Chapter 19
HOW WELL ARE MY INVESTMENTS DOING?
Measuring Investment Results
 
 
 
 
 
 
 
 
There are many ways to measure the results of your investing activities. You may have heard about the high performance of a particular mutual fund, or that a certain stock has a good return on equity (ROE) or return on investment (ROI). All of these have some bearing on the measurement of investment results. But right now, we are more concerned with a more personal aspect of return: How can you calculate the return on your money?
performance
Measure of return as proportion of original investment, or measure of investment performance in an average period.
Figuring out the return on your money can be a difficult and time-consuming task. Most of the difficulty is of the grunt-work variety. You may have to track down all the pieces of paper that tell what happened when. Some of the difficulties are interesting and can lead to a deeper understanding of the investment process.
We will begin by giving you an overview of what goes into calculating the rate of return on a single security and on a portfolio. We will explain how the different ways of thinking about averages affect the measurement of investment results. Building on the simplest concept of return, we illustrate how time and cash flows in and out of an investment portfolio factor into measurements of performance. This will lead into a comparison of time-weighted and money-weighted measures of return. Finally, we explain how these two very different measures are used in the reporting of investment performance. As the saying goes, knowledge is power. And that is probably why you are reading this book.

The Basics of Return

Starting with some amount of hypothetical capital, the simplest measure of return is just the change in dollar value: “How much did I make (or lose)?” At a gut level, this is what we all want to know when we turn to the financial pages or open the envelope containing a statement from our broker.
At the most elementary level, return is simply the profit or loss on an investment. If you bought 200 shares of IBM at $100 per share and sold them for $106, your gross return is
($106 - $100) X 200 = $1,200
This $1,200 return was earned on an investment of $20,000. The percentage return is
058
Many people, when they see a “%” sign, automatically assume that it refers to a rate. But this is not so. We haven’t yet mentioned how long the stock was held. Adding to the potential for confusion, the word return is often used as shorthand for a rate of return, usually an annual percentage. If we assume that the shares were held for exactly one year, then the annual rate of return is exactly 0.06, or 6 percent. This number takes into account how quickly the return was earned. It is expressed as a return for some unit of time, most often a year. Other frequently used time periods include quarterly, monthly, weekly (seven-day), and even daily rates of return. By convention, they are usually expressed as percentages, but remember that they need not be.
Rates for longer time periods are also frequently quoted by mutual funds: 2-year, 5-year, and even 10-year rates of return. Frequently, these numbers are annualized to represent how a fund or other investment did in an average year of the 2, 5, or 10 years in question.

An Example of How to Calculate Return and Monthly Rate of Return for a Two-Stock Portfolio

Let’s say you have a two-stock portfolio with no cash, just 100 shares of XYZ Corporation and 200 shares of ABC Corporation. Say XYZ’s closing price last month was $75 per share and ABC’s was $55 per share. Last month’s total portfolio value was $18,500 = ($75 X 100) + ($55 X 200).
Annualizing Rates of Return and the Meaning of Average
Annualizing is a way of converting a rate of return with a period of either less than a year or more than a year into a number that represents its hypothetical return for a year. If the rate you start with is for less than a year—say it’s a monthly return—then annualizing tells you what the return would be if it continued to grow at the same rate for exactly a year. If the rate you start with is for a period greater than a year, then annualizing shrinks the time period to a year, giving you an idea of what the portfolio earned in an average year of the full period.
 
In a way, it’s unfortunate that both of these conversions are called annualizing, because they are very different. In the first case, you are extrapolating, or stretching, from a shorter period to a year, while in the latter case, you are interpolating, or shrinking, from a longer period to a year. As we will soon see, there are also two fundamentally different ways to shrink or stretch rates of return—arithmetic and geometric averaging.
 
We explained that annualizing a return of longer than one year provides you with a measure of the rate of return for an average year of the period. This raises an important question: Just what is meant by an average year? Sometimes, when people speak about averages, they have in mind the commonest, most frequent value of something. Statisticians call this type of average the mode. Other times, people think of the average as the value that sits halfway between the thing with the highest value and the thing with the lowest value. Statisticians refer to this type of average as the median. Usually, however, when people think about averages, they have in mind what the statisticians call an arithmetic mean. To get an arithmetic mean of a set of prices, you just add up all the prices and divide by the number of prices.
Your current month’s statement arrives from your broker. How did you do? It shows that XYZ’s closing price has advanced to $85 per share, a nice move for a month, while ABC is off a little, down to $53 per share. Nothing else happened Your portfolio is now worth
$19,000 = ($85 X 100) + ($53 X 200)
This is a gross return or profit of $600. (This is a so-called paper profit, or unrealized gain.)
Rounded to the nearest hundredth of a percent, or basis point, your monthly rate of return is
059
We are unlikely to be content with just knowing the monthly return of our account. We’d like to know how this month’s return compares to returns in previous periods for this portfolio and to previous and current returns of other investments. We’d also like to make projections of how much the portfolio might be worth in the future at various rates of growth.
Avoiding a Common Pitfall
Remember to use consistent numbers when doing your calculations. End of month ÷ end of previous month is a good way to figure out a monthly return, but end of month ÷ beginning of month is bad. Why? You’re not using a full month of data; you’re leaving out any price movement from the end of the previous month to the beginning (opening price) of the current month. This could be a significant source of error.
It is most important when comparing rates of return of an investment over different time periods, or when comparing rates for different portfolios, to make sure that they are stated on a comparable basis. Obviously, it would make no sense to compare a monthly return with an annual return. One of the numbers needs to be adjusted in some way to account for the difference in time periods. By convention, returns are frequently annualized to make a common basis for comparing investment results obtained over differing time periods.
If we are to accurately compare rates of return for different investments, we must take into account the time units or periods of the different investments’ rates of return. Periods can be of any length. Typical measurement periods include daily, seven-day, monthly, month-to-date, quarterly, quarter-to-date, annual, year-to-date, 2-year, 3-year, 5-year, 10-year, 20-year, and since inception.
since inception
Measure of investment performance since a fund opened.
How Not to Compare Fund Performance
Some rates of return are described as since inception. This refers to the lifetime performance of a fund or other security. Sometimes these numbers are annualized; sometimes they are not. Since different securities have been around for different periods of time, it would be a serious mistake to compare their unannualized lifetime returns. Even if they are annualized, the comparison is often not very meaningful, because it ignores the effects of the environment in which those results were obtained.

Annualized Returns: Arithmetic (Simple) or Geometric (Compound)?

Whether you start with a period shorter or longer than a year, there are two ways to annualize a rate of return. The simpler way is to calculate an arithmetic mean return, also called a simple return. Arithmetic means are useful, but they are not good at representing rates of growth. For this we need a different kind of number, a geometric mean. Geometric means take into account the effect of compounding, or exponential growth. Compounding occurs when you get money (e.g., interest or dividends) from an investment and put it back into the portfolio, letting it grow alongside the original investment. Over time, the effects of exponential growth can dwarf the value of the original investment, so it is very important to understand the difference between arithmetic mean returns, also known as simple returns, and geometric returns, also known as time-weighted returns.

Example of an Arithmetic Return Calculation

To calculate an arithmetic rate of return, all you need is the starting and ending values for the period, although a good calculator wouldn’t hurt. Let’s say that we’re reviewing a monthly statement from a broker. At the beginning of the month, the portfolio had a value of $10,000. At month’s end, the portfolio was worth $10,200. Divide the ending value by the starting value:
060
This means that the portfolio is now 1.02 times larger than it was one month ago. Expressed as a percentage, it is 102 percent. This means that it has grown by 2 percent during the month. Of course, the portfolio’s value could have been higher or lower during the month, but right now our only concern is with the beginning and ending points.
In the current example, we can annualize the arithmetic return of a 2 percent monthly increase simply by multiplying 2 percent times 12 to get 24 percent for the year. Nice and simple, but potentially misleading. It does not take into account the money that is being generated by the portfolio each month, which could be reinvested in something, perhaps earning the same rate of return as the initial capital investment. Arithmetic returns assume that you earn nothing on this interim money, while geometric returns assume that you earn the same rate of return on this new money as on the initial investment. The former accurately describes certain kinds of investment situations, where one has no opportunity to reinvest new money. The latter accurately describes a more typical situation in the world of no-load mutual funds, where money being reinvested into the account from interest or dividends gets pretty much the same treatment as money already in the account. Taxes may make a difference—distributions in taxable accounts are treated as income or capital gains. In general, you cannot know what return will be available on the cash that comes into your portfolio—this is known as reinvestment risk, which will be described in the next chapter.
Table 19-1 How a Net Loss Can Look Like a Gain (key numbers in boldface)
061
Beyond the frequently inaccurate assumption of no return on intra-period earnings, there is an even more serious problem with arithmetic returns: They can lead you to believe that you are up 15 percent per year for two years, when in fact you have lost 10 percent of your money. How? Let’s say you start with $100 and earn 80 percent in the first year. At the end of the first year, you have $180. In the second year, however, you lose 50 percent, going down to $90. Even though you’ve lost 10 percent of your money, the arithmetic mean annual return for the two-year period is 15%. (See Table 19-1.)

Example of a Geometric Return Calculation

When interest or dividends are reinvested, a geometric return should be calculated. A good calculator or financial software is indispensable for working with geometric returns.
With arithmetic returns, you might need to do only a single multiplication or division. In contrast, geometric returns are combined by multiplying return relatives. A return relative is simply the ratio of the current end-of-period value of the portfolio, security, or index to its previous end-of-period value.
Working on a Geometric Chain Gang
If, for example, your rate of return for a given month was 2 percent, the return relative would be 102 percent or 1.02. To annualize, we raise 1.02 to the twelfth power:
1.0212 = 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 X 1.02 = 1.2682 (to the nearest basis point)
You can annualize a monthly rate using return relatives. If the monthly rate of increase is 0.02, or 2 percent, the return relative is 1.02, or 102 percent. Multiply 1.02 together 12 times (1.0212). This repeated multiplication is sometimes referred to as a geometric chain.
A point worth remembering: You must always remember to subtract 1 after annualizing a return relative to get back to a percentage change (what you’re subtracting is the starting value, which is always 1, or 100 percent). So you raise 1.02 to the twelfth power, subtract 1, and you get 26.82 percent. This is the geometric return, also known as the effective annual rate of return. Note that it is 2.82 percent larger than the arithmetic return of 24 percent. The extra 2.82 percent comes from the compounding of returns.
effective rate (effective annual rate of return)
Yield to maturity.
Sometimes you may want to calculate your effective annual return from starting and ending values separated by some number of years. This is easy to do—with a decent calculator. For example, consider an investment that, after eight years, is worth $20,000. Its value at the beginning of the eight-year period was $8,000. To figure the compound annualized (i.e., geometric) rate of return, just take the endpoint value of $20,000 and divide by the starting value of $8,000 to get the return relative of 2.5 for the whole eight-year period. Instead of multiplying this number eight times, you need to find the number that multiplied by itself eight times equals 2.5—in other words, you must take the eighth root of 2.5. This is approximately 1.1214, the annualized return relative. To get the annual return, you still need to subtract 1: giving 0.1214, or in percentage terms, 12.14%, your annualized geometric rate of return.
Rule of 72
There is a nice shortcut for figuring how long it takes money to double at various rates of return (its usual use is for working with compound interest, but it works just as well for working with any return calculation). It is called the rule of 72. To figure the approximate doubling time of an annual rate of return, just divide 72 by that rate, leaving off the percent. For example, the calculation for a 6 percent annual return would be
062
You can check that result by raising 1.06 to the twelfth power (or if you don’t have that kind of calculator, just multiply 1.06 12 times). The answer is (approximately) 2.012, which for most purposes is a very good approximation. Don’t use the rule of 72 for calculations that must be precise or for rates of return less than 2 percent or greater than 24 percent.
A Practical Hint for Calculating Effective Annual Rate of Return
How, exactly, do you calculate the eighth root of 2.5? Even good calculators don’t have a special eighth-root button. In this particular example, even a cheap calculator can do the job, because you can get an eighth root by taking the square root three times in a row, which is nice for this example and for other examples in which the number of periods is a power of two. But it won’t help with three years, five years, and so forth.
 
On a good calculator or a spreadsheet, you just use the exponent key (or function), yx . When you want an eighth root, you are raising 2.5 to the one-eighth power. Expressed as a decimal number, one-eighth is 0.125. So you would need to calculate 2.5 .125. On a good calculator, you would just input 2.5 [function key(s) for yx].125. And on a spreadsheet? Depending on which spreadsheet program you use, you would simply enter 2.5^.125 or 2.5**.125 into a spreadsheet cell.
 
Remember—to get the annual return, you must subtract 1.
From the preceding examples, you can readily see why it is so important to know whether a given rate of return was calculated on an arithmetic or geometric basis. We have seen that an arithmetic rate of return can grossly mislead an investor into thinking that he or she is making money when in fact the portfolio’s value is standing still or, worse, rapidly diminishing. Geometric returns are far more accurate in evaluating past performance. For this reason the CFA Institute insists, in its widely adopted Global Investment Performance Standards (GIPS®), that investment managers (e.g., mutual fund portfolio managers) use geometric returns. While geometric returns are more difficult to calculate than arithmetic returns, they give you a much more reliable idea of what the portfolio is actually doing. Even geometric returns, however, must be used carefully, especially when extrapolating from a short period to a much longer time frame (see the discussion of extrapolation risk in Chapter Twenty).

Time-Weighted Returns versus Money-Weighted Returns

If we want to know how well our investments did in an average period of time, we calculate what is known as a time-weighted rate of return. If, however, we care about what happened to an average dollar or other unit of money, we calculate a dollar or money-weighted rate of return.
The King and the Wizard: A Tale of Chess and Geometric Returns
Perhaps the oldest story illustrating the hazards of projecting geometric returns too far into the future concerns a king and the wizard who invented chess. The king, thrilled by the wizard’s invention, offers to grant any wish the wizard might have. The wizard, who we can infer was a crafty sort, says that all he would like is a single grain of rice for the first square of the chessboard, two grains for the second square, four grains for the third, and so on, each succeeding square to be paid for with twice as many grains of rice as its predecessor until all 64 squares have been accounted for. The king, apparently not familiar with the fecundity of geometric progressions, quickly grants this wish, muttering to himself in some versions of the story about what a good deal he has gotten from this silly wizard! He calls for his servants to bring forth the required number of grains. For the first 10 squares, 2,047 grains of rice are counted out—about a bowl’s worth. For the next 10 squares, more than 1,000 times as much rice is counted out. Each succeeding square requires more rice than all the previous squares combined. Eventually, it becomes clear to the king that the whole kingdom’s supply of rice is inadequate. (In fact, the total number of grains of rice necessary to fulfill this contract equals 264 minus 1, which is more than 18,000,000,000,000,000,000 [18 quintillion] grains of rice!)
 
At this point in the story, the king reneges on the deal. Some say he opted instead to execute the sly wizard as punishment for his overabundant cleverness. This tale can serve as a warning against extrapolating geometric returns too far out into the future. They always look great on paper, but continued long enough, other factors (e.g., the world’s supply of rice, the king’s patience) run out and further growth prospects evaporate. So the next time your broker, insurance salesperson, or financial consultant tells you about the wonders of compound interest, you can tell them that you are well aware of the wonders—and the dangers—of excessive reliance upon exponential growth curves.
Neither of these approaches tells the full story. In general, the time-weighted approach gives a more accurate measure of an investment manager’s skill. In contrast, the money-weighted approach provides more information about the actual accumulation of money in a portfolio. It also provides a formula for aggregating multiple portfolios run by the same investment manager, so it can be used to safeguard against selective presentation of performance results.
An example should help explain things. Suppose that at the beginning of the year you had $100,000 in your brokerage account in the form of 2,000 shares of the DEF company, trading at $50 per share. After six months, DEF has risen in price to $55, and you decide to buy an additional 2,000 shares with money transferred from another account. You now have 4,000 shares, whose current market value is $220,000 and whose cost to you was $210,000 (for simplicity, we will ignore transaction costs). In the second half of the year, the price of DEF rises to $66 per share, with a market value of $264,000. What is your rate of return?
Time-weighted return: From the standpoint of time, the second half-year period has twice the rate of return of the first half-year period:
063
The total increase for the period is
064
This tallies with the product of the two six-month periods:
(1.1 X 1.2) - 1 = 32%
Linking the periods in this way is referred to as forming a geometric chain. For this reason, the time-weighted return is sometimes called a geometric chain time-weighted rate of return.
Notice that there is an extra 2 percent, compared with the arithmetic sum of 10% + 20%. This extra 2 percent comes from the compounding effect—in the second half of the year, you are getting 20 percent not just on the original 100 percent, but also on the 10 percent you made in the first half.
If you are wondering how a time-weighted return takes into account the effect of money being added to or subtracted from the portfolio, the answer is that it doesn’t. It ignores the effect of deposits and withdrawals completely. All it “cares” about is the effect of time. While this is fine for understanding how well a stock or mutual fund is performing, it ignores what for some purposes is more important to you, namely, understanding how an average unit of money that you invest in the stock or fund has performed. To see the combined effect of the stock or fund manager’s performance with the timing of your deposits and/or withdrawals, we must look at money-weighted returns.
Money-weighted return: If you hadn’t put in additional money at midyear, you would have 32 percent more money at year-end than at the beginning. In fact, you invested only $100,000 at the beginning of the year, putting in another $110,000 at midyear. So you had a total cost of $210,000, but not all of this money was invested for the full year. There are a number of ways to figure out the return on an average dollar invested. One way is to subdivide the year into the two six-month periods. In the first six months, you earned 10 percent on $100,000. In the second six months, you earned 20 percent on $220,000. This second period counts, or weighs, 2.2 times as much as the first period in figuring the return on an average dollar.
065
This number represents the (arithmetic) average return on a dollar in a six-month period. It needs to be annualized to get the money-weighted return for the full year.
To get a more accurate picture of the return on an average dollar, financial analysts calculate an internal rate of return. The idea is to find the rate of return that equates two sets of cash flows. For a simple example, let’s find the rate of return that makes $1,000 increase to $4,000 in 12 years. By the rule of 72, we know that money invested at 12 percent doubles roughly every six years. Investing for 12 years at 12 percent would quadruple your money. Therefore, 12 percent is the internal rate of return that equates $1,000 today with $4,000 twelve years hence.
internal rate of return
A measure of return on investment that takes into account both the timing and the amounts of individual cash flows.
In working this simple example, we’ve snuck in another key concept used by financial analysts: present value. What is the present value of a future cash flow? That depends on the rate of return used in the calculation. In the preceding example, the present value of $4,000 received 12 years from now is $1,000 if we assume a 12 percent rate of return. If we assume a lower rate of return, say 6 percent, then the present value would increase to $2,000, reflecting the fact that you need more money at the starting point if your rate of growth is smaller.
The standard way of calculating investment performance is to use time-weighted rates of return for individual portfolios. Money weighting is used, but not for calculating investment manager performance. The reason? Investment managers seldom have control over the timing of cash inflows and outflows. Since they are beyond the manager’s control, it would make little sense to credit or penalize the manager for their occurrence.
Nevertheless, inflows and outflows of cash do impact a manager’s strategy and, indirectly at least, his or her performance. Usually, they act as a drag on performance. Cash tends to come into the portfolios of top-performing managers. While management companies love to accumulate assets, the fast accumulation of cash in a portfolio can make it extraordinarily difficult for a manager to maintain a particular strategy. The manager might want to take some profits, but is already “awash in cash” and struggling to find investments that meet his or her objective. Likewise, when a fund manager is doing poorly, investors rush to the exits, forcing the manager to sell at a time when it might make more sense to buy.
Despite its limitations, time weighting is the best available way to measure an investment manager’s performance. But this doesn’t mean that money weighting has no role to play. In fact, it has an important role in keeping performance reporting fair and accurate. Money weighting is the standard way that investment managers are supposed to aggregate collections of portfolios with similar investment goals. These standards are intended to prevent a manager from selectively highlighting a few small portfolios with good time-weighted returns while hiding poor performance in other portfolios.

Complicating Factors

In real life things are seldom as simple as the idealized examples of this chapter. Most brokerage accounts consist of more than just two securities and have some spare cash lying around. Trades may occur, accounts may be assessed fees or interest charges for stocks purchased on margin, securities may undergo any of a number of capital events during the month, and cash may be deposited into and/or withdrawn from the account at different times during the month.
Let’s look at each of these complicating factors in turn:
Multiple securities and spare cash. Multiple securities present no special problem, just a few more multiplications and additions to calculate end-of-previous-month and end-of-current-month values. Similarly, cash in the account presents no intrinsic difficulty. For purposes of calculating return, it can be regarded as just another security. The cash may sit idle, in which case it neither adds to nor subtracts from the account return (but can impact its rate of return). Or it may be swept into a money market account, leaving you with slightly more cash at the end of the month than at the beginning. In that case, it makes a small but positive contribution to your return, though it can still lower your rate of return if your securities are doing well. In rising markets, excess cash is frequently frowned upon by investors, most of whom try to stay fully invested. On the other hand, in nervous markets, investors are sometimes encouraged to be selective or defensive and to raise cash.
fully invested
A portfolio that has used (almost) all of its cash to buy securities or other assets.
Trades may occur. Assuming for the moment that no new cash or securities enter or leave the account, trading that takes place within the account will produce a calculation similar to our simple example, although perhaps lacking its symmetry. If, for example, we had sold half of the 200-share position in ABC Corporation sometime during the month for $54 per share, we would have an end-of-month account value that was $100 higher than given in the example.
Accounts may be assessed fees or margin interest charges. Such fees lower both the return and the rate of return. The impact of explicit fees on return is easy to measure. In contrast, there are the hidden trading costs we discussed in Chapter Eight, such as bid/asked spread, dealer markup, and payment for order flow, all of which can impact your return without showing up on your statement.
Securities may undergo any of a number of capital events. Stock is subject to splits, reverse splits, dividends, rights offerings, mergers, and spin-offs; bonds can offer interest income or return of principal (either at maturity or at call date); bonds can also default in payment of either interest, principal, or both; mutual funds have dividend distributions and capital gains distributions.
Cash or securities may be deposited into and/or withdrawn from the account. This is a tough one, from a calculation standpoint. When money comes into or leaves an account, you have a brand-new ball game. Previously, you had a unique starting point or base amount from which to calculate return. But each time you add or subtract money from the account, you create a new base amount from which return is to be judged. This is where money-weighted returns can be useful.
Calculation of raw investment performance tells you how much an investment made or lost, but does not provide a context for evaluating that investment in the light of your unique risk profile. Such an evaluation requires an understanding of how risk relates to performance, the subject of our next chapter.