asset allocation, individual security selection and market timing decisions.  Results from a study of 91 large U.S. pension plans from 1974-1983 indicated that asset allocation (investment policy) was the dominant factor in determining the portfolios performance [1].  The results are summarized in Figure 1 which shows the contributions of total portfolio return were asset allocation (94%), individual security selection (4%), and market timing (2%).

Asset allocation is the process of selecting and dividing one's wealth among broad asset classes or categories.  In the past, only three categories were considered (cash, equities and bonds).  Today a fully diversified portfolio can include equities, fixed income instruments, international equities, real estate, precious metals, and cash equivalents.  Asset allocation involves analyzing the risk return characteristics of each asset class rather than the individual securities that comprise each asset class.  This is an important distinction and lends credibility to the use of passive rather than active strategies. 

There are four steps used to designing a portfolio [3].  The first step is choosing which asset class will be included in the portfolio.  Second, the long-term percentages to allocated to each class must be determined.  Third, one must identify the ranges that each class can be modified to achieve an optimal mix.  Lastly, the securities from each asset class must be selected.  Traditionally, steps 1 and 2 are known as the investment policy or asset allocation, step 3 as market timing, and step 4 as security selection.

All asset classes should be initially considered for the portfolio and then individual class can be eliminated if sound reasoning prevails. All classes must be analyzed to determine their impact on the overall portfolio since changes in economic conditions can affect asset classes in different directions and magnitudes.

Modern portfolio theory states that an investor would choose the market portfolio of risky assets to combine with risk free asset.  Thus, a good starting point for long-term asset class percentages would be the percentage of world wealth in the overall market.  However, the asset allocation must be a good fit for the investors.  Every allocation is slightly different depending on each investor’s investment objective, total assets, cash flow needs, tax situation, risk tolerance, and time horizon.  The goal is to determine the best asset class allocation that will meet the investment objective at the lowest risk.

The third step involves setting a range at which the investor is willing to adjust each asset class percentage.  These ranges can be used to increase or decrease allocation to the asset class given short-term expectancies.  A word of caution: this procedure involves timing the market, something which empirical evidence has shown cannot be done on a consistent basis as a means to achieve superior returns.  As such, this step is best used to adjust portfolio allocations based on individual investor portfolio preferences rather than as a means of trying to time the market.

The fourth step involves security selection.  As noted prior, security selection only contributes 4% to the overall portfolio performance.  One of the important concerns with individual security selection for an investor employing a passive investing strategy is to keep transaction costs and management fees as low as possible.  This is the reason that Dimensional Fund Advisors (DFA) funds are employed by Karmikel Investments.

DFA offers mutual funds designed to capture the behavior of an entire asset class, index funds engineered to capture specific dimensions of worldwide returns, and fixed income strategies with a variable annuity structure.  In addition, DFA maintains very low management fees and has even achieved negative trading costs over an extended period in the U.S. small company stocks.  Utilizing DFA’s broad-based funds and low cost management style allows investors to achieve broad global diversification at minimal trading and management fee costs.  For more information on DFA, see DFA funds elsewhere on this site.

Why do so many investors subscribe to active investment strategies and overemphasize security selection or market timing?  Part of the answer is tradition.  Money management was founded on the belief that one could beat the market and achieve superior returns by stock selection and market timing.  In fact, many managers believe that without superior returns, money managers would not exist.  In reality, time and again, research concludes that managers under-perform the market and those who do outperform the market in one period have an equally likely chance of under-performing in the next period.

Other reasons investors incorrectly choose active management is simply the seductive appeal.  It is the fun of it, the challenge of it, or the hope of "beating the other guy."  While particular money managers may beat an index three years in a row, studies repeatedly reveal that it is merely due to random chance and not superior stock picking or market timing.  To illustrate these odds, you may think of a coin toss: if you flip a coin three times, your chances of getting heads three times in a row is 12.5%.

Let's look at efficient portfolios.  To illustrate how efficient portfolios are created we will look at an example comprising three main asset classes: T-bills, stocks, and bonds [4].  For graphical presentation, the efficient frontier is plotted in mean-standard deviation region.  First, we will show the individual asset classes, then the two-asset efficient frontier, and finally the three asset efficient frontier.

First, we need some statistics to plot the efficient frontiers.  Table 1.1 presents hypothetical statistics relating to three asset classes.  For a more rigorous explanation of the detailed calculations involved in creating Table 1.1, see the modern portfolio theory section elsewhere on this website. 

Table 1.1

 

 

 

 

 

Mean

Variance

Std. Dev.

Stocks

11.06

103.34

10.17

Bonds

8.79

14.65

3.83

T-bills

6.98

2.17

1.47

 

 

 

 

 

Correlation

 

 

Stocks

Bonds

T-bills

Stocks

1

.42

-.39

Bonds

.42

1

-.31

T-bills

-.39

-.31

1

 

 

Covariance

 

 

Stocks

Bonds

T-bills

Stocks

103.34

16.184

-5.798

Bonds

16.184

14.65

-1.736

T-bills

-5.798

-1.736

2.17

 

 

 

 

Plotting the statistics from Table 1.1 creates the two-asset efficient frontiers displayed in Figure 2 (below).  The individual stock, bond, and T-bill securities are represented by a dot and are labeled accordingly.  The lower arc represents the full range of combinations of stocks and T-bills and the upper twin humped line represents the stock/bond efficient frontier.  Notice that the two-asset portfolio dominates the single-asset portfolio in several places.  

  Figure 3 shows the three-asset portfolio frontier.  The 100% T-bill and the 100% stock positions do not change.  The only way to adjust them would be short selling.  The segments close to the 100% T-Bill and 100% stocks do not change much.  As you get to the center of the segment, the three-asset portfolio dominates the two-asset portfolio the most.

This is a simplified example but it none-the-less provides illustration of diversification and what an efficient portfolio means.  As the possibility of securities expands and the number of assets included in the portfolio grows the computations become more rigorous.  However, the concepts are exactly the same and an efficient frontier can still be calculated and diagramed.

Table 1.2 below shows actual statistics for range of risk tolerance portfolios from 1973 through 2001 [2].  Notice that the standard deviations and expected returns increase from left to right as the portfolios become more aggressive.  Also notice that the highest and lowest annual returns can and do lie outside of the one standard deviation range of the annualized return.  This clearly shows that while the standard deviation is a good measure of portfolio risk, it is based on regression analysis, which is an average of the deviations from the average return, not an absolute range for each of the annual portfolio returns.  In other words, actual returns can and will be less than and greater than the standard deviation.  However, over a longer period of time, the average risk will trend according to the standard deviation.  This table shows, in general, how different levels of aggressiveness affect a portfolio's risk and return characteristics.

Table 1.2  

 

Fixed

Conservative

Moderate

Normal

Aggressive

Equity

Equity

0%

20%

40%

60%

80%

100%

Fixed Income

100%

80%

60%

40%

20%

0%

1973-2001Statistics

 

 

 

 

 

 

Annualized Return

8.9

10.3

11.6

12.8

14.0

15.1

Annualized Std. Dev.

2.5

3.8

6.2

8.9

11.7

14.4

Lowest Annual Return

-1.1

-1.4

-9.4

-18.0

-26.0

-33.4

Highest Annual Return

24.2

28.5

35.9

43.7

51.7

60.4

Growth of $1

$11.83

$17.08

$24.11

$33.27

$44.72

$58.84

Figure 4 shows the portfolios from Table 1.2 in the mean variance standard deviation region.  The purpose of Table 1.2 and Figure 3 is to give an example of general portfolio characteristics for common risk tolerances.  The equity and fixed income categories can be broken down into subcategories.  Equities are broken down into U.S. and international equities.  U.S. equities can be further broken down into different capitalization stocks (small, medium, and large) and value stocks.  International stocks include different capitalization (small, medium, and large) and emerging market stocks.  Fixed income categories include subcategories of different investment grades and maturity dates.  The idea is to allocate assets broadly among all asset classes and at weights that create an efficient portfolio.

To summarize, we have illustrated that asset allocation is the single most important determinate of portfolio performance.  Active managers attempt to use superior stock selection and market timing to beat the market but a steady stream of empirical research shows that these managers only conform to statistical chance.  We presented an example of a two-asset and a three asset efficient frontier.  We also presented a general look at the risk return characteristics of broadly diversified portfolios at various risk tolerances.

Works Cited

1. Brinson, Gary P., L. Randolph Hood, Gilbert L. Beebower, “Determinants of Portfolio Performance”, Financial Analyst Journal, July-August 1986, pp. 39-44.

2. Dimensional Fund Advisors data

3. Gibson, Roger C., Asset Allocation: Balancing Financial Risk, Homewood, Dow Jones-Irwin, Inc., 1990.

4. Rudd, Henry and Henry Clasing, Jr.  Modern Portfolio Theory: The Principles of Investment Management. Orinda: Andrew Rudd, 1988.

Copyright © 2008 Karmikel Investments - ALL RIGHTS RESERVED

 

Home

My Account

Investment Philosophy

Modern Portfolio Theory

Active vs. Structured Management

Portfolio Performance

Tax Management

Dimensional Funds

Consolidated Custom Reporting

Service Provider Coordination

Risk Management

Wealth Transfer, Legacy Planning & Charitable Giving

Frequently Asked Questions

Services Provided

Fee Structure

Contact Us

Links

In The News