Types And Sources
In the U.S., there are two main types of income: as reported and operating. As reported income, sometimes called Generally Accepted Accounting Principal (GAAP), is income from continuing operations and it excludes discontinued and extraordinary income. Both of these terms are defined by the Financial Accounting Standards Board (FASB) under GAAP. As reported earnings represent the longest-monitored earnings series available today.
Operating income, by contrast, excludes unusual items from that value, and companies began reporting it in 1988. The intent for operating earnings is to show how much companies make from their operations (making widgets), excluding corporate expenses and unusual items. Operating income is not defined under GAAP by the FASB; however, companies are required to reconcile how they get from their operating income figures to their as reported income figures. This permits individual companies to interpret what is (and what is not) unusual. The result is a varied interpretation of items and charges, in which the same specific type of item may be included in operating earnings for one company and omitted from operating earnings for another company. Additionally, during difficult times, the term ‘unusual’ appears to be used more liberally.
S&P uses methodology to ensure items included (or excluded) from operating earnings are consistent across sector lines, in order to permit issue comparison. In general, as reported earnings are less than operating earnings, since they exclude corporate expense and write-offs. However, on occasion, the situation occurs where there are unusual income items, which reverses the normal trend.
Things get a bit more complicated for index-level estimates because there are two methodologies used for calculating the S&P 500 index level estimate: bottom-up and top-down. Bottom-up estimates come from the covering equity analyst for their specific issue. We then build up the index level estimate by combining all 500 issue level estimates, after each issue’s estimate is adjusted for its index weight. These are the more commonly used estimates on the street. On an issue level, bottom-up estimates are typically the ones used to determine if a company met expectations or not.
Top-down index-level estimates are usually supplied by economists and strategists on an index and (sometimes) sector level. They use a broad array of indicators and matrices based on economic and market history to project the index level earnings estimate. The economists or strategists never come down to the issue level and say how much a company will make, but instead make their estimates from the top down. They may adjust the index, sector, or group estimate based on overall economic beliefs, such as expected employment, tax rates, government costs, and expenditures. Economists will typically create estimates that project over many years, while strategists will generate them for just a few years (similar to equity analysts).
|S&P Dow Jones Indices|
|S&P 500 Quarterly Earnings Per Share|
Expectation Versus Reality (Or Normalized Versus Methodology)
Operating earnings per share (EPS) is the dominant focus on the street. Equity analysts make their projections based on operating EPS, and most news reports cite them. In this case there are different versions of estimates. The two groups of estimates each have different functions and roles, similar to fundamental and technical analysis, and each group has its strengths. On an index level, a brokerage house may aggregate the S&P 500 earnings estimate based on the estimates supplied to it from their own covering analyst. Since estimates on issues differ, brokerage house index levels estimates differ as well (since their contributing analysts differ).
On a larger scale, several companies gather estimates and calculate a consensus estimate, which is widely reported in the press. S&P Capital IQ (CIQ) is one of the providers that gathers the estimates, so it can be used as an example. The CIQ numbers represent what the covering equity analysts are predicting at a certain point in time, meaning it is closely tied to the analysts’ current expectations and to the market. CIQ takes steps to ensure that when it calculates a consensus estimate for an issue, all of the analysts are predicting the same situation, meaning that if there is an item being excluded or included in the earnings, this is being done uniformly. For that reason, there may be 25 covering analysts but only 22 are used (typically the estimates are uniform).
The consensus number is the estimate which is used to determine a beat or miss. Generally, this estimate also shapes how the market initially reacts. Because of the nature of the market, analysts’ approach sometimes change their estimates, and an item in one industry may not be an item in another industry. Further, even within an industry there can be differences. Again, the strength and intent of the number is to show what analysts who cover that stock are currently thinking. Once the results are released, the number used for CIQ history is the one that matched up with the estimate. Therefore, if an item was excluded in the estimate, it will be excluded from the history. This process is known as ‘normalizing’
S&P Indices (which is separate from CIQ) uses a methodological approach to the S&P 500 earnings. In general, an item is included or excluded across all of the issues in the index. This permits comparisons of sectors and industries, and most of all, companies. Therefore, the result is also compatible with the index price, dividends, balance sheet items, etc. The strength of that number is its defined methodology, uniformity, and comparability–you know exactly what you are getting (there is no black box). Typically, the index level number is used on a higher level to analyze markets, sectors, and industries.
On an issue level, it permits comparison for screening and relative positions. However, many brokerage houses will note or modify certain S&P data on the lower-end (issue or group) so their specific approach to an issue can be incorporated (obviously, each house is slightly different). Since the index data are uniform and methodology-based, this adjustment can be done. The data can be thought of, in the context of building a house, as a series of 2’x8’ logs, each made with the same specifications. For most of the house, those uniform pieces would be exactly what one would want, but for parts of the house that log would need to be crafted. The good news is that the exact specifications of the log are known, so ‘crafting’ is possible.
|S&P Dow Jones Indices|
|S&P 500 Quarterly Earnings Per Share|