Bottom-Up, Top-Down Approach to Fundamental Analysis

<< INVESTMENT RATIOS

The Industry Life Cycle >>

Investment Analysis & Portfolio Management (FIN630)

Lesson # 14

ECONOMY AND MARKET ANALYSIS Contd...

A Framework for Fundamental Analysis:

Under either of these fundamental approaches, an investor will obviously have to work with

individual company data. Does this mean that the investor should plunge into a study of

company data first and then consider other factors such as the industry within which a

particular company operates or the state of the economy, or should the reverse procedure be

followed? In fact, each of these approaches is used by investors and security analysts when

doing fundamental analysis. These approaches are referred to as the "top-down" approach

and the "bottom-up" approach.

Bottom-Up Approach to Fundamental Analysis:

With the bottom-up approach, investors focus directly on a company's basics, or

fundamentals. Analysis of such information as the company's products, its competitive

position, and its financial status leads to an estimate of the company's earnings potential

and, ultimately, its value in the, market.

Considerable time and effort are required to produce the type of detailed financial analysis

needed to understand even relatively small companies. The emphasis in this approach is on

finding companies with good long-term growth prospects, and making accurate earnings

estimates. To organize this effort, bottom-up fundamental research is often brokers into two

categories, growth investing and value investing.

Value versus Growth:

Growth stocks carry investor expectation of above-average future growth in earnings and

above-average valuations as a result of high price/earnings ratios. Investors expect these

stocks to perform well in the future, and they are willing to pay high multiples for this

expected growth. Recent examples include Microsoft, Cisco Systems, and Intel.

Value stocks, on the other hand, feature cheap assets and strong balance sheets. Value

investing can be traced back to the value-investing principles laid out by the well-known

Benjamin Graham, who coauthored a famous book on security analysis in the 1930s that has

been the foundation for many subsequent security analysts.

Top-Down Approach to Fundamental Analysis:

The top-down approach is the opposite of the bottom-up approach. Investors begin with the

economy and the overall market, considering such important factors as interest rates and

inflation. The next consider future industry prospects or sectors of the economy that are

likely to do particularly well (particularly poorly). Finally, having decided that macro

factors are favorable to investing, and having determined which parts of the overall

economy are likely to perform well, individual companies are analyzed.

There is no "right" answer to which of these two approaches to follow. However,

fundamental analysts can be overwhelming in its detail, and a structure is needed. This text

takes the position that the better way to proceed in fundamental analysis is the top-down

approach: first, analyze the overall economy and securities markets to determine if now is a

Investment Analysis & Portfolio Management (FIN630)

good time to commit additional funds to equities; second, analyze industries and sectors to

determine which have the best prospects for the future, and finally analyze individual

companies.

Thus, the preferred order for fundamental security analysis used here is (1) the economy and

market, (2) the industry/sector, and (3) The company .this approaches is used in Part IV

,which explains fundamentals security analysis in detail. Here we consider only the

justification for this approach.

Economy/Market Analysis:

It is very important to access the state of the economy and the outlook for primary variables

such as corporate profits and interest rates. Investors are heavily influenced by these

variables in making their every day investment decisions. If a recession is likely, or under

way, stock prices will be heavily affected at certain times during the contraction.

Conversely, if a strong economics expansion is under way, stock prices will be heavily

affected, again at particular times during the expansion. Thus, the status of economic

activity has a major impact on overall stock prices. It is, therefore, very important for

investors to access the state of the economy and its implications for the stock market.

In turn, the stock market impacts each individual investor. Investors cannot very well go

against market trends. If the markets goes up (or down) strongly, most stocks are carried

along. Company analysis is likely to be of limited benefit in a year such as 1974, when the

stock was down 25 percent. Conversely, almost all investors did well in 1995 regardless of

their specific company analysis, because the market was up about 37 percent as measured

by the S & P 500.

Industry / Sector Analysis:

After completing an analysis of the economy and the over all market, an investor can decide

if it is a favorable time to invest in common stocks. If it is, the next step should industry or

sector market analysis. King identified an industry factor as the second component (after

market movements) affecting the variability in stock returns.

Individual companies and industries tend to respond to general market movements, but the

degree of response can vary significantly. Industries undergo significant movements over

both relatively short and relatively long periods. Industries will be affected to various

degrees by recession and expansions. For example, the heavy good industries will be

severely affected in recession. Consumer's goods will probably be much less affected

during such a contractionary period. During a severe inflationary period such as the late

1970s and very early 1980s, regulated industries such as utilities were severely hurt by their

inability to pass along all price increases. Finally, new "hot" industries emerge from time to

time and enjoy spectacular growth. Examples include synthetic fuels and genetic

engineering.

Company Analysis:

Although the first two steps are important and should be done in the indicated order, great

attention and emphasis should be placed on company analysis. Security analysts are

typically organized along industry lines, but the reports that they issue usually deal with one

or more specific companies.

The bottom line for companies, as far as most investors are concerned, is earnings per share.

There is a very close relationship between earnings and stock prices, and for this reason

Investment Analysis & Portfolio Management (FIN630)

most attention is paid to earnings. Dividends after all are paid out of earnings. The

dividends paid by companies are closely tied to earnings, but not necessarily the current

quarterly or even annual earnings.

A number of factors are important in analyzing a company. However, because investors

tend to focus on earnings and dividends, we need to understand the relationship between

these two variables and between them and other variables. We also need to consider the

possibilities of forecasting earnings and dividends.

Because dividends are pad out of earnings, we will concentrate on earnings in our

discussion of company analysis. Earnings are the real key to the fundamental analysis of a

common stock. A good understanding of earnings is vital if an investor is to understand, and

perform, fundamental analysis.

Assessing the Economy:

A basic measure of the economy is Gross Domestic Product (GDP), defined as the market

value of final goods and services produced by an economy for some time period. GDP

numbers are prepared quarterly and released a few weeks following the end of the quarter.

These numbers constitute a basic measure of the economic health and strength of the

economy.

We can measure and observe GDP on both a nominal and real basis. The shaded areas

indicate officials' periods of recession. Note the ups and downs up to about 1990 and the

long upward movements from early 1991 into 2000, the greatest economic expansion in US

history.

GDP is revised twice in the first three months after its initial release. The bureau of

Economic Analysis initially releases an advance estimate of quarterly GDP in the first

month following quarter end. In the second month, it provides a preliminary estimate, and in

the third month, it provides a final estimate. The cycle then starts over again. Over the last

30+ years, the average revision of GDP growth from the advance to the final estimate has

been about two-third of a percentage point. It should e noted that almost 90 percent of the

time the advance estimate correctly predicts the direction of quarterly change in real GDP

growth.

What Determines Aggregate Stock Prices?

We examined the variables that are used to estimate the intrinsic value of stocks with the

dividend discount model----dividends are required rate of return----and with the P/E ratio

model-----earnings and the P/E ratio. The same models apply to the aggregate stock market

as represented by a market index such as the S&P 500 index.

To value the stock market using the fundamental analysis approach explained in chapter 10,

we use as our foundation the P/E ratio or multiplier approach, because a majority of

investors focus on earnings and P/E ratios. Estimates of index earnings and the earnings

multiplier are used in equation 13-1. As explained in chapter 10, this model uses a forward

P/E ratio. We will use the S&P500 Index as our measure of the stock market:

P0 = P0/E1×E1;

Where;

E1 = expected earnings on the S&P500 index

Investment Analysis & Portfolio Management (FIN630)

P0/E1 = the forward price earning ratio or multiplier

We consider each of these variables in turn.

The Earning Stream:

Estimating earnings for purposes of valuing the market is not easy. The item of interest is

the earnings per share for a market index or, in general corporate profits after taxes.

Corporate profits are derived from corporate sales, which in turn are related to GDP. A

detailed, top down fundamental analysis of the economy/market would involve estimating

each of these variables, starting with GDP, then corporate sales, working down to Corporate

earnings before taxes, and finally to corporate earnings after taxes. Each of these steps

involves various levels of difficulty.

Looking at real (inflation-adjusted) earnings growth, we would expect it to correlate closely

with real GDP growth in the long run and, in fact, of the last 30 or so years, real GDP

growth has averaged about 3.1% annually, whereas real earnings for the S&P500 index has

averaged 2.7% annually. Therefore, when estimating real earnings growth for the future, the

best guide may be expected real GDP growth.

It is reasonable to expect corporate earnings to grow, on average, at about the rate of the

economy as a whole. However, for the last years of the 20th century, operating earnings per

share for the S&P500 grew an average of 10.2% per year versus a rate of 5.56% for

economic growth. This simply illustrates how difficult it is to forecast earnings accurately.

Extenuating factors can cause some divergences, for example, share repurchases by firms;

any increase the rate of earnings growth relative to historical rates.

Since earnings have to be allocated over fewer shares, as firm repurchase shares,

earnings/share increases. Estimate are that this could add anywhere from one-half to one

and one-half% points to the growth rate of real earnings.

Which Earnings Should We Use?

Note that an annual EPS for the S&P500 index, as obtained at www.spglobal.com, can be

instructed in various ways. For example as of June 02, all four quarters for 02 were

estimates-of 4 quarter total of $51.23. One year ahead estimate could involve the last two

quarters of 02 and the first two quarters of 03. The past years EPS numbers could be taken

to be the calendar year 01, or the last two quarters of 01 and the first two quarters of 02.

This is further complicated by the fact that for the S&P500, S&P provides both top-down

and bottom up estimate and both "as reported" estimates and operating estimates.

Furthermore, S&P is now providing its "core earnings" for the S&P500 index, which

focuses on companies after tax earnings generated from their Principal businesses. S&P has

determined that the primary reasons the core earnings and as-reported earnings differ are

pension income and stock options grant expenses, with the treatment of pension gains

having a very significant impact. The differences between these two numbers can be

substantial.

The Multiplier or P/E ratio:

The multiplier to be applied to the earnings estimate is as important as the earnings

estimate, and often more so. Investors sometimes mistakenly ignore the multiplier and

Investment Analysis & Portfolio Management (FIN630)

concentrate only on the earnings estimate. But earnings growth is not always the leading

factor in significant price changes in the market. Instead, low interest rates may lead to high

P/E ratios, which in turn may account for the majority price changes.

The multiplier is more volatile than the earnings component and, therefore, it is even more

difficult to predict. Consider figure 13-3, which shows the P/E ratio for S&P500 index,

since 1947. A trend line has been included to show the general upward swing of P/E ratio

across time. Also, 3 different levels of P/Es are identified with the bars showing that at

different times (the 60s and early 70s, the late 70s and early 80s and the late 80s and the

early 90s), P/Es tended to cluster together.

The P/E ratio begin to rise in the early 50s, reached a peak by 1960, and declined to the 16

to 18 area and remained around the level through 1972. as inflation heated up in 1973, and

interest rates rose, the multiplier started to decline and by 1974 it was around 7, less than

half its previous level, a drastic cut for such a short time. Therefore, what was considered

normal (about17), in the 60s and early 70s was not the norm in the late 70a and early 80's?

Most investors did not estimate P/E ratio this low for this length of time. The lesson from

this analysis is obvious: investor cannot simply extrapolate P/E ratios, because dramatic

changes occur over time. Perhaps the most that can be said is that in the post war period;

P/E ratios of broadly based indices have ranged from an average of 7 to an average of about

17. For the S&P500 composite index, the average P/E for the period 1920 to 2001 was

approximately 17, and for the period 1950 to 2001 it was 17. The variation however can be

dramatic. The P/E ratio for the S&P500 was 7.8 in 1978 and 32.6 in 1998. By 2000, the P/E

had decreased to 26.

P/E ratios are generally depressed when the interest rates and the rates of inflation are high,

such as around 1980-81. P/E ratios tend to be high when inflation and interest rates are low,

such as the period of the mid to-late-1990s, when P/E ratios were at quite high levels by

historical standards. When earnings are growing and the upward profit trend appears to be

sustainable, investors are willing to pay, more for today's earnings. Think of the following

relationships between interest rates and P/E ratios. In 1982, yields on 10- year treasury

bonds were approximately 13%, and the P/E ratio on the S&P500 index was around 11. In

the late 1990s interest tares were around 65 and the P/E ratio for 1998 was 32.6, and for

1999, 30.5. In 2001, it was 46.

Investors must be careful when using P/E ratios to place them in the proper context. P/E

ratios can refer to historical data, an average for the year, or for a prospective period as such

as the year ahead. Obviously, a significant difference can exist between P/E ratios

calculated using these different definitions. Furthermore as noted earlier, various definitions

of earnings for an index such as the S&P500 are available.

Using the Business Cycle to Make Market Forecasts:

Earlier we established the idea that certain composite indexes can be helpful in forecasting

or ascertaining the position of the business cycle. However, stock prices are one of the

leading indicators, tending to lead the economy's turning points, both peaks and troughs.

What is the investor who is trying to forecast the market to do? This leading relationship

between the stock prices and the economy must be taken into account in forecasting likely

changes in stock prices. Stock prices generally decline in recessions, and the steeper the

recession, the steeper the decline. However, investors need to think about the business

cycle's turning points months before they occur in order to have a handle on the turning

Investment Analysis & Portfolio Management (FIN630)

points in the stock market. If a business cycle downturn appears to be likely in the future,

the market will also be likely to turn down some months ahead of the economic downturn.

We can be somewhat more precise about the leading role of the stock prices. Because of this

tendency to lead the economy, the total return on stocks ( on an annual basis) could be

negative (positive) in the years in which the business cycle peaks( bottoms). Stock prices

have almost always rising as the business cycle is approaching a trough. These increases

have been large, so that investors do well during this period. Furthermore, stock prices often

remain steady or even decline sharp rise as the bottom is approached, a period of steady

prices or even a decline typically occurs. The economy, of course, is still moving a head

based on the above analysis.

Based on the above analysis:

1. If the investor can recognize the bottoming out of the economy before it occurs, a

market price can be predicted, at least based on past experience, before the bottom is

hit. In previous recessions since World War II, the market started to rise about half

way between GDP starting to decline and staring to grow again.

2. The market average gain over the 12 months following its bottom point at about

36%.

3. As the economy recovers, stock prices may level off or even decline. Therefore, a

second significant movement in the market may be predictable, again based on past

experience.

4. Based on the most recent ten economic slumps in the 20th century, the market P/E

usually rises just before the end of the slump. It then remains roughly unchanged

over the next year.

The value of being able to analyze business cycle turning points as an aid to market timing

is obvious. Investors would have increased their returns, over the entire sweep of US

economic history, by switching into cash before the business cycle peaks and into stock

before the cycle reaches its trough. It is particularly important to switch into stocks before

business cycle troughs. However, as our discussion in Chapter 11 about market timing

indicated, the chances of timing the market successfully on a regular basis are small.

Using the Fed's Model to Make Market Forecasts:

The Fed has developed a market forecasting model that has captured considerable attention

both because of its relative accuracy over some time periods and because of its simplicity.

This model has a simple-premise--- because investors can and do easily switch between

stocks and bonds, based on the asset with the higher yield, stock returns will tend to restore

an equilibrium relationship between the two assets.

To measure bond yields, the Fed uses the yield on 10-year Treasuries. Of course, this

number can be observed on an updated basis every day. To measure stock yields, the

earnings yield is used---earnings divided by stock price, using the S & P500 index. The

earnings figure used is a forward 12-month earnings estimate based on operating earnings.

Thus, on January 1, 2004, we would use an estimate of operating earnings for the S&P500

Index for the next 12 months through the end of the year. On April 1, 2004, we would use

an estimate of the next 12 month earnings through April 1, 2005.

Investment Analysis & Portfolio Management (FIN630)

The Fed model can be used to formulate decision rules in the following ways:

When the earnings yield on the S&P500 is greater than the 10-year Treasury yield,

stocks are relatively attractive.

When the earnings yield is less than the 10-year Treasury yield, stocks are relatively

unattractive.

An alternative way to use the Fed model is to estimate the "fair value" level of the S&P500

Index and compare it to the actual current index value.

If the estimated fair value of the market is greater than the current level of the

market, stocks are undervalued.

If the estimated fair value of the market is less than the current level of the market,

stocks are overvalued.

Finally, note that the Fed model implies that the reciprocal of the yield on 10-year

Treasuries is an estimate of the S&P500's equilibrium P/E ratio. That is,

Equilibrium estimate of the S&P500 P/E Ratio = 1/Yield on 10-year treasuries

If the S&P 500's actual P/E ratio is less than the estimated equilibrium P/E ratio,

equities are relatively attractive.

If the S&P500's actual P/E ratio is greater than the estimated equilibrium P/E ration,

equities are relatively unattractive.

Potential Problems with the Fed Model:

The Fed Model has the great virtue of simplicity and has given some useful signals, but it is

not without its problems and limitations, which are important to note.

1. As noted above, the model implies a linear relationship between the reciprocal of the

Treasury bond yield and the estimated equilibrium P/E ratio. This means that with a

Treasury bond yield of 4 percent, the predicted equilibrium P/E ratio is 25.

However, with a Treasury bond yield of 2 percent, predicted P/E is 50, and at 1

percent it is 100. Therefore, it is highly probable that the Fed model is not as reliable

when interest rates are unusually low, because the implied linear relationship

overstates the estimated equilibrium P/E ratio.

2. The model relies on the estimated earnings for the S&P500 Index for the next 12

months. There are different estimates of this number, involving top-down, bottom-

up and S&P's core earnings, and they are revised often. Therefore, it is difficult to

determine exactly which number to use at any given time.

3. The model is derived by assuming that the yield on 10-year treasuries can be

substituted for the required rate of return on equities and for the return on equity on

the S&P500 Index. Historically, this has often not been the case.

In conclusion, the Fed model may offer some valuable insights to investors trying to

forecast the future direction of the stock market, but the model should be used with care. It

is by no means a simple solution to the forecasting problem, and could easily mislead

investors, particularly when interest rates are unusually low.

Table of Contents: