Over the course of the past several years, I've noticed an increasing number of people referencing "Operating Earnings" as a proxy for S&P 500 (SPY) profitability, often without any regard for either the definition or implications of this measure. The problem is that S&P's Operating Earnings are overly optimistic and arguably misleading.
Firstly to clear up any possible confusion, the term "operating earnings" (along with operating income and operating profit) is commonly used interchangeably with EBIT (earnings before interest and taxes). By factoring out the effects of capital structure (interest) and the sometimes volatile tax components (both of which are unrelated to day-to-day operations), EBIT is designed to give an idea of how a company performed operationally. This makes it a useful tool for examining the prospects for continuing profitability. But this is not what S&P is referring to when it uses the term "Operating Earnings."
Instead, S&P uses the following definitions:
Actual As Reported Diluted earnings, sometimes called Generally Accepted Accounting Principal (GAAP) earnings, is income from continuing operations. It excludes both discontinued and extraordinary income. Both of these terms are defined by Financial Standards Accounting Board (FASB) under GAAP. As Reported earnings represent the longest monitored earnings series available today.
Actual Operating Basic earnings then excludes 'unusual' items from that value. Operating income is not defined under GAAP by FASB. This permits individual companies to interpret what is and what is not 'unusual'. The result is a varied interpretation of items and charges, where the same specific type of charge may be included in Operating earnings for one company and omitted from another. S&P reviews all earnings to insure compatibility across company and industry groups.
Basically, S&P takes Diluted As-Reported Earnings (which are analogous to diluted EPS), removes the effects of dilution, and then removes unusual items. Unusual items include events such as: fixed-asset write-downs, inventory impairment, goodwill impairment, gains and losses on the sale of assets, etc.
There are a number of reasons why I believe As-Reported Earnings are a much better metric, each of which I will detail below.
- Unusual items actually aren't unusual for the broad market. This is the most glaring issue with Operating Earnings. Given the number and diversification of companies in the S&P 500, it shouldn't be surprising to have unusual charges at one company or another; an event, which is abnormal for an individual can be entirely normal for a large population. For example, you might be surprised to learn that your friend's friend is an Olympic athlete, but it would not be the least bit surprising to learn that someone is an Olympic athlete. It's the same thing with companies. A major asset impairment is unlikely for any one company, but across the entire market it's reasonable to expect to find several major impairments in any given year. They will probably occur at different companies from year-to-year (just as there are new Olympic athletes), but there will be a certain level of impairment charges that persist somewhere in the market. For this reason, it does not make sense to exclude unusual items when analyzing the S&P 500.
- Operating Earnings have historically exceeded As-Reported Earnings by 10-15%. This upward bias implies that unusual losses are systematically excluded to a greater extent than unusual gains. The problem here arises when investors take Operating Earnings and apply a "normal" P/E of around 15-17 to it, forgetting that 15-17 times earnings is only considered to be normal because of a long history of P/E ratios, which are based on As-Reported Earnings. Stated differently, a multiple of 15-17 has been the historical norm for As-Reported Earnings, however because Operating Earnings typically exceed As-Reported by 10-15%, a multiple of 15-17 is 10-15% above normal for Operating Earnings. Currently, Operating Earnings are 96.81 compared to As-Reported Earnings of 86.41, a difference 10.40 per share (Source: S&P Data). If you apply a multiple of 15-17 to Operating Earnings, you will overvalue the S&P 500 by roughly 150-175 points.
- Diluted vs. Basic. This point is straight forward. As seen in S&P's definitions, the As-Reported Earnings number takes into account the effect of potential dilution from things like employee stock options, warrants, convertible debt/preferreds, etc, while Operating Earnings do not. This adds to the upward bias of Operating Earnings. Generally, a prudent approach favors the use of diluted earnings over basic. This is a practice strongly supported by Ben Graham.
- Write-downs can enable higher future earnings. When an asset is written down, depreciation estimates are redone based on the now lower asset value. This generally lowers future depreciation expense and allows a company to report higher profitability following an impairment charge. Similarly, Cost of Goods Sold is based on the balance sheet value of inventory. If inventory was previously written down, future COGS will be lower, leading to higher gross margins. It is fundamentally flawed to exclude the write-down itself, but include the higher earnings that the write-downs later allow for.
- To examine the previous point from a slightly different perspective, consider that during the 2008 financial crisis, there were massive amounts of losses at banks and financial institutions due to bad mortgage debt. A lot of these losses are omitted by Operating Earnings since they are classified as unusual (Operating Earnings in Q4 2008 were $23 per share higher than As-Reported, compared to the current quarterly difference of ~$2.50 per share). The problem is that during the years preceding the financial crisis, banks and financial institutions were able to report record profits from the very same mortgages that were later written off, and these profits are still included in Operating Earnings. The losses are ignored, yet the profits they enabled are still included. It makes little sense that a bank can issue large amounts of bad debt, realize huge profits, and then have the subsequent (and completely justified) write-offs omitted. If the profits are included, the losses need to be as well, and Operating Earnings do not do this.
It's worth mentioning that S&P reviews earnings to "insure compatibility across company and industry groups," however they provide limited transparency regarding the specific adjustments they make. Not only that, but I feel some slight trepidation about one of the ratings agencies, who so infamously and thoroughly botched credit ratings prior to the GFC, adding a layer of subjective judgment to earnings figures.
However, regardless of any adjustments the S&P makes, it's clear that based on the differential between As-Reported and Operating, significant amounts of unusual losses are consistently excluded from Operating Earnings, and as the above points illustrate, this is not sound methodology.
When viewing special charges, Ben Graham issued the following caveat in The Intelligent Investor:
Suppose that any time a company had a loss on any part of its business it had the bright idea of charging it off as a "special item," and thus reporting its "primary earnings" per share so as to include only its profitable contracts and operations? Like King Edward VII's sundial, that marked only the 'sunny hours.'
This is exactly what Operating Earnings do. To exclude unusual charges across the entire market is to ignore a significant portion of the economic costs associated with corporate earnings. Furthermore, as I pointed out, unusual items cease to be unusual in the context of a broadly diversified index like the S&P 500. Consequently, my personal feeling is that S&P's Operating Earnings are unduly inflated and should not be used (especially in conjunction with a "normal" P/E multiple).