Changes in brokerage analyst earnings per share (EPS) estimates, known as estimate revisions, are the most powerful force impacting stock prices. The importance of estimate revisions was known as early as the late 1960s. Yet, it wasn't until the ground breaking article by Leonard Zacks, Ph.D., entitled "EPS Forecasts - Accuracy Is Not Enough," was published in the Financial Analysts Journal in 1979 that the ability to use estimate revisions to produce significant returns was documented for the practical use of a greater audience.

In this exclusive Special Report, Leonard Zacks, co-founder of Zacks Investment Research and Ben Zacks' brother, explains how analysts' earnings estimate revisions and changes in recommendations drive stock prices. He then shows you nine strategies using this critical information provided in the Zacks Advisor that let you build portfolios that will outperform! Whether you have $5,000 to invest or $5,000,000, this valuable report could help you avoid some costly mistakes and make some extremely profitable investments.

It is not common knowledge outside of the industry, but brokerage firms pay over 3,000 certified stock market analysts more than $1 Billion dollars a year to research, then estimate, the earnings of publicly traded companies. When analysts change their estimates, Zacks records estimate revisions. Zacks' philosophy is that these estimate revisions are the most powerful force impacting stock prices.

It is said that the value of a stock is simply what someone else will pay for it. While this is certainly true, over the last 50 years financial economists have worked to move beyond this simple statement and provide us with a basic theory of how value is determined for stocks.

Today, most academics and professional investors agree that the core technique for determining the value of a stock, or the value of almost any other asset, should be discounted cash flow. In the case of stocks, the cash received by the owner of the stock is the quarterly dividend paid by the company up to the date when the stock is sold, at which time the cash received is the price at which the stock is sold. So, if you know the dividend and the price at which you will sell the stock in the future, you need only apply a discount rate to these cash flows to obtain the current value of the stock.

This, of course, is by no means easy since most investors have no idea when they will sell the stock or at what price. The academics respond to this dilemma by saying that the value at which you sell the stock in the future will be determined by the buyer, who will use his own discounted cash flow to determine the price he will pay for the stock. As this iterative process is applied, as explained in Chapter 1 of most MBA investment textbooks, one obtains what is known as the dividend discount model. This model simply says that the current value of a stock is equal to the discounted value of ALL future dividends.

To apply the model, the dividends paid by the stock must be projected into the future many, many years (actually an infinite number). This is the standard way that most professional investors (not to mention the academics) would begin talking about how to determine the value of a stock. For example, the investor, using this valuation approach might say that the dividend would grow 12 % a year for 5 years, 10% a year for an additional 5 years and then grow 6 % per year forever. By applying an appropriate discount rate to this infinite stream of dividends, one then determines the current value of the stock.

Okay, so what does this have to do with EPS? Well, since most companies pay a fixed percentage of their earnings as dividends, the dividend discount model immediately becomes the earnings discount model. And, since we are talking about the price of a share of stock, earnings become earnings per share, or EPS. So, as any first year MBA student could recite, the current value of a stock is proportional to the discounted stream of future EPS.

In other words, both academics and professional investors agree that the key numbers that determine a stock's price are estimates of future EPS. It then follows that changes in stock prices are the result of changes in EPS estimates. Oh, if it were that simple.

The first question is "whose estimates?" and the second question is "estimates for which time period?" Certainly prices are not responding to changes in your own estimates (unless you are buying large blocks of thinly traded stock). However, because of the organizational structure of the investment industry, prices respond strongly to the changing EPS estimates made by analysts at the brokerage firms. To understand why this is true, we need to first understand the role of the brokerage analyst in the investment industry.

Predicting EPS is not easy. To predict EPS, you must predict future sales and also the margin (net income/sales). To predict future sales you must fully understand a company's competitive position in its industry and be able to accurately predict the growth rate of that industry. To predict changes in margins you must understand the details of the company's cost structure. Possible acquisitions and new lines of business must also be considered. This is not a simple task and could very well be a full-time job. In fact, it is the full-time job of two types of people; the manager of strategic planning at the company, and the brokerage analysts who follow the company.

Brokerage analysts are professionals who belong to the Association of Investment Management and Research (AIMR, previously called Financial Analysts Federation). The AIMR has about 33,000 members and about 26,000 are actually analysts. The rest are portfolio managers, salespeople, traders, a few brokers, etc.

Of the 26,000 analysts, 4,000 work for brokerage firms and 22,000 work for professional investment organizations such as mutual funds, investment advisors, trust departments, and pension funds. The 22,000 operate on the "buy side." They buy research from the 4,000 "sell side" analysts who produce the research. These 4,000 "sell-side" analysts are employed by about 250 brokerage firms and in total follow 6,000 companies.

These "sell side" brokerage analysts are the "gurus" of the investing world. They are wined and dined by company management and given just the slightest bit of insider information, or at least hints as to whether their predictions are "on target." Often, these analysts may have worked for one of the companies in the industry in the past and have continued to follow the company as company management has changed over the years. They may have a more realistic perspective on a company's growth possibilities and competitive position in its industry than company management.

The brokerage analyst's job is threefold. He or she is expected to a) predict future EPS for the company, b) maintain a recommendation (buy, sell, hold) for the company, and c) identify investment banking opportunities among companies.

In major brokerage firms, an analyst's entire job may be to "follow" 20 companies, and for those 20 companies, the analyst knows a great deal about future EPS. In contrast, a "buy side" analyst at an investment organization (e.g. a mutual fund) may be responsible for 200 companies. To keep up on these 200 companies, the "buy side" analyst may spend much of his time talking to 10 "sell side" analysts who in total follow the 200 companies. To pay for these conversations and associated reports, the buy side analyst, who works for the money management firm, causes his firm to execute trades with the brokerage firms that employ the "sell side" analysts. So the analysts of the money management firms buy research from the brokerage analysts.

When a brokerage analyst changes his EPS estimate, he makes the change known, using fax, phone and the Internet, to his firm's sales force and to all of the "buy side" analysts who are purchasing his research. The sales force and the buy side analysts make the change known to portfolio managers and stock prices begin to move as the portfolio managers begin making buy/sell decisions.

For any given stock there may be from 1 to 40 brokerage analysts "following the company" and making EPS estimates. For nearly 20 years, Zacks has been tracking these individual analyst estimates and creating consensus EPS estimates.

Zacks summarizes the EPS estimates of these individual brokerage analysts to produce four types of consensus estimates:

1. Zacks Consensus Estimate. This is the average of all EPS estimates made within the last 120 days. Older estimates are not included.
2. 30-Day Consensus. This is the average of all EPS estimates made within the last 30 days.
3. Median Estimate. This is the middle of the estimates (i.e. half the estimates are above the median and half are below).
4. Smart Consensus. This is the average of the analysts with the best historical track record of forecasting EPS for the company.

Zacks calculates these four definitions of consensus estimates for each of the next four fiscal quarters, for each of the next three fiscal years and as a growth rate over the next five years.

Most of the services Zacks offers individual investors are based on the Zacks Consensus Estimate. Our services to brokers and institutional investors include the 30-day Consensus, the Median Consensus, and the Smart Consensus. The value of these other definitions is that they may give you more accurate expected EPS growth rates and P/E ratios, which are key inputs of most fundamental stock valuation models.

Estimate Revisions

The real value of brokerage analyst EPS estimates is not in the consensus but in the changes in the individual brokerage analyst EPS estimates.

At any time, brokerage analysts will have an estimate for EPS for each of the next few quarters, each of the next few years and an estimate for the growth rate in EPS over the next three to five years.

When the analysts change any of these estimates, Zacks records an estimate revision and identifies the analyst who made the change. We receive daily electronic files from almost all of the 250 brokers and, during an average week, we record over 25,000 estimate revisions made by these brokerage firm analysts. These estimate revisions are delivered to Zacks Advisor subscribers through our daily e-mail alerts.

Theoretically, when analysts increase their EPS estimates, investors' expectations of future earnings, which translate into future dividends, increase and the value of the company and its stock price increases. This relationship has been measured hundreds of times by academics ever since Zacks released its history tapes to universities for use in academic research.

The general conclusion is that, on a coincident basis, this is one of the strongest relationships in finance. When the consensus estimates increase, stock prices usually increase. When consensus estimates decrease, stock prices usually decrease. However, this relationship is relative to the overall market. If the market is going up, then those stocks for which analysts are raising their estimates will go up more than the market, and those stocks for which analysts are not raising or lowering EPS estimates will go up less than the market.

Conversely, when the market is going down, those companies with upward estimate revisions will drop less while those with downward estimate revisions will drop more.

These are statistical relationships that are true for groups of companies. For a single company, the chance that this will hold may be 65% to 70%. But then the chance of winning in Las Vegas is 45%. So the difference between Las Vegas and the market is a 20% edge. If you would like to read some of the academic articles written on this subject, check out the bibliography at www.zacks.com by clicking on Academic Research.

So if one could predict future estimate revisions, one could predict stock price changes. One effective way to predict future revisions is to analyze patterns in EPS surprises.

When a company reports quarterly EPS that differs from the consensus estimate of EPS, the stock price generally reacts immediately, or at least within a few hours. This change is too fast for most individual investors to exploit.

The real impact of the EPS Surprise occurs over the next few weeks as the brokerage analysts dig through the quarterly financials, talk to company management and decide why/if they were wrong. They may then change their EPS estimates for the next quarter or for the full year. They have learned something from the EPS report. It is these changes in EPS estimates that have a long-term impact on the stock prices.

Almost all companies report quarterly financials and generally release their quarterly income statement and partial balance sheet to the news wires a few days to a few weeks after the end of their fiscal quarter (80% of the fiscal quarters end in March, June, September and December). They then report to the SEC EDGAR system a few weeks later. Zacks monitors the news wires on a daily basis and within one hour from the time a company first releases its quarterly EPS, we compare it to the Zacks consensus, calculate an EPS surprise and do an analysis of the surprise. The surprise is defined as (actual EPS - Consensus EPS) / Consensus EPS. Both the actual EPS and the consensus EPS are adjusted by Zacks to exclude non-recurring items.

Zacks EPS Surprises, using both the full consensus and the 30-day consensus, are published daily in the Wall Street Journal during EPS reporting season. The EPS release date can be found at www.zacks.com under "Earnings Calendar" and is sent to you for the stocks in your portfolio as part of our daily e-mail alert service.

Zacks was the first firm to calculate consensus EPS estimates of quarterly earnings in 1981, and to create the concept of EPS Surprises. Ernesta Tallet Kepsla, a Zacks analyst in 1982, was the first to analyze the price response to EPS Surprises. Prior to her work, academics used trend lines through historical quarterly EPS to create a prediction of next quarter's EPS and then used that prediction to calculate what was called Standardized Unexpected Earnings (SUE).

Zacks replaced the concept of SUE with EPS Surprises by using analyst estimates as the forecast against which actual EPS was compared. We like to think that Zacks was instrumental in moving the idea of estimate revisions and EPS surprises into the mainstream of investing. Zacks was also the first firm to develop a quantitative stock selection model based on estimate revisions.

The Zacks Rank is a quantitative model based on trends in estimate revisions and EPS surprises that classifies stocks into five groups, #1, #2, #3, #4, and #5. The Zacks Rank can be used to predict future movements in stock prices.

Ben and I first approached institutional investors in 1980 with the Zacks Rank and the belief that trends in estimate revisions and EPS surprises could predict future stock prices. Though the institutional investors responded with disdain to our ideas and insisted that the market was efficient, we persisted. Each week we calculated the Zacks Rank and followed the stocks at the top of the Rank and those at the bottom. Over time, our evidence grew.

The portfolio of the Zacks Rank #1 stocks has had an annualized return of 35.1% per year since 1980 vs. the S&P return of 17.4% during the same period. The actual performance of the Zacks Ranking System during this 19-year period is shown below. This is not a backtest. It is the performance of portfolios created each month using published Zacks Ranks as of the end of the previous month.

The Zacks Rank is based on four factors: Agreement, Magnitude, Upside and Surprise. Agreement is the extent to which all brokerage analysts are revising their EPS estimates in the same direction. Magnitude is the size of recent changes in current fiscal year and next fiscal year consensus estimates. Upside is the deviation between the most accurate EPS estimates and the consensus. Zacks monitors the track record of almost 4,000 brokerage analysts, so we know, for any company, who has been the most accurate in forecasting EPS in the past. Surprise is based on the last few quarterly EPS surprises.

These four factors are combined into a composite score, which is then used to create the Zacks Rank. Generally 5% of the companies will have a Zacks Rank = 1, 15% will have a Zacks Rank = 2, 60% will have a Zacks Rank = 3, 15% will have a Zacks Rank = 4, and 5% will have a Zacks Rank = 5. Over the years the Zacks Rank has substantially outperformed other published ranking systems such as the Valueline Rank.

If the most powerful force is changes in brokerage analyst EPS estimates and these changes are incorporated into the Zacks Rank, what is the role of brokerage firm recommendations in the Zacks philosophy?

After tracking the stock picking success of thousands of brokerage analysts for more than five years, we are now convinced that there is valuable information in changes in broker recommendations and that by using them in conjunction with the Zacks Rank, investors can benefit. Ben Zacks uses both the Zacks Rank and patterns in broker recommendations when constructing the Zacks Focus List.

When using analyst recommendations, you must distinguish between two types of recommendations, 1) the analyst recommendation and 2) the brokerage firm buy list recommendations. A brokerage firm may employ 20 analysts, each of whom is following 20 stocks, so the firm may have recommendations on 400 stocks. But most brokerage firms go beyond this and create an investment committee that may consist of the Director of the Research Department (where the analysts work), the economist, and a senior portfolio manager to pick the best of the best. This committee selects a few of the stocks recommended by their analysts and creates a "buy list" for the firm. This is a portfolio that the brokerage firm presents to the world as its best picks.

Zacks is the only firm in the world to track the full performance of brokerage analysts. We track both their ability to predict EPS and also their ability to make buy/sell/hold decisions. Jointly with the Wall Street Journal, each year Zacks publishes the All-Star Analyst Survey that identifies the top analysts in each of 50 industries. We identify the top 5 in each industry based on their ability to predict EPS and also the top 5 based on their ability to recommend stocks.

We make these analyst track records available to Advisor subscribers as part of the Zacks abstracts of brokerage research reports, and also identify the current stocks recommended by All-Star Analysts as a Zacks Advisor portfolio on our web site.

Each quarter, Zacks Investment Research, Inc. also tracks the performance of the "buy lists" of 15 major brokerage houses and reports the results in the Wall Street Journal. We have a track record for not only each individual analyst employed by the firm, but also for the firms "buy list." Our general conclusions, after over 5 years of results, are:

1. You can make money from the Brokerage firm buy list.
2. You can make money by actively trading on changes in analysts' stock recommendations.
3. You can make money by following the picks of the All-Star Analysts.

In summary, we have found, after years of statistical analysis, that estimate revisions, EPS surprises and changes in broker recommendations can be used to construct portfolios of stocks that will outperform equal weighted indices.