Asset Allocation Update - Part I

Over the next two months, we will present a review of asset allocation issues. A common mistake we have seen others make is to focus too much attention too early in the process on different optimization techniques and the model portfolios they produce. To avoid that, we will defer our discussion of these subjects to next month, and instead start with some equally important issues that often don't get the attention they deserve.

What Drives Portfolio Returns?

Let's start with a list of the factors that determine the return you earn on your investment portfolio. They include the following:

• The asset classes in which you will consider investing your money, and the way your define them.
  
  
• The percentage of your total portfolio that you invest in these asset classes.
  
  
• The extent to which, and the circumstances that will trigger departures from these basic weights.
  
  
• Whether you will implement this strategy using active investment managers or index funds.
  
  
• The specific funds you will use to implement your strategy (e.g., based on their costs and tax effectiveness, and, if actively managed, your perception of the manager's relative skill).
  
  
• How often you will rebalance your portfolio to its target asset class weights. Along with your answers to questions 1, 2, and 3, your answer to this question is critical to your ability to control the riskiness of your portfolio (because less frequent rebalancing runs the risk of being more exposed to adverse events than you planned when you established your basic allocation strategy)

While all of these questions are critical, this review will only focus on the first two questions, which together constitute the basic "asset allocation" challenge.

Asked to define the meaning of "asset allocation", most people would give answer similar to question two, and say that it has something to do with the way you divide your investments between different groups of similar assets. Unfortunately, this overlooks question one, which is arguably far more important in terms of its impact on the returns you earn and the risks you take. Here's an example of what we mean: Ask five of your friends to identify the different asset classes in which they've considered investing. You are almost certain to hear answers that include "growth stocks", "value stocks", "bonds", "small caps", "international stocks" and occasionally "real estate." As we have written many times, the problem with these answers is that they are either too narrow or too broad. Given that the reason you diversify your investments across different asset classes is to minimize the riskiness of your portfolio, you want to avoid investments whose returns tend to move too closely with the other investments in your portfolio (statistically, you want to invest in asset classes whose returns have a low correlation with the returns on other asset classes in the portfolio). Given this, the problem we have when someone says "growth stocks", "value stocks", and "small caps" is that from our perspective their returns are all have relatively high correlations with each other, which makes them all members of the same asset class: domestic equities. In other words, the answer above contains not six, but rather at least four different asset classes: domestic equities, international equities, bonds and real estate.

The second issue that is raised by this answer is that, unlike the three flavors of domestic equities, some of the other asset class definitions are too broad rather than too narrow. For example, assuming "bonds" means "domestic bonds", we see not one, but at least three different asset classes: real return bonds (that protect you against inflation); investment grade bonds (that protect you against deflation); and high yield (also known as "junk") bonds that are more problematical (more on that later). By now you've got the point (actually, you probably got it a while back, but have been bearing with us because you're polite): While the actual allocation of your portfolio to different asset classes is important, the definition of the asset classes you will even consider is critical.

How Important is Asset Allocation?

But let's put that behind us, and move on to the next logical question. Just how important is the allocation of your assets between different asset classes? This is one of our favorite questions, because it is so frequently answered incorrectly. Unfortunately, there is no simple answer. Consider two investors, who (to simplify matters), have to answer two questions: how to allocate their assets between three asset classes, and whether to implement this strategy using index funds or actively managed funds. How important is asset allocation (as opposed to manager or security selection) to the returns they achieve? There are four ways to answer this question, and they are all correct.

If the two investors choose different asset allocations, but both implement their strategies using the same index funds, then asset allocation accounts for 100% of the difference in the returns they achieve after ten years. Similarly, if they have the same asset allocations and implement them through the same index funds, asset allocation again accounts for 100% of the returns they achieve.

On the other hand, suppose they both have the same asset allocations, but choose different actively managed funds to implement their common strategy. In this case, asset allocation would account for zero percent of the difference in the returns their portfolios achieve after ten years. All of the difference would be due to some combination of manager selection (by our two investors), stock picking skill (by the managers of the funds each one invests in), and the costs and taxes incurred by the respective funds.

So far, each of these answers has been pretty straightforward. The far more difficult situation is the fourth one, in which our two investors have different asset allocation strategies and choose different actively managed funds to implement them. In this case, the answer has been the source of quite a bit of controversy and disagreement between academics and industry players. The key disagreement is over the right measure you should use to answer the question. Two key studies on this issue have been conducted. The first is titled "Does Asset Allocation Policy Explain 40%, 90%, or 100% of Performance?", by Roger Ibbotson and Paul Kaplan. The second is "The Contribution of Asset Allocation to Portfolio Performance", by Wolfgang Drobetz and Friederike Kohler. The first study used ten years of data from the United States on the performance of balanced mutual funds (that invested in different combinations of bonds and stock), while the second used seven years of data on balanced mutual fund performance from Germany and Switzerland. Fortunately, both of these studies reached similar conclusions.

One way to measure the impact of asset allocation is to see how well a fund's basic asset allocation strategy explained the returns it earned from year to year. To do this, each study performed a regression analysis, in which the independent variable was the weighted performance of the basic asset allocation (e.g., if stock was 60% of the fund, and earned 10%, while bonds were 40%, and earned 5%, the asset allocation measure for the year would be 8%), and the dependent variable was the actual performance of the fund. As you might guess, the range of answers was wide. The 90% confidence range for one study was 47% to 94%, while for the other it was 58% to 96%. But what does this really tell us? In actual fact, it doesn’t tell us much at all. Some funds apparently stuck quite closely to their basic asset allocation policy, while others did not. What this measure doesn’t tell us is whether these active management departures from funds’ basic asset allocations ended up benefiting or hurting investors.

To answer that question, you need to use a different approach, and both studies did so. For each fund they compared the compound annual return earned over the study period by the base case asset allocation to the compound annual return actually earned by the fund (both before costs and taxes). If ratio between the two returns was less than 100%, active management had added value; if it was greater than 100%, active management had destroyed value. The results of this analysis were not pretty (if you are an active manager). In the first study, the 90% confidence range was from 86% to 113%, and in the second it was from 101% to 180%. In other words, most funds studied (especially those in Germany and Switzerland) were destroying value through active management. Therefore, by this measure, asset allocation policy was responsible for almost all the returns earned by investors.

Should we also conclude from these two studies that the American mutual fund managers were better than their German and Swiss counterparts? We don’t think so. A key difference between the two studies was the number of asset classes between which the mutual funds studied divided their assets. In the American case, there were only three; domestic bonds, domestic equities, and cash. In the German/Swiss case, more asset classes were used, including foreign stocks and bonds. In general, the correlations of returns within an asset class are likely to be higher than the correlations of returns between asset classes. Therefore, as the number of asset classes in which you invest increases, the importance (to your returns) of asset allocation relative to security selection is likely to rise. In our eyes, this goes a long way toward explaining the relative underperformance of the German/Swiss fund managers relative to their American counterparts. Another way of looking at this same question is to compare the average performance of top quartile and bottom quartile active managers in different asset classes. Where the difference between the two is relatively small (e.g., in domestic bonds and public equities), asset allocation choices should have a much bigger impact on realized returns than manager (i.e., stock selection) choices. Conversely, the relative importance of asset allocation should be lower when substantial investments are made in asset classes where the performance gap between top and bottom quartile managers is large (e.g., hedge funds or private equity funds).

So, in answer to our original question, "how important is asset allocation?" our conclusion is that in most cases it is likely to be the key determinant of the long-term returns you will realize on your investments.

Intuitively, How Do You Do It?

Okay. Now that we've defined it and shown why it is important, let's move on to how you should go about dividing your assets between different asset classes. Before we introduce any numbers, let's start with some basic principles. First, every investor faces the challenge of balancing downside protection (against capital loss) with upside potential (for high returns). Second, research has shown that in general, people are more sensitive to losses than they are to gains of the same magnitude. Third, your need for downside protection is also a function of the length of your investment horizon (how long before you'll need the money) and whether or not you are making regular withdrawls from your savings (as would be the case, for example, if you were gradually drawing down your savings to pay your bills during retirement). The shorter the time before you need the money, and the more you're planning on taking out along the way, the more downside protection you need. Fourth, the degree of mismatch between your current and expected savings and your financial goals can (absent a change elsewhere) tends in practice to create situations where the amount of downside protection you want is less than the amount you can afford (in terms of foregone returns on higher risk assets). In other words, there is usually a tradeoff between your lifestyle, your financial goals, and your asset allocation. As with so many other things in life, there is no free lunch here either!

Finally, different asset classes provide different degrees of downside protection and upside potential. Broadly speaking, our "home market" (we think of this as the market in which returns are denominated in the same currency as our liabilities) can be in one of three states: normal, high inflation, or deflation. In the normal state, we don't need as much downside protection as we do in the other states, and look to equity type investments to generate high returns for us. In the inflationary state, we look to asset classes like real return bonds and commodities (and possibly foreign bonds and real estate, but more on that later) to protect the purchasing power of our capital. Finally, in the deflationary state, we look to investment grade bonds to preserve our capital while maximizing our real returns.

Hopefully, by now we've accomplished our two main goals in writing this article: we’ve confirmed the importance of asset allocation, and given you a better intuitive understanding of how to do it well. So at this point you might want to go and get a cup of tea (or whatever) before we move on to our next section, in which we'll begin to look at different asset classes in more depth, and using more numbers.

A Closer Look At Different Asset Classes

We are now about to embark on a short tour of the different asset classes in which you may invest. Before leaving, however, we'd like to say a few words about some of the statistical animals we'll encounter in the coming paragraphs. First, all of the data you'll be seeing will be real (yes, quite…), rather than nominal. By that we mean, all of the numbers will be net of inflation. We do this because one goal of investing is to preserve, and ideally increase your purchasing power over time. Given this, the fact that an asset class earned 12% in nominal terms is meaningless unless you know what inflation was over the same period. If it was 5%, you earned a 7% real return. If it was 14%, you earned a real return of negative (2%).

Second, we will express returns two ways. Arithmetic returns are the simple average of a series of returns earned over a given time period. Geometric returns are the compound annual return earned over the same period. If the same return is earned each year, then the arithmetic and geometric means are the same. However, if different returns are earned in each period, then the two average returns will be different. For example, an investment that returns 15%, (10%), and 10% over three years has an arithmetic average return of 5% per year, but a geometric, or compound average annual return of about 4.42% per year. This example also illustrates a larger point: the greater the volatility of annual returns, the more the geometric average return will tend to be below the arithmetic average return. So why do we use both of these terms? In any given year, the arithmetic return is the best measure of the return you are likely to earn. This is therefore the return that we use in our optimization models, which we'll discuss next month. It is also the return we'll use when we develop estimates of how the returns different asset classes may earn in the future. However, over a longer holding period (which is a more accurate description of the situation faced by most investors), geometric returns better describe the returns that were actually realized in the past. So we'll use them too to provide perspective.

Finally, as we have discussed in the past, no single statistic gives a good picture of the riskiness of a given asset class. So we'll use four of them. We can hear the groans now! Let me explain. A lot of entry level finance textbooks define risk as the extent to which returns over a given period of time are distributed around the arithmetic average return for the period. In stats-speak, this is known as the variance or the standard deviation of returns (the latter is the square root of the former, but we won't belabor the point…). The problem is, when asked to define financial risk, most people don't talk about the standard deviation of returns. Typically, they'll say something like "risk is the chance I'll lose my money", or "risk is the probability I'll fall short of achieving my goals." In other words, in most people's minds, risk is not the symmetrical concept that textbooks assume it is when they equate it with the standard deviation of returns. So we need to take some other factors into consideration when we talk about the riskiness of an asset class.

The first we've already mentioned. It is the extent to which the returns on a given asset class vary with those on other asset classes. In stats-speak, this is called the covariance, or correlation of returns. If the latter is equal (at one extreme) to 1.0, they two series move in tandem in the same direction (not good for your portfolio). If (at the other extreme), the correlation of returns between two asset classes equals (1.0), they also move in tandem, but in opposite directions (surprisingly, not as good for a portfolio as you might think). And if their correlation equals zero, their returns are completely unrelated. Intuitively, to minimize the downside risk in a portfolio, you'd like to have some asset classes that have low positive or negative correlations with one another.

There is, however, an important catch. When we say that a correlation of zero means that the returns on two asset classes are completely unrelated, we're assuming that both return series are "normal" distributions. These are the familiar bell curve found in every introductory statistics textbook that has ever been written. In fact, once you assume that any set of data is normally distributed, a whole world of different statistical tests opens up to you. For example, you can say with confidence that about 68% of the time, the return on a normally distributed asset will fall within the range defined by the arithmetic average, plus or minus one standard deviation, and about 95% of the time, it will fall between the mean plus or minus two standard deviations. Unfortunately, not many financial assets have returns that are normally distributed. It is for that reason that we have to drag two other statistical terms into our conversation.

The first is called "skewness", and it measures the extent to which a given distribution of returns is "off center" or "tilted" compared to a normal distribution. Specifically, a normal distribution has a skewness of zero. A skewness less than zero indicates that more returns fall below the arithmetic average than above it, while a positive skewness indicates just the opposite. The returns on many asset classes (but not all of them) are negatively skewed. Generally speaking, we don't like negative skewness (actually, to be technically correct, what we really don’t like is coskweness, but we’re not going to go there in this article!), although we accept the fact that sometimes we have to accept it as one of the prices of earning higher returns.

The second statistical measure we need to use to assess asset class risk is called "kurtosis". This measures the extent to which a given distribution of returns is taller or shorter than the normal distribution. If it is taller, it is said to have "excess kurtosis". Practically, this means that because relatively more returns are clustered closer to the average return than would be the case if the distribution was normal, it must also be the case that relatively more returns also lie in the tails (that is, the extreme ends) of the distribution (that is, in order for the two distributions to have the save average return). In contrast, distributions with less than normal kurtosis are shorter than the normal distribution, but with fewer than normal returns located in the tails. In both cases, standard deviation will not accurately describe the dispersion of returns around the average. In the case of excess kurtosis, more returns will occur at the extremes than standard deviation would predict, while just the opposite would be the case if kurtosis were less than normal.

When it comes to judging the riskiness of an asset class, kurtosis and skewness have to be looked at together. The riskiest situation is one with excess kurtosis and negative skewness. In this case, you are likely to get more and larger downside surprises than you bargained for (assuming past returns are a reasonable guide, in a statistical sense, to what you can expect in the future). On the other hand, excess kurtosis also can be a good thing when skewness is positive -- in which case, you'll get more pleasant surprises than you would if the returns were normally distributed.

The final point we'd like to make before embarking on our asset class tour is that along the way we're going to be looking at how different asset classes performed during different periods of time. For example, we'll look at the 70s, 80s, and 90s, which represented periods of high, moderate, and low inflation in many countries. We'll also look at two shorter crisis period, covering the equity market crash of 1987 and the collapse of Long Term Capital Management in 1998. We'll also look at the geometric average returns different asset classes delivered over the 1971-2002 period, or the longest set of data we have available. We hope this will help you develop a better intuitive feeling for the real returns different asset classes have delivered under varying historical circumstances. We also hope it will put into better perspective the future return estimates we will present for each asset class.

So let's begin the tour, which is organized on a scale of more or less increasing risk and return. This month, we will four different fixed income asset classes: real return bonds, investment grade bonds, high yield bonds, and foreign currency bonds. Next month, we will cover domestic and foreign property (real estate), commodities, domestic, foreign, and private equity, and absolute return (also known as hedge fund) strategies. We will conclude with a discussion of the models we use to combine these different asset classes into model portfolios, and summarize the results of these efforts.

Real Return Bonds

Fixed income investments potentially expose you to different types of downside risk. The first type is usually called market risk. Because the future stream of cash flows you receive when you buy a bond is fixed (i.e., the coupon payment on most bonds doesn’t change over time), an increase in interest rates will cause the present value of the bond to decline (that is, the amount for which you could sell the bond today will decline in value when interest rates rise). This decline in the value of your capital also reduces the total rate of return you receive on your investment. For example, if a bond with a coupon of five percent experiences a n eight percent loss due to a rise in interest rates, your total return on it would be negative three percent. Two underlying factors can cause market interest rates to increase: a rise in inflation, or a rise in real interest rates. There are a number of ways you can limit your exposure to downside market risk. First, you can invest in bonds with shorter rather than longer maturities. All else being equal, the longer the maturity of a bond, the bigger the reduction in its present value that will result from a rise in interest rates. Of course, the flip side of this statement is also true: if interest rates decline, bonds with longer maturities will experience a larger increase in value than bonds with shorter maturities.

The other way you can protect yourself from market risk is to find a way to eliminate your exposure to changes in the rate of inflation. Until recently, this was very difficult to do. However, in the past decade, more and more governments have begun to issue what are known generically as "real return bonds" (e.g., TIPS or Series I Savings Bonds in the U.S.). In the UK, these are known as "index-linked bonds." The unique feature of these instruments is that investors are guaranteed to receive a constant real rate of return if inflation increases. Depending on the way the bond is structured, they may even provide some protection against deflation. For example, in the United States, Treasury Inflation Protected Securities (TIPS) guarantee that the principal value of the bond (which is adjusted with inflation) will not fall below its face value, even if a prolonged period of deflation suggests that this is what should be done to maintain its real return. As a result, the real return on these bonds would actually rise during a prolonged deflation, though not by less than the rise in real return on investment grade bonds (see below). To put it slightly differently, real return bonds protect both principal and interest against inflation, and (depending on their structure) sometimes principal against deflation. By comparison, nominal return bonds (that is, any bond that isn’t a real return bond) protect neither principal nor interest payments against inflation, but protect both of them against deflation.

On the other hand, real return bonds still leave an investor exposed to changes in the real rate of interest. For example, during periods when the economy is growing quickly, demand for capital and real rates of interest can rise, causing the capital value of real return bonds to fall. The opposite can happen during recessions. Still, because they eliminate some, but not all of the risk associated with a change in nominal interest rates (which can be caused by changes in expected inflation and/or real rates), real return bonds should be less volatile (that is, have lower standard deviations) than nominal bonds of comparable maturity.

Real Return Bonds are currently available in five of the six functional currencies covered by The Index Investor. They were first issued by the government of the UK in 1981, Australia in 1985, Canada in 1991, Sweden in 1994, the US in 1997 and France in 1998. So far, Japan has not issued any (which is logical, as they are in a prolonged period of deflation). Current real yields on these instruments are quite closely grouped (reflecting a highly efficient global fixed income market), and range from 2.92% in Canada to 1.79% in the Eurozone. The average yield for all five regions is 2.33%.

Looking more closely at TIPS in the United States, their arithmetic average annual real return since they were first issued in 1997 has been 3.89% (through the end of 2002). The standard deviation of these returns is 4.27%. This is lower than the standard deviation on the different types of nominal bonds we will look at, but it is not zero; there is still some risk in holding real return bonds. Since they have been issued, returns on TIPS have had a low correlation with the returns on other asset classes; in other words, they provide substantial diversification benefits to a portfolio. However, the skewness and kurtosis of the return distribution for TIPS are not pretty. The former is (.52), while the latter is high at 3.11. In other words, since they were first issued six years ago, there have been more and bigger negative real return surprises than positive ones for holders of TIPS. Frankly, we think these data vastly overstate the risk of holding TIPS. They seem to reflect two factors, only one of which will persist in the future.

As we have mentioned, the factor that will persist is changes in the real rate of interest over the business cycle. In this regard, a point to keep in mind is that since real return bonds first introduced in the United States in 1997, the economy has generally been growing, and most of the surprises have therefore been on the downside (that is, increases in the real rate of interest). This may have exaggerated the statistical riskiness of this asset class. The factor that will change is the excess volatility associated with the introduction of this asset class, and investors’ initial learning process. For example, real return bonds have been since 1981 in the United Kingdom, and the skewness and kurtosis of their returns are more in line with other government bonds (that is, positively skewed, with low kurtosis).

In light of all these considerations, the estimate of future risk and returns that we will use in our asset allocation models for real return bonds is an annual real arithmetic return of 2.50%, and a standard deviation of 2.50%. So, to summarize the pros and cons of investing in this asset class:

Investment Grade Bonds

In addition to market risk, investors in fixed income instruments may also take on credit risk (that is, the risk that the issuer of the bond will default, and you will lose your capital value). To help investors judge this risk, most bonds are rated by a credit rating agency. Typically, bonds receiving the top four ratings (e.g., AAA, AA, A, and BBB) are considered "investment grade", while those with lower ratings are politely called "high yield", and less politely called "junk bonds." Theoretically, the only bonds that don’t have any credit risk are those issued by governments, since governments don’t go bankrupt. However, history has proven that some governments do go out of business (reluctantly, and usually with a lot of bloodshed), which, from a bondholder’s perspective, is the same thing. So, to be more accurate, the lack of credit risk associated with government bonds is only guaranteed for the time horizon over which you believe you can predict potentially disastrous political events.

Broadly speaking, there are three types of investment grade bonds. The first are issued by governments. The second are issued by corporations, but not secured by any assets (technically, these are known as debentures). The third are bonds secured against assets, the most common of which are mortgage bonds (though the volume of other "asset-backed bonds" secured against things like credit card receivables has been growing in recent years).

As previously noted, investment grade bonds are particularly attractive during deflationary periods, which increases the real value of both the interest and principal payments received by bondholders (assuming said gains aren’t offset by increasing defaults brought on by deflation). During inflationary periods, the real value of bond interest (assuming fixed payments) and principal declines, and bondholders sometimes earn negative real returns (particularly if they hold long term instruments). In normal periods, with low inflation and steady GDP growth, bondholders generally earn total returns that lag behind those on other asset classes.

| Global Asset Class Returns | Equity Market Valuation Update | Asset Allocation Update - Part I | Model Portfolio Update | The Economic Outlook: May 2003 |