Read and give your opinion on the following text:
ANALYSIS TOOL #1: ANALYZING ANALYSTS’ DATA
Disregarding day-to-day fluctuations, stock prices typically reflect the market’s expectations for the underlying company’s future earnings growth. Other things being equal, shareholders make money when expectations rise, and lose when they decline.
Stock market analysts came under fire in recent years, first for advising us to buy ridiculously overpriced tech stocks at the market peak in 2000, then in 2001 for urging us to buy Enron shortly before the en-ergy-trader collapsed, and again in 2002 when government investiga-tions revealed that some analysts were advising investors to buy stocks that they themselves believed were losers.
Despite their low repute, analysts’ buy/sell ratings and earnings forecasts are the best measure of the market’s expectations for a specific stock. Although their buy/sell ratings per se won’t help you make mon-ey, you can find important information embodied in analysts’ ratings and estimates and in their research reports.
The Sentiment Index described later in this chapter will help you determine if market expectations best qualify a stock as a value or as a growth candidate. Analysis of analysts’ earnings growth forecasts and of the firm’s surprise history will help you to further define the stock as a viable growth or value candidate. Revenue forecasts are a powerful new tool that you can use to validate the reasonableness of earnings growth forecasts.
Who Are the Analysts?
Stock analysts come in two varieties: buy-side and sell-side. In-vestment bankers, including most full-service brokerages, hire sell-side analysts to research stocks of interest. Originally, brokerages derived most of their income from commissions on stock sales, hence the term, sell-side analysts. These days, investment banking accounts for the li-on’s share of full-service brokerage income, but the sell-side label is still applied.
Brokerages employ scores of analysts. Each typically covers a specific industry such as semiconductor equipment or restaurants. Ana-lysts write research reports on their industry as a whole, and on specific companies within the industry. The analysts devise sales and earnings forecasts, buy, hold, or sell recommendations, and target prices for com-panies they follow. They update their forecasts and recommendations after each company’s quarterly report is released and at other times as events warrant. Sell-side analysts ratings and reports are widely distrib-uted, and third parties such as First Call, Zacks Research, and Multex tabulate their ratings and estimates and publish them in the form of an-alysts’ consensus ratings and forecasts.
Mutual funds, pension plans, and other institutional players read the sell-side analysts’ reports, but many also employ their own analysts. These buy-side analysts do their own research and arrive at their own opinions about a company’s future prospects. The buy-side analysts’ reports are rarely publicized.
All of the ratings and forecasts that we hear about or see compiled, come from sell-side analysts. Analysts publish an in-depth report describing the business model, industry, and competitive situation when they begin coverage of a new company. After that, most analysts’ reports are short updates, typically responding to an earnings report or other news affecting the company’s outlook. Each report or update includes the analysts’ current buy/sell recommendation (rating), as well as earn-ings forecasts for upcoming quarters and for the current and next fiscal years. The report also provides background information justifying changes in the ratings or forecasts.
The point of an analyst’s report is to advise clients whether to buy or sell the company’s stock. However, it’s not that simple. For in-stance, an investor planning on holding for five years might be buying a stock that an investor with a shorter horizon is selling. So analysts de-veloped a variety of rating variations, and, to further confuse the issue, each brokerage has its own terminology.
For instance, Goldman Sachs will put a stock on its Recommended List if it thinks it’s going up in short order, while Merrill Lynch might label it a short-term buy, and Prudential Securities would say strong buy. The only way you can be sure of a specific ratings definition is to consult the brokerage’s rating explanation, sometimes included with the report and sometimes not.
It’s not as complicated as it sounds, because in my experience, all analysts’ ratings other than strong buy equate to sell. Regardless of the spin, anything short of strong buy means that the analyst is not ex-cited about the stock’s prospects and probably wouldn’t add it to his own portfolio.
“Sell” Is a Four-Letter Word
Sell-side analysts are real people like you and me who happen to have very well-paying jobs. You can’t blame them for wanting to hold onto those jobs. Their employers, mostly brokerages, derive much of their profits from investment banking, that is, working with companies when they issue more stock, make acquisitions, borrow money, and so forth. How much money is involved? Say an investment banker brings a new company public by underwriting its IPO. The underwriting fee is negotiable of course, but think 7 percent. So a deal is worth 7 percent of $150 million, or $10.5 million, if the new company issues 10 million shares at an initial offering price of $15. When one company acquires another, both firms hire investment bankers to advise them on the transaction, involving fees running into tens of millions of dollars.
With that much money involved, the competition among invest-ment banks to land these juicy contracts is intense. Naturally, the client, say a new company going public, picks the bank it believes will do the best job, meaning the one that will get the most shares sold at the highest price, and equally important, keep the share price up after the IPO. The latter is important because company insiders personally own tons of shares they will eventually want to sell. That’s where the analyst comes into the picture. A highly regarded analyst’s strong buy recommenda-tion can make a big difference in a stock’s trading price. According to a January 29, 2001, Wall Street Journal story, analysts receive “bonuses of several hundred thousand dollars for helping their firms win big un-derwriting deals.” You can connect the remaining dots on your own.
Most public corporations represent potential investment banking business of some sort. Executives at these companies usually have in-centives to keep their firm’s share price up. They may be on bonus plans tied to the share price, have stock options, or own shares outright. It’s understandable that most take it personally when an analysts’ sell rec-ommendation tanks their company’s stock price. Since they’re in a po-sition to get even by diverting investment banking business to another firm, analysts get the picture, and most don’t see anything to gain by ad-vising selling a stock.
Consequently, most analysts don’t issue sell ratings. Instead they say, hold, neutral, or market perform, and the pros know that means sell.
On occasion, analysts do want to advise selling, but some bro-kerages don’t allow sell ratings. So, the policy of the analyst’s brokerage house determines whether a stock will be rated hold or sell, not the ana-lyst’s view of the stock’s prospects. As a rule of thumb, interpret hold, sell, and strong sell ratings as sell.
Analysts use terms such as buy, accumulate, long-term buy, out-perform, and the like, to specify ratings between strong buy and sell. These ratings mean that the analyst isn’t sure which way the stock is headed, at least in the near future. Note: Standard & Poor’s analysts are the exception to the hold means sell rule. S&P does issue sell ratings, and a hold recommendation means exactly that; don’t sell if you own it, but don’t buy it either.
Firms such as First Call, Zacks, and Multex compile the analysts’ individual buy/sell ratings into a consensus figure. They do that by first assigning each brokerage’s individual ratings into one of five categories: strong buy, buy, hold, sell, and strong sell. They assign each category a numerical weight as shown in Table 4-1.
The compiler tabulates and averages the individual analyst’s ratings. For instance, if three analysts all rate a stock strong buy, the total equates to 3, and the average (total divided by number or ratings) is 1. The consensus rating for a stock with one strong buy (1) and one hold (3) would be 2 (4/2), equating to a buy, even though neither of the analysts actually rated the stock a buy. The consensus is 2.7 (16/ 6) if three analysts say hold (9), two say strong buy (2), and one rates the stock a strong sell (5). Table 4-2 illustrates how the ratings are shown on Yahoo’s Research page.
Compiling consensus ratings this way enables you to get a sense of the ratings trend by comparing older ratings to the current value. For instance, you’d see that analysts are getting more excited about a stock if last month’s rating was 2.2 (weak buy) and this month it’s 1.8.
TABLE 4-1 Numeric values for analysts’ buy/sell ratings.
Category Numeric Value
Strong Buy 1
Strong Sell 5
TABLE 4-2 Analysts consensus rating of 2.0, equating to buy for Clayton Homes, even thought none of the six analysts following the stock had rated it buy.
Strong Buy 3
Strong Sell 0
Some may differ on the meaning attached to particular consensus values, but here’s my rule of thumb:
1.0 to 1.5: Strong Buy
1.6 to 2.4: Buy
2.5 to 3.5: Hold
3.6 to 5.0: Sell
Do Strong Buys Outperform Sells?
You’d logically assume that you’d make money buying strong buys and lose money investing in hold or sell rated stocks. However there’s little evidence to support that assumption.
Research on the subject is inconclusive. Some studies show that holds outperform strong buys and others draw the opposite conclusion. One reason strong buys don’t outperform sells is simply that analysts issue many more buys than sells and analysts can make some really dumb calls.
Consider ADC Telecommunications, a maker of broadband equipment for telephone, cable TV, and wireless communications systems. I’ll track the advice given by one analyst who initiated coverage of ADC in February 2000, just before the collapse of the then still soaring telecommunications equipment market. The analyst calls related here are real, but I gave him a made-up name, Andy Analyst, because this analyst was just one of 20 or so making similar calls. Each entry lists ADC’s share price as of the report date.
February 7, 2000 @ $17.25 (All prices adjusted for stock splits.)
Telecom equipment was a hot sector when Andy initiated coverage with an outperform rating calling ADC a “reasonably valued vehicle to gain exposure to the increased levels of investment by communications service providers.” Shortly thereafter, Andy upgraded ADC to buy.
May 19, 2000 @ $29.16
Andy, noting ADC’s strong second quarter results, increased his price target to $38, commenting that the shares, “currently trading at 50 times our calendar 2000 EPS estimate,” were trading “at a discount to the current and expected growth rate.” Can you imagine that? A stock changing hands at a 50 P/E based on current fiscal year forecasted earnings is trading at a discount.
August 18, 2000 @ $42.38
On the day its share price peaked, Andy raised ADC’s price target to $55 after the firm reported results “crushing our top line estimate … and surpassing our $0.15 EPS estimate.” “ADC continues to grow at rates well in excess of its historical guidance of 25 percent to 30 percent increases,” gushed Andy.
October 6, 2000 @ $25.19
With ADC’s share price down 40 percent, and word spreading that some of its customers could fail, Andy publishes a report headlined: “ADCT: Recent Concerns Overdone,” and noted that he had recently raised capital spending forecasts for the telecom sector. Much of the telecom sector capital spending came from upstart telephone companies funded by recent IPOs that by that time were running out of money and couldn’t raise more.
November 29, 2000 @ $19.06
With the stock down 55 percent, Andy stubbornly maintained his buy based on “another solid quarterly financial performance … slightly ahead of expectations.” Andy wasn’t concerned that many of ADC’s telecom customers were facing bankruptcy, because, as he put it: “ADC endorsed consensus estimates for fiscal 2001.”
January 22, 2001 @ $14.25
With news of the plunge in telecom equipment spending now widespread, Andy maintained his buy even after reporting that ADC said it now “expects its current quarter earnings to come in about 50 percent below previous forecasts.”
February 16, 2001 @ $12.06
With the stock down more than 70 percent from its August 2000 peak, and down 30 percent from when he started coverage, Andy finally threw in the towel, changing his recommendation to neutral, which we were supposed to know meant sell. What caused the change of heart? On the day before, Andy noted that rival equipment maker Nortel said that it expected its current quarter results to come in “substantially below current estimates.”
The moral of the story: You’re on your own; you can’t rely on analysts to make your buy and sell decisions for you. But don’t go away, there is important information to be found in analysts’ ratings.
Number of Analysts
Each analyst following a stock works for a different brokerage or investment banker. One brokerage may employ thousands of stock-brokers, and each of those brokers may have dozens, if not hundreds of individual clients. So each analyst’s report potentially reaches tens of thousands of investors. Equally significant, analysts’ research reports circulate to mutual funds and other big buyers.
How many analysts is enough? Most well known larger companies have between 20 and 35 analysts. For instance, in early 2002, Cisco Systems had 33 analysts, Merck had 26, AOL Time Warner had 30, and Microsoft had 28 analysts. Smaller firms that have already attracted interest will have somewhere between 7 or 8 to 15 or so analysts. For instance, Chico’s FAS, a fast-growing women’s clothing store operator, had 11 analysts when I checked. Firms that hadn’t yet garnered much attention may only have coverage from one, two, or three analysts.
The number of analysts following a stock is significant, but its interpretation depends on your perspective. For instance, value inves-tors seek out stocks given up for dead by the growth crowd, who make up the bulk of the market. All other things being equal, analysts go where the action is and usually drop coverage when a company’s stock goes into the tank. So the lack of analyst coverage signals a potential value candidate.
However, often all other things aren’t equal. Sometimes analysts continue covering a down-and-out stock but not because of investor in-terest. Investment bankers figure that the distressed company may have to raise cash by selling off operating divisions, selling bonds, bank bor-rowings, and the like. They know that the resulting fat consulting fees will go to investment banks that stuck with the firm and continued to provide analyst coverage during their dark days.
That’s why, in January 2002, down and outers such as Xerox and Lucent Technologies still had coverage by 11 and 29 analysts, respec-tively. But you can still spot out-of-favor stocks if you take a closer look at the ratings. For instance, here’s the ratings distribution for the 29 an-alysts that were following Lucent:
Strong buy: 3
Strong sell: 0
Most of the analysts covering Lucent were advising selling the stock (holds and sells), and only three issued strong buys. So there wasn’t much enthusiasm for Lucent, even though 29 analysts covered the company.
The analysts’ ratings tell you a lot about the market’s expectations for the company. For instance, you’d interpret the information differently if 20 out of the 29 analysts covering Lucent said strong buy, instead of only three advising buying Lucent’s shares.
I devised a rule of thumb for using the analysts’ ratings to gauge the market’s enthusiasm, or excitement for a stock, at least from the an-alysts’ perspective. I call it the Sentiment Index. You calculate the Sen-timent Index by adding points for strong buy ratings and subtracting points for holds and sells. Regular buys are ignored.
Strong buy: Add the number of “strong buy” ratings.
Hold, sell, and strong sell: subtract a point for each of these ratings.
For example, a stock with three strong buys and no other ratings would score 3. A stock with three strong buys and three buys would still score 3 because the buys aren’t counted. A stock with three holds would score –3. A stock with three holds and one strong sell would score –4.
Interpret negative scores as meaning analysts want nothing to do with the stock. Scores of 9 and above reflect strong enthusiasm.
I’ll give you an example of how the index works in practice. Back in March 1999, Nortel was trading at $15 (all prices split adjusted). With 26 analysts, Nortel had plenty of coverage, but only seven rated Nortel strong buy, four rated it hold, and the rest were at buy. So Nortel’s sentiment score stood at 3.
By October 1999, Nortel’s share price had almost doubled to $29. Out of 28 analysts covering Nortel, 11 rated it strong buy, two were at hold, and the balance rated it buy. Nortel’s sentiment score had moved up to 9, a relatively high number.
By April 2000, Nortel’s share price had soared to $59, and its sentiment index hit 14.
In June, with its stock trading at $63, Nortel’s sentiment score hit 16, which turned out to be its peak level.
In July 2000, Nortel’s share price hit $71, but its sentiment score had slumped to 13.
Nortel’s share price peaked at $82 in August, before beginning its unrelenting slide to $5 in August 2001. By January 2002, Nortel’s sentiment score registered –21, a notably unenthusiastic reading.
Consider ADC Telecom, another high-flyer that I mentioned in an earlier example. I started following ADC in June 1999, when its stock was trading at $12, pretty much where it had been for the previous three years. With 10 of 22 analysts rating ADC a strong buy, and 5 at hold, ADC’s sentiment score totaled 5.
By December 1999, ADC’s share price had moved up about 25 percent, and its sentiment score had moved up to 7.
ADC’s sentiment score peaked at 17 in August 2000, the same month that its stock topped out at $41.
By July 2001, ADC, with a score of –6, was trading at $6, on the way to its $3.50 September 2001 low.
Not all high-flying stocks that I tracked hit teen sentiment levels in their heydays. Market maker Knight Trimark, for instance, doubled in price between March and July 1999. Its sentiment score peaked out at 9, and that was in August 1999 when the stock was already well off its July high.
Here’s how I suggest applying the index—
Stocks with negative scores are clearly in the doghouse, and your best bets. Value priced stocks with scores between zero and 2 reflect weak sentiment and may also be value candidates.
Scores significantly below zero (e.g., –4) reflect strong negative sentiment, and that doesn’t bode well for growth stocks. Sentiment scores of 9 or higher reflect high risk, but that doesn’t mean that they won’t trade higher, as the Nortel example proves.
Scores between –2 and 8 are acceptable, but preliminary evidence indicates that growth stocks with lower scores, in the zero to 2 range, have more upside potential than stocks with higher scores.
Inspiration for the Sentiment Index
My inspiration for the sentiment index came from CNET.com’s (investor.cnet.com) momentum rating. CNET calculates its momentum rating by awarding four points for each strong buy and buy, and sub-tracting two points for each hold and six points for each sell or strong sell. The main difference between CNET’s approach and mine is that I don’t count buy signals, and I give the same weight to holds and sells. CNET interprets high scoring stocks as buy candidates, where I interpret high scores as signaling risk.
Analysts consensus earnings forecasts are the single most important factor influencing stock prices. Changes in consensus forecasts often precipitate major stock price moves. You can find analysts’ forecasts on most major financial sites, but at this writing, Yahoo shows you more data than other sites, and displays all of the information on a single page, making it easier and faster to access. Consequently, I’ll use Yahoo’s format to explain how to use consensus forecasts.
Earnings Growth Forecasts
You would think that the consensus forecasts reflect sophisticated statistical processing of the analysts’ raw forecasts. But in fact, consensus forecasts—the numbers that determine whether a company’s stock moves up or down on report day—are simple averages of the individual forecasts.
For example, say four analysts publish forecasts for a company, and three expect $1 per share, while the fourth predicts a break-even quarter, that is, no earnings. The average of the four estimates is $0.75, even though no one expects the company to earn $0.75.
Yahoo displays consensus earnings per share (EPS) forecasts for the company’s current quarter, next quarter, current fiscal year, and next fiscal year (see Table 4-3, which is based on Yahoo information). Yahoo shows you the number of analysts making estimates, the high and low estimates, and the year-ago reported EPS for each period. For example, 12 analysts had made forecasts ranging from $0.28 to $0.31 per share for Accredo Health’s March 2002 quarter, averaging to $0.30, the consen-sus forecast. Accredo had reported $0.18 per share earnings in the year-ago, March 2001 quarter. Accredo’s June quarter marks the end of its fiscal year.
Note: often the year-ago earnings shown on earnings estimate reports don’t match the income statement earnings because analysts typ-ically go along with the reporting company’s preference to use pro for-ma earnings rather than those calculated through generally accepted accounting principles.
TABLE 4-3 Analysts consensus earnings and revenue forecasts for Accredo Health. Get a quote like this on quote.yahoo.com and then select Research from the dropdown menu.
This Qtr. (3/02) Next Qtr. (6/02) This Year (6/02) Next Year (6/03)
Avg. Estimate 0.30 0.30 1.09 1.59
# of Analysts 12 12 12 11
Low Estimate 0.28 0.27 1.05 1.42
High Estimate 0.31 0.38 1.15 1.90
Year-Ago EPS 0.18 0.18 0.66 1.09
Avg. Estimate $176 mil. $158 mil. $651 mil. $859 mil.
# of Analysts 2 2 11 7
Low Estimate $176 mil. $154 mil. $594 mil. $735 mil.
High Estimate $177 mil. $162 mil. $774 mil. $1.4 bil.
Year-Ago Sales $124 mil. $124 mil. $462 mil. $651 mil.
Sales Growth 41.9% 26.8% 40.9% 31.9%
There’s considerable information that can be gleaned from the earnings estimates data.
The difference between the high and low analysts’ estimates, 31 cents versus 28 cents, is typical for a current quarter. Accredo’s June 2002 quarter and both fiscal years’ estimates show wider spreads, sig-naling that those forecasts will likely move closer over time. A wide spread (e.g., 5 cents or more) close to announcement date indicates the likelihood of a significant earnings surprise.
Analysts expected Accredo’s earnings to grow 65 percent year-over-year in its June 2002 fiscal year (1.09 vs. 0.66), and another 46 percent the following year (1.59 vs. 1.09).
Growth investors should focus on stocks with a least 15 percent forecast year-over-year earnings growth. Accredo’s strong year-over-year earnings growth forecasts qualified the stock as an attractive growth candidate.
Value stock candidates will likely have low or nonexistent forecast earnings growth. Consensus growth forecasts exceeding 5 percent signal relatively high expectations. Accredo’s strong earnings growth forecasts would disqualify it as a value candidate.
Forecast EPS Trend
Consensus earnings forecast trends are even more significant than the forecasts themselves.
The forecast trend is the current forecast for a period, say the current fiscal year, compared to forecasts for the same period one, two, or three months ago. A positive trend in forecasts tells you that analysts are becoming increasingly optimistic about the company’s prospects, and a positive earnings surprise is likely. Conversely, a negative trend raises the specter of further forecast reductions and a negative surprise at report time.
The best value candidates will show flat or negative forecast trends. Positive forecast trends signal increasing enthusiasm, which means that it’s probably too late for value investors.
Stocks with flat (no trend) or positive forecast trends are valid growth candidates. But growth investors should avoid stocks with negative forecast trends.
Yahoo’s EPS Trend report (see Table 4-4) shows consensus estimates going back 90 days for each of the two quarters and fiscal years covered. Pay most attention to the fiscal numbers because the quarterly results often fluctuate for a variety of short-term reasons. Ignore $0.01 changes.
TABLE 4-4 Earnings forecast trend report for Accredo Health.
This Qtr. (3/02) Next Qtr. (6/02) This Year (6/02) Next Year (6/03)
Current 0.30 0.30 1.09 1.59
7 Days Ago 0.30 0.28 1.07 1.44
30 Days Ago 0.30 0.28 1.07 1.44
60 Days Ago 0.26 0.27 0.99 1.28
90 Days Ago 0.25 0.24 0.93 1.16
Analysts had been consistently increasing their estimates for Accredo. The magnitude and consistency of Accredo’s positive earnings forecast momentum was stronger than most you’ll find, and reinforced Accredo’s standing as a strong growth candidate.
99 CENTS Only Stores’ consensus forecast trend (Table 4-5) is more typical.
99 CENTS Only Stores’ current fiscal year’s forecasts had barely moved during the preceding three months, and its next fiscal year’s forecast had moved up only 2 cents. 99 CENTS Only Stores’ flat EPS trend signals less enthusiasm than Accredo’s positive trend, but it wouldn’t disqualify it as a growth candidate, and would also qualify it as a value prospect.
TABLE 4-5 99 CENTS Only Stores’ consensus earnings forecast trend.
This Qtr. (3/02) Next Qtr. (6/02) This Year (12/02) Next Year (12/03)
Current 0.23 0.25 1.11 1.34
7 Days Ago 0.23 0.25 1.11 1.34
30 Days Ago 0.22 0.25 1.10 1.30
60 Days Ago 0.22 0.25 1.10 1.30
90 Days Ago 0.23 0.25 1.11 1.32
TABLE 4-6 InFocus’ consensus earnings forecast trend.
This Qtr. (3/02) Next Qtr. (6/02) This Year (6/02) Next Year (6/03)
Current 0.15 0.19 0.90 1.25
7 Days Ago 0.15 0.19 0.91 1.25
30 Days Ago 0.11 0.19 0.91 1.25
60 Days Ago 0.24 0.27 1.17 1.37
90 Days Ago 0.25 0.27 1.18 1.42
InFocus’ earnings forecast trend (Table 4-6) tells a gloomier story. InFocus’ earnings forecasts had been steadily trending down, and its fiscal year forecasts were both down substantially from 90 days before. InFocus’ negative forecast trend was a red flag warning of further cuts and/or a negative earnings surprise.
Long-term Earnings Growth
Many analysts, as a matter of course, estimate a company’s long-term (usually 5 years) average annual earnings growth. Those forecasts are averaged and listed on many sites. The company’s consensus long-term growth forecast is the G in PEG, the valuation method favored by many growth investors (see Chapter 5). Although widely followed, the long-term growth forecasts are not tracked for accuracy. (PEG is the acro-nym for the ratio of the stock’s P/E divided by the expected earnings growth.) Think about it! Have you ever heard of anyone looking up a par-ticular analyst’s long-term earnings growth forecast for say, Microsoft, from 5 years past, and then comparing it to what actually happened? Me neither!
Even so, long-term consensus forecasts work as an expectations gauge. High average annual growth forecasts reflect high expectations, and vice versa. The earnings growth rate for stocks making up the S&P 500 Index has averaged around 15 percent annually over the past few years. Using 15 percent as a base, forecast annual growth rates below 10 percent can be said to imply below average, or low expectations, while those above 20 percent reflect high expectations.
Long-term growth forecasts of 10 percent or less signal value candidates. Growth investors, however, should stick with stocks reflecting at least 15 percent long-term growth expectations.
An earnings surprise is the difference between analysts consensus forecasts and the company’s reported earnings. If the reported earnings come in below forecasts, it’s a negative surprise, and a positive surprise if earnings come in above forecasts.
Surprises are usually quantified in cents, as in, “a two-cent positive surprise.” Absent other overriding factors, a negative surprise of any amount drives the share price down, often sharply. Most companies routinely report a one or two-cent positive surprise, so that amount is not really a surprise and doesn’t move prices much.
Positive surprises of four or five cents—or more—usually do move the share price up, although not nearly as much as a negative sur-prise forces it down. A big positive surprise, say 10 cents, can have a more pronounced effect. Surprisingly, the surprise percentage isn’t as important as the number of cents. A 4-cent shortfall, say a company re-ported $4.04 instead of the expected $4.08, is about as significant as if the company reported $0.08 instead of $0.12.
Although the stock price reacts immediately, a significant surprise can have a longer lasting effect because the event forces analysts to reevaluate their earnings forecasts for coming quarters. For instance, analysts almost always increase their estimates following a large positive surprise.
Although a stock reacts to the magnitude of the surprise in cents rather than percentage, some investors believe that in the event of a pos-itive surprise, the surprise percentage does foretell future price action. That is, stocks with higher percentage surprises gain more in the ensuing months than stocks with lower percentage surprises.
You can see a company’s recent surprise history on most financial sites. Yahoo displays the estimated earnings, actual reported earnings and the cents and percentage surprise for each of the last four reported quarters.
Here were Mylan Laboratories’ last four quarters’ surprises listed in chronological order (oldest first) as of February 2002: +0.02, +0.09, +0.09, +0.06
From the data, it appeared that Mylan, at least in the past year, had been a habitual positive surpriser. Further research might give you a different slant, but based only on its surprise history, it looked likely that Mylan would surprise again on the upside when it reported its March 2002 results.
On the other hand, a history of negative surprises signals risk. Nothing is for sure in the stock market, but a habitual negative surpriser is more likely than the average firm to announce another negative surprise.
Value vs. Growth
Some research shows that growth stocks drop more, percentage-wise, than value-priced stocks in the event of a negative surprise. That’s logical since growth stocks’ valuations imply high expectations, and value stocks, by definition, are low expectation stocks. So for a value stock, a negative surprise is really no surprise since most players already view the company as a loser. Conversely, growth investors expect their picks to surprise on the upside, so a negative surprise is a real surprise.
Research results are mixed on positive surprises. Some studies show that growth stocks outperform value stocks in the event of a positive surprise, while other research shows the opposite result.
Although sometimes it may seem so, analysts just don’t pull their earnings forecasts out of thin air. Rather, they set up detailed earnings models, starting with estimated sales, and then deduct their estimated costs to arrive at their earnings forecast.
Earnings forecasts have been available for years, but the sales (revenue) forecasts used to derive the forecasted earnings have been hard to come by. That changed in 2001 when Yahoo added consensus revenue estimates to its Research page report. The availability of sales forecasts is an important breakthrough that, so far, has gone unnoticed by many investors.
Consensus sales forecasts are most important when analyzing growth companies. Growth investors often don’t realize that a company’s recent earnings growth was driven by a large acquisition or another one-time event and won’t be repeated. Reviewing sales forecasts would alert you to that instance. Other times, slowing sales growth estimates are your first clue that once hot earnings growth is about to slow.
You can see that reflected in Mentor Corporation’s sales fore-casts for its March and June 2002 quarters (Table 4-7). Mentor’s year-over-year sales growth had totaled 20 percent, 22 percent, and 28 per-cent, respectively, in the three quarters prior to March 2002. But ana-lysts were only forecasting around 11 percent year-over-year growth in its March and June 2002 quarters.
Analyzing revenue growth is covered in detail in Chapters 11 and 12.
TABLE 4-7 Mentor Corporation’s revenue (sales) forecasts.
This Qtr. (3/02) Next Qtr. (6/02) This Year (3/02) Next Year (3/03)
Avg. Estimate 0.47 0.47 1.63 1.86
# of Analysts 4 4 4 4
Low Estimate 0.44 0.45 1.57 1.80
High Estimate 0.48 0.49 1.65 1.90
Year-Ago EPS 0.43 0.42 1.40 1.63
Avg. Estimate $87 mil. $90 mil. $323 mil. $364 mil.
# of Analysts 2 1 1 1
Low Estimate $87 mil. $90 mil. $323 mil. $364 mil.
High Estimate $88 mil. $90 mil. $323 mil. $364 mil.
Year-Ago Sales $79 mil. $81 mil. $269 mil. $323 mil.
Sales Growth 11.0% 11.5% 20.2% 12.6%
In October 2000, a new SEC rule, Regulation FD (Fair Disclo-sure) outlawed selective disclosure of all material financial information. Prior to Reg. FD, corporations would routinely disclose important infor-mation, such as whether they would likely meet, beat, or come in below existing earnings forecasts to favored analysts. Other analysts might hear the same news later, if at all. Individual investors were shut out.
Reg. FD shut down that game and now most firms announce such changes in guidance via a press release, in a conference call open to the public, or both. Consensus forecasts change immediately since all analysts get the same news at the same time. Consequently, changes in guidance have the same effect on a stock price as an earnings surprise. However, so far as I know, nobody is tracking management guidance changes in the same way that surprises are tabulated.
I may not follow analysts’ buy/sell advice because of their posi-tive bias, but I still avidly read their research reports. Some don’t say much, but many are filled with essential information about the compa-ny’s business plan, the problems it’s encountering, and the analysts’ take on the competition and industry trends. That’s valuable information that would require days of research if you did it on your own.
Further, you can often deduce that a buy rating really means sell by reading the report. For instance, in late June 2001, investment banker Credit Lyonnais reiterated its add recommendation for Global Crossing, then trading at around $9, but at the same time cut its price target from $25 to $12. Or consider ABN Amro’s mid-August 2001 report reducing its 2002 and 2003 revenue and earnings estimates for Microsoft, while reiterating its add advice. Whether by reducing the target price, or cut-ting earnings estimates, an analyst is signaling reduced expectations, and reduced expectations translate to sell.
Multex Investor offers research reports from dozens of brokerages. Some reports are free if you sign up for a 30-day trial with the broker issuing the report. But most research reports are for sale at prices ranging from $10 to $75 each. At those prices you can run up a big bill if you research many companies. However, many Web brokers offer free research reports to their customers. I suggest starting with your broker’s offerings, and even consider opening a second account with a different broker to access additional free reports.
You won’t fare well following analysts’ buy/sell advice, but their recommendations and forecasts gives you information about the market’s enthusiasm for any stock and can help you qualify stocks as vi-able value or growth candidates. Further, analysts’ research reports can help you to understand a company’s business and competitive standing, and often yield clues revealing the analysts’ real view of the company’s business prospects.
ANALYSIS TOOL #2: VALUATION
How much is a stock worth? If we knew, making money in the stock market would be easy. All we would have to do is buy stocks cur-rently trading below their value, and then simply sit back and wait for them to move up to their “correct value.” Of course it’s not that easy. Unless it pays a meaningful dividend, a stock has no value, other than what another investor is willing to pay.
That said, there are a plethora of stock valuation schemes in use. Many originated when stocks did pay significant dividends. These mea-sures originally valued stocks by calculating the present value of their expected future dividends. That made sense, but over time dividends faded in importance, and now most investors buy stocks for capital ap-preciation and don’t consider dividends in their evaluations. You’d think that given that shift, analysts would have found new ways of val-uing stocks. Many have, but many others simply replaced dividends with expected earnings or cash flow and continue using the same formu-las. That makes sense from an academic perspective, but I doubt that you will find anyone willing to buy your shares at a price calculated by those methods.
In fact, stocks trade at whatever price investors are willing to pay today. The greed, excitement, fears, expectations, and enthusiasm that determine today’s value are impossible to quantify. What you can do is evaluate the reasonableness of the expectations reflected in today’s stock price. This section describes two ways to do that.
Growth at a reasonable price The first, determining the earnings growth rate implied by a stock’s current trading price, although employed by many professionals, is unknown to most individual investors. The second, growth at a rea-sonable price (GARP), applies only to growth stocks and is arguably the valuation formula most widely followed by individual investors and pros alike. That truth undoubtedly goes a long way toward explaining why so many got it so wrong in the 2000/2001 tech debacle. After reading this chapter, all sane investors will no doubt glue the implied growth formula onto their foreheads. However, implied
growth only conveys what is true today. You’ll have to calculate target prices (see Chapter 6) to find out what happens next.
Benjamin Graham, sometimes called the father of value invest-ing, proposed a practical and easily calculated formula for estimating the intrinsic value of a growth stock in his pioneering treatise on fundamen-tal analysis, “Security Analysis,” co-written with David Dodd and first published in 1934. Don’t be put off by the algebraic formulas that fol-low. They’re included to justify the result. In the end, all you’ll have to do is look up implied growth on Table 5-1. Graham and Dodd defined intrinsic value as:
Intrinsic Value = Eps x [8.5 + (2 x forecast annual earnings growth %)]
(where Eps is the TTM earnings per share)
Put into words, Graham said that a company’s intrinsic value is its latest annual earnings multiplied by a factor equal to 8.5 plus twice the projected earnings growth rate.
Later, Graham modified the formula to account for the notion that stock valuations vary inversely with prevailing interest rates. That is, stocks tend to trade at higher valuations when interest rates are low, and vice versa (see Chapter 2). Graham used AAA (highest quality) cor-porate bond rates as a proxy for prevailing interest rates. The AAA cor-porate bond rates were around 4.4 percent when he first devised the formula, so the revised version looks like:
Intrinsic Value = Eps x (4.4/AAA) x [8.5 + (2 x forecast annual earnings growth %)]
(where AAA is the current yield of AAA-rated corporate bonds)
For example, if a company’s latest earnings were $1 per share, the bond yield was 7.2 percent, and analysts forecast 20 percent average annual earnings growth over the next five years, the intrinsic value would be:
Intrinsic Value = $1.00 x (4.4/7.2) x [8.5 + (2 x 20)] = $29.64
The intrinsic value is $29.64, based on November 2001’s 7.2 percent corporate bond yield.
Graham’s intrinsic value calculation, per se, is interesting, but it isn’t of much practical value since it hinges on analysts’ long-term earn-ings growth forecasts. While analysts strive to accurately predict a com-pany’s current quarter’s earnings, they’ll undoubtedly revise their forecasts for the next quarter based on the current quarter’s results. Con-sequently, their long-term growth forecasts are likely to be considerably off the mark.
However, Graham’s formula can be very insightful used another way. If you substitute the current stock price for intrinsic value, and im-plied earnings growth for forecast growth, and then do some algebraic manipulation, you get:
Implied growth rate = P/E (AAA/8.8) –4.25
Implied growth, as I’ve defined it, is the long-term average annual earnings growth that the company would have to achieve to justify its current P/E.
To gain further insight, assume for the moment that the AAA corporate bond rate is 8.8 percent. Then the formula simplifies down to:
Implied growth rate = P/E –4.25
For example, using the simplified formula, a P/E of 50 implies a growth 46 percent average annual earnings growth rate.
According to Graham’s formula, the implied growth rate corresponding to a particular P/E moves in tandem with the corporate bond rate. For example, the market will support higher P/Es if interest rates drop. Table 5-1 shows how it works out. You can use the table to look up the long-term average annual growth rate corresponding to your stock’s P/E.
The current AAA corporate bond rate is available at www.neatideas.com/aaabonds.htm.
Most tech stocks sported P/E ratios well in excess of 50 during the 1998/1999-tech boom. For instance, Cisco Systems, using its July 2000 fiscal year earnings and its July 30, 2000, closing share price, sported an off-the-chart 175 P/E. As you can see in Table 5-1, a 100 P/E implies that the market is expecting earnings growth in the 68 to 96 percent range.
TABLE 5-1 Implied annual EPS growth rates for various AAA corporate bond rates.
P/E 6% 7% 8.8%
10 2% 4% 6%
15 6% 8% 11%
20 9% 12% 16%
25 13% 16% 21%
30 16% 19% 26%
35 19% 23% 31%
40 23% 27% 36%
50 30% 35% 46%
60 37% 43% 56%
80 50% 59% 76%
100 68% 75% 96%
What’s a reasonable annual earnings growth rate expectation? Here’s a look at some well-known tech stocks. I’ll start with sales growth, since in the end, that is what determines earnings growth. Table 5-2 shows each company’s average annual sales growth for the most re-cent five years and for the five years before that. Table 5-3 illustrates how those sales growth rates translated to earnings growth.
TABLE 5-2 Average annual sales growth. (Growth rates are as of the last reported ﬁscal year data available on 12/01/01.)
Company Last 5 years Prior 5 years
Microsoft 23% 38%
Cisco 19% 86%
Qualcomm 27% 55%
Sun Micro 21% 16%
IBM 4% 1%
Dell 43% 57%
Intel 16% 33%
TABLE 5-3 Average annual earnings growth. (Growth rates are as of the last reported ﬁscal year data available on 12/01/01.)
Company Last 5 years Prior 5 years
Microsoft 25% 38%
Cisco 24%* 72%
Qualcomm 119%* losses
Sun Micro 12% 21%
IBM 20% -8%
Dell 57% 43%
Intel 25% 38%
* 2001 ﬁscal year excluded because the company had negative EPS. In these instances, growth data covers the four years ending in ﬁscal 2000.
Most of these companies saw their growth slow in the most recent five years compared to the earlier period. You can draw your own conclusions. My take is that 20 percent to 30 percent annual earnings growth is a realistic expectation for mature tech companies, and 30 percent to 40 percent is reasonable for younger firms.
You can look up the current AAA corporate bond rate on the Financial Forecast Center (www.neatideas.com/aaabonds.htm). What corporate bond rates should you assume for the future? Use history as your guide. Table 5-4 show historical ranges dating to the 1920s. You can draw your own conclusions, but my take is that barring a period of runway inflation, rates are likely to hover in the 6 percent to 9 percent range.
Table 5-1 gives you the earnings growth rate implied by your stock’s P/E. It’s up to you to determine the reasonableness of the implied rate. However, a little common sense goes a long way.
For instance, say that you’re a value investor looking at a candidate with a 5 percent implied annual earnings growth rate. Assume that you think the firm will see earnings growth in the 10 percent to 20 percent range when it recovers from its current plight. For now, you don’t
TABLE 5-4 Historical AAA corporate bond rates.
Years Low High
1920-29 4.6% 6.4%
1930-39 2.9% 5.2%
1940-49 2.5% 3.0%
1950-59 2.6% 4.6%
1960-69 4.2% 7.7%
1970-79 7.1% 10.8%
1980-89 8.4% 15.5%
1990-94 6.7% 9.6%
1995-99 6.2% 8.5%
2000-01 7.0% 8.0%
* Source: Moody’s via Financial Forecast Center (www.neatideas.com).
care whether the market ends up pricing the company as a 10 percent or 20 percent grower because either is well above its current valuation.
On the other hand, say that you’re a growth investor looking at a hot stock priced to grow earnings 50 percent annually. With a 50 percent growth rate priced in, what happens to its stock price if it only achieves 40 percent growth?
Growth at a Reasonable Price
Many growth investors don’t spend much time worrying about the subtleties of stock valuation. Instead, they adhere to a “keep it sim-ple” philosophy. In their view, valuation boils down to earnings and earnings growth.
These investors look for a balance between price and expected earnings growth. Specifically, they want to buy growth at a reasonable price (GARP). The reasonable price is determined by comparing a stock’s P/E to the company’s expected annual earnings growth rate.
PEG and Fair Value
A stock is said to be fairly valued when its P/E equals its growth rate, undervalued when trading at a P/E below its expected growth, and overvalued when trading above. For example, if a company is expected to grow earnings 25 percent annually, its stock is fairly valued when its P/E is 25. It’s undervalued when trading at P/Es below 25, and overvalued when trading at P/Es higher than 25.
PEG is the acronym for the ratio of the stock’s P/E divided by the expected earnings growth.
P/EPEG = Forecast Annual EPS Growth
A PEG of 1 translates to fair value. PEGs below 1 signify undervalued and above 1, overvalued.
Presumably, defining fair value as the condition when P/E equals the growth rate is based on a mathematically elegant principle, but I’ve never been able to find any basis for the relationship. Possibly, just that the concept is widely followed gives it validity.
Most players adjust the fair value definition to market conditions. Few growth stocks traded at PEGs below 2 during the 1998/1999momentum market, so growth investors changed their PEG definition of fair value to 2 (PE equal to twice the earnings growth rate) instead of 1. Doing that isn’t as silly as it sounds. In the end, P/E measures the mar-ket’s enthusiasm for a stock, and most stocks trade at higher valuations during a bull market.
While the definition sounds precise, calculating PEG isn’t cut and dried. The only factor universally agreed is P, the latest closing stock price.
The “E” in P/E
Everyone agrees that the E in P/E is 12-months’ earnings, but which 12 months, and which earnings? Some use the last four quarter’s earnings, adhering to generally accepted accounting principles (GAAP), while others prefer the pro forma earnings favored by many reporting companies and by their analysts because it omits a variety of charges and thus is higher than the GAAP earnings. Most analysts and many money managers prefer, however, to use analysts consensus forecast earnings for the current year. Since we’re talking about growth stocks, forecast earnings are higher than historical earnings, thus reducing the P/E. Besides, using forecast earnings avoids the GAAP versus pro forma debate, since analysts always forecast pro forma earnings.
Pro Forma Earnings
Some companies highlight pro forma earnings instead of GAAP earnings in their quarterly reports. The pro forma earnings calculation omits certain costs that the reporting company deems not representative of its operating perfor-mance. There are no standards defining which expenses should or should not be included in pro forma earnings. It’s up to the discretion of the reporting company.
The earnings growth rate, the G in PEG, could be historical long-term earnings growth, but most participants use analysts consensus fore-cast growth. Here again, there’s room for discussion. Some practitioners use analysts’ five-year average annual earnings growth forecasts, while others prefer the current, or the next fiscal year’s, year-over-year fore-cast earnings growth.
The growth money managers and analysts that I interviewed use PEG, despite the inaccuracies, because it’s close enough. They say that they’re not calculating PEG down to decimals. If the P/E is 20, and fore-cast earnings growth is 40 percent, the stock is undervalued. The logic still works, even if the company ends up growing earnings at 30 percent annually instead of 40 percent.
Realistic Earnings Growth Estimates
Successful emerging growth companies often chalk up supercharged earnings growth in their early years. Sales are growing rapidly, but more important, many are near the breakeven point, and gross profits are just beginning to exceed fixed costs. As revenues grow, higher percentages of gross profits fall to the bottom line, driving earnings up faster than sales.
Eventually, sales growth slows to levels similar to its market sector. The timing depends on the particulars, but in my experience that happens sooner than most market analysts expect.
Once a company is past that initial growth spurt, earnings growth, although volatile quarter to quarter, trends down toward the lev-el of sales growth. That happens because a company can only grow earn-ings faster than sales by increasing its profit margins, and margin expansion opportunities diminish over time. That said, for the best com-panies you could probably expect annual earnings growth to exceed sales growth by 10 to 20 percent on an ongoing basis. That means that a well-run company may be able to grow earnings from 22 percent to 24 percent annually on only 20 percent sales growth.
Table 5-5 shows historical long-term sales growth rates for a few representative industries. The best companies exceed these industry averages by taking market share from weaker competitors.
Estimating Maximum Growth Using ROE
Many analysts say that return on equity (ROE) defines a firm’s maximum earnings growth rate. That concept has mathematical validity, and is fully explored in Chapter 11. However, it assumes that the com-pany will not increase its profit margins or raise additional funds through borrowing or by selling more stock.
While high ROE is desirable, many firms do manage to increase profit margins and raise additional funds, and it’s not clear that in practice ROE works well to define a company’s maximum annual earnings growth.
TABLE 5-5 Historical Average Annual Sales Growth.
Industry Average Annual Longterm Sales Growth
Educational Services 16%
Food Processing 7%
Healthcare Info Systems 18%
Household Products 12%
Life Insurance 0%
Medical Services 14%
Ofﬁce Equipment 9%
Retail Stores 10%
Securities Brokerage 16%
Semiconductor Equipment 13%
Telecom Equipment 10%
Wireless Networking 15%
The valuation formulas described here do not take dividends into account. That’s okay for the most part because relatively few stocks pay significant dividends. Some do, however, and in these instances, the div-idend payout should be considered when valuing the stock. One way to estimate the value added by dividends is to divide the annual dividend payout by the AAA corporate bond rate. For instance, if a company is paying $1.00 per share annually, and the corporate bond rate is 7 percent, the value added by the dividend is $14.28 ($1.00/0.07). That equation as-sumes that the dividend payout will continue indefinitely at the same lev-el. Dividends growing over time would warrant a higher valuation.
Market analysts all too often ignore the earnings growth expectations built into the current price when they tell us to buy their favorite stocks. But you can check the reasonableness of their recommendations yourself. Simply look up the current AAA corporate bond rate on the Web and then find the growth rate implied by a stock’s P/E in Table 5-1.
ANALYSIS TOOL #3: ESTABLISHING TARGET PRICES
Many professional money managers compute a target price, the price they expect to sell it at if all goes well, before they buy a stock. The target price defines the potential profit on the investment, and if it isn’t high enough to justify the risk, they don’t buy the stock.
Computing target prices is not the same as determining the rea-sonableness of a stock’s current price (described in Chapter 5). Instead, the target price calculation forecasts a stock’s trading range at some fu-ture time. How far depends on your goals and investing style. Value in-vestors usually find themselves analyzing distressed companies, and they don’t know when a candidate will regain its footing. So it’s neces-sary for them to look three to five years ahead. Growth investors usually have shorter timeframes and may only look 12 months to 18 months ahead, typically to the end of the next fiscal year.
The beauty of the target price method is that it doesn’t matter. The accuracy of your target prices depends on the accuracy of your assumptions, not the time span. Once you understand the process, you can vary the number of look-ahead years to suit your needs.
Why is setting target prices important? Consider an example. Say that you’re analyzing two stocks, both in similar businesses and both are currently trading for $30 per share.
Now assume that after analyzing relevant factors, you determine that if all goes as you predict, two years from now Stock A will be trad-ing between $35 and $40, while Stock B will be trading in the $60 to $70 range. Your analysis could be wrong, of course, or events may not go as expected, but given that information, most would agree that Stock B pre-sents the better opportunity.
The target price approach, unlike other valuation methods, doesn’t use analysts’ earnings forecasts in the computation. On the other hand, consensus sales (revenue) forecasts, if available, can be very helpful.
Because it relies on historical performance, the target price approach is most effective analyzing companies that have been in business long enough to amass a significant track record, say seven or eight years minimum.
Developing target prices involves seven steps. Don’t be put off by that. You can do the whole calculation in less than 10 minutes.
Forecast sales in the target year.
Estimate proﬁt margin in the target year.
Compute the target year net income by multiplying your sales forecast (Step 1) by your estimated proﬁt margin (Step 2).
Estimate the number of shares outstanding at the end of the target year.
Estimate the target year EPS by dividing your net income forecast (Step 3) by your estimated number of shares outstanding (Step 4).
Estimate the likely high and low P/E when the target year’s results are announced.
Compute high and low target prices by multiplying your estimated EPS (Step 5) by your forecast high and low P/Es (Step 6).
I’ll demonstrate the process by estimating pharmaceutical maker Alpharma’s target prices for early 2005, after its 2004 fiscal year results would have been announced. I wrote this chapter in April 2002, so I was estimating Alpharma’s target prices a little less than three years into the future. Alpharma had been a steady grower, but it made several missteps in 2001, dumping its share price and making it a potential value candidate.
FIGURE 6-1 MSN Money’s Key Ratio 10-Year Summary report for Alpharma. From MSN Money’s main page, get a stock quote, then choose Financial Results under Research, and click on Key Ratios. You’ll also need the 10-Year Financial Summary report. Click on Statements and then select 10-Year Summary from the Financial Statements dropdown menu.
The analysis is best-done using fiscal year data. MSN Money is the best resource because its Financial Statement and Key Ratio 10-Year Summary reports list sales, profit margins, shares outstanding, and P/E data going back 10 years. Figure 6-1 shows MSN Money’s Key Ratio 10-Year Summary report for Alpharma.
(1) Forecast Target Year Sales
You have two resources at your disposal to help you forecast your target year sales: (1) MSN Money’s 10-Year Summary lists annu-al sales going back 10 years, and (2) Yahoo’s Research report shows analysts consensus sales (revenues) forecasts for the current and next fiscal year.
You can use the 10-year sales history by itself to forecast future sales growth, you can use the consensus sales forecasts, or you can use both resources.
The most appropriate approach depends on the circumstances. Yahoo’s consensus forecast is often your best resource if you’re looking only a year or so ahead, and a combination of the two is usually best if you’re looking further ahead. Circumstances often call for exceptions to these guidelines, as the Cisco Systems example later in this chapter illustrates, so use common sense.
To demonstrate the process, I’ll estimate Alpharma’s 2004 sales using its 10-year sales history alone, and then in combination with the consensus sales forecasts.
USING HISTORICAL SALES HISTORY
MSN Money listed Alpharma’s sales going back to 1992, as shown in Table 6-1.
Alpharma’s sales grew in every year except 1996. Sales in-creased by about $137 million in 1999, and by another $177 million in 2000, before the firm stumbled in 2001. The most recent years before the stumble are probably the best predictors of future growth, so I aver-aged the 1999 and 2000 sales growth figures to come up with $157 mil-lion. I rounded that number down to $150 million to be conservative, and I settled on $150 million as representative of Alpharma’s typical (normalized) annual sales growth.
I figured that Alpharma was likely to spend 2002 retrenching and reorganizing and it wouldn’t see significant growth until 2003. So I estimated 2002 sales at $975 million, and then estimated $150 million annual sales growth in 2003 and 2004, bringing estimated sales in 2004 to $1,275 million.
TABLE 6-1 Alpharma’s historical sales as shown on MSN Money’s 10-Year Financial Summary Report.
Fiscal Year Sales
ADDING REVENUE FORECASTS TO THE MIX
Looking up Alpharma’s consensus revenue forecasts on Ya-hoo’s Research report (quote.yahoo.com, get quote, then select Re-search) showed that analysts were projecting sales in 2002 of $1,100 million, considerably above my $975 million estimate. Consensus fore-casts for 2003 were $1,370 million, which was $195 million above my estimate. Yahoo’s forecasts cover only the current and next fiscal years, so I added my normalized $150 million annual growth estimate to the analysts’ 2003 figure, and that brought the estimated 2004 sales forecast up to $1,520 million.
Researching recent news stories about Alpharma, I found that the analysts 2002 consensus sales estimates were based on Alpharma’s own forecasts. I assumed that Alpharma’s management was probably too op-timistic and that I was probably too pessimistic. I averaged my $1,275 million original forecast with the analysts’ $1,520 million figure to come up with $1,400 million (rounded) estimated sales for fiscal 2004.
(2) Estimate Net Proﬁt Margin
The best way to estimate a company’s future net profit margins (net income divided by sales) is by reviewing its historical performance. Table 6-2 summarizes Alpharma’s historical profit margins. The mar-gins were erratic until 1997, and then they climbed steadily from 3.5 percent in 1997 to 6.6 percent in 2000, before turning into a loss in 2001.
Alpharma’s 2001 setback represents the “problem” that value investors typically look beyond to forecast target prices after the firm recovers.
Given the 2001 setback, I estimated that Alpharma’s profit margins would also be depressed in 2002 but would recover in 2003 and return to historical levels by 2004. I forecast that Alpharma’s 2004 margin would at least reach its 1999 figure of 5.3 percent, and I used that number for my 2004 estimated profit margin.
TABLE 6-2 Alpharma’s net proﬁt margin history.
Fiscal Year Net Proﬁt Margin (%)
(3) Compute Net Income
Net income is sales multiplied by profit margin. I’ve forecasted Alpharma’s 2004 sales and profit margin, so multiplying the two figures gives Alpharma’s estimated net income for 2004.
Net Income = Sales x Profit Margin
2004 Net Income—$1,400 million x 5.3 percent = $74.2 million
Alpharma’s estimate 2004 net income is $74.2 million.
(4) Estimate Shares Outstanding
Most firm’s number of outstanding shares increases annually be-cause they issue stock to raise money, make acquisitions, or provide em-ployee stock options. You can use the 10-year history of shares outstanding shown on MSN Money’s 10-year Financial Summary re-port to gauge the historical annual share inflation and estimate the num-ber the number of shares outstanding at the end of your target year, in this case, 2004 (Table 6-3).
TABLE 6-3 MSN Money lists Alpharma’s number of shares outstanding at the end of each ﬁscal year.
Fiscal Year Shares Outstanding (Millions)
MSN Money showed that Alpharma had 37.9 million shares out at the end of 2000, and 44.3 million shares out by the end 2001, but only
29.9 shares out in 2000. I figured that the 2000 figure was probably an error and I ignored it. Alpharma appeared to be a habitual share inflator in recent years. I estimated that it would add around two million shares annually, bringing the 2004 year-end total to 50 million shares.
(5) Convert to EPS
Once you have the estimated net income and the number of outstanding shares, you can compute the EPS by dividing the estimated income found by the number of shares:
EPS = Net Income/Shares Outstanding =