As I write, Pfizer's
It can't possibly grow its earnings fast enough to justify paying 50 times every dollar earned in the last 12 months, can it? That's lunacy for a company that size, right?
Heck, yeah, it can. In fact, trying to analyze whether a company is cheap or expensive based solely on its P/E is like trying to determine whether a car is fast based on its paint color. Yeah, fire-engine red tends to be associated with fast cars, but not always. I seem to recall the Ford Pinto coming in a pretty sweet cherry red, and there were buildings that were quicker off the line than a Pinto.
More than meets the eye
In Pfizer's case, the stated earnings under generally accepted accounting principles (GAAP) include a non-cash writedown of impaired goodwill to the tune of $4 billion. Back this amount out, and Pfizer's P/E drops to a much more reasonable 19. But an investor just looking at the number would come away with a singular determination: Pfizer is insanely expensive.
We get asked all the time what the highest P/E we would consider when looking at a company for investment. The question's on the right track, but it's the wrong question. There's no right answer for this, because someone who is looking just at P/E for investing ideas is not nearly doing enough. Maverick Tube
In all cases, there is much, much more to the story than a simple "here are the earnings, here is the ratio" situation. What is correct is that in many, if not most, cases, higher P/E companies have much larger expectations built into them than lower P/E ones. But P/E misses a lot of things.
In fact, I've gotten some great investing ideas by looking for companies with the highest P/E ratios at times, because a company that, for example, has some large one-time expenditures will suddenly have its earnings number distorted, and will seem much more expensive than it is. This is something, by the way, that is an ideal place for pro forma numbers. But do not, under any circumstances, accept company pro forma numbers unless there is every indication possible that its management are straight dealers. Pro forma, since it is an unaudited number, can become a great way for companies to make their performance not only better than it was, but also better than it had any right to be.
It's too bad unscrupulous managements took what should have been a useful tool and made it yet another data point to distrust, but there you go.
People who spend too much time obsessing over P/E multiples are in severe danger of missing the boat. In fact, given that so many people look obsessively at the P/E ratio, you may have an advantage by discounting it, especially for companies that have had extraordinary events take place in the last year. Also, remember this, while GAAP does a great deal to normalize accounting standards between companies, in the end it is still fraught with vagaries and judgment calls. Earnings are little more than an opinion, while cash is fact. This is why we at the Fool have extolled the virtues of concentrating on operating and free cash flow over earnings. A company either has created coin or it hasn't. There's not nearly as much room for opinion.
What are some of the things that can distort earnings?
Extraordinary events. UST had to pay a billion-dollar legal settlement this past year for anti-competitive practices. This legal burden has wiped out any reported earnings, giving it a P/E of N/A. But is UST going to get socked for a billion in legal liabilities every year? Not a chance. As such, UST's reported earnings are way out of whack from where they would normally be. Same with Pfizer and its $4 billion in goodwill writedowns -- it distorts earnings, but not ongoing earnings.
On the flip side, although I don't know what's coming up next for the housing market, I'd say that New Century has just been through the best market conditions for current earnings that it will ever see. Counting on a repeat, or several of them, is most likely folly.
Working capital changes. Earnings are an accrual construct. A company can ship a product and not get paid for it for a long time, but still count the revenues it will at some point receive. Some companies have gotten in trouble when they've taken this very simple theory and been a bit too aggressive with it (known as "channel stuffing"). Earnings substantially higher than operating cash flows with big working capital adjustments are generally considered of poorer quality.
The capital expenditure fairy. Earnings do not take into account current capital expenditures. Companies only get charged for depreciation of capital goods against current earnings. This is proper, but companies also have some leeway into how they can depreciate.
Many companies depreciate much faster for their tax returns than they do for the financials they file with the SEC. Why? It benefits them to have a lower reported earnings for the IRS since they won't have to pay as much tax, while it benefits them to have as high earnings as possible to report to investors, because people pay multiples off those earnings. Companies with big reported earnings but consistently massive capital requirements are not as financially strapping as they may let on.
Non-cash expenses. There are now some 250 companies expensing stock options, but those that do not are reporting dramatically higher earnings than will ever be available to their common shareholders. Even worse are those that reprice options.
Before you buy into a company, take a look at its history -- if it has ever repriced stock options, you may take its general and administrative account on the income statement and throw it out the window. A company with big stock option packages for three straight years and then a $200 million charge in Year 4 for repricing doesn't have to go back and refile its financials for the three previous years, even though by all rights that $200 million consisted of expenses incurred for employees over all four years, not one "extraordinary" one.
Derivatives and pensions. In particular, financial firms use a great deal of derivatives, and though there are rules in place to value them, derivatives are in no small measure marked to a model or, as Warren Buffett calls it, "marked to myth." No one can give a good answer for how much they've made, but under GAAP, companies have to.
Pensions are similarly marked to an actuarial model instead of to real performance. In 2002, General Electric's
Long tail cycles. Three hundred sixty-five days is an artificial length of time from a business perspective. It is also dizzyingly short. Worrying what a company made in the last year may be overstretching the point. What you want to know is how much a company will make in a normal year. That requires noodle work, and that's why people got in trouble in the late 1990s, thinking telecommunications companies would grow to the sun, because their earnings were growing (and they were, even if some cheated badly.) Berkshire Hathaway
This is not to say that the P/E isn't important. It really, really is. A company that cannot earn more than its costs won't be in business for long without substantial sustenance from the capital markets. But that much of the market treats the P/E as the be-all and end-all in measures of how expensive a stock is means that those who wish to dig a little deeper could find values in unexpected places.
Bill Mann, TMFOtter on the Fool Discussion boards
The old folksinger lays it down, not for long, no longer ignored. Bill Mann holds shares in Berkshire Hathaway and has beneficial interest in UST and Pfizer. The Motley Fool is investors writing for other investors.